[00:00:07] <mage2> would the fact that im on solaris 10 x86 change the name of the file im looking for
[00:00:41] * gisburn downloads a HappyTreeFriends "smoochie" in the meantime to torture a little sentient being.
[00:04:07] <sickness> I go to sleep, nite all :)
[00:05:16] <timeless> does anyone here have a debug kernel handy? (or a symbol'd kernel)
[00:05:34] *** mage2 has quit IRC
[00:06:49] *** Oski has joined #opensolaris
[00:07:18] <timeless> interesting
[00:07:21] <timeless> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6478246
[00:07:30] <timeless> the code i'm playing with is very close to panic'ing code
[00:08:00] <timeless> ok, how do i find out what version of a file is in snv_49?
[00:09:15] *** coffman has joined #opensolaris
[00:09:20] <boyd> Good question
[00:09:24] <boyd> (Morning, all)
[00:09:35] <jamesd_> hi boyd
[00:09:41] <timeless> jamesd: ok, i'm happy
[00:09:43] <timeless> http://cvs.opensolaris.org/source/diff/on/usr/src/uts/common/inet/ip/ip_ire.c?r2=1.146&r1=1.145
[00:09:54] <boyd> Hey jamesd_
[00:10:09] <timeless> i'm pretty sure that change is what fixed the panic i have today
[00:11:07] *** solarisjon has joined #opensolaris
[00:11:51] <schily> dclarke: does star link now?
[00:12:09] *** GmanZZZ is now known as Gman
[00:12:10] 
[00:12:26] *** solarisjon has left #opensolaris
[00:12:27] <timeless> is it possible for me to follow a bug or something?
[00:12:32] <timeless> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6474802
[00:12:33] *** Oski has left #opensolaris
[00:12:35] <timeless> is my crash
[00:15:57] <coffman> yeah great.
[00:16:09] <coffman> patch 121017 is dead locking my system
[00:16:11] <coffman> ffs
[00:16:31] * timeless ponders
[00:16:43] <coffman> uh
[00:16:52] <coffman> now its working again
[00:16:54] <timeless> i guess i can hack [zone]/etc/hosts so that it doesn't let me ping global's localhost :)
[00:22:10] <gisburn> can anyone explain why % stty -icanon min 1 time 0 -inpck  # makes an application over ssh much more responsive ?
[00:22:44] <timeless> can someone do: svcs|grep graphical-login/cde-login
[00:23:42] <boyd> What are you looking for?
[00:23:48] <timeless> x11
[00:23:49] <jamesd_> bash-3.00#  svcs|grep graphical-login/cde-login
[00:23:50] <jamesd_> online         Oct_06   svc:/application/graphical-login/cde-login:default
[00:23:50] <jamesd_> bash-3.00#
[00:23:53] <timeless> which i had a whlie ago and lost
[00:24:05] <boyd> try svcs cde-login
[00:24:23] <boyd> svcs without params won't list it unless it's enabled
[00:26:55] *** sommerfeld has joined #opensolaris
[00:26:55] *** ChanServ sets mode: +o sommerfeld
[00:27:04] <timeless> gisburn: wow
[00:29:23] <Gman> morning dudes! :)
[00:30:43] <gisburn> timeless: "wow" .. what ?
[00:31:38] <gisburn> Gman: morning... happy "chicken execution day"!
[00:31:57] <timeless> oh, it's morning
[00:32:10] <timeless> sommerfeld: so, the problem i was chasing is already fixed in current
[00:32:12] <Gman> who are the chickens?
[00:32:15] * gisburn wishes he could see Gman's face when he reads the chicken thing...
[00:32:58] * Gman has admittedly been giggling through the latest mails on opensolaris/cab-discuss
[00:33:25] *** yippi has quit IRC
[00:33:26] <noyb> maybe he means, turkey not chicken
[00:33:40] <Gman> i can sympathize with josh, because that's how i felt about arc too
[00:33:43] <Gman> [like casper]
[00:34:05] * timeless wonders who josh is
[00:34:07] *** doownek has joined #opensolaris
[00:34:15] <doownek> Hi
[00:34:25] <Gman> timeless, you need to catchup on your email ;)
[00:34:29] <Gman> hey doownek
[00:34:40] <gisburn> oh crap
[00:34:47] <timeless> i don't think i'm on whatever mailinglist you're reading
[00:34:51] <Gman> the process isn't absolutely spot on for opensolaris yet, but the idea is definitely there
[00:34:56] <Gman> timeless, opensolaris-discuss or cab-discuss
[00:35:03] <Gman> there's a thread sort of mixed up in those 2 lists
[00:35:13] <gisburn> oh crap crap crap
[00:35:14] * timeless doesn't have time for that :)
[00:35:26] <Gman> timeless, good call
[00:35:27] <doownek> I've a Solaris 10 3/05 and windows but I reinstall windows then I lost the multiboot, How can I reintall the multiboot please ?
[00:35:58] <sommerfeld> i'm 200 message behind on cab-discuss and 2700 behind on opensolaris-discuss
[00:36:07] <Gman> sommerfeld, painful
[00:36:11] <Tpenta> is that all?
[00:36:54] <boyd> sommerfeld: "Mark all as read" works well for me :)
[00:37:05] <sommerfeld> feels too much like cheating :-)
[00:37:28] <doownek> no solution for my problem ?
[00:37:45] <Gman> doownek, i don't personally know because i don't have that configuration - did you try google?
[00:37:54] *** RaD|Tz has joined #opensolaris
[00:38:12] <doownek> yes but I nothing found :-(
[00:38:15] <RaD|Tz> is there a channel specifically for nexenta?
[00:38:18] * Gman finishes all the unread mail for opensolaris on monday mornings when he does the summaries
[00:38:29] <elektronkind> doownek: boot off the installation cd into failsafe mode, and research the "installgrub" command - google for info on it.
[00:38:30] * boyd grumbles about a certain blog that's syndicated on planetsolaris which hardly ever mentions solaris or anything related to it
[00:38:59] <doownek> elektronkind: the version 3/05, it's not grub
[00:39:01] <Gman> i stopped reading planetsolaris a while ago, because it was pretty broken
[00:39:06] <boyd> RaD|Tz: #gnusol I think
[00:39:08] <Gman> i wish we had planet.opensolaris.org
[00:39:13] <boyd> Yeah
[00:39:18] <RaD|Tz> thanks boyd
[00:39:26] *** ruben has joined #OpenSolaris
[00:39:28] <ruben> Hi, I have a big problem, by error I changed the vfstab with a bad option, now when I start the server I see 'run fsck etc', how can I mount the / as rw to change vfstab?
[00:39:30] *** RaD|Tz has left #opensolaris
[00:39:36] <Gman> instead of the current set of feeds spread across the communities - i doubt anyone looks at those
[00:39:43] <timeless> out of curiosity
[00:39:47] <timeless> can a zone core dump?
[00:39:48] <gisburn> Any C99 guru here ?
[00:39:51] <timeless> err panic
[00:39:55] * gisburn looks at Tpenta
[00:39:56] <timeless> i presume it can't
[00:40:10] <boyd> Gman: I think I'd find planetsolaris about 100 times more useful if it wasn't spammed with that one blog so much
[00:40:23] <gisburn> timeless: panic would AFAIK fatal for the whole system. Zones are just chroot on steroids
[00:40:28] <elektronkind> doownek: ah, not grub. hmm.
[00:40:38] <elektronkind> doownek: then use the installboot command
[00:40:46] *** MindChild has joined #opensolaris
[00:40:58] <elektronkind> doownek: from the man page: installboot /usr/platform/`uname -i`/lib/fs/ufs/pboot /usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/c1t0d0s2
[00:41:01] <boyd> ruben: You may be able to boot to single-user and fix it, or boot from install media
[00:41:19] <ruben> boyd: how can I boot to single-user?
[00:41:21] <elektronkind> doownek: this done after you boot into single user mode from the install cd, of course
[00:41:24] <doownek> thanks, I will try
[00:41:29] <boyd> ruben: x86 or sparc?
[00:41:33] <doownek> x86
[00:41:34] <ruben> it is my problem and mount rw the /
[00:41:38] <ruben> x86
[00:41:45] <doownek> I try now
[00:41:47] <mlh> the CAB mailman archives is only available to list subscribers, but the CAB forum (which I presume is the same thing) is open to all
[00:41:50] * gisburn looks at Tpenta
[00:42:01] * gisburn stares at Tpenta
[00:42:06] <boyd> ruben: get to the grub menu, hilight the normal option, press "e"
[00:42:13] <mlh> could someone please open up the mailman archives?  I hate forums
[00:42:22] <boyd> ruben: then select the line with "multiboot" in it and press "e"
[00:42:28] *** Netwolf has quit IRC
[00:42:34] <boyd> ruben: then add -s to the end
[00:42:34] * gisburn uses his voodo puppet of Tpenta and some nails to get Tpenta's attention...
[00:42:37] *** MindChild has quit IRC
[00:42:48] <boyd> ruben: Then choose (i think) "b" to boot
[00:42:53] <ruben> boyd: uhmm, okey
[00:43:10] * boyd wonders if the failsafe option does that, since he's never used it
[00:43:27] * Tpenta stares back at gisburn
[00:43:36] * boyd fires up parallels to check
[00:43:50] <boyd> Ah.. it does
[00:44:08] <boyd> ruben: Ok, so you could just choose the "failsafe" option in the grub menu :)
[00:44:16] *** eboutilier has quit IRC
[00:44:29] <ruben> hehehe, more easy :)
[00:44:55] <Gman> boyd, heh, see what you mean
[00:45:06] <Gman> mlh, yeah, i think tha'ts pointless
[00:45:07] <boyd> Gman: :)
[00:45:23] <Gman> mlh, send a mail to cab-discuss?
[00:45:27] * Gman will back you
[00:45:53] <boyd> Are they not on pipermail?
[00:46:12] <gisburn> Tpenta: Do you have any document or URL which describes the math functions defined by C99 ?
[00:46:30] <boyd> You could look at the draft standard
[00:46:35] <Gman> boyd, they are, you just have to be subscribed to it, to be able to read them
[00:46:39] *** lacaAFK is now known as laca
[00:46:44] <boyd> Gman: !?
[00:46:53] <Gman> yeah, it's the only list like that
[00:47:00] *** besonen_mobile__ has quit IRC
[00:47:01] <ruben> thanks boyd :)
[00:47:08] <mlh> Gman: ok, ta I will
[00:47:15] *** besonen_mobile__ has joined #opensolaris
[00:47:22] <boyd> Hey that failsafe option is pretty cool. I should have tried it before :)
[00:47:24] <boyd> ruben: np
[00:47:37] <gisburn> boyd: and the draft standard is... where ?
[00:48:18] *** libkeiser is now known as natas
[00:49:59] *** natas is now known as libkeiser
[00:50:06] <stevel> $ hg clone ssh://anon at hg dot opensolaris.org/hg/onnv/onnv-gate
[00:50:13] <stevel> it's up
[00:50:20] <boyd> Yay!!!!
[00:50:27] <ProfMikey> oink!
[00:50:30] <sommerfeld> woohoo
[00:50:33] <boyd> Well done stevel
[00:50:37] <ProfMikey> hi team
[00:51:01] <stevel> thanks boyd
[00:51:06] <stevel> sorry it took so long
[00:51:14] *** koolniczka has left #opensolaris
[00:51:17] <Gman> sweet stevel !
[00:51:18] <ProfMikey> 23:48:09 [Users #Evadia]
[00:51:19] <Gman> nice work
[00:51:28] <ProfMikey> aww
[00:51:31] <ProfMikey> wrong btn
[00:51:32] <boyd> gisburn: Google found me this: http://clc-wiki.net/wiki/The_C_Standard  and this: http://www.dinkumware.com/manuals/default.aspx?Page=#Standard%20C%20Library
[00:51:33] <stevel> now let's stress test it ;-)
[00:51:40] <Gman> stevel, anon working on svn yet?
[00:51:45] <stevel> gman: should be
[00:51:47] <boyd> Sheesh! :)
[00:51:53] <ProfMikey> remind me not to chat when I am drnk
[00:51:53] <stevel> i fixed it around noon-ish PDT
[00:52:00] <Gman> stevel, cool, just need to get dermot to come in...
[00:52:14] <astinus> hi guys :)  could someone quickly clarify a couple of questions about ZFS please? I currently use a RAID5 array of 6 * 250GB disks, another RAID5 of 200G disks, and another RAID-5 of 120G disks .. so I lose a fair bit (n-1 formula). They're on a Linux box I'm looking to migrate to Solaris. Would a ZFS pool be able to take multiple disk sizes?
[00:52:19] <Tpenta> gisburn: not to hand
[00:52:21] <stevel> http://www.opensolaris.org/jive/thread.jspa?threadID=15748&tstart=0
[00:52:26] <stevel> for more info
[00:52:36] <sommerfeld> stevel: is that: RSA key fingerprint is 63:b9:7a:f3:dc:a4:47:a7:a5:50:ef:05:78:39:8d:ad.
[00:52:36] <sommerfeld> ?
[00:53:17] <sommerfeld> astinus: should work fine.  cleanest way is to s/raid-5/raid-Z/ -- one raid-z group of 6x250, one of Nx200, etc.,
[00:53:27] <stevel> yeah - you'll have to accept the RSA host key first
[00:53:32] <boyd> sommerfeld: Good point, it's worth posting that somewhere
[00:53:34] <sommerfeld> astinus: note that with compression you typically get back the parity disk
[00:53:52] <Gman> stevel, dude, opensolaris-announce!
[00:54:21] <astinus> sommerfeld:  So I'll end up with probably more storage than before, decent snapshot ability and all the other magic ZFS providers which Reiser doesn't? Sounds like a winner :D
[00:54:29] <stevel> gman: thought about it, but i thought it would be fine to announce it on tools-discuss.  i wasn't sure if opensolaris-announce was appropriate or not
[00:54:31] <sommerfeld> flash crowd!
[00:54:40] <Gman> stevel, yeah, i think it is
[00:54:48] <Gman> it's an important milestone
[00:55:07] <stevel> it's only a read-only mirrour at this point though.. it's not the full read/write
[00:55:07] <Tpenta> +1
[00:55:11] <coffman> uh
[00:55:13] <Gman> so what
[00:55:28] <sommerfeld> stevel: updated nightly?
[00:55:35] <stevel> sommerfeld: updated real-time :-D
[00:55:44] <sommerfeld> ooh.  better than onnv-clone then.
[00:55:47] <Tpenta> The onnv-gate repository is a live MIRROR of the internal Teamware
[00:55:48] <Tpenta> onnv-gate workspace.  It will be updated as putbacks happen real-time.
[00:55:58] * Tpenta shudders as to what this means for encumbered bins
[00:55:59] <boyd> Hell yes, definitely announce it on -discuss. It's better than "what machine should I buy"
[00:56:22] <stevel> i'll let the tools-discuss and #opensolaris people hammer at it for today and we'll see how well it holds up
[00:56:37] <stevel> if it works, then i'll announce osol-announce tomorrow
[00:56:42] * boyd grabs the hg client on this box to hammer away :)
[00:56:46] <stevel> since it hasn't seen extensive load testing yet
[00:56:59] <Gman> stevel, and /topic it ;)
[00:57:16] <stevel> hah. that would be commandeering the topic for my own personal interest ;-)
[00:57:33] <timeless> /usr/sbin/pmconfig: cannot open "/dev/pm": No such file or directory
[00:57:36] <timeless> Console login service(s) cannot run
[00:57:39] <timeless> oops :)
[00:57:55] <Gman> stevel, give me op then ;)
[00:57:58] <gisburn> stevel: we have a commit mailnglist ?
[00:58:03] *** stevel sets mode: +o Gman
[00:58:16] <gisburn> stevel: http://mail.opensolaris.org/pipermail/onnv-notify/2006-October/date.html , right ?
[00:58:26] <stevel> gisburn: i've setup a notification list (onnv-notify at opensolaris dot org) that you can subscribe to to get putback notifications
[00:58:37] * gisburn grabs stevel ...
[00:58:41] <gisburn> ... and hugs him.
[00:58:49] <Tpenta> so do we have any trigger that will notify us if we need to rebuild encumbered-bins?
[00:58:53] <gisburn> Yipppiiieee!
[00:58:58] <stevel> tpenta: nope, sorry.
[00:59:07] <gisburn> Tpenta: use email and filtering.
[00:59:09] <Tpenta> that sould get painful
[00:59:13] <stevel> tpenta: my plan is to continue making the weekly deliveries, and just do closed-bins once a week
[00:59:20] <astinus> Last question from me, since you all look so busy  .. 'Marvell Technology Group Ltd. MV88SX6081 8-port SATA II PCI-X Controller (rev 07)' is my controller - and I understand some of the Nevada builds have support for this .. saw somewhere it was in Thumper?  Could someone confirm it'll work?
[00:59:30] <stevel> we'll (meaning you, me, and kupfer) have to figure out a better way of doing closed-bins - but i didn't want that to hold this back
[00:59:33] <Tpenta> probably ok; ummm perhaps if any of the putbacks puts back into usr/closed?
[00:59:40] <gisburn> Tpenta: umpf... wait for the ksh93 putback and then remind me to write a script for you.
[01:00:21] *** alanc_away is now known as alanc
[01:00:22] <sommerfeld> astinus: I know that thumper has a Marvell SATA chip in it
[01:00:33] <sommerfeld> (well, a half dozen 8-port marvells)
[01:00:34] *** Gman changes topic to "ON Hg Mirror Live, please test : http://tinyurl.com/wgrd8 | SXCR: 50, ON build: 50, ON nightly: 20061016"
[01:01:01] <astinus> sommerfeld:  I *think* my controller will work with the marvel88sx driver
[01:01:24] <astinus> sommerfeld:  But removing Linux after backing up ~2TB of data is a big chore, wanted to be damn certain before I start :)
[01:01:28] * Gman subscribes to onnv-notify too..
[01:01:36] <boyd> astinus: It rings a bell... search "
[01:01:39] <sommerfeld> astinus: indeed
[01:01:48] <boyd> search "zfs-discuss" for the subject "marvel cards.. as recommended"
[01:01:55] <gisburn> stevel: should I send a notification to opensolaris-discuss@ ?
[01:01:57] <astinus> boyd:  thank yoy :)
[01:02:17] <quasi> astinus: that sounds like the one supposedly in thumpers
[01:02:17] <stevel> gisburn: no, not yet
[01:02:35] <astinus> quasi:  In which case, I made a wise decision buying it 12 months ago :P
[01:02:36] <stevel> i'd like the tools-discuss and #opensolaris crowd to hammer on it first.  i'll send a wider announcement tomorrow
[01:02:38] <boyd> astinus: I think there is some -rev dependencies
[01:02:53] <stevel> think of it as your little sneak-preview bonus for hanging out here ;-)
[01:03:53] <gisburn> stevel: minor idea: would it be possible to feed the data as self-designed XML as attachmen to the emails, too ?
[01:03:58] <dclarke> I am happy to report that star 1.5a75 builds
[01:04:00] <gisburn> I am NOT asking for RSS
[01:04:02] <dclarke> $ star --version
[01:04:02] <dclarke> star: star 1.5a75 (sparc-sun-solaris2.11)
[01:04:27] <stevel> gisburn: hrm. possibly, haven't thought much about that.  though RSS would definitely be nice
[01:04:49] <Gman> attachments? ewww.
[01:05:07] * Gman is trying to get rid of attachments in the jds-notify mails
[01:05:20] <gisburn> Gman: the idea is to allow scriptsable and reliable processing of the emails
[01:05:39] <stevel> gisburn: make a note of that on the tools-discuss list
[01:05:43] <gisburn> stevel: please avoid using RSS - it's IMO useless and a pain to use.
[01:05:50] <Gman> and reduce the readability for people who subscribe who are just using a mail reader
[01:05:57] <stevel> gman: perhaps we can offer both
[01:06:13] <gisburn> Gman: the XML thing would be a binary attachment.
[01:06:13] <Gman> yeah
[01:06:23] <Gman> gross.
[01:06:27] <gisburn> or multipart/alternative
[01:06:32] <sommerfeld> gisburn: how does publishing information via multiple paths hurrt you?
[01:07:02] <Gman> you're sending more data down a stream that 99% of the people won't use
[01:07:34] *** mkhl has quit IRC
[01:07:41] <gisburn> sommerfeld: it is Ok if it gets published via multipe paths, including a mailinglist - but currently opensolaris.org appears to suffer from the "RSS everywhere" madness which requires to run special readers to fetch the info
[01:08:09] <stevel> okay time for me to run
[01:08:17] <stevel> bbl
[01:08:18] *** _william_ has left #opensolaris
[01:08:40] <Gman> cya stevel
[01:09:22] <gisburn> sommerfeld: and RSS has no long-term archive and the trafiic generated by the RSS readers is higher than a normal mailinglist.
[01:09:29] <Tpenta> PSARC/2006/587 just got approved
[01:09:32] <Tpenta> :)
[01:09:37] <astinus> boyd:  damn, you were right .. there's a problem with rev07 yet to be resolved, although the Linux driver has a workaround
[01:09:53] <boyd> :(
[01:10:01] <astinus> that really sucks :(
[01:10:18] <gisburn> Tpenta: which case is that ?
[01:10:33] <Tpenta> ksh.kshrc
[01:10:40] <gisburn> hurray!
[01:10:41] <gisburn> :-)
[01:10:42] <gisburn> HURRAY!
[01:10:44] <gisburn> ....
[01:10:46] <gisburn> wait
[01:10:53] <gisburn> Tpenta: no change request ?
[01:11:01] <Tpenta> arc agreed that there are no remaining open architectural issues (ie the contents of teh file are not an arc issue)
[01:11:04] <jbalint> is there anything in dbx to print utf16 strings?
[01:11:04] <boyd> Does anyone understand the order of the cases on the left of http://www.opensolaris.org/os/community/arc/caselog/2006/
[01:11:07] <boyd> ?
[01:11:35] *** nwf has quit IRC
[01:11:46] <astinus> presumably rev07 can't be flashed to rev09 ? :/
[01:11:49] <schily> dclarke: does star link now?
[01:12:19] <Gman> boyd, nfi
[01:12:38] <doownek> I just tried but after use installboot, I have no active partition, then I insert the windows cd and use the command fiboot c: and now all partitions boken :-(
[01:12:45] <gisburn> Tpenta: I am wondering whether jek3 (or what is name was) would be interested to come to giessen and teach students of the 1st semester "vi" ... =:-)
[01:12:54] <doownek> fixboot c:
[01:13:01] <Gman> mmm, loading the arc pages is pretty slow.
[01:13:18] <icon> hrmm
[01:13:23] <alanc> dunno about the ARC case copying, but in most communties the left sidebar ordering is the order created unless you use the painful manual system to reorder them
[01:13:26] <icon> i remember my introduction to vi... 'this is vi, learn it'
[01:13:46] <alanc> if you're not teaching them vi, then they haven't learned to use Unix!
[01:13:57] *** nwf has joined #opensolaris
[01:14:03] <icon> and so, i learned it :)
[01:14:12] <alanc> (not that they should have to use vi, just be able to in case of emergency)
[01:14:28] <boyd> Mine was "Here, program this in C on unix you have 3 weeks"
[01:14:38] <icon> haha
[01:14:45] <icon> that sounds familiar except it was haskell
[01:14:55] <icon> years ago when the language wasnt documented at all hardly
[01:15:10] <gisburn> icon: the problem is tha students should learn shell programming. and the "vi" mode is... uhm... tricky for people who are not familar with informatics and think that the moveable thing in front of a PC is a cupholder.
[01:15:13] <icon> had to write a regex interpreter... that was a really tough semester
[01:15:45] <icon> gisburn: im a bit of an old schooler when it comes to what students should and shouldnt learn
[01:16:12] <icon> if a student doesnt have the gumption to learn the environment/learn the tools, then well, im not very interested in teaching them
[01:16:34] <astinus> If I wanted to request a driver get 'fixed' (http://www.opensolaris.org/jive/thread.jspa?messageID=56889) how would be best to go about that? I'm not sure its a bug, its more a workaround :X
[01:16:39] <gisburn> icon: mmhhh... but I don't like to drive the 1st semester students mad with an editor mode they can't handle.
[01:16:53] <icon> thats not to say i would just cut them loose with no guidance, but i certainly wouldnt teach VI 101 :D
[01:17:15] <Tpenta> vi is easy. it has two modes. Typewriter mode and Beep mode.
[01:17:22] <clee> Tpenta: indeed :)
[01:17:24] <gisburn> heh
[01:17:28] <icon> gisburn: well, have you thought of simpler editors like ee or pico?
[01:17:29] <icon> haha
[01:17:40] <Tpenta> that's an old rec.humour.funny joke
[01:18:05] * gisburn mourns the old days
[01:18:10] *** axisys has joined #opensolaris
[01:18:17] <icon> honestly, as long as they arent using FSF GNU Emacs OS (tm) I wouldnt worry too much if the work was being done :)
[01:18:23] * gisburn remebers archie
[01:18:48] <icon> ah yeah, those days stuck in front of an old vt100 term cursing at why vi beeps incessantly
[01:19:07] <gisburn> heh
[01:19:12] <Tpenta> http://www.netfunny.com/rhf/jokes/91q1/beepmode.html
[01:19:29] <icon> haha
[01:19:31] <icon> nice Tpenta
[01:19:44] *** woland_ has quit IRC
[01:19:57] <axisys> hmm.. we were doing a benchmark on x2100 .. at 8000 qps sol10u2 cpu was 0% idle when freebsd 6.x was below 50% cpu idle
[01:20:13] <axisys> qps = dns queries per sec
[01:20:46] <axisys> looks like freebsd was outperforming sol10u2
[01:20:52] <icon> are you sure the bind releases were the exact same?
[01:21:01] <icon> same settings etc.?
[01:21:02] <axisys> icon: exat same release..
[01:21:23] <icon> axisys: have you done any profiling on the solaris box to see what the bottlenecks were?
[01:21:25] <axisys> icon: sol using dns as a separate zone
[01:21:46] <axisys> icon: no.. not sure how/where
[01:21:56] <icon> also is this a stock freebsd install, or has it been tweaked?
[01:22:07] <axisys> freebsd has been tweaked
[01:22:14] <icon> ie: running GENERIC, no sysctl's set etc.
[01:22:19] <icon> axisys: thats hardly fair :)
[01:22:32] <axisys> icon: i know.. hence looking for suggestions
[01:22:52] <axisys> icon: i impression was sol 10 would change parameters dynamic
[01:23:20] <icon> axisys: it makes you coffee too if you ask it nicely
[01:23:21] <axisys> s/i/my/
[01:23:33] <axisys> icon: :-)
[01:23:59] <icon> i would start looking at prstat output
[01:24:26] <icon> freebsd maintains their own branch of bind
[01:24:46] <icon> which is tweaked for fbsd, solaris does much the same, but there are differences
[01:25:32] <icon> short answer: theres more than meets the eye here. do some digging and find out exactly why bind is so much slower on your sol machine
[01:26:42] <axisys> freebsd 6.1 single thread (performs bad w/ multithread)
[01:26:53] <axisys> both OS with 2M buffer thread
[01:27:03] <axisys> so some tuning done on sol 10
[01:27:19] * astinus sighs contentedly
[01:27:30] * icon grumbles
[01:27:31] <axisys> freebsd works real bad w/ threads
[01:27:47] <icon> fbsd works just fine with threads when it is setup correctly
[01:28:04] <icon> its dependant on hardware, scheduling settings, and whether you are using kernel preemption
[01:28:25] <gdamore> its gotta be better than netbsd threads. :-)
[01:28:30] <icon> haha
[01:28:33] <icon> or lack thereof? :D
[01:28:34] <axisys> ISC suggested not to use thread w/ BIND
[01:28:44] <axisys> i guess there is a reason why
[01:28:51] <axisys> for fbsd
[01:28:54] <icon> have you looked at the bind source anytime recently?
[01:28:56] <gdamore> netbsd has threads.  oh, wait, that's the plural.  netbsd has "thread". :-)
[01:29:01] <icon> hahah
[01:29:04] <icon> gdamore++
[01:29:16] <axisys> icon: nawp.
[01:29:32] <icon> axisys: well, bind is like most other isc products. its big, its old, its nasty
[01:29:33] <axisys> icon: i am looking for suggestions basically to make sol outperform fbsd
[01:29:49] <icon> axisys: to be honest, pushing over 8000 qps is insane
[01:29:52] <axisys> icon: if it is my call i would use something else ;-)
[01:30:00] <gdamore> FWIW I've just started using ISC bind 9.3 on netbsd.  i used the default compilation out of pkgsrc.  NFI if it is configured with threads.  i hope not.  libpthread doesn't work with concurrency >= 1 on netbsd
[01:30:07] <gdamore> s/>=/>/
[01:30:22] <gisburn> icon: you never saw what our poor solaris name servers have to handle.
[01:30:32] <icon> gisburn: how much are you guys pushing?
[01:30:47] <gisburn> icon: thank god that there is nscd on the client machines.
[01:30:57] <icon> yup
[01:31:01] <axisys> icon: i am not trying to see why at 8000 fbsd is around 50% cpu idle
[01:31:02] <astinus> Another newb question:  what version of Xorg is currently shipping with Nevada, and is there an easy way to look at what version ships with each release?  Does it have stuff like Xinerama for multiple head setups, and how's the nVidia driver - Twinview work okay?
[01:31:07] <icon> recursive ns's suffer alot on client load
[01:31:14] <gisburn> icon: I don't know... but thing about 12000+ machines campuswide.
[01:31:22] <gisburn> s/thing/think/
[01:31:26] <icon> gisburn: ouch
[01:31:33] <astinus> geeeez :X
[01:31:47] * astinus tries to do the math on his fingers and fails :S
[01:31:59] <icon> gisburn: no load balancing?
[01:32:02] <gdamore> icon: to really show good solaris performance beating those other desktop OS', you need more processors.  solaris historically wins on near linear SMP scalability, but at the bottom end of the curve (1 or 2 processors) fsd or loonix are often faster
[01:32:24] <icon> true
[01:32:39] <icon> fbsd will more than likely beat out solaris on bind tests
[01:32:41] <gisburn> icon: we use older E4500 with 8CPUs to do DNS and NIS+
[01:32:43] <icon> id be suprised if it didnt
[01:33:01] <icon> but he should be able to narrow that margin a bit with a correct setup
[01:33:09] <gdamore> atinus.  i don't know about Nevada, but nVidia Twinview works like a charm on my U20 running S10 03/05.
[01:33:12] <icon> gisburn: thats a good use for those guys
[01:33:31] <icon> crap that reminds me
[01:33:36] <astinus> gdamore:  thanks :)
[01:33:36] <icon> i need to update the driver on my opteron box
[01:33:38] *** Gman sets mode: -o Gman
[01:33:48] <astinus> gdamore:  If you can't tell, I'm kinda new to this :P
[01:33:54] <gdamore> on 8 cpus, a good MT program will run better on Solaris than on nearly anything else.
[01:33:59] <icon> im running snv46 like a champ on my opteron/quadro 540
[01:34:11] <icon> yup
[01:34:14] <gdamore> no worries.
[01:34:17] <astinus> gdamore:  I've been a developer for Gentoo for longer than I care to remember, going "proper UNIX" is a shock to the system.
[01:34:41] <icon> nah
[01:34:44] <gdamore> heh.  i think you're probably not the only recovering loonix user here.
[01:34:46] <astinus> But I think its time to grow up, plus I'm vying for a placement at Sun next year, so the experience will be good.
[01:34:50] <axisys> icon: i am not to see why at 8000 fbsd is around 50% cpu idle
[01:35:05] <axisys> icon: while sol10u2 is 0% idle
[01:35:15] <icon> axisys: different os' different settings
[01:35:25] <icon> if you want a true benchmark, remove the variables
[01:35:28] <movement> richlowe: ping
[01:35:28] <gdamore> probably lock contention on freebsd. :-)
[01:35:39] <icon> or learn how to tweak the hell out of solaris ;)
[01:35:49] <icon> axisys: does it cap at 8k?
[01:36:00] <icon> or is that just the max load you put it under?
[01:36:07] <gdamore> 50% cpu idle doesn't mean freebsd is running faster.  it could mean you have threads that aren't runnable due to lock contention
[01:36:14] <icon> yup
[01:36:19] <icon> which fbsd is reknown for
[01:36:19] <gdamore> and which caps at 8k, solaris, or freebsd?
[01:36:19] *** ruben has quit IRC
[01:36:30] <icon> it still suffers from giant locking in several places on the kernel
[01:36:46] <astinus> axisys:  Are you looking at CPU idle using top?
[01:36:55] <gdamore> yes, and its not going to be easy for that to get fixed. :-)
[01:36:58] <astinus> axisys:  If so, your box could still be maxed on I/O *shrug*
[01:37:15] <icon> probably IOWAIT
[01:37:24] <astinus> I seem to hit that a lot with NFS
[01:37:25] <gdamore> or (see above) you have threads waiting on contended locks.
[01:37:26] *** sickness has quit IRC
[01:37:28] * astinus scratches his head.
[01:37:41] <icon> hrmm
[01:37:44] <icon> you know
[01:37:51] <icon> does anyone here run a small nas ?
[01:37:54] <astinus> sure
[01:38:01] <icon> what do you use?
[01:38:02] <gdamore> i'm using a cobalt with nbsd for my nas.
[01:38:06] <icon> ahh
[01:38:09] <icon> i was wanting to be lazy ;)
[01:38:17] <icon> i have too much hardware laying around at home as it is
[01:38:33] <gdamore> i've heard very good things about the Buffalo Linksys Pro.  but you have to hack it to add NFS.
[01:38:35] <astinus> icon:  I have two, a mini-ITX fitted with 4 * 500GB
[01:38:48] <icon> ive been thinking of putting a dual core chip in the opteron, moving it to a server and picking up a sunray
[01:38:57] <icon> astinus: how do you keep that guy cool?
[01:39:01] <icon> that sounds like a space heater
[01:39:23] <gdamore> miniITX probably has a cool via chip in it. ;-)
[01:39:30] <astinus> icon:  Its actually really good. One 92mm fan in the front, low RPM, over the disks, one 80mm out the back
[01:39:33] <icon> well, the opt has a 5 disk array in it. i could just add a few disks and up to a larger psu
[01:39:42] <icon> nice
[01:39:49] <icon> use zfs on it
[01:39:52] <gdamore> myself, I'm planning on getting a ReadyNAS soon, which has good performance, and supports NFS out the gate.
[01:39:59] <icon> hrmm
[01:40:06] <astinus> my larger system is a Dual 3GHz Xeon (EM64T) with two Marvell 8-port PCI-X SATA II controllers, and a bunch of 250GB and 200GB drives
[01:40:07] <icon> how about sunrays?
[01:40:09] <icon> good, bad?
[01:40:13] <astinus> that's what I'm looking to put Solaris on today :)
[01:40:15] <gdamore> but i'm not planning on using Solaris on the NAS directly.  I don't want to suck ~200W just for a home NAS
[01:40:19] <icon> im sick of huge workstations
[01:40:28] <icon> yeah thats my issue
[01:40:31] <gdamore> sunrays are awesome.  i have lots of experience with them.
[01:40:33] <icon> an itx would work, but head could be an issue
[01:40:39] <icon> gdamore: hows the price on them?
[01:40:50] <icon> sexy boxes, but ive never bought one for myself
[01:40:54] <gdamore> if you buy from sun, dirt cheap.  sub $300 i think
[01:40:59] <gdamore> (new)
[01:41:00] <icon> no way
[01:41:11] <icon> pxe boot?
[01:41:12] <astinus> gdamore:  Ah, you can probably answer a simple question .. Something like a 2FS with dual-head support, is that done using Xinerama, and does the window manager that Solaris uses know about varying geometries? ie: I want to use a 24" and a 20" screen, is it going to crap out on me?
[01:41:26] <gdamore> no pxe.  you can't run anything but the custom sunray firmware on them.
[01:41:32] <icon> ahh
[01:41:37] <icon> drawbacks?
[01:41:48] <astinus> icon:  no pxe by the sounds :P
[01:41:51] <axisys> icon: if the queries up'd to 10000 systems get too busy to collect the statistics correctly
[01:41:52] <icon> haha
[01:41:56] <icon> well, other than that ;)
[01:41:57] <gdamore> astinus: it will probably "just work", but to tell the truth I've not tried the 2FS yet.  want one.
[01:42:09] <axisys> icon: so we stayed at 8k qps
[01:42:14] <icon> ahh
[01:42:21] <icon> i would do some trending
[01:42:33] <icon> see how it handles with incremental load
[01:42:37] <astinus> I find dual-head really boosts my productivity .. I'm a student so I can't afford twin 24's :P
[01:42:38] <icon> response times etc.
[01:42:46] <icon> bah
[01:42:51] <icon> i still like vt's
[01:42:51] <gdamore> sun rays are awesome as long as you don't want good streaming video or 3D graphics.  then they suck.
[01:42:59] * astinus nods
[01:43:00] <icon> gdamore: how bad?
[01:43:13] <gdamore> depends on your network, server, and what the data you're streaming is.
[01:43:20] <icon> i would at least be able to still watch ... porn ... ;)
[01:43:21] *** deather_ has joined #opensolaris
[01:43:23] <gdamore> 3D graphics is software only.
[01:43:30] <icon> 3d isnt a big deal
[01:43:39] <gdamore> yeah, low res free porn clips work fine. :-)
[01:43:43] <astinus> icon:  we have a lab of pretty old Sun Ray kit here .. they're slooooooooow, but thats only because the servers that drive them are a bit under-spec'd
[01:43:44] <icon> haha
[01:44:06] <icon> do they support gigeth?
[01:44:13] <astinus> don't think so...
[01:44:16] <gdamore> no.  100 mbps only.
[01:44:32] <icon> damn
[01:44:39] <icon> max res?
[01:44:44] <gdamore> we have a version in the lab that has gigE.  but it is a prototype we built only.
[01:44:52] <gdamore> max res is limited to 1920x1200.
[01:45:01] <astinus> 25-30 rays running off two servers .. they're Sun Fire V240's with twin 1GHz processors (each) and 3GB RAM (each) .. get a class in there using all 30 rays and tell them all to open a PDF, then *laugh* as everything gets slower than Vista on a K6-25 500mhz
[01:45:08] <gdamore> (or maybe 1920x1600?  let me check the source.  hang on a sec.)
[01:45:15] <astinus> gdamore:  x1200
[01:45:24] <alanc> SunRay 2FS is dual-head in the box, like many recent video cards, so it's invisible to the X server, not Xinerama
[01:45:25] <astinus> gdamore:  which is the native res of the 24's you guys sell
[01:45:45] <gdamore> i don't work for Sun. :-)  and yes, I have sun ray source (firmware side only)
[01:45:48] <astinus> alanc:  awesome, very nifty. So it'll magically handle weird geometries then?
[01:45:49] <alanc> the X server things all screens are rectangular though, so trying to mix and match strange geometries may confuse it
[01:45:58] <astinus> *nod*
[01:46:01] <astinus> mmkay
[01:46:06] <gdamore> no, it has problems with weird geometries.
[01:46:15] <alanc> haven't tried it myself - our 2FS is still on order
[01:46:27] <gdamore> basically, this is one area where at least up to SRSS 3.1, sun ray is brain damaged.  it has a fixed set of resolutions it can support.  the limitation is in the protocol.
[01:46:42] <astinus> I know if you run a recent GNOME or KDE with Xorg and nVidia / Xinerama, metacity / whatever KDE uses can handle screwed up geometries pretty good
[01:47:01] <icon> hrmmm
[01:47:10] <gdamore> but if you use xinerama, that's okay.  the problem is just the resolution of a single screen
[01:47:14] <astinus> alanc:  When you get the 2FS, give it a try if you remember me :) I'd be real interested to know.
[01:47:35] <astinus> gdamore:  *nod*
[01:47:54] <gdamore> heh.  i have the list of resolutions in my editor now.  they added 1400x1050 for our Comet laptop-sunray.
[01:48:03] <alanc> last I heard you couldn't get the 2FS to expose it's heads individually to the X server, so you couldn't Xinerama them
[01:48:09] <gdamore> and the max is 1920x1200x70Hz
[01:48:16] <astinus> admittedly, the 24's are 1920x1200 and the 20's are 1600x1200 - which still makes a perfect rectangle :D
[01:48:25] <astinus> So that should work out okay
[01:48:28] <icon> how well does audio work?
[01:48:31] *** kleppari has joined #opensolaris
[01:48:36] <gdamore> again, I have no idea how the 2FS does it.  they have to have changed the protocol, and I've not gotten a more recent code drop.
[01:48:45] <astinus> But I think mixing in 17s with 1280x1024 resolution would screw it up.
[01:49:08] * astinus hrms
[01:49:08] <gdamore> Maybe it will pan the smaller screen.  That's how TwinView works.
[01:49:27] <astinus> Yeah :)  Ah well, something for the Todo list :D
[01:49:41] <gdamore> audio is OK.  not great, but not bad.
[01:49:58] <astinus> gdamore:  music stream OK then?
[01:49:59] <gdamore> the newer protocols have some stuff to make it usable for VoIP.  I've not tried that.
[01:50:16] <gdamore> yeah, streaming music is ok.  but it won't compete with a good MP3 player for quality.
[01:50:24] <icon> hrmm
[01:50:28] <icon> well
[01:50:32] <alanc> Sun Ray does have panning code for when you hotdesk from one with a bigger monitor to one with a smaller monitor
[01:50:39] <gdamore> the old builtin speakers in Sun Ray 1g were pretty crummy.
[01:50:52] <gdamore> yes, you can even _force_ a panning configuration without hotdesking.
[01:51:02] <icon> well, id at least use external speakers ;)
[01:51:02] <astinus> gdamore:  'good MP3 player' - as it happens I just got a m-Cody, very sweet  (http://www.advancedmp3players.co.uk/shop/product_info.php?products_id=1493)
[01:51:13] <icon> im using a roku
[01:51:16] <icon> long live daap :D
[01:51:19] <astinus> all OLED'd
[01:51:42] <gdamore> I had a version of our firmware that would pan a lower resolution to the higer resolution, so you could use internal and external monitors simultaneously.  but it was never shipped because they screwed up the EDID wiring on the external monitors on Comet.
[01:52:13] <icon> hrmm
[01:52:17] <icon> im tempted to pick one up
[01:52:23] <astinus> yeah
[01:52:25] <gdamore> i'm not much of a music afficianado.  i'm enough of a philistine that i don't even own an MP3 player, or have any .mp3's on my harddrive
[01:52:29] <icon> at 250$ it would almost be retarded not to
[01:52:48] <icon> id love to move the opteron into a closet where it belongs
[01:53:00] <astinus> icon:  a well cooled closet? :P
[01:53:33] <gdamore> there will be some new features coming into SunRay 2FS that are nice.  VPN, and remote access.
[01:54:12] <icon> astinus: hvac actually ;)
[01:54:13] <gdamore> (we have had those features in our version of the code in Comet for a long time now.  Even FIPS 140-2 link layer encryption with wireless.)
[01:54:18] <icon> hrmm
[01:54:27] <icon> is 2fs the latest model?
[01:54:34] <gdamore> from Sun, yes.
[01:55:16] <gdamore> I think they have some models based on the same chip (Au1550) but that ahve builtin screens.  I've not looked lately.  but they will be a lot more $$.  just get the 2FS and BYOM.
[01:55:23] <alanc> the Sun Ray 2, 2FS & 270 all came out within the last few months
[01:55:26] <gdamore> (BYOM = bring your own monitor)
[01:55:35] <icon> 270?
[01:55:41] <alanc> the 2 is single head, 2FS dual head, 270 has the built-in LCD
[01:55:58] <icon> ahh
[01:55:58] <gdamore> I think 270 == 17" LCD.
[01:56:08] <icon> same perf out of each one?
[01:56:13] <gdamore> (sort of like the 170, but based on the same au1550)
[01:56:18] <astinus> icon:  expect perf is dependant on the backend
[01:56:24] <astinus> icon:  more than anything else
[01:56:25] <gdamore> yes, same basic perf.  they all have the same core, and memory configs.
[01:56:27] <alanc> yeah - it's the Sun Ray 170 case with a Sun Ray 2 series board
[01:56:36] <icon> hrm
[01:56:46] <gdamore> and all use the radeon 7000 video chip.
[01:56:47] <icon> looks like the 2fs has an optic on it...
[01:56:53] <gdamore> yes, but its 10/100 only
[01:57:00] <astinus> 100BASE-FX if memory serves
[01:57:04] <icon> ahh just fx
[01:57:13] <alanc> www.sun.com/sunray has pictures of each
[01:57:34] <icon> ahh cool
[01:57:37] <icon> what protocol does it use?
[01:57:38] <icon> just x11?
[01:57:41] <alanc> Sun Ray protocol
[01:57:51] <gdamore> kinda lame, IMO, but the just use the on-chip 100mbps MAC and add a fiber phy, instead of actually putting on a gigE PCI chip.  which they could do.
[01:57:58] <gdamore> alanc++.
[01:58:01] <icon> ive never run rays before... how does one get that going with a solaris server?
[01:58:06] <astinus> alanc:  sun employee? :)
[01:58:08] <alanc> the X server runs on your Solaris server, and uses a proprietary protocol to send the images to the Sun Ray
[01:58:25] <icon> ahh
[01:58:28] <gdamore> there is some special software you have to run on the server.  its pretty straight-forward.
[01:58:42] <icon> so im guessing things like dri and gl are out of the question?
[01:58:47] <gdamore> absolutely.
[01:58:57] <delewis> Xvideo isn't, though :-)
[01:59:10] <icon> gah
[01:59:18] <icon> brb
[01:59:19] <alanc> GL can be done in software, so non-accelerated
[01:59:21] <gdamore> Xvideo uses radeon 7000 YUV decode.  I don't think they've done anything more than that
[01:59:24] <alanc> astinus: yes
[01:59:27] *** deather has quit IRC
[01:59:42] <jerome__> does b49 support the LSI megaraid sata?
[02:00:16] <gdamore> we were going to do streaming MPEG4 in the Comet next-gen product, but NFI if it will ever ship.  kinda back burnered at the moment.
[02:00:39] <alanc> I believe that there's also some people using VirtualGL to render 3D on real video cards and then image grab and display the output on Sun Rays, but that's still kinda slow for more than a small image
[02:01:50] <gdamore> yes.  rendering pixel by pixel will be very, very slow on sun ray.
[02:02:16] <gdamore> its one reason why we wanted to do gigE.  scientific users want to stream very high-resolution and "uncompressible" images.
[02:02:32] <gdamore> there is wavelet compression in the newer versions of the protocol, but you lose some image quality using it.
[02:02:43] <astinus> alanc:  Like your blog, esp. the 'Just like running Windows' post in July :P
[02:02:52] * astinus smiles
[02:02:55] <alanc> astinus: thanks
[02:02:59] <alanc> heh
[02:03:08] <astinus> alanc:  so you're the poor sod shouldered with dealing with Xorg :P
[02:03:27] <alanc> yep
[02:03:32] <gdamore> btw, alan, did you notice my offer of help -- anything i can do to help with getting the sparc stuff switched over?
[02:03:46] *** ShadowHntr has joined #opensolaris
[02:03:58] <alanc> I saw your e-mail and forwarded to the person tackling the problem currently
[02:04:20] <gdamore> ok, thanks. even if you just need an external "partner" or "customer" to lean hard somewhere, let me know.
[02:04:25] *** piwi has joined #opensolaris
[02:04:29] *** bengtf has quit IRC
[02:04:41] <alanc> we spent yesterday afternoon determining that the Xorg SPARC PCI support doesn't work well with the schizo controller in the SB1500 he picked as his victim machine
[02:05:00] <alanc> had lots of fun starting Xorg, watching it poke registers on the PCI bus and making the hardware reset
[02:05:14] <astinus> Any of you guys happen to own a Samsung laptop, one with a Core Duo per chance?
[02:05:15] <alanc> not even giving the kernel a chance to panic and dump - instant reboot 8-)
[02:05:30] <astinus> alanc:  'Oops' quite literally then
[02:05:30] <ShadowHntr> astinus: i'd like one, does that count? ;)
[02:05:35] <gdamore> heh.   I think this is a general problem with the design of X -- it should not be touching registers directly IMO.
[02:05:50] <alanc> yeah, but the PCI-rework project in Xorg isn't done yet
[02:05:51] <ofu> astinus: my girlfriend got an R65
[02:05:51] <gisburn> alanc: I hope you were clever enougth to boot the macgine via NFS
[02:06:00] <alanc> gisburn: certainly not
[02:06:02] <icon> hrmm okay last ray question
[02:06:06] <gisburn> alanc: why ?
[02:06:15] <alanc> because we're not that clever
[02:06:20] <icon> given the choice of having an ultra 40 on your desk or a ray connected to the u40, which would you take, and why ;)
[02:06:28] <gisburn> alanc: umpf
[02:06:32] <icon> and now that i think about it, let me tack on an a/s/l at the end of that :D
[02:06:59] <alanc> I'd go with the Ultra 40 on my desk, but that's because I have to do OpenGL stuff on x86
[02:07:04] <gdamore> dunno.  i might put the 40 in a closet somewhere.  but it depends on usage pattern.
[02:07:16] <gisburn> alanc: hook-up a terminal to the serial port and add checkpoints to the PCI code which send the marks to the serial port
[02:07:17] <astinus> a ray connected to an U40 because then my desk is tidier, I have more room for reference books, there's less noise in my office and peoples jaws can hit the floor shortly followed by gasping and "Wow your computer is small" ..
[02:07:37] <astinus> (because they're used to hulking desktops running Windows)
[02:07:55] <alanc> I have been tempted to replace my SB2500 with a Sun Ray since I really only use it as a display terminal, running Firefox, Thunderbird, xchat and gnome-terminal ssh'ed to all the machines I'm really working on
[02:08:18] <gisburn> alanc: can I get the Sb2500 ? Pleeaaasssseeeee...
[02:08:21] <gdamore> i used a sun ray back while working at sun, and i was quite happy with it.  but i never needed 3D or streaming media.
[02:08:21] <astinus> i *almost* got a SB1000 the other day
[02:08:26] <gisburn> alanc: PURRRRR.. PURRRRR....
[02:08:28] <icon> thats exactly what my usage is too
[02:08:34] <icon> light video, streaming music while working
[02:08:36] <alanc> gisburn: before tackling this, Jay spent 6 months working on x86 suspend and resume, where debugging was often limited to looking at LED's on the motherboard - I'm sure he knows how to debug
[02:08:39] * jamesd_ would settle for  blade 1500  ram...
[02:08:44] <icon> im a coder, all i really do is code ;)
[02:08:53] <gisburn> alanc: jay hobson ?
[02:08:57] <alanc> Cotton
[02:09:06] <gdamore> then go with the ray.  better yet, pick up a Comet and you can carry it around with you.  :-)
[02:09:17] <alanc> we had 2 Jays
[02:09:24] <icon> its been a bit since ive looked at the comets
[02:09:28] <icon> are you a comet guy gdamore ?
[02:09:35] <gdamore> (disclaimer: i work for Tadpole and did almost all the Comet firmware work)
[02:09:38] <astinus> there's not enough Alex's in Sun .. blogs.sun.com/alex doesn't go anywhere!
[02:09:40] <gisburn> alanc: "have" or "had" ?
[02:10:03] <icon> ahh how cool
[02:10:03] *** jmcp has joined #opensolaris
[02:10:04] <gdamore> icon: I think that's an unqualified "yes". :-)
[02:10:22] <alanc> also wondering if we should just work on an fbdev-like driver to bypass the PCI probing and just map /dev/fb as a dumb frame buffer as a first pass to get most cards up and running quickly
[02:10:27] <alanc> gisburn: had
[02:10:57] * gisburn mourns the bad changes in the world
[02:11:06] <gdamore> alanc: yes.  for most stuff you could just add a few rendering primitives to the drivers (blit, fill, and cursor) and actually have decent 2D performance
[02:11:11] <alanc> change is inevitable
[02:11:12] <jmcp> alanc: I'm in favour of that - it would reduce the barrier to entry
[02:11:27] <jmcp> and improve the initial user experience
[02:11:27] * gisburn considers to build a clock makde out of glass like the one in Terry Pratchett's "Thief of time" ...
[02:11:43] <alanc> it would reduce the dependence on us waiting for SPARC graphics to port drivers or release specs
[02:11:48] <icon> gdamore: so why would i go with a comet instead of a ray? ;D
[02:12:05] *** ofu_ has joined #opensolaris
[02:12:15] <gdamore> its mobile, has 802.11a/g wireless, and includes VPN support. :-)
[02:12:18] <gisburn> alanc: change is not invitable. Read TP's "Thief of time"
[02:12:31] <icon> as in laptop mobile?
[02:12:33] <alanc> it's lighter than carrying around your own UPS for the Sun Ray when you want to move to a different place
[02:12:34] <Tpenta> i only read that a few weeks back
[02:12:37] <gdamore> yes.
[02:12:41] *** ofu has quit IRC
[02:12:45] <gdamore> (to say nothing of the monitor. :-)
[02:12:49] <alanc> gisburn: It's been a year or two since I read that
[02:13:07] <Tpenta> rereading neuromancer at the moment
[02:13:24] <gdamore> it comes in 12" or 15".  but you can use an external monitor.  Unfortunately the external screen can only show the same thing as the internal (though you can use a bigger workspace and have the smaller monitor pan around).
[02:13:38] *** lloy0076 has joined #opensolaris
[02:13:51] <gisburn> alanc: We just need to build a clock.... the glass clock... and the all changes... are... frozen. 1.00PM always, over and over again only 1.00PM for eternty
[02:14:13] <gdamore> maybe some time i'll get the code changes for the 2FS, and then I can add dual monitor support properly.  right now i don't know the details in the protocol for how 2FS does it.
[02:14:40] <alanc> but then ksh93 would be frozen outside the tree and never integrated!
[02:15:17] * gdamore has gotten along well enough without ksh93 so far...
[02:15:18] * gdamore ducks
[02:15:38] <gisburn> alanc: The tree and overone working on it would be frozen, too. And think in temporal facts - there will no longer be "never" or "tomorror", just 1.00PM
[02:16:24] <alanc> yeah, I use tcsh myself interactively, but have lots of scripts written for ksh
[02:16:43] <gdamore> gisburn: are you _sure_ you aren't part of the NetBSD development team?
[02:16:47] * gisburn considers to have some fun with torturing gdamore
[02:16:58] <gisburn> gdamore: uhm..
[02:17:07] <gisburn> gdamore: .... AFAIK no. why ?
[02:17:14] <gdamore> stuck in time....
[02:17:31] <gisburn> heh
[02:17:39] <gisburn> gdamore: no cookies for you.
[02:17:46] <gdamore> (there are people in NetBSD land seriously concerned about sun2 portability, and working on a port to pdp10...)
[02:17:49] <lloy0076> OMGWTFBBQ
[02:17:50] <gisburn> and I make good cookies
[02:18:00] <gdamore> darn.
[02:18:09] <lloy0076> LGPL Clause 3: You may opt to apply the terms of the ordinary GNU General Public License instead of this License to a given copy of the Library.  To do this, you must alter all the notices that refer to this License, so that they refer to the ordinary GNU General Public License, version 2, instead of to this License.  (If a newer version than version 2 of the ordinary GNU General Public License has appeared, then you can specify that
[02:18:11] * lloy0076 sigh
[02:18:28] <lloy0076> Which essentially means you can make a GPL version of any LGPL code and stick thy tongue out.
[02:18:54] <delewis> or you could just ignore the license, and continue writing/using programs.
[02:19:17] <gdamore> GNU is really big on auto-upgrade of "lesser" or "older" licenses to whatever their latest GPL is.  its one reason why i seriously regret ever using GPL for anything in my past.
[02:19:32] <gisburn> gdamore: I am famoour for my baileys cookies! :-)
[02:19:41] <gisburn> s/famoour/famous/
[02:19:52] <gdamore> i'll settle for just the bailey's at the moment. :-)  and maybe some coffee.
[02:20:22] <icon> gdamore: i think you just sold a comet 15
[02:20:54] <gdamore> awesome.  too bad i don't get a commission. :-)
[02:21:09] <gisburn> (the tricky part is to avoid putting too much baileys inside... or... *hicks* ... drink ... *hicks* the *hicks* baileys *hicks* while *hicks*  making the *hicks* *hicks*  cookies *hicks* *hicks* *hicks* )
[02:21:40] <gisburn> :-)
[02:21:47] <icon> gdamore: how does the perf compare to a ray connected to 100 duplex?
[02:22:36] <gdamore> Comet is same basic performance as Sun Ray 1g.  so slightly slower for streaming video/audio, but you're unlikely to notice a difference.
[02:22:42] <gdamore> (slower than a 2FS i mean.)
[02:23:01] <icon> hrmm
[02:23:04] <dclarke> hello ..
[02:23:05] <gdamore> the 2FS will be faster at VPN connections, when it has support for them, though.  much faster for IPsec.
[02:23:12] <boyd> Hi, dennis
[02:23:17] <icon> what kind of hardware is in those guys?
[02:23:24] <Tekni> wee... "errors: 38594 data errors"
[02:23:27] <dclarke> I return
[02:23:36] 
[02:23:38] <dclarke> happily with a fresh and complete build of star
[02:23:47] <gdamore> the comet/sunray 1g has a 100Mhz copernicus (uSparc iiep variant).  the 2FS has an AMD alchemy Au1550
[02:23:51] <dclarke> sorry .. I had to step away
[02:24:03] <icon> ahh
[02:24:04] <boyd> dclarke: He was asking about whether star liked ok now, I think
[02:24:07] <dclarke> I will post my update to the star maillist
[02:24:23] <icon> gdamore: what does retail run on a comet15? not seeing a price anywhere...
[02:24:25] <rydis> gdamore: To be honest, the pdp10 port was probably more of a lark/nostalgia trip by Ragge, the NetBSD-VAX maintainer, when the pdp-10 emulators became available.
[02:24:27] <dclarke> yep .. linked perfectly once I killed off an older release that I had
[02:24:37] <dclarke> my fault totally .. I'm a jerk
[02:24:40] <gisburn> rydis: erm, no.
[02:24:58] <gdamore> i think those comments could be applied to _most_ of the NetBSD ports. :-)
[02:24:59] <gisburn> rydis: there are "PDP on a chip" systems used by the US military
[02:25:06] <gdamore> (except for Ragge's name. of course)
[02:25:22] <boyd> dclarke: Nah, I think jerk implies something about behaviour towards others.
[02:25:31] <rydis> gisburn: Hardly -10s. Not that I'd /know/, not being a US military guy.
[02:25:42] *** coffman has quit IRC
[02:26:40] <rydis> (XKL did make single-chip-ish 10:s, and might still, but the NetBSD port came slightly after the emulators.)
[02:27:22] <gdamore> time to go pick up the kids.  back later.
[02:27:33] <icon> later
[02:31:51] *** lktm_ has joined #opensolaris
[02:32:13] <lktm_> Hi
[02:32:34] <lktm_> could someone me help me with opensolaris licensing questions?
[02:32:44] <dclarke> sure
[02:32:47] <dclarke> its free
[02:32:55] <dclarke> or do you mean the CDDL thing ?
[02:32:58] <Tpenta> remembering that no-one here is a lawyer
[02:33:01] <lktm_> i mean CDDL
[02:33:10] <dclarke> CD DownLoad
[02:33:15] <lktm_> and free on english means so many things
[02:33:19] <dclarke> nope .. thats not it
[02:33:40] <lktm_> in spanish isnt the same "gratis" than "libre"
[02:33:41] <alanc> you can ask, but we can't promise an answer until we know the question
[02:33:48] <lktm_> lol, ok :P
[02:33:51] <Tpenta> if any of us feel competant to answer, I'm sure one of us will
[02:33:54] <lktm_> i have modified pam_ldap
[02:33:55] <lktm_> from opensolaris
[02:34:04] <dclarke> in any case .. no one here is a lawyer and thus any answer you get is also free and worthless .. possibly
[02:34:05] <lktm_> to have it working on solaris 8
[02:34:16] <lktm_> and support password expiration etc..
[02:34:29] <lktm_> could i redistribute that code and binaries?
[02:34:34] <lktm_> always both
[02:34:37] <lktm_> could i?
[02:34:39] <alanc> sure
[02:34:46] <Tpenta> i think so
[02:34:47] <alanc> that's what the CDDL is all about
[02:34:50] <astinus> lktm_:  pam_ldap is LGPL
[02:35:14] <alanc> the one in Solaris is probably not LGPL
[02:35:21] *** stevel has quit IRC
[02:35:27] <lktm_> weed@ultra:~$ head sources/pam_ldap/ldap_acct_mgmt.c
[02:35:28] <lktm_>  * CDDL HEADER START
[02:35:31] <lktm_> yess
[02:35:34] <lktm_> solaris has another one
[02:35:36] *** sahafeez has quit IRC
[02:35:41] <astinus> good call, I didn't know that.
[02:35:51] <lktm_> :)
[02:36:08] <lktm_> ok, and does anyone remember if
[02:36:17] <lktm_> libsldap and libldap are also CDDL?
[02:36:22] <alanc> there's probably other ones - after Sun invented PAM, other OS'es copied it and wrote their own modules, and somehow the PAM modules for LDAP support got unimaginatively named pam_ldap on multiple OS'es when they're not the same code, just the same idea
[02:37:06] <astinus> alanc:  annoying isn't it :P
[02:37:29] <lktm_> im having real troubles to extract just libldap and libslapd from opensolaris tree
[02:37:42] <astinus> "Which frickin version of pam_ldap am I working with today..." seems like a good blog post subject, mmmmm :D
[02:37:53] <lktm_> there is any chance to get their source on separate packages?
[02:37:55] <alanc> they need to think up faniciful names, like "Firebird" to avoid picking the same names as other projects...oh, wait, didn't help there either
[02:38:29] <sommerfeld> lktm_: with the just announced mercurial repository it should be possible to do a subtree checkout
[02:38:43] <astinus> alanc:  and even when Fire*fox* gets thought up, they then say, "Right, even though this is open source, you can't use our name for your own binaries." so now there's Iceweasel :/
[02:38:44] <lktm_> yes
[02:38:46] <lktm_> but makefiles
[02:38:48] <lktm_> and all
[02:39:03] <lktm_> are screwed as all that source is part of ON
[02:39:18] *** jcsmith has quit IRC
[02:39:21] <alanc> ON is not designed for easily pulling parts out into standalone releases, but it's been done before
[02:39:31] *** jcsmith has joined #opensolaris
[02:39:32] <lktm_> i did with pam_ldap
[02:39:33] <sommerfeld> if we manage to get teamware-style subtree checkouts...
[02:39:42] <sommerfeld> or, I should say, sparse-tree checkouts
[02:39:43] <lktm_> but libldap and libsldap are harder...
[02:39:44] <alanc> there's a set of "heritage tools" on sourceforge that come from the opensolaris sources
[02:40:24] <astinus> hm, another newb question:  say I've got a system which is intermittently connected to the network. I want it to used shared authentication principles; so normally it'll authenticate against a server to verify users, but if there's no network, is there any way to cache previous logins? ie:  astinus logs into laptop while its connected, then goes to a meeting, there's no network there, but he can still login?
[02:40:25] <lktm_> i was just wondering if someone managed an easy way to extract parts of the tree
[02:40:30] <lktm_> instead of manual hacking
[02:40:52] *** jcsmith has quit IRC
[02:40:59] <lktm_> astinus ldapclient does
[02:41:06] <lktm_> it contains ldapcachemgr daemon
[02:41:09] *** jcsmith has joined #opensolaris
[02:41:14] <lktm_> but cache isnt persistent
[02:41:46] <lktm_> you will loose it after killing the ldapcachemgr proc
[02:42:02] <lktm_> and if you are using nss over ldap
[02:42:23] <lktm_> will be a real headache, as it will probably need to resove a lot of user names
[02:42:30] <lktm_> on each "ls" os "ps"
[02:42:35] <lktm_> to find owners
[02:42:43] <astinus> hm :/
[02:42:45] <lktm_> or even get timeout on hosts resolutions
[02:42:57] <lktm_> if desired host wasnt cached
[02:43:18] <elektronkind> lktm_: isn't that supposed to be nscd's job anyway?
[02:43:30] <elektronkind> the caching, that is
[02:43:32] <lktm_> its done twice
[02:44:02] <lktm_> nscd doesnt care how data is retrived, ej, ldap, nis, nis+, files
[02:44:10] <lktm_> it just cache it
[02:44:17] <elektronkind> right
[02:44:23] <lktm_> and nis or nis+ doesnt have an aditional cache
[02:44:25] <elektronkind> so why the secondary ldap cache?
[02:44:37] <lktm_> but ldapclient implementation has it
[02:44:52] <lktm_> maybe to avoid overheading ldap servers
[02:44:59] <lktm_> on machines without nscd
[02:45:14] <elektronkind> ldap servers are designed to hand tons and tons of small queries though
[02:45:19] <elektronkind> s/hand/handle
[02:45:35] <lktm_> really i dont know why
[02:45:44] <lktm_> but the truth is that it exists
[02:45:45] <lktm_> lol
[02:46:00] *** kimc has left #opensolaris
[02:46:39] <lktm_> root@samba-eads:~> ps -ef | grep cache
[02:46:39] <lktm_>     root  2865     1  0 05:19:45 ?        0:00 /usr/lib/ldap/ldap_cachemgr
[02:48:15] <lktm_> thanks for the help with the CDDL stuff
[02:48:41] <lktm_> gonna play around little bit more with libldap and libsldap
[02:48:44] <lktm_> bye
[02:48:52] *** lktm_ has quit IRC
[02:50:25] *** besonen_mobile__ has quit IRC
[02:52:53] <astinus> I see Nevada b48 ISOs available, whats the easiest way to get b50 for a new install?
[02:53:54] <gdamore> download SXCR from opensolaris.org/os/downloads
[02:54:12] <astinus> ah, gdamore - just the person. I saw you weighed in on some ipw3945 threads
[02:54:50] <gdamore> can't prove it. :-)
[02:54:56] *** qbit has joined #opensolaris
[02:54:59] <astinus> was there any movement lately on ipw3945 support?
[02:55:51] <gdamore> i wouldn't know.  intel chips aren't available without intel processors/motherboards, so they aren't very interesting to me.
[02:56:02] <astinus> *nod*
[02:58:58] <gdamore> done for now.  back in a a while (probably a few hours)
[03:00:36] <astinus> see ya :)
[03:02:24] *** gtc_ has quit IRC
[03:11:40] *** besonen_mobile__ has joined #opensolaris
[03:14:56] *** jmcp has quit IRC
[03:17:07] *** lloy0076 has quit IRC
[03:18:04] *** jsubl2 has joined #opensolaris
[03:20:12] *** sommerfeld has quit IRC
[03:24:06] <Error_404> wow, glassfish sure does chew through ram
[03:24:51] <jamesd_> java is a  conspiracy of the ram manufactuors
[03:25:34] <ShadowHntr> heh.
[03:29:11] <Error_404> heh
[03:43:16] *** mikaeld has quit IRC
[03:43:17] *** stefanjo has quit IRC
[03:43:17] *** AbeFroman has quit IRC
[03:43:17] *** sparc-kly_ has quit IRC
[03:43:17] *** SymmWork has quit IRC
[03:43:17] *** libkeiser has quit IRC
[03:43:17] *** PerterB has quit IRC
[03:43:19] *** cypromis has quit IRC
[03:43:19] *** richlowe has quit IRC
[03:43:19] *** asyd has quit IRC
[03:43:19] *** Trident has quit IRC
[03:43:20] *** miffe_ has quit IRC
[03:43:20] *** Tpenta has quit IRC
[03:43:20] *** schily has quit IRC
[03:43:21] *** Vanuatoo has quit IRC
[03:43:22] *** TBCOOL has quit IRC
[03:43:22] *** xzi11a has quit IRC
[03:43:22] *** dduvall has quit IRC
[03:43:23] *** B|nTaRa has quit IRC
[03:43:23] *** phips has quit IRC
[03:43:23] *** jacotton has quit IRC
[03:43:23] *** pjlv has quit IRC
[03:43:23] *** astinus has quit IRC
[03:43:23] *** nightswim has quit IRC
[03:43:23] *** doownek has quit IRC
[03:43:23] *** Odin- has quit IRC
[03:43:25] *** Stric has quit IRC
[03:43:25] *** milo has quit IRC
[03:43:25] *** the-decider has quit IRC
[03:43:25] *** DataStream has quit IRC
[03:43:25] *** noyb has quit IRC
[03:43:26] *** switch has quit IRC
[03:43:26] *** kirma has quit IRC
[03:43:26] *** paulproteus has quit IRC
[03:43:26] *** Kernel86|Laptop has quit IRC
[03:43:26] *** mlh has quit IRC
[03:43:27] *** kleppari has quit IRC
[03:43:28] *** gisburn has quit IRC
[03:43:28] *** sparkleytone has quit IRC
[03:43:29] *** timeless has quit IRC
[03:43:29] *** Andrew____ has quit IRC
[03:43:30] *** Sporq has quit IRC
[03:43:30] *** Tekni has quit IRC
[03:43:31] *** dvorak has quit IRC
[03:43:31] *** yusufg has quit IRC
[03:43:31] *** boyd has quit IRC
[03:43:32] *** quasi has quit IRC
[03:43:32] *** trygvis has quit IRC
[03:43:32] *** paul has quit IRC
[03:43:32] *** heffnerd has quit IRC
[03:43:32] *** cormac has quit IRC
[03:43:33] *** prg3 has quit IRC
[03:43:34] *** jcsmith has quit IRC
[03:43:34] *** piwi has quit IRC
[03:43:35] *** jwtodd has quit IRC
[03:43:35] *** dunc has quit IRC
[03:43:35] *** slowhog has quit IRC
[03:43:35] *** trs81 has quit IRC
[03:43:35] *** laca has quit IRC
[03:43:35] *** Method has quit IRC
[03:43:36] *** kFuQ has quit IRC
[03:43:37] *** jteo has quit IRC
[03:43:37] *** icon has quit IRC
[03:43:38] *** lulf has quit IRC
[03:43:38] *** andersmo has quit IRC
[03:43:38] *** Triskelios has quit IRC
[03:43:38] *** Hunger- has quit IRC
[03:43:38] *** Risky has quit IRC
[03:43:39] *** Ezechial_ has quit IRC
[03:43:39] *** Doc has quit IRC
[03:43:39] *** IvanR_ has quit IRC
[03:43:39] *** j0sephi has quit IRC
[03:43:40] *** twincest has quit IRC
[03:43:41] *** nwf has quit IRC
[03:43:41] *** besonen_mobile__ has quit IRC
[03:43:41] *** ShadowHntr has quit IRC
[03:43:41] *** axisys has quit IRC
[03:43:42] *** Sieghard has quit IRC
[03:43:43] *** steleman has quit IRC
[03:43:43] *** Kernel86 has quit IRC
[03:43:44] *** r3boot has quit IRC
[03:43:44] *** Berny has quit IRC
[03:43:44] *** hali has quit IRC
[03:43:44] *** Trinisan has quit IRC
[03:43:44] *** junks has quit IRC
[03:43:45] *** mustang has quit IRC
[03:43:45] *** jsubl2 has quit IRC
[03:43:45] *** qbit has quit IRC
[03:43:46] *** deather_ has quit IRC
[03:43:46] *** ProfMikey has quit IRC
[03:43:47] *** jamesd_ has quit IRC
[03:43:47] *** phalenor- has quit IRC
[03:43:47] *** Gman has quit IRC
[03:43:47] *** alanc has quit IRC
[03:43:47] *** medar has quit IRC
[03:43:48] *** rachel has quit IRC
[03:43:48] *** neoxed has quit IRC
[03:43:48] *** kb7sqi has quit IRC
[03:43:49] *** polk__ has quit IRC
[03:43:50] *** g4lt-mordant has quit IRC
[03:43:50] *** xinkeT has quit IRC
[03:43:50] *** alanc-away has quit IRC
[03:43:50] *** sparc-kly|WORK has quit IRC
[03:43:50] *** fgd has quit IRC
[03:43:50] *** uzi has quit IRC
[03:43:51] *** james_ has quit IRC
[03:43:53] *** ofu_ has quit IRC
[03:43:53] *** LeftWing has quit IRC
[03:43:53] *** Seawolf99 has quit IRC
[03:43:54] *** bondolo has quit IRC
[03:43:54] *** jcea has quit IRC
[03:43:54] *** Auralis has quit IRC
[03:43:54] *** drifter747 has quit IRC
[03:43:54] *** svoboda has quit IRC
[03:43:56] *** Jiko_ has quit IRC
[03:43:56] *** delewis has quit IRC
[03:43:56] *** dclarke has quit IRC
[03:43:57] *** esproul has quit IRC
[03:43:57] *** kloczek has quit IRC
[03:43:57] *** Dar_HOME has quit IRC
[03:43:57] *** simford has quit IRC
[03:43:57] *** udos has quit IRC
[03:43:58] *** crib has quit IRC
[03:43:59] *** sniffy has quit IRC
[03:43:59] *** jengelh has quit IRC
[03:43:59] *** gdamore has quit IRC
[03:44:00] *** jbalint has quit IRC
[03:44:00] *** postwait has quit IRC
[03:44:00] *** kalif has quit IRC
[03:44:00] *** Bart_M has quit IRC
[03:44:00] *** lasseoe has quit IRC
[03:44:00] *** elektronkind has quit IRC
[03:44:01] *** Justin has quit IRC
[03:44:01] *** clee has quit IRC
[03:44:02] *** cnuke has quit IRC
[03:44:02] *** Wez has quit IRC
[03:44:02] *** Drone has quit IRC
[03:44:02] *** soultan has quit IRC
[03:44:03] *** mazon has quit IRC
[03:44:03] *** pjd- has quit IRC
[03:44:03] *** rydis has quit IRC
[03:47:18] *** Tpenta has joined #opensolaris
[03:47:19] *** B|nTaRa has joined #opensolaris
[03:47:19] *** Stric has joined #opensolaris
[03:47:19] *** milo has joined #opensolaris
[03:47:19] *** switch has joined #opensolaris
[03:47:19] *** the-decider has joined #opensolaris
[03:47:19] *** kirma has joined #opensolaris
[03:47:19] *** DataStream has joined #opensolaris
[03:47:20] *** noyb has joined #opensolaris
[03:47:20] *** Odin- has joined #opensolaris
[03:47:20] *** paulproteus has joined #opensolaris
[03:47:20] *** nightswim has joined #opensolaris
[03:47:20] *** pjlv has joined #opensolaris
[03:47:20] *** jacotton has joined #opensolaris
[03:47:20] *** phips has joined #opensolaris
[03:47:20] *** dduvall has joined #opensolaris
[03:47:20] *** xzi11a has joined #opensolaris
[03:47:21] *** TBCOOL has joined #opensolaris
[03:47:21] *** Vanuatoo has joined #opensolaris
[03:47:21] *** schily has joined #opensolaris
[03:47:21] *** astinus has joined #opensolaris
[03:47:21] *** doownek has joined #opensolaris
[03:47:21] *** jsubl2 has joined #opensolaris
[03:47:21] *** besonen_mobile__ has joined #opensolaris
[03:47:21] *** qbit has joined #opensolaris
[03:47:22] *** jcsmith has joined #opensolaris
[03:47:22] *** ofu_ has joined #opensolaris
[03:47:22] *** piwi has joined #opensolaris
[03:47:22] *** ShadowHntr has joined #opensolaris
[03:47:22] *** kleppari has joined #opensolaris
[03:47:22] *** deather_ has joined #opensolaris
[03:47:22] *** axisys has joined #opensolaris
[03:47:23] *** nwf has joined #opensolaris
[03:47:23] *** gisburn has joined #opensolaris
[03:47:23] *** ProfMikey has joined #opensolaris
[03:47:23] *** LeftWing has joined #opensolaris
[03:47:23] *** lulf has joined #opensolaris
[03:47:23] *** jwtodd has joined #opensolaris
[03:47:23] *** Seawolf99 has joined #opensolaris
[03:47:23] *** dunc has joined #opensolaris
[03:47:23] *** bondolo has joined #opensolaris
[03:47:23] *** jcea has joined #opensolaris
[03:47:23] *** slowhog has joined #opensolaris
[03:47:23] *** trs81 has joined #opensolaris
[03:47:23] *** Auralis has joined #opensolaris
[03:47:23] *** drifter747 has joined #opensolaris
[03:47:23] *** svoboda has joined #opensolaris
[03:47:23] *** jamesd_ has joined #opensolaris
[03:47:24] *** irc.freenode.net sets mode: +o Tpenta
[03:47:24] *** laca has joined #opensolaris
[03:47:24] *** sparkleytone has joined #opensolaris
[03:47:24] *** timeless has joined #opensolaris
[03:47:24] *** medar has joined #opensolaris
[03:47:24] *** phalenor- has joined #opensolaris
[03:47:24] *** Gman has joined #opensolaris
[03:47:24] *** alanc has joined #opensolaris
[03:47:24] *** Sieghard has joined #opensolaris
[03:47:24] *** Jiko_ has joined #opensolaris
[03:47:24] *** delewis has joined #opensolaris
[03:47:25] *** Method has joined #opensolaris
[03:47:25] *** dclarke has joined #opensolaris
[03:47:25] *** Kernel86|Laptop has joined #opensolaris
[03:47:25] *** steleman has joined #opensolaris
[03:47:26] *** esproul has joined #opensolaris
[03:47:26] *** Andrew____ has joined #opensolaris
[03:47:26] *** Kernel86 has joined #opensolaris
[03:47:26] *** kloczek has joined #opensolaris
[03:47:26] *** kFuQ has joined #opensolaris
[03:47:26] *** rachel has joined #opensolaris
[03:47:26] *** neoxed has joined #opensolaris
[03:47:26] *** mlh has joined #opensolaris
[03:47:26] *** kb7sqi has joined #opensolaris
[03:47:26] *** Dar_HOME has joined #opensolaris
[03:47:26] *** r3boot has joined #opensolaris
[03:47:26] *** simford has joined #opensolaris
[03:47:26] *** udos has joined #opensolaris
[03:47:26] *** jteo has joined #opensolaris
[03:47:26] *** Sporq has joined #opensolaris
[03:47:26] *** Tekni has joined #opensolaris
[03:47:27] *** crib has joined #opensolaris
[03:47:27] *** polk__ has joined #opensolaris
[03:47:27] *** icon has joined #opensolaris
[03:47:27] *** Berny has joined #opensolaris
[03:47:27] *** sniffy has joined #opensolaris
[03:47:27] *** g4lt-mordant has joined #opensolaris
[03:47:27] *** andersmo has joined #opensolaris
[03:47:27] *** boyd has joined #opensolaris
[03:47:27] *** soultan has joined #opensolaris
[03:47:27] *** jengelh has joined #opensolaris
[03:47:27] *** Triskelios has joined #opensolaris
[03:47:27] *** gdamore has joined #opensolaris
[03:47:28] *** xinkeT has joined #opensolaris
[03:47:28] *** jbalint has joined #opensolaris
[03:47:28] *** alanc-away has joined #opensolaris
[03:47:28] *** sparc-kly|WORK has joined #opensolaris
[03:47:28] *** Hunger- has joined #opensolaris
[03:47:29] *** hali has joined #opensolaris
[03:47:29] *** Risky has joined #opensolaris
[03:47:29] *** postwait has joined #opensolaris
[03:47:29] *** uzi has joined #opensolaris
[03:47:29] *** james_ has joined #opensolaris
[03:47:29] *** Ezechial_ has joined #opensolaris
[03:47:29] *** twincest has joined #opensolaris
[03:47:29] *** Doc has joined #opensolaris
[03:47:29] *** fgd has joined #opensolaris
[03:47:29] *** junks has joined #opensolaris
[03:47:29] *** mustang has joined #opensolaris
[03:47:29] *** IvanR_ has joined #opensolaris
[03:47:29] *** j0sephi has joined #opensolaris
[03:47:29] *** Trinisan has joined #opensolaris
[03:47:29] *** cormac has joined #opensolaris
[03:47:29] *** paul has joined #opensolaris
[03:47:29] *** dvorak has joined #opensolaris
[03:47:30] *** quasi has joined #opensolaris
[03:47:30] *** yusufg has joined #opensolaris
[03:47:31] *** prg3 has joined #opensolaris
[03:47:31] *** heffnerd has joined #opensolaris
[03:47:31] *** trygvis has joined #opensolaris
[03:47:31] *** Drone has joined #opensolaris
[03:47:31] *** clee has joined #opensolaris
[03:47:31] *** kalif has joined #opensolaris
[03:47:31] *** mazon has joined #opensolaris
[03:47:31] *** Wez has joined #opensolaris
[03:47:31] *** Bart_M has joined #opensolaris
[03:47:31] *** lasseoe has joined #opensolaris
[03:47:31] *** pjd- has joined #opensolaris
[03:47:31] *** Justin has joined #opensolaris
[03:47:31] *** rydis has joined #opensolaris
[03:47:31] *** elektronkind has joined #opensolaris
[03:47:31] *** cnuke has joined #opensolaris
[03:47:34] *** yongsun_ has joined #opensolaris
[03:47:35] *** yongsun_ has quit IRC
[03:47:35] <rydis> sudo has a configuration file; sudo (in recent versions, IIRC) cleans up the environment by itself.
[03:47:37] *** mikaeld has joined #opensolaris
[03:47:46] *** stefanjo has joined #opensolaris
[03:47:46] *** AbeFroman has joined #opensolaris
[03:47:46] *** sparc-kly_ has joined #opensolaris
[03:47:46] *** SymmWork has joined #opensolaris
[03:47:46] *** libkeiser has joined #opensolaris
[03:47:46] *** PerterB has joined #opensolaris
[03:47:46] *** cypromis has joined #opensolaris
[03:47:46] *** richlowe has joined #opensolaris
[03:47:47] *** asyd has joined #opensolaris
[03:47:47] *** miffe_ has joined #opensolaris
[03:47:47] *** Trident has joined #opensolaris
[04:01:14] *** halton has joined #opensolaris
[04:04:59] <axisys> rydis: i dont think sudo is the problem, i just tried `su newuser' and still cant run xclock after that .. getting same X11 rejection
[04:05:53] <axisys> i ran echo $DISPLAY as me and as the new user .. same env
[04:09:25] <alanc> you need to share xauth cookies between the two users
[04:10:23] <alanc> if the new user can read your $HOME/.Xauthority (say if they're root and your home dir isn't NFS mounted) then have them setenv XAUTHORITY /home/originaluser/.Xauthority
[04:10:30] <alanc> otherwise you need to play with xauth
[04:10:41] <alanc> google for the "Remote X Mini HOWTO"
[04:11:32] <dclarke> any standards gurus around ? specifically when it comes to dealing with the lonlong_t typedef in sys/types.h that does not compile neatly with the -Xc strict ANSI compliance option in Studio 11 ?
[04:11:57] *** LeftWing has quit IRC
[04:12:46] *** LeftWing has joined #OpenSolaris
[04:13:11] <dclarke> okay .. never mind ..
[04:13:18] * dclarke goes back to work
[04:15:29] <axisys> alanc: i was hoping for a solution parameter on sshd_config or ssh_config :-)
[04:16:11] <Triskelios> axisys: you can ssh to localhost
[04:17:05] <axisys> Triskelios: dont follow.. this is what i do : ssh -X remoteserver, su newuser, xclock fails
[04:17:36] <axisys> alanc: googling ur string
[04:17:37] <Triskelios> axisys: because newuser has no access to the X server
[04:17:39] <delewis> axisys: the problem is your .Xauthority isn't being made available to the other user.
[04:18:05] <delewis> in the case of root, you can always do the quick hack, XAUTHORITY=/home/<user>/.Xauthority; export XAUTHORITY
[04:18:13] <Triskelios> right
[04:18:23] <axisys> delewis: any ssh parameters can drop the .Xauthority somewhere like /tmp automaticly with a switch?
[04:18:30] <alanc> replace "su user" with "ssh localhost -l user"
[04:18:47] <axisys> alanc, Triskelios .. let me try that
[04:20:12] <axisys> alanc, Triskelios .. that won't work cuz this test user is `oracle' and it cannot ssh .. ssh is PAM ==> Securid enabled..
[04:20:15] <rydis> Otherwise, something like xauth list $DISPLAY|sudo su user -c "xauth merge -" before the su might help, too.
[04:21:00] <axisys> rydis: hmm.. let me try that
[04:21:40] <rydis> (Or just do the merge manually after you've logged in.)
[04:22:40] *** LeftWing has quit IRC
[04:23:21] *** LeftWing has joined #OpenSolaris
[04:23:34] <alanc> or get someone to port the pass-my-xauth-cookies PAM module to Solaris (don't remember the real name, sorry - think Fedora has it)
[04:23:35] <Triskelios> rydis: I used to do that before I knew about XAUTHORITY, heh
[04:24:24] <Triskelios> as mentioned before, there might be support in sudo for this. there's also a su wrapper called 'sux'
[04:24:58] <rydis> Triskelios: The good thing about it is that it works across machines which don't share disk, too. ;)
[04:25:24] <Triskelios> rydis: right, which is why we use sux now
[04:25:45] <Triskelios> (which is basically identical to doing it manually)
[04:26:44] <axisys> Triskelios: i rather use sux .. for the sake of avg users
[04:26:57] <axisys> Triskelios: let me google that..
[04:27:28] <axisys> alanc: pass-my-xauth-cookies .. do u know how to implement that on solaris
[04:28:03] <Triskelios> axisys: just fix the module so that it builds (if it currently doesn't)
[04:29:45] *** solarisjon has joined #opensolaris
[04:29:48] <axisys> Triskelios: i am using pam_auth_radius module
[04:30:05] *** solarisjon has left #opensolaris
[04:30:18] <axisys> which talks to the radius daemon of our rsa securid server and then take rsa securid auth to validate user
[04:30:29] *** piwi has quit IRC
[04:30:50] <Triskelios> you need to add another module in the stack after authentication
[04:30:54] <axisys> so i guess pam_auth_radius (opensource) should need a fix
[04:31:10] <Triskelios> no
[04:31:13] <axisys> Triskelios: currently i dont .. but i c what u r saying
[04:31:18] <Triskelios> pam modules are all seperate
[04:31:24] <axisys> Triskelios: ofcourse
[04:32:22] <Triskelios> you probably don't have the module alanc mentioned right, so you need to get that, but you don't have to touch the other module
[04:33:15] <alanc> PAM lets you mix and match modules, so you can have both pam_auth_radius and pam_xauth_cookies in use
[04:33:18] <axisys> Triskelios: currently this is all i have "sshd auth required      pam_radius_auth.so"
[04:33:30] <axisys> alanc: ofcourse..
[04:33:44] <alanc> but first you would have to find the Linux code for pam_xauth_cookies (or whatever it's called), build it for Solaris, and install it
[04:34:01] *** besonen_mobile__ has quit IRC
[04:34:05] <axisys> alanc: i always get the wording between required/sufficient mixed up :-)
[04:34:32] *** karrotx has joined #opensolaris
[04:34:34] <alanc> looks like "pam_xauth" is the name of the Linux module: http://www.kernel.org/pub/linux/libs/pam/Linux-PAM-html/sag-pam_xauth.html
[04:34:36] <axisys> alanc: i guess i should stack "sshd session optional pam_xauth_cookies" underneath?
[04:34:40] *** LeftWing has quit IRC
[04:34:47] *** LeftWing has joined #OpenSolaris
[04:34:55] <axisys> assuming i have the module name right and vailable
[04:35:09] <alanc> I don't know - I always have to go ask the PAM guys for help or crack open the manuals when I edit pam.conf
[04:35:23] <axisys> alanc: lol .. same here
[04:35:46] <alanc> Dammit Jim, I'm an engineer, not a sysadmin!   I don't know how to configure crap, just build it!
[04:35:50] <alanc> 8-)
[04:36:06] <axisys> alanc: wish i have one of those "pam" guys
[04:36:37] <alanc> it does help to have your office about 100 feet down the hall from one of the original developers 8-)
[04:37:09] <axisys> alanc: must be nice
[04:37:33] <timeless> heh
[04:37:50] * timeless just wants to be ablet to set a passwd for a user
[04:37:59] <gdamore> i imagine it has plusses and minuses.  i'm not sure i would want to be the guy who had an office 100 feet down the hall from me. :-)
[04:38:15] <timeless> Re-enter new Password:
[04:38:15] <timeless> Oct 18 19:24:49 nexenta passwd[2519]: passwdutil.so: can't get domain
[04:38:20] <timeless> Permission denied
[04:38:41] <axisys> alanc: according to he url u sent me.. "su session  optional  pam_xauth.so"
[04:38:52] <axisys> ofcourse i need to check on the module name
[04:39:33] <alanc> gdamore: he comes down the hall to yell at us for the PAM abuse in xscreensaver in return, so it evens out
[04:39:54] <timeless> heh
[04:39:55] <gdamore> heh.
[04:40:06] <alanc> and of course, anytime someone gets a new laptop, you know whose door they end up at begging for xorg.conf help, so it goes around
[04:40:38] * timeless frowns
[04:40:42] * axisys taking notes alanc is xorg expert
[04:40:51] <axisys> lol
[04:40:52] <timeless> isn't nis something that normal end users don't use?
[04:41:01] <alanc> Dammit Jim, I'm an engineer, not a sysadmin!   I don't know how to configure crap, just build it!
[04:41:04] <timeless> this seems ot be from nis_getpwnam
[04:42:25] <alanc> though I have got the build working now...
[04:42:38] <gdamore> i dunno.  am i a normal user?
[04:43:05] <timeless> if you work for sun, you aren't normal
[04:43:20] *** karrotx has quit IRC
[04:43:21] <gdamore> i don't work for sun.
[04:43:30] <alanc> build_32/xorg-server-1.1.99.901/hw/xfree86/Xorg:        ELF 32-bit LSB executable 80386 Version 1, dynamically linked, not stripped
[04:43:30] <alanc> build_64/xorg-server-1.1.99.901/hw/xfree86/Xorg:        ELF 64-bit LSB executable AMD64 Version 1, dynamically linked, not stripped, no debugging information available
[04:43:37] <gdamore> (though I once did.)
[04:43:43] <alanc> (1.1.99.901 is Xorg 7.2 RC1)
[04:43:54] *** karrotx has joined #opensolaris
[04:44:29] <gdamore> how has the new reorganized Xorg 7.x packaging worked out for you, alan?  it looked like a nightmare to me.
[04:44:53] *** nwf has quit IRC
[04:45:29] <alanc> it's been nice in that it allows us to move a little bit at a time instead of all at once, but sure takes a lot longer to build once you add up 200+ runs of configure and thousands of runs of libtool
[04:46:28] *** nachox has joined #opensolaris
[04:46:31] <gdamore> yeah.  its not like building software doesn't take long _enough_
[04:46:44] <timeless> you could be building openoffice
[04:46:53] <alanc> it was nice when we got our first dual opteron and suddenly all of X built in half an hour
[04:46:56] <gdamore> netbsd has their own build framework to eliminate the extra configure for answers they already know.
[04:47:00] <delewis> god, printing Postscript on a non-Postscript printer is awful.
[04:47:13] <alanc> when I first started at Sun, one of the first builds of X I did took 24 hours
[04:47:18] <gdamore> only if you don't like to read forth?
[04:47:19] <timeless> cups/ghostscript?
[04:47:40] <delewis> thankfully, HP had enough of a brain in 94' to add support for an add-in module that would support Postsript
[04:47:43] <delewis> Postscript*
[04:47:51] <alanc> though that was on a SPARC 2 which was 10 years old by then - most people had Ultra 10's on their desk by then
[04:47:51] * gdamore confesses never to have done a full build of X since the mid 90's.
[04:47:52] <delewis> timeless: it's still slow as hell
[04:47:59] <delewis> I'm using the foomatic drivers to convert PCL to Postscript
[04:48:10] <delewis> 1pg/30 seconds roughly :-)
[04:48:28] <gdamore> convert PCL to PostScript?  not the other way around?
[04:48:31] <delewis> and the fact it's over parallel isn't helping matters, either -- I don't have the 10/100 module
[04:48:51] <delewis> gdamore: nope, HP LaserJet 4s use PCL, unless you purchased the extra SIMM module that added Postscript support
[04:49:00] <gdamore> heh.  I remember my first PS printer was an HP LJetIIP with an addin postscript cartridge.  slow as hell.
[04:49:15] <LeftWing> Hah, I have the PostScript SIMM for my LJ4. =D
[04:49:18] <delewis> well, *supposedly* this thing can print ~ 30pgs/minute
[04:49:23] <delewis> LeftWing: bastard. :-)
[04:49:29] <delewis> how fast is it with the PS SIMM?
[04:49:31] <LeftWing> I also have 24MB RAM and a JetDirect.
[04:49:35] <delewis> yeah
[04:49:44] <delewis> I'm planning on getting the JetDirect and maxing out the memory
[04:49:53] <delewis> it's cheaper than buying a new LaserJet, considering I got this one free today :-)
[04:50:03] <LeftWing> I haven't really played much with it in a while, I don't print often.  I don't think I ever got around to getting a PPD that would actually send PS to the thing from Solaris at the time.
[04:50:12] <gdamore> i must highly recommend the Brother series of printers ... I have a Brother HL5170DN, which includes postscript, 10/100 network, and _a duplexer_ for about $300.
[04:50:19] <delewis> LeftWing: it "just works" for me
[04:50:30] <delewis>  /usr/sadm/admin/bin/printmgr and *poof*
[04:50:31] <delewis> :-)
[04:50:33] <LeftWing> delewis: It works fine in PCL mode. ;P
[04:50:44] <delewis> ah
[04:50:59] <LeftWing> All the built-in PPDs are for foomatic/etc converters.
[04:51:09] <delewis> and foomatic is slow as hell :-/
[04:51:14] <LeftWing> It was pretty quick for me
[04:51:49] <delewis> might be the lack of memory or the parallel cable that's killing the speed then.
[04:51:52] <LeftWing> I've always had the Ethernet module though.
[04:51:54] <gdamore> i've had problems with foomatic accuracy with some postscript files.  Certain PDFs printed from Acrobat didn't render properly.
[04:52:21] <delewis> I just printed the first section in the AMD64 "user programming" spec -- looks great at 600dpi
[04:53:12] <astinus> Okay, I have a weird data migration problem
[04:53:24] *** gnu2it2 has joined #opensolaris
[04:53:46] <astinus> Old server is running Linux, has a hardware RAID controller with 250GB drives. Contains 1.65TB of data across the 8 drives, which are RAID-5
[04:53:55] <astinus> New server will run Solaris, using ZFS
[04:54:10] <astinus> I've bought three new 250GB drives, and plan to setup a storage pool and use RAID-Z
[04:54:13] <gdamore> heh.  The "afe" driver is going to have to get a PSARC fast track.  The amount of time/effort being spent on getting this code into Solaris amazes me.
[04:54:24] <gnu2it2> will solaris 10 run ok on a ultra 5 with 440mhz and 1 gb ram ?
[04:54:32] <gdamore> Sun has probably spent lke $10k just on man hours so far.
[04:54:40] *** jamesd_ has quit IRC
[04:54:43] <gdamore> gnu2it2: yes, it should.
[04:55:05] <gdamore> it might be a bit slow, but 440mhz is, well, slow, these days
[04:55:06] <astinus> If I set that up, then NFS ~500-600GB of data across .. I can pull two drives out of the Linux box (hardware RAID controller supports that, it just takes ages to remove a disk from an active array)
[04:55:06] <delewis> gnu2it2: I certainly hope so
[04:55:10] <delewis> well
[04:55:18] <delewis> the processor in the Ultra 5 isn't really the problem
[04:55:27] <delewis> it's the fact the shitty cmdk chipset doesn't do DMA.
[04:55:28] <astinus> Can I then insert these extra 250's into Solaris, add them to RAID-Z thus expanding capacity, and rinse and repeat until I've got all 1.6TB across?
[04:55:41] <delewis> which means even they slightest amount of IO on that system is going to make it crumble
[04:55:50] <gnu2it2> looks to be about $100 ebay upgrade from 277mhz and 128k,, sound like ok deal?
[04:55:57] <Triskelios> gnu2it2: I have an Ultra 5 2 feet away from me with 384M RAM
[04:56:07] <delewis> gnu2it2: have you already purchased on Ultra 5?
[04:56:16] <delewis> (if not, I'd recommend you take a look at the Ultra 2 instead)
[04:56:21] <gnu2it2> yes,, had it couple years
[04:56:26] <delewis> much better piece of hardware, as it uses SCSI and supports dual processors
[04:56:26] <delewis> ah
[04:56:32] <delewis> well, you can't go wrong then, I suppose :-)
[04:56:35] <timeless> woohoo, after replacing nsswitch.conf w/ nsswitch.dns i can ssh into my zone!
[04:56:38] <gdamore> ultra 2 was a sweet box.  god, i remember when i had one on my desk.
[04:56:43] *** switch has quit IRC
[04:57:02] <timeless> now all i need to do is make a backwards tunnel so that my zone can have internet access :(
[04:57:03] <delewis> between the Ultra 2 and SB1000, I don't know which one I love more as a workstation.
[04:57:11] <gdamore> then i left qualcomm, went to work for sun, and got a Sparcstation 5.  what a let down.
[04:57:18] <delewis> I've got an SB1000 sitting beside me now -- last workstation Sun ever made :-)
[04:57:20] *** switch has joined #opensolaris
[04:57:29] <axisys> Triskelios: hey i was playing w/ the sux .. did it work for u as it is?
[04:57:35] <Tekni> anyone in the market for a DAC-SATA-MV8 8port PCI-X SATA II card?
[04:57:36] <axisys> Triskelios: i get error message like this
[04:57:37] <gdamore> what's wrong with the U40?
[04:57:39] <axisys> Triskelios: /dev/fd/4: TERM=xauth -q remove localhost:21.0 2>/dev/null;: is not an identifier
[04:57:44] <movement> Subject: [tools-discuss] ON Mercurial repository mirror
[04:57:45] <movement> awesome
[04:57:52] <astinus> gdamore:  does the above sound feasible?
[04:58:03] <delewis> gdamore: Ultra 40 actually might not he that bad, but the SB1000 is built like a tank and it uses internal FC for the drives (along with SCSI for the CDROM and floppy)
[04:58:22] <delewis> Sun hasn't put FC back into the workstation since the SB1000/2000
[04:58:25] <gdamore> and sucks power like an electrolux.
[04:58:46] <delewis> it's not *that* bad, compared to the IBM p640 sitting behind me with 2xPOWER3s :-)
[04:59:19] <gdamore> our Bullfrog II (a "mobile" SB2500) has an external 300W power supply.  this is for a "mobile" unit. :-)
[04:59:36] <delewis> of course, the Bullfrog is basically a briefcase :-)
[04:59:52] <delewis> a very large one at that
[04:59:53] <gdamore> yeah, kinda.  have you seen one?
[04:59:58] <delewis> gdamore: only pictures
[05:00:09] <delewis> the fold-down keyboard is pretty slick :-)
[05:00:24] <gdamore> that's the "hardened" version.  i've not seen one of those.
[05:00:25] <delewis> though, it might look a bit weird whipping something like that out at the local Starbucks :-)
[05:00:30] *** jamesd has joined #opensolaris
[05:00:30] *** ChanServ sets mode: +o jamesd
[05:00:31] <delewis> gdamore: interesting
[05:00:33] <Tekni> i like my ultra 40
[05:00:49] <gdamore> the original commercial version was based on a "whitebook" (Acer?) 17" laptop.
[05:01:19] <delewis> it's amazing that Tadpole can fit an SB2500 into that kind of package.
[05:01:21] <gdamore> except it has a big expansion chassis added to accomodate a full length PCI card.  (Mostly for Sun PCi3)
[05:05:03] <gdamore> there are some "interesting" challenges in trying to convert sun platforms into mobile units.
[05:05:12] <gdamore> i'm looking forward to a design based on niagra though.
[05:05:35] <delewis> gdamore: is Tadpole looking in that direction?
[05:05:49] <delewis> or do they plan on waiting for the Niagra2 (which should have much better floating-point support)
[05:05:50] <gdamore> i cannae say.
[05:05:53] <delewis> ah :-)
[05:06:22] <delewis> s/support/performance/
[05:06:25] * delewis needs coffee
[05:06:43] <gnu2it2> would a Sun Ultra 2 Enterprise Dual 2x300MHz 512MB 18GB CDROM be better,, about $75 w/freithg?
[05:06:51] <delewis> gnu2it2: *much*
[05:07:00] <delewis> the Ultra 5 doesn't even come close to comparing to that
[05:07:18] <movement> anyone have any messages from onnv-notify??
[05:08:03] <noyb> not even close...  as I write this on my sad little U5.  Wish I had a pair of U2's fully stacked like that.
[05:08:32] <delewis> there's something to be said for doing an ON build (with a 4.0 load average on 2x750MHz UltraSPARC-IIIs) and not even feeling the hit.
[05:08:59] <delewis> not to mention I'm running JDS, with a Firefox session, and Thunderbird open :-)
[05:09:07] * gdamore is jealous.  my Bullfrogs are out for repair, and all I have for sparcs are 500MHz class systems.
[05:09:25] <delewis> gdamore: unfortunately, these USIIIs aren't the CU models :-(
[05:09:53] <delewis> the 900MHz USIII CU is supposedly 2 times as fast, roughly.
[05:09:55] <gdamore> i get really, really annoyed that i have to restart firefox all the time.  my opteron is zippy enough, but firefox eats all memory if not killed every so often
[05:10:12] <gdamore> i don't think the Cu units are that much faster.  they are a bit, but not 2x, I think.
[05:10:27] <delewis> 26576 dlewis    194M   47M sleep   47    4   0:08:55 0.7% acroread/1
[05:10:29] <delewis>  26159 dlewis    152M   48M sleep   47    4   0:06:52 0.2% firefox-bin/4
[05:10:29] <delewis>  26106 dlewis    115M   35M sleep   47    4   0:05:49 0.0% thunderbird-bin/3
[05:10:32] <delewis>  26033 dlewis    108M 8880K sleep   47    4   0:00:17 0.0% nautilus/2
[05:10:33] <delewis>  26031 dlewis     87M   12M sleep   47    4   0:00:25 0.0% gnome-panel/2
[05:10:41] <delewis> not surprising, those are the 5 processes with the largest size :-)
[05:12:20] <gdamore> i am still irked that the dual core USIIIi chips were killed.   we had an early unit based on them (a prototype) that _rocked_.  in a 15" case.  it used less power wth dual cores than the current single core systems.
[05:12:44] <timeless> do i want to know why gnome panel is using 87m?
[05:12:56] <delewis> timeless: no, and I'm not even using that many applets
[05:13:05] <Error_404> because gnome's a disgusting bloated mess?
[05:13:12] <delewis> workspace switcher, task manager, and the clock
[05:13:15] <delewis> and that's about it.
[05:13:22] <timeless> is that prstat?
[05:13:30] <delewis> prstat -s size -n 5
[05:13:46] <gdamore> that's why i use xfce.
[05:13:51] <timeless>    890 timeless  159M   42M sleep   59    0   0:00:25 0.0% java/20
[05:13:55] <delewis> 22:13:12  pgout/s ppgout/s pgfree/s pgscan/s %ufs_ipf
[05:13:56] <delewis> 22:13:17     0.00     0.00     0.00     0.00     0.00
[05:13:59] <delewis> surprisingly :-)
[05:14:02] <timeless> i wonder why i have java running
[05:14:02] <delewis> with 1GB of memory
[05:14:10] <delewis> (that's sar -g 5 5) -- swap statistics.
[05:14:26] <timeless> # sar -g 5 5
[05:14:27] <timeless> zsh: command not found: sar
[05:14:48] <delewis> which sar
[05:14:50] <delewis> /usr/sbin/sar
[05:15:26] <timeless> oh
[05:15:26] *** jamesd has quit IRC
[05:15:34] <timeless> one of my panels is this blue shield w/ a ? mark
[05:15:37] <timeless> which is java
[05:15:49] <delewis> that's Update Manager, probably
[05:15:51] <delewis> (kill it)
[05:15:56] <timeless> http://www.sun.com/service/sunupdate/
[05:16:13] <delewis> I always kill the network status & sound applet
[05:16:18] <timeless> does it really require me to register in order to not get any useful updates?
[05:16:27] *** jamesd has joined #opensolaris
[05:16:27] *** ChanServ sets mode: +o jamesd
[05:16:29] <delewis> the sound applet might be useful, but I'm using my SB1000 remotely via SSGD
[05:16:31] <gdamore> i wish sun wouldn't write key long lived software in java.
[05:16:50] <gdamore> there are enough pigs in the system.
[05:16:51] <delewis> gdamore: it's not really that bad, except in the case of the update tools.
[05:17:09] <delewis> the only things that are really written in Java tend to be the administration GUIs (dhcpmgr, printmgr, and SMC)
[05:17:18] <delewis> the first 2 are actually fairly decent, but SMC..
[05:17:27] <gdamore> i wrote our wifi gui in java, and i now wish i had done it in C/gtk+.  except GTK wasn't an option when I first wrote it
[05:17:42] <delewis> Tcl/Tk :-)
[05:18:22] * timeless wonders if 'remove from desktop' stops the update manager or just hides it
[05:18:29] * gdamore throws a brick at dewlewis
[05:18:34] <Auralis> hides
[05:18:34] <delewis> hehe
[05:18:48] <delewis> why do I get bricks thrown at me constantly? :-)
[05:20:04] * timeless sighs
[05:20:10] <delewis> throw one of those Bullfrogs my way, instead
[05:20:19] <timeless> my ssytem is configured to have a network proxy
[05:20:30] <timeless> but this silly java app didn't pick it up
[05:20:49] <gdamore> i can't throw a bullfrog very far.  they weigh 28lbs.
[05:20:54] <delewis> ouch.
[05:21:08] <delewis> I feel sorry for the soldier that has to carry one of those around.
[05:21:23] <gdamore> they get mounted on humvees. :-)
[05:21:32] <timeless> so how do i kill this thing if there's no preference or menu item for it?
[05:21:36] <gdamore> we sell thin clients for the soldiers. :-)
[05:21:40] <Error_404> me too, but for reasons quite aside from having to lug around a big computer
[05:21:45] <delewis> gdamore: nice :-)
[05:24:56] * timeless wonders what /usr/share/lib/jmdk/jdm is
[05:25:33] <nachox> does it have a man page?
[05:25:42] *** jsubl2 has quit IRC
[05:26:22] *** kloczek has quit IRC
[05:27:03] <timeless> SUNWjdmk-base it has a package
[05:27:16] * timeless tries to remember how to ask the package manager what a package is/does
[05:27:18] <jbalint> looks like some JMX/snmp package
[05:27:29] <timeless> application SUNWjdmk-base Java DMK 5.1 minimal subset
[05:30:09] * sparc-kly_ is away: tooo sleeeeeeeeeeeeeeeeeeeeeeeeeeep
[05:31:45] *** rodrickbrown has joined #opensolaris
[05:31:51] <gdamore> i notice that it looks like tcl got into ON.
[05:32:00] <gdamore> or at least the WOS
[05:32:03] <rodrickbrown> so does Sun plan to bring BlackBox to market?
[05:32:04] <timeless> ok, i still have no clue what it is, but it seems to be 60mb of my system
[05:32:17] <rodrickbrown> man that's a really novel idea but how practical is it this day and age
[05:32:40] <gdamore> i suspect it is _eminently_ practical in some places where it is hard to get a data center.
[05:33:00] <timeless> seems like fema should get some of them
[05:33:26] <gdamore> yeah, although i would not be surprised if they already have something like them.
[05:34:26] <jamesd> rodrickbrown, yes its going to go to market they are giving it a major push.... especially for companies that need to build out fast...
[05:34:51] <rodrickbrown> I remember rumors about google doing somehting like this
[05:34:57] <rodrickbrown> did Sun work with Google on this?
[05:35:15] <jamesd> rodrickbrown, check my blog it has all the details...   http://uadmin.blogspot.com
[05:35:35] <nachox> jamesd: needing the extra hits? ;)
[05:35:46] <jamesd> no..  google talked about it,  sun is going to sell it...  if google does it, its going to just be  for google.
[05:36:09] <jamesd> nachox, i can always use extra hits, just like nachos can always use extra cheese
[05:36:43] * nachox eats jamesd in a weird turn of events
[05:38:44] <gdamore> heh.  i see you have got your hands on niagra 2.
[05:38:50] <jamesd> rodrickbrown, what would be easier...  having a helicopter drop a blackbox on the roof, or  installing 8 racks full of systems...
[05:38:53] <gnu2it2> what is a bullgrog?
[05:38:58] <jamesd> gdamore, only in my dreams...
[05:39:03] <rodrickbrown> what's so virtualized about this setup though ?
[05:39:49] <jamesd> rodrickbrown, because its all in one box.. its a stretch to call it virtualized...
[05:40:05] <rodrickbrown> exactly
[05:40:10] <gdamore> i think virtualized has become another marketspeak thing.  like e-words and i-words, and (for Sun) java this or that.  (e.g. Java Desktop System, which has very very little to do with Java)
[05:40:28] <delewis> gdamore: nonsense, both have large memory footprints :-)
[05:40:31] <jamesd> but the idea is cool... even if  sun's maketing has got a few screws loose...
[05:40:33] <richlowe> I had assumed that the stuff *in* the box was in some way ... well.
[05:40:35] <richlowe> I dunno.
[05:40:40] <richlowe> more than in a damn shipping crate. :)
[05:40:59] <movement> hey richlowe, sent you email
[05:41:00] <delewis> richlowe: they've got some cabling infrastructure inside of it, as well as quick ways to move the racks around inside of it
[05:41:19] <jamesd> it is... its a ready to go systems...  just hook up.. and its ready to run...
[05:41:50] <gdamore> jamesd: just saw on your blog 10 things to do when the internet is down.  ROTFL.
[05:41:59] <delewis> of course, you have to ask how "resilent" the servers are when transported.
[05:42:17] <rodrickbrown> delewis, they are bolted odwn
[05:42:22] <delewis> unless they're throwing NEBS-certified equipment in there
[05:42:23] <jamesd> delewis, they say it can and has survied a drop of 6"....
[05:42:26] <rodrickbrown> if securly racked I dont see any issues
[05:42:47] <gdamore> the problem is cpus in sockets, and connections internally.  they have to have some kind of protection as well.
[05:42:54] <rodrickbrown> we just moved about 1000 servers from NY to NJ
[05:42:57] <rodrickbrown> not a single issue :-)
[05:43:18] <jamesd> i would be more worried about harddrives than cpus...  being dropped and vibrated
[05:43:20] <richlowe> movement: reading it now. :)
[05:43:26] *** nwf has joined #opensolaris
[05:44:05] <movement> richlowe: hoping you find the energy :)
[05:50:55] <gdamore> any PSARC experts around?
[05:52:20] <rodrickbrown> gdamore, no try #intel
[05:53:01] <richlowe> what counts as expert?
[05:53:15] <richlowe> like, are you look for a member, or just information, or?
[05:53:18] <gdamore> someone who can help me determine if a case qualifies for self review
[05:53:34] <richlowe> sure as hell ain't me then :)
[05:54:06] <gdamore> basically, my "afe" driver is in the process of being putback, but the question is whether or not addition of a driver qualifies as creating a new interface.  if it does, then a review (a fasttrack) would be needed.  otherwise it can self review.
[05:54:44] <richlowe> I would sincerely hope your sponsor would know.
[05:55:10] <rodrickbrown> I cut class today first time all year
[05:55:16] <gdamore> the sun employee who's working this is not necessarily fully savvy yet on this (alan duboff) -- i.e. i am not sure he has ever put anything back into ON yet.
[05:55:24] <rodrickbrown> one of my programming courses and I felt good about it :-)
[05:55:56] <gdamore> i felt much, much better about cutting english and history than i did programming classes.  mostly because I actually enjoyed the programming courses. :-)
[05:56:14] <gdamore> that was decades ago, though. :-)
[05:57:06] <gdamore> well, i suppose i should actually try to get some work done now.  gotta debug my spiflash driver (netbsd) for the meraki mini.
[05:57:19] <rodrickbrown> I enjoy programming but i'm so far advance than most
[05:57:38] <gdamore> then you need to take more interesting courses. :-)
[05:57:49] <rodrickbrown> I'm truly bored out my mind, most of my time spent in class is helping the teacher by helping other students
[05:57:56] <richlowe> well, atheros had one, but atheros was a whole new 'thing'.
[05:57:57] <rodrickbrown> gdamore, they wouldnt let me skip :-)
[05:58:01] *** maverickbna has joined #opensolaris
[05:58:05] *** ShadowHntr has quit IRC
[05:58:10] <richlowe> and that's the only recently added just-a-driver I can think of.
[05:58:15] <richlowe> try asking Tpenta or sommerfeld.
[05:58:33] *** maverickbna is now known as ShadowHntr
[05:58:34] <gdamore> i'll ask when they're around
[05:58:48] <timeless> richlowe: anyway, thanks for the help last night
[05:58:59] <richlowe> gdamore: Hm, even better, ask arc-discuss then everyone can see the answer :)
[05:59:04] <timeless> it was a bug in zones that someone already reported and fixed in the source tree
[05:59:21] <richlowe> timeless: the real bug, or the fun with UFS?
[05:59:25] <timeless> the real bug
[05:59:28] <richlowe> timeless: I think the UFS thing was exactly the CR you linked to, by the way.
[05:59:42] <richlowe> but I could easily be very wrong.
[06:00:23] *** laca has quit IRC
[06:00:44] <timeless> well i filed that one and it looks like i neglected to include the link :(
[06:01:01] <timeless> fwiw the secondary bug i never could trace
[06:01:07] <Tpenta> ask me what?
[06:01:28] <timeless> the one wear deadman fires because the dump didn't finish in 3mins
[06:01:50] <timeless> iow after all that time, i never did get coreing to work :(
[06:02:03] <timeless> i had to stick to mdb
[06:02:39] <timeless> kinda strange, because  tracking the main problem i was able to use :b effectively
[06:02:52] <timeless> but i couldn't get deadman:b to work right
[06:02:58] <Tpenta> I believe that the additino of a network driver can be done closed auto-approved; I generally submit those with a rider of "if anyone disagrees wit this qualifying for self review, I'll happily make it a fast track"
[06:03:49] <Seawolf99> need help. I am trying to upgrade ON from snv_46 to current current build, but when I run the bfu command I get a message "BIND 8 has been removed from ON; BIND 9 is available from SFW.
[06:03:49] <Seawolf99> Run /ws/onnv-gate/public/bin/migrate_bind9 to migrate to BIND 9", and bfu is aborted
[06:05:38] <gdamore> Tpenta: thanks.  I submitted the question to arc-discuss already. :-)
[06:06:17] <Tpenta> i saw, I'll reply
[06:07:58] <rodrickbrown> I heard a dirty rumor that sun is moving to grub on x86?
[06:08:02] <rodrickbrown> err
[06:08:04] <rodrickbrown> sparc
[06:08:08] <Tpenta> ummmm
[06:08:15] <Tpenta> i'd be surprised
[06:10:30] *** drifter747 has quit IRC
[06:10:45] <Error_404> why would they? the only reason you'd use grub is lack of a real firmware loader
[06:11:25] <rodrickbrown> that's what I thought not sure why they would drop it outer than cost savings?
[06:13:36] <Error_404> it would cost more, since OFW already has boot facillities
[06:13:39] <Tpenta> ok, replied to that mail
[06:13:44] <Error_404> porting grub over would cost money
[06:15:06] <rodrickbrown> http://www.rodrickbrown.com/docs/scripts/ypcat_ng.pl
[06:15:24] <rodrickbrown> might be useful if you spend time debugging nis/netgroups
[06:16:20] * timeless frowns
[06:20:11] <timeless> swift% /usr/sbin/ping 127.0.0.2
[06:20:18] <timeless> should that work ?
[06:20:34] <timeless> i suspect i've managed to misconfigure something :(
[06:21:16] <rodrickbrown> ?
[06:21:26] *** gba has joined #opensolaris
[06:21:30] <rodrickbrown> do you have 127.0.0.2 defined?
[06:21:47] <timeless> not specifically
[06:22:26] <gdamore> if it responds, then yes, you have something misconfigured.
[06:23:03] <timeless> timeless@boffo:~$ ping 127.0.0.2
[06:23:04] <timeless> PING 127.0.0.2 (127.0.0.2) 56(84) bytes of data.
[06:23:26] <gdamore> on grub and sparc: that would be a really, really poor choice IMO.  with the exception of the graphical boot menu, grub is very, very inferior to OFW in many ways.  they could use grub to get a secondary boot menu, but what would the point be?
[06:23:41] <timeless> i'm used to systems where 127.* will respond in general
[06:23:58] *** gm152 has joined #opensolaris
[06:24:12] <gdamore> such a system is probably misconfigured.  or  someone took a very lazy shortcut in the IP stack.
[06:24:14] *** silly_girl22 has joined #opensolaris
[06:24:37] <gdamore> (but for most systems, comparing a 32-bit address is no harder than comparing a 8-bit class A prefix)
[06:25:34] <gdamore> the world would have been a much, much better place if Intel had switched to IEEE 1275 (OFW) years ago...
[06:26:16] <gdamore> like right before they introduced PCI.  that would have been a good time. :-)
[06:26:37] <Auralis> OFW is not convoluted enough for x86, can't use it
[06:28:34] <timeless> didn't x86 recently switch to a replacement (whicih apple uses)? :)
[06:28:58] <Error_404> EFI
[06:29:03] <gdamore> Solaris x86 switched to grub.  Then there is EFI.  Which is something else, just as convoluted
[06:29:06] *** Tekni has quit IRC
[06:29:07] <gdamore> (as BIOS)
[06:29:14] <rodrickbrown> EFI is ncie
[06:29:18] <Error_404> and from the looks of it, apple will continue to be the only ones using it
[06:29:27] <Error_404> and IA-64, so SGI too i guess
[06:29:34] <Error_404> and HP
[06:29:49] <gdamore> i thought IA-64 was basically dead now.
[06:29:58] <rodrickbrown> nope
[06:30:01] <Error_404> don't tell HP
[06:30:27] <gdamore> well, i suppose it is all relative.  compared to PA-RISC, Itanic looks positively "promising".
[06:30:31] <rodrickbrown> alot of vendors are still selling IA64 Unisys, HP, Intel, Bull, Cray, SGI
[06:30:52] <Error_404> gdamore: heh
[06:30:53] <mustang> for very small values of 'selling'
[06:31:05] <gdamore> yeah, and HP/Dell still offer Alpha systems too.
[06:31:13] <gdamore> s/Dell/Compaq/
[06:31:54] <nachox> siemens is also selling ia64
[06:32:03] <mustang> they don't offer them. you have to ask them really nicely to get them to sell you one.
[06:32:07] <gdamore> i'll now the architecture is totally dead when NetBSD fully supports it. :-)
[06:32:14] <Error_404> lol
[06:32:19] <gdamore> s/now/know/
[06:32:24] <mustang> :)
[06:32:51] <nachox> netbsd alreaady suppports sparc
[06:32:52] * nachox hides
[06:33:02] <gdamore> that being said, i'm hacking NetBSD MIPS code at the moment.
[06:33:05] <Auralis> but not fully
[06:33:11] <gdamore> no US3i system support.
[06:33:15] <gdamore> no sparc64 SMP support.
[06:33:16] *** cturk has joined #opensolaris
[06:33:57] *** cturk has quit IRC
[06:34:13] <gdamore> before i got pissed off at internal politicking and such, i was going to work on NetBSD us3 support.  but i have largely abandoned NetBSD work in favor of Solaris.
[06:34:51] <richlowe> Tpenta: I believe it relates to zfsboot.
[06:34:53] <richlowe> (the grub thing)
[06:35:15] <gdamore> zfsboot?  can you have ZFS on a root drive?
[06:35:57] <gdamore> holy crap.  upgrading from S10 3/05 to S10 6/06 on SB100 takes _hours_.  i'm still waiting for my machine to finish.
[06:36:10] <gdamore> (using locally installed DVD drive)
[06:36:13] <richlowe> Tpenta: but you've seen 2006/525, and we haven't.
[06:36:13] <ShadowHntr> gdamore: shoulda seen my Ultra 5.
[06:36:13] <richlowe> so.
[06:36:20] <Auralis> that is fully normal for upgrading
[06:37:04] <gdamore> i figured it would be faster than scrapping all my downloaded software in /opt (SUNWspro, blastwave, etc.)  I think I misjudged.  reinstalling fresh and redownloading would have been faster.
[06:37:31] <dclarke> gdamore: its a tough call either way
[06:38:05] <gdamore> i have 16M down/2M up cable.  i think its not that close a call, in retrospect.
[06:38:34] <g4lt-mordant> gdamore, it is when you consider that the eri is less than useful ;P
[06:38:38] <gdamore> although, i am blessed by being able to do just do pkg-get -i <some packages> for most things. :-)
[06:38:41] <richlowe> it could easily be better not to use the local DVD too.
[06:38:43] <nachox> dclarke: http://blogs.sun.com/anay/entry/tamarack in case you havent seen it already
[06:38:52] <richlowe> ideally, you wouldn't use the local disk, either.
[06:38:57] <richlowe> or, really, an blade 100 in general ;)
[06:39:01] <g4lt-mordant> richlowe, what, and miss that blazing ATA100 speed? ;P
[06:39:18] <gdamore> heh.  its the only machine i can install s10u2 on at the moment.
[06:39:36] <jerome__> damn I forgot
[06:39:39] <gdamore> unless i want to dredge up a 300MHz U10 or a 333MHz voyager IIi.
[06:39:45] <nachox> g4lt-mordant: welcome to the 21st century, there is sata ;)
[06:39:50] <jerome__> is it this:
[06:39:55] <jerome__> zfs create zfs c0t1d0s0 c0t2d0s0
[06:39:58] <jerome__> or that
[06:40:11] <jerome__> zfs create zfs c0t1d0p0 c0t2d0p0
[06:40:14] <g4lt-mordant> nachox, not on the SB100.  is there even a sparc compatible SATA card?
[06:40:42] <gdamore> i don't know about "card", but we have some SATA chips in some of our newer SPARC systems.
[06:40:55] <gdamore> don't ask me what make/model, 'cause i don't know right now. :-)
[06:41:15] * g4lt-mordant hands gdamore a soldering gun.  when you get done with yours, you can do mine too
[06:41:15] <gdamore> (and you need our drivers, because SATA isn't supported on older Solaris)
[06:41:49] <dclarke> nachox: why am I looking at http://blogs.sun.com/anay/entry/tamarack  ??
[06:42:02] <gdamore> trust me.  you do _NOT_ want to hand me a soldering gun.  that's more dangerous than giving a hardware guy the root password. :-)
[06:42:22] <nachox> dclarke: it's what will replace vold, you were asking how fast that would happen
[06:42:47] <dclarke> ah
[06:42:50] <dclarke> thanks
[06:42:55] <dclarke> I needed the context
[06:43:01] <dclarke> okay ..
[06:43:17] <dclarke> and now .. I'll go pass out unconscious for a few hours
[06:43:26] <gdamore> speaking of ofw, i would _kill_ for IEEE 1275 firmware instead of the crappy redboot firmware this NetBSD box has.
[06:54:44] <dclarke> http://maps.google.ca/?hl=en&ie=UTF8&z=18&ll=46.774219,-65.003663&spn=0.001885,0.003616&t=k&om=1
[06:54:55] <dclarke> parking lot in the middle of no where
[06:54:57] <rodrickbrown> ok folks i'm off to bed
[06:55:03] * dclarke slightly OT
[06:55:12] <dclarke> nighty
[06:55:30] <gdamore> g'nite rodrickbrown.
[06:56:44] <gdamore> dclarke: it looks like there is a small building just south, but its hard to see.  is this an entrance to some underground complex. :-)
[06:57:08] *** nwf has quit IRC
[06:59:57] <dclarke> its middle of nowhere
[07:00:02] <dclarke> totally
[07:01:00] <dclarke> here : http://www.blastwave.org/dclarke/stuff/google/index.html
[07:01:19] <dclarke> just every now and again when I take a moment to watch a compile run ..
[07:01:27] <dclarke> I scan the planet for "interesting" places
[07:05:09] <asyd> \_o<
[07:05:39] <jamesd> dclarke, have you called and emailed all your sun buddies, you could offer to test a blackbox for them and  submit it to the harsh  candian winters..
[07:06:12] <dclarke> my electric bill is big enough
[07:06:12] *** Teknix has joined #opensolaris
[07:06:38] * Tpenta wonders if they'll be offered on try/buy ;)
[07:07:05] *** besonen_mobile__ has joined #opensolaris
[07:07:08] <dclarke> this is odd
[07:07:14] <dclarke> http://maps.google.com/?ie=UTF8&om=1&z=18&ll=55.754219,37.627911&spn=0.001537,0.003616&t=k
[07:07:23] <dclarke> a rip in the map ..
[07:07:28] <jamesd> Tpenta, only if you make enough money  to buy it out right will your credit ever be approved
[07:07:31] <dclarke> in the middle of Moscow
[07:08:38] <Tpenta> it's a rip in the fabric of space-time
[07:08:57] <dclarke> where is Douglas Adams when we need him ?
[07:09:08] <dclarke> oh yeah .. never mind .. he's dead
[07:09:11] <timeless> with the dolphins :)
[07:09:11] <Tpenta> if yo zoom back one, you will see another one to the north of it
[07:09:18] <dclarke> I noticed
[07:09:22] <timeless> so long, fare well, thanks for all the fish
[07:09:29] <dclarke> looks to be flare caused by sunlight
[07:09:51] <dclarke> definately a problem for a high resolution satellite image
[07:10:00] <Tpenta> too sharp for that i would have thought
[07:10:06] <dclarke> although this looks to be a aircraft image
[07:10:13] <gdamore> i wonder how big the google database is.  i once examined blue marble, i think i did the math that at 500m/pixel it was something like 130TB.
[07:10:24] <dclarke> well .. no .. the best quality optics from the military
[07:10:25] <Error_404> dclarke: s/sattelite/kremlin
[07:10:36] <dclarke> could spot a pack of cigs
[07:10:57] <gdamore> there is a clear change from satellite imagery to aircraft imagery.  big difference in quality
[07:10:58] <dclarke> the whole concept of diffraction limited optics was pushed to the limit indeed
[07:11:05] *** mikaeld has quit IRC
[07:11:48] <dclarke> http://maps.google.com/?ie=UTF8&om=1&z=18&ll=55.7696,37.613196&spn=0.001536,0.003616&t=k
[07:11:57] <dclarke> thats really a lot of error there
[07:12:10] *** sahafeez has joined #opensolaris
[07:12:13] <Error_404> what are the soviets hiding?
[07:12:36] <dclarke> down in Australia : http://maps.google.com/?ie=UTF8&om=1&z=18&ll=-46.662264,169.021107&spn=0.001874,0.003616&t=k
[07:12:46] <dclarke> an amazing looking landscape there
[07:12:57] <gdamore> i did some very coarse mapping using xplanet and blue marble data (8km/pixel) for netbsd at http://www.netbsd.org/gallery/xplanet
[07:13:10] <Error_404> dclarke: alien landing strip
[07:13:37] <gdamore> someone went nuts with a lawnmower. :-)
[07:13:51] <dclarke> zoom out and look to the south west
[07:13:56] <dclarke> amazing beach
[07:14:03] <dclarke> pristine beach there
[07:14:18] <dclarke> http://maps.google.com/?ie=UTF8&om=1&z=18&ll=-46.666111,169.056936&spn=0.001874,0.003616&t=k
[07:14:21] <dclarke> bloody amazing
[07:14:23] <Tpenta> hate to tell you dennis, that's new zealand
[07:14:32] <dclarke> yeah .. I knew that
[07:14:39] <dclarke> I was waiting for you the smack me !
[07:14:41] <LeftWing> We don't talk about New Zealand.
[07:14:46] <dclarke> oh ?
[07:14:51] <dclarke> oh .. sorry
[07:14:58] <dclarke> social faux pas I guess
[07:15:25] *** gba has quit IRC
[07:16:16] <Tpenta> http://maps.google.com/maps?f=q&hl=en&q=828+pacific+highway+gordon+nsw+australia&ie=UTF8&z=18&ll=-33.753643,151.151105&spn=0.003051,0.005364&t=k&om=1 <- where I am now
[07:16:18] <gdamore> wow.   i bet those waves are _enormous-
[07:16:38] *** Trident has quit IRC
[07:17:08] <dclarke> you under the green arrow ?
[07:17:16] <Error_404> wave
[07:17:23] <Tpenta> well I'm on the other side of the building to the arrow
[07:17:24] <LeftWing> Tpenta: That's the building with the 3D metal blue Sun logo out the front?
[07:17:32] <Tpenta> I was on the balcony to the left about 15 minutes ago
[07:17:35] <Tpenta> yup
[07:17:51] <dclarke> can you stick yer head out a window and wave ?
[07:17:56] <dclarke> I'll hit refresh !
[07:17:59] <Error_404> apparantly sun's got a campus in burnaby
[07:18:00] <Tpenta> nope
[07:18:06] <LeftWing> I drive past that thing on the way into North Sydney for SOSUG. =P
[07:18:10] <dclarke> don't laugh .. its in the works ya know ... live feeds
[07:19:03] <gdamore> heh.  i'll watch for google making a buy offer on some big aircraft company.   :-)
[07:19:17] <dclarke> http://maps.google.com/maps?f=q&hl=en&q=828+pacific+highway+gordon+nsw+australia&ie=UTF8&om=1&z=13&ll=-33.753675,151.151142&spn=0.072647,0.171661&t=k
[07:19:25] <dclarke> geez .. thats a city there
[07:19:37] * gdamore would like to visit oz someday.
[07:19:38] <dclarke> we don't have anything like that in my neck of the woods
[07:19:53] <Error_404> dclarke: what, a city?
[07:20:04] <dclarke> buildings
[07:20:09] * LeftWing wonders if Iridium sats have cameras attached.
[07:20:09] <Error_404> heh
[07:20:15] <Tpenta> you aint even close to the CBD yet
[07:20:20] <Error_404> LeftWing: don't ask.
[07:20:26] <dclarke> actually .. a few years ago my nearest neighbor would have been cows
[07:20:27] <LeftWing> Error_404: Why not? =P
[07:20:38] *** sickness has joined #opensolaris
[07:20:49] <Error_404> you know about that whole "redefinition of torture" thing?
[07:20:50] <Error_404> ...
[07:21:00] <dclarke> I could walk out my back door .. and keep walking for fifteen minutes and hit nothing but trees
[07:21:05] <LeftWing> I don't quite follow you, Error_404. ;P
[07:22:51] <dclarke> look at this : http://maps.google.com/?ie=UTF8&om=1&z=8&ll=51.337476,-68.703003&spn=1.746736,3.702393&t=k
[07:23:05] *** Seawolf99 has left #opensolaris
[07:23:05] <gdamore> woo hoo.   my spiflash strategy() didn't panic this time.
[07:23:05] <dclarke> clear evidence of a massive meteor impact
[07:23:39] *** bougie has joined #opensolaris
[07:23:43] <LeftWing> dclarke: Or some sort of synthetic structure... ALIENS PEHRAPS!?
[07:23:54] <Tpenta> interesting
[07:24:02] <dclarke> worse than that .. the French !
[07:24:07] <LeftWing> LE' ACCUSE!
[07:24:12] <g4lt-mordant> dclarke, yeah, but it missed: DC is still there
[07:24:21] <dclarke> its dead obvious I think
[07:24:31] <dclarke> a circle like that .. nearly perfect circle
[07:25:33] *** nachox has quit IRC
[07:25:55] <Error_404> lol
[07:27:27] <dclarke> before I moved into town .. this is where I lived : http://maps.google.com/?ie=UTF8&om=1&z=13&ll=44.063413,-78.137727&spn=0.062785,0.1157&t=h
[07:27:45] <dclarke> now I am surrounded by bloody houses ..
[07:28:00] <timeless> so how would i make 127.0.0.2 work in a zone? i tried :
[07:28:00] <timeless>   <network address="127.0.0.1/8" physical="lo"/>
[07:28:01] <timeless>   <network address="127.0.0.2" physical="lo"/>
[07:28:09] <dclarke> but .. there is nothing here for miles with more than three floors
[07:32:56] <dclarke> now I'm really heading off to sleep
[07:32:59] <dclarke> ta !
[07:33:15] *** MadMethod has joined #opensolaris
[07:34:22] *** silly_girl22 has quit IRC
[07:36:37] <gdamore> aargh.  i wish od had a flag to change the endianness for -x.
[07:39:48] <richlowe> Hm, that was foolish.
[07:39:53] *** jamesd has quit IRC
[07:40:10] <richlowe> nothing like replying off-list, demarking the fact you've done so.  But then killing the text that caused you to want to, but forgetting to re-add the Cc.
[07:41:40] <richlowe> Tpenta: do you have any kind of intended schedule for putting the -nd stuff back properly?
[07:42:21] <Tpenta> not yet, I'm running with the stuff on steve's timetable to iron out nits at the moment
[07:42:36] <richlowe> That's fine, I'm just trying to schedule some other thinking around it.
[07:43:17] *** Method has quit IRC
[07:43:18] *** jamesd has joined #opensolaris
[07:43:18] *** ChanServ sets mode: +o jamesd
[07:43:18] <richlowe> I'll Cc you on the mail, then that statement may make some sense. :)
[07:50:22] *** bougie has quit IRC
[07:55:23] *** jamesd has quit IRC
[07:55:27] *** Disorganized has joined #OpenSolaris
[07:57:13] *** dwc- has joined #opensolaris
[08:04:05] <richlowe> actually, no I won't, I got several paragraphs in then realized the scheme is unworkable.
[08:05:35] *** jamesd has joined #opensolaris
[08:05:36] *** ChanServ sets mode: +o jamesd
[08:06:46] <Tpenta> what were you trying to do?
[08:06:59] <richlowe> I'm trying to come up with a scheme to automatically handle the closed-bins (from a user POV)
[08:07:21] <richlowe> such that the bringover step in nightly (post Danek's hg away nightly), will bringover new closed bins if the closed tree is not present.
[08:07:26] <richlowe> s/away/aware/
[08:07:54] <richlowe> the problem there is finding a neat way not only to snag everything, but to try to avoid pulling what would currently be the same bins over and over and over for no good reason.
[08:07:57] <Tpenta> i suspect you *really* don't want to go there
[08:08:20] <richlowe> Given that I already am "there", care to explain why?
[08:08:43] <Tpenta> there is already a lot of hackery to build without usr/closed in there
[08:09:06] <richlowe> and more to refresh the closed-bins would be seen as uncouth to those not having to deal with it?
[08:09:09] <richlowe> To be blunt "So?"
[08:09:36] <Tpenta> are you talking about adding stuff into the build system or something else?
[08:09:43] <richlowe> nightly, only nightly.
[08:09:48] <richlowe> nightly without -n runs bringover.
[08:09:55] <richlowe> Danek had a copy of nightly that was aware of mercurial for that.
[08:10:06] <Tpenta> oh, I see
[08:10:07] <richlowe> refreshing the closed-bins in that same step seems not only logical, but just about the only sane way to go about it.
[08:10:35] <Tpenta> IF we are pushing out a closed-bins on every mercurial update
[08:10:45] <richlowe> You aren't, but you should.
[08:10:51] <Tpenta> the problem is that you are looking at about 3 hours or so build time on each of sparc/x86 for thise
[08:10:52] <richlowe> That was also part of my thinking.
[08:11:13] <richlowe> I have another idea there, which I know I've mentioned to you before, but it has large areas of fog when seen from outside Sun.
[08:11:35] <Tpenta> I am thinking that any putback involving usr/closed should perhaps have a flag day
[08:12:07] <richlowe> Well, I've mentioned previously that I can't see any barrier to generating the closed bins from the gate's proto area.
[08:12:12] <richlowe> sure, it's only nightly, but that's better than weekly.
[08:12:19] <richlowe> and given build times for 4 builds, not that much different.
[08:12:32] <Tpenta> it's actually 6 builds
[08:12:49] <Tpenta> you need to do an open build to find out what is delivered by the current open tree
[08:12:53] <Tpenta> as it changes
[08:12:58] <richlowe> my fear of that suggestion is that you have nearly a years worth of work that suggestion effectively kills. :)
[08:13:03] <richlowe> well, not really kills.
[08:13:12] <richlowe> you'd effectively be using the gate nightly as your 'closed' build, in the same process.
[08:13:36] <Tpenta> I *have* been talking to danek about using the nightlies
[08:13:39] <richlowe> I'm not sure if the gate keeps both builds though, given the multi-proto stuff got backed out (I think)
[08:13:44] *** lisppaste3 has quit IRC
[08:13:54] <Tpenta> but you still need an open build AND he's currently not doing debug and non-debug nightlies
[08:14:09] <Tpenta> the gate does not do an open build
[08:14:13] <richlowe> Yeah, that was some of the fog you can't see through from outside Sun.
[08:14:14] <richlowe> :)
[08:14:25] <richlowe> I know it doesn't, but that cuts your N builds down to N/2.
[08:15:15] <richlowe> but the actual meat of this, is that I'm not sure of a sane distribution mechanism either way.
[08:15:22] <Tpenta> danek has plans to do the debug/non-debug into nightly, that will cut the generatiotime down to under an hour
[08:15:40] <richlowe> The way I do it is to drop the closed bins into mercurial, but mid-way through writing the mail explaining that, I realized why it could never work well.
[08:15:56] <richlowe> (*everything* changes each time the closed bins are generated, so you gain nothing)
[08:16:15] <Tpenta> yup
[08:16:22] <Tpenta> you'd be just as well doing wgets
[08:16:23] *** halton has quit IRC
[08:16:48] <richlowe> well, you'd be slightly worse off in that you could get identical stuff multiple times.
[08:16:52] <richlowe> but not much worse off, no.
[08:17:15] <Tpenta> or an rsync
[08:17:27] <Tpenta> and things get *really* messy if you are running an old build
[08:18:00] <richlowe> Building old source, or the build machine on an old build?
[08:18:07] <Tpenta> old source
[08:18:11] <richlowe> build old source, you would not have nightly bringover.
[08:18:14] <richlowe> your closed bins wouldn't be refreshed.
[08:18:17] <Tpenta> you know what is really frustrating?
[08:18:21] <richlowe> thus what they contain is your own responsibility.
[08:18:39] <Tpenta> do you know how much of closed you need to build in order to do an open build to see what files open delivers?
[08:18:56] <richlowe> 1 thing on x86, somewhere around 3  on sparc.
[08:19:05] <richlowe> libc_i18n on both, and a couple of sparc things I forget
[08:19:06] <Tpenta> yea
[08:19:09] <richlowe> libdisasm/dis and something else.
[08:19:28] <Tpenta> teh sparc disasm stuff and libi18_n.a
[08:19:44] <richlowe> oh, the rest of the prtdiag-ish bits opened I guess.
[08:20:38] <richlowe> handling flag days is going to be a lot of fun, by the way.
[08:20:44] <Tpenta> on x86 it's just the two libc_i18n.a's and on sparc it's that and the disasm stuff, yea
[08:21:11] <Tpenta> all the other stuff should not really cause issues if it is a little out of sync
[08:21:14] <Tpenta> hmmmmm
[08:21:19] <richlowe> I suspect the first one will lead to a lot of angry people for a few days.
[08:21:38] <richlowe> between open and closed, it's easier.
[08:21:45] <richlowe> between consolidations it's not very good *now*
[08:21:48] <richlowe> and it only gets worse.
[08:21:54] <Tpenta> the update_on stuff needs to be opened up along with making it work with an install dvd
[08:22:45] <Tpenta> I *suspect* that as long as libc_i18n.a and the disasm stuff doesn't change, for the most part, you are going to be pretty safe
[08:22:51] <Tpenta> hmmmmm
[08:23:50] <Tpenta> those are the only libs that get used during a build, everything else should stand on it's own, UNLESS we break kernel driver interfaces
[08:23:55] <Tpenta> I think
[08:24:12] <richlowe> there aren't many closed drivers that are actually *important*
[08:24:14] <richlowe> beyond sparc ata.
[08:24:17] <richlowe> (I forgot that one, sorry)
[08:24:33] <richlowe> as long as the breakage doesn't panic, and you're not on a larger sparc machine, you probably wouldn't actually lose much.
[08:24:42] <Tpenta> you don't need sparc ata to do the build
[08:24:50] <richlowe> Yeah, but if you then BFU you lose your disk.
[08:25:16] <richlowe> but again, closed-bins refreshes for such things are easy.
[08:25:17] <Tpenta> no, what i mean is that it should be safe to use a "recent" encumbered-bins tarball
[08:25:23] <richlowe> It is.
[08:25:35] <richlowe> I did it by accident a *lot*, that's why I started putting them in hg, and started looking into this.
[08:25:36] <Tpenta> as long as the things we link against don't change
[08:25:45] <Tpenta> I might have to do a blog on this
[08:25:46] <richlowe> there's been times I've used month-old closed bins accidentally.
[08:26:20] <Tpenta> you are only going to get issues where you have a serious flag day, and then genberally only on the stuff required to do the build
[08:26:29] <richlowe> I've been told (I forget by who, by the way) that update_nonON, and similar *did* work against a net image.
[08:26:34] <richlowe> the problem is SXCR is a long long time coming.
[08:26:39] <richlowe> so you still have to get that stuff out sooner.
[08:26:41] <Tpenta> they do
[08:26:58] <Tpenta> take it from me, I have run them against different images than the one at nimpk-17.eng
[08:27:17] <richlowe> I've also been bugging people about those scripts for pretty much a year now. :)
[08:27:43] <richlowe> at one point, I'd thought Bonnie said they'd be looked at, but I'm assuming they never were.
[08:27:58] <Tpenta> I see another quest for moi
[08:28:47] <richlowe> there's also the bfu handling of flag days thing.
[08:28:52] <richlowe> which the release notes say there's a fix for.
[08:28:57] <richlowe> but they've said that since June 2005. :)
[08:29:16] <Tpenta> i think if we can get the update_nonon stuff out, that should address much of that
[08:29:50] <timeless> woohoo
[08:30:00] <timeless> ok, i can use apt-get update in my nextenta zone :)
[08:31:10] *** Disorganized has quit IRC
[08:31:51] <timeless> Setting up sunwusb (5.11.50-1) ...
[08:31:52] <timeless> Cannot open (/dev/ksyms): No such file or directory.
[08:32:06] * timeless sighs, well, nobody's perfect
[08:32:25] <richlowe> Tpenta: 4865419
[08:32:31] *** gnu2it2 has quit IRC
[08:33:05] <richlowe> ... that's totally not the CR it should be.
[08:33:12] <richlowe> at least, not from the public info.
[08:33:13] <richlowe> hang on.
[08:33:40] <richlowe> actually, the Release notes say that *is* the CR, but it's talking about something else in the actual description.
[08:34:37] <richlowe> Tpenta: and getting the upgrade_noon stuff out only addresses it to the extent the required packages are also available.
[08:34:40] <richlowe> Tpenta: I'd imagine *that* is the hard part.
[08:37:36] <Tpenta> ok, it looks like the part that needs work is the auxilliary scripts for bfu
[08:38:24] <richlowe> well, as I said, that CR's description doesn't match what the release notes suggest it should be.
[08:38:29] <richlowe> so I'm confused by it overall currently
[08:38:38] <richlowe> (though fixing FASTFS and friends to not need setting should be fairly easy)
[08:38:47] <Tpenta> the important stuff is in the evaluation
[08:38:48] * richlowe just does bldenv -t env/something 'bfu nightly-nd'
[08:39:04] <richlowe> (I set ARCHIVEPATH=$CODEMGR_WS in the env files)
[08:44:29] <Tpenta> sorry i have some stuff i need to deal with
[08:45:40] <richlowe> heh, don't we all :)
[08:46:42] <Error_404> i don't
[08:46:45] <Error_404> *shrug*
[08:46:50] <Error_404> my life's pretty low stress
[08:53:43] *** halton has joined #opensolaris
[08:53:58] *** Kernel86|Laptop has quit IRC
[08:54:06] *** Kernel86|Laptop has joined #OpenSolaris
[08:54:39] *** Cyrille has joined #opensolaris
[09:01:31] *** slowhog has quit IRC
[09:05:49] *** mikaeld has joined #opensolaris
[09:07:58] *** Dar_HOME is now known as Dar
[09:17:23] *** Sieghard has quit IRC
[09:21:49] * gdamore is very happy he figured out weird SPI bus addressing problem.
[09:27:31] *** jamesd_ has joined #opensolaris
[09:44:58] *** jamesd has quit IRC
[09:52:07] <jteo> gdamore: kudos
[09:52:30] *** kimc has joined #opensolaris
[09:53:37] <kimc> good morning
[09:58:57] <timeless> gettimeofday()
[10:01:02] *** dunc has quit IRC
[10:01:13] <kimc> coffee rdy
[10:02:34] *** sahafeez has quit IRC
[10:02:51] <timeless> oh right, in snv_49/nautilus, is the view list (instead of icons) option supposed to work?
[10:02:54] *** ShadowHntr has quit IRC
[10:02:55] <timeless> it doesn't work for me
[10:10:05] *** Gman has quit IRC
[10:10:06] *** alanc has quit IRC
[10:10:58] *** Gman has joined #opensolaris
[10:11:00] *** alanc has joined #opensolaris
[10:18:59] *** kFuQ has quit IRC
[10:29:06] *** Tpenta has quit IRC
[10:32:17] *** spike723 has joined #opensolaris
[10:32:29] <spike723> I'm having trouble changing my screen resolution in nexenta
[10:32:49] <spike723> I don't see /etc/X11/xorg.conf on my system, this ia a fresh install
[10:33:19] <trygvis> that's correct, it depends on xorg to detect and configure itself
[10:33:20] *** mazon is now known as Mazon
[10:33:22] <Cyrille> isn't there an Xorgconfig or similar exec?
[10:33:45] <trygvis> you can run X with a special argument to have it write out the detected config and tweak that
[10:33:59] <spike723> how about dexconf?
[10:36:07] *** medar has quit IRC
[10:36:15] *** ShadowHntr has joined #opensolaris
[10:36:45] *** bengtf has joined #opensolaris
[10:37:38] *** Sieghard has joined #opensolaris
[10:38:11] *** damienc has joined #opensolaris
[10:39:32] <spike723> how do I get to console?
[10:39:39] <spike723> I have to kill gdm or something?
[10:41:01] <trygvis> svcadm -t disable gdm; and kill X
[10:41:01] <trygvis> kill or log out
[10:41:09] <nightswim> there should be an option that says "console login"
[10:41:14] <spike723> svcadm is a solaris tool?
[10:41:27] *** jmcp has joined #opensolaris
[10:41:29] <Cyrille> yes
[10:45:38] *** maverickbna has joined #opensolaris
[10:45:46] *** ShadowHntr has quit IRC
[10:46:27] *** maverickbna is now known as ShadowHntr
[10:50:43] <kimc> if I log into the console then logout, and the windowing system doesn't restart, is there a way to restart it short of rebooting ?
[10:51:06] <trygvis> svcadm restart gdm perhaps
[10:51:16] <kimc> trying,,
[10:52:47] <kimc> nope
[10:53:18] <kimc> gdm was not running and using svcadm to restart didn't change things
[11:04:17] *** janus_man has joined #opensolaris
[11:08:01] <kimc> looks like its Ok now.. that after b49 crashed while: find / -name wget -print
[11:08:19] <kimc> hope its not bad hardware..
[11:11:58] *** ShadowHntr has quit IRC
[11:19:59] *** MattMan has joined #opensolaris
[11:32:08] <ofu_> wow, Bug ID:    1085341... Submit Date   04-MAR-1992, Release Fixed    solaris_nevada(snv_39)
[11:41:10] *** MattMan is now known as MattAFC
[11:48:14] *** MattAFC is now known as MattMan
[11:52:36] <timeless> kimc: i had that problem on beos (bad fs driver) :)
[11:54:14] *** Joker^ has joined #opensolaris
[11:54:54] <kimc> thanks for the help.. time to roll, cu all
[11:54:59] *** kimc has left #opensolaris
[11:57:40] *** dunc has joined #opensolaris
[12:05:03] *** Joker^ has quit IRC
[12:21:56] *** calumb has joined #opensolaris
[12:26:50] *** bondolo has quit IRC
[12:30:06] *** SymmWork has quit IRC
[12:33:25] *** Capricorn^80 has joined #opensolaris
[12:47:50] *** trs81 has quit IRC
[12:49:46] *** Gman is now known as GmanAFK
[12:57:48] *** coffman has joined #opensolaris
[12:59:33] <musasabi> Is there an easy way to get (Open)Solaris installed and booting mostly from a network? The machine has CDROM + can boot from USB sticks, and I'd like to have it booting from USB stick and then mounting everything from a non-solaris NFS server.
[13:00:22] <oxygene> uh, "non-solaris nfs server"... there are certain nfs servers that don't quite live up to the specs
[13:01:10] <oxygene> other than that, it shouldn't be too hard. did you consider booting via pxe, too?
[13:01:36] <musasabi> oxygene: I have Linux and NetBSD server machines perhaps one of them might be able to service NFS properly enough.
[13:02:37] <musasabi> PXE booting works for me (if the USB-stick + mounting) is not easier. But how would I install Solaris to the nfs-root? (if not using the USB-stick)
[13:03:16] <musasabi> All the guides seem to be assuming a Solaris server and solaris specific commands for creating the nfs-root.
[13:03:28] *** crib|nb has joined #opensolaris
[13:04:59] <quasi> musasabi: http://www.samag.com/documents/s=9598/sam0504e/ may be an option
[13:05:04] <oxygene> well, regular install, there you can setup remote mountpoints at some place - I'm not 100% sure if that allows nfs-root installs, but that's the first way I'd try it
[13:06:35] <musasabi> ok, will read the article and also try the remote mountpoints if it supports that.
[13:08:48] *** crib|nb has quit IRC
[13:09:00] *** crib|nb has joined #opensolaris
[13:10:59] *** Capricorn^80 has quit IRC
[13:12:49] *** tsoome has joined #opensolaris
[13:19:05] *** LeftWing has quit IRC
[13:22:46] *** LeftWing has joined #OpenSolaris
[13:23:30] *** lloy0076 has joined #opensolaris
[13:28:10] *** crib|nb- has joined #opensolaris
[13:32:31] *** dunc has quit IRC
[13:33:53] *** crib|nb has quit IRC
[13:37:23] *** crib|nb- has quit IRC
[13:46:51] *** LeftWing_ has joined #OpenSolaris
[13:47:02] *** LeftWing has quit IRC
[13:54:21] *** salamanders has joined #opensolaris
[13:56:05] *** laca has joined #opensolaris
[14:00:46] *** qbit has left #opensolaris
[14:00:48] *** GigaE has joined #opensolaris
[14:01:34] *** calumb is now known as calLNCH
[14:01:49] <GigaE> can somebody with snv_48 or higher please run
[14:01:50] <GigaE> 1) ld -znointerp
[14:01:50] <GigaE> 2) which ld
[14:01:50] <GigaE> and tell met the results ?
[14:02:36] <GigaE> the reason is that I get an error with this flag when building xen on snv_46 and I encountered the following link when googling for the error:
[14:02:37] <GigaE> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6465623
[14:03:02] <GigaE> which says that snv48 fixed it ; I hesitate before upgrading to 48
[14:03:12] <GigaE> and simply want to verify this
[14:03:51] <GigaE> according to the Xen solaris web site snv 44 is enough though
[14:10:18] <icon> morning all
[14:13:23] <rodrickbrown> hi
[14:18:24] <GigaE> I am repeating my question:
[14:18:33] <GigaE> can somebody with SXCR snv_48 please run from the command line:
[14:18:33] <GigaE> "ld -znointerp"
[14:18:52] <GigaE> I simply want to know if a bug was fixed
[14:22:14] <PerterB> hmm, I have snv_43 BFU'd to opensol-20060904 ... can't remember if that drop was before or after b48
[14:23:00] <GigaE> PerterB, thnks
[14:23:07] <GigaE> Anyone else ?
[14:24:33] <GigaE> PerterB, any idea: as i undersatnd, the ld from /usr/ccs/bin/ld is NOT from a package buf fron the OS kernel ; am I right ? any idea?
[14:24:50] <GigaE> s/buf/but
[14:25:01] *** crib|nb has joined #opensolaris
[14:28:05] *** Tpenta has joined #opensolaris
[14:30:55] *** GmanAFK has quit IRC
[14:30:55] <PerterB> /usr/ccs/bin/ld comes from the SUNWtoo package
[14:30:55] *** MattMan is now known as MattAFC
[14:32:05] *** gm152 has quit IRC
[14:32:26] <GigaE> PerterB, I don't undesrtand something: I am looking at http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6465623
[14:33:58] <GigaE> I see that they say there that in snv_48 it was fixed
[14:33:58] <GigaE>  solaris_nevada(snv_48)
[14:33:58] <GigaE> But I see you are right and it is part of SUNWtoo
[14:34:12] <GigaE> So if I have snv_46 and upgrade to the latest SUNWtoo , it will probably be OK
[14:35:56] <GigaE> I have this SUNWtoo : 11.11,REV=2006.07.12.16.20 , Aug 24 2006 17
[14:41:27] *** echelog has joined #opensolaris
[14:42:58] *** calLNCH has quit IRC
[14:48:03] * jmcp gives up on trying to get macosx to boot reliably on his laptop
[14:49:12] <rodrickbrown> get a macbookpro :-)
[14:49:23] <jmcp> not until the core2-duos come out
[14:50:08] <jmcp> rodrickbrown: I want 64bit
[14:51:53] <rodrickbrown> powerbooks are 64bit
[14:52:55] *** |tsoome| has joined #opensolaris
[14:53:30] <jmcp> yes, but they cannot -as yet-run Solaris
[14:54:03] *** vigos has joined #opensolaris
[14:55:25] *** deather_ is now known as deather
[14:56:10] *** Teknix has quit IRC
[14:57:35] *** |tsoome| has quit IRC
[14:58:40] *** GigaE has quit IRC
[14:58:41] *** tsoome has quit IRC
[14:59:58] *** PosixCompliant has joined #opensolaris
[15:04:08] <PosixCompliant> where is SXCR 50 ? I go to the Distribution Page , http://www.opensolaris.org/os/about/distributions
[15:04:15] *** lloy0076 has quit IRC
[15:04:16] <PosixCompliant> then I slecet solaris express
[15:04:50] <PosixCompliant> I select : Get Solaris Express for
[15:04:51] <PosixCompliant> Noncommercial Use
[15:05:18] <PosixCompliant> and then DVD : but I reach snv_48 download , and not SXCR 50 !!
[15:05:26] <PosixCompliant> where am I wrong ?
[15:07:18] *** karrotx has quit IRC
[15:08:28] <Cyrille> http://www.genunix.org/mirror/index.html
[15:08:58] *** tsoome has joined #opensolaris
[15:10:23] *** kFuQ has joined #opensolaris
[15:10:43] <PosixCompliant> Cyrille, isn't this opensolaris ? I am talking about SXCR !
[15:11:30] <Cyrille> yes, sorry, I misunderstood you.
[15:12:06] <PosixCompliant> When I log into this IRC I see SXCR 50 ! but I cannot find it anywhere ... !
[15:12:13] *** calumb has joined #opensolaris
[15:12:20] *** calumb has quit IRC
[15:12:29] <PosixCompliant> SXCR=Solaris Express Comunity Rlease
[15:18:00] <Cyrille> PosixCompliant: if you cheat and "manually" update the link from the distributions page you can get to it. I.e by using http://javashoplm.sun.com/ECom/docs/Welcome.jsp?StoreId=7&PartDetailId=Sol-Express_b50-x86-SP-G-B&TransactionId=try
[15:21:39] *** rodrickbrown has quit IRC
[15:21:40] *** axisys has quit IRC
[15:23:53] *** jcea has quit IRC
[15:29:22] *** Uraeus has joined #opensolaris
[15:29:54] *** nachox has joined #opensolaris
[15:29:56] <Uraeus> hi, stupid question here. I am building packages using the jds-CBE system. But I can't seem to understand where the finished pkg packages ends up
[15:29:59] <Uraeus> any ideas?
[15:31:33] <icon> any sun ray users around?
[15:31:53] <icon> Uraeus: well if you are feeling particularly desperate, just do a 'find / |grep pkg'
[15:32:05] *** dunc has joined #opensolaris
[15:32:20] <icon> icon: otherwise, id read the docs
[15:32:30] <icon> s/icon/Uraeus
[15:33:09] <Cyrille> the finished packages usually end in /<CBE install>/packages/PKGS (at least they used to at some point
[15:33:28] <Uraeus> Cyrille, ok thanks
[15:33:32] <Cyrille> and what's the Sun Ray question?
[15:33:55] * dclarke awakes
[15:35:17] *** lloy0076 has joined #opensolaris
[15:35:38] * quasi is planning to buy a SunRay, but just needs to convince his supplier that they should rather keep the EUR43.20 powercord
[15:36:04] <jteo> wb dclarke.
[15:36:36] * dclarke stares blankly
[15:36:53] <dclarke> morning ..
[15:37:02] *** crib|nb has quit IRC
[15:37:25] *** crib|nb has joined #opensolaris
[15:38:08] *** svoboda has quit IRC
[15:40:35] <icon> i was curious how they performed compared to x11
[15:40:48] <icon> considering beefing up the opteron box and throwing it in the closet
[15:41:02] <icon> the disk array in there keeps the fans on nonstop and it gets a bit loud
[15:41:05] <jengelh> icon: Worse.
[15:41:12] <icon> jengelh: oh?
[15:41:29] <jengelh> Some sunray1s that lure around here... you can't even play a 29fps movie with it
[15:41:36] <jengelh> GL I have not even tried
[15:41:56] <jengelh> (it's because 29fps requires more than 100MBit ethernet can do)
[15:42:03] <jengelh> 29,25,whatever
[15:42:12] <icon> well i was looking at the 2 or 2fs
[15:42:30] <icon> at 249 its more than a little tempting
[15:42:47] <PosixCompliant> Cyrille, thnks , this works
[15:43:00] <PosixCompliant> is there something for DVD ?
[15:43:11] <dclarke> I have tried and failed at using SunRays with graphics intense software
[15:43:13] <jengelh> well chogging up all bandwidth for video/GL -- your coworkers sharing the ethernet segment won't be happy :D
[15:43:21] <Cyrille> PosixCompliant: I guess same, take the dvd link from the distro page and fix it
[15:43:34] <dclarke> you need each SunRay on its own port on the ethernet switch
[15:43:39] <jengelh> sunray2 is also only 100Mbit
[15:44:06] <dclarke> and software like CAD/CAM stuff will not make users happy at all
[15:44:11] <icon> dclarke: thats no proble m
[15:44:26] <dclarke> okay .. I never tried video or anything like that
[15:44:27] <icon> hrmm
[15:44:46] <icon> what about dealing with sunstudio or other gui heavy stuff?
[15:45:00] <jengelh> gui-heavy is rather low-bandwidth
[15:45:07] <dclarke> thats no problem
[15:45:11] <icon> hmm
[15:45:18] <icon> how responsive is it?
[15:45:19] <dclarke> in fact .. a non-issue
[15:45:19] <jengelh> To put it in words: video > 60mbps, audio = 14 mbps, other work max 3 mbps
[15:45:24] <dclarke> instant
[15:45:27] <icon> on 100mbps, x11 is almost the same as native
[15:45:32] <dclarke> well hold on a sec ..
[15:45:45] <dclarke> how long is the distance from the server to the SunRay ?
[15:45:47] <icon> well this is at home, so its not a large setup
[15:45:53] <icon> same segment
[15:45:57] <dclarke> oh .. okay
[15:46:04] <dclarke> I tested SunRay over internet
[15:46:08] <icon> straight to the same switch
[15:46:12] <dclarke> it was less than responsive
[15:46:17] <icon> i imagine
[15:46:30] <dclarke> works well enough .. but not instant ..
[15:46:35] <icon> im just wanting to make some changes to the opteron machine that will likely break the sound barrier
[15:46:40] <dclarke> more like .. 250ms delay minimum for everything
[15:46:46] <jengelh> sunray is basically a strangely encoded x11 :)
[15:46:51] <icon> ahh
[15:46:57] <jengelh> aka xdmcp for enterprise
[15:46:58] <dclarke> highly optimal X11
[15:47:11] <icon> hrm
[15:47:27] <jengelh> well given that there is not really a free srss anywhere...
[15:47:36] <jengelh> makes it a hard-to-decompose protocol
[15:47:44] <jengelh> s/free/open/
[15:53:50] *** crib|nb has quit IRC
[15:55:38] <icon> do you think a 640x480 video at 29fps would work on a full open 100 line?
[15:56:37] <icon> one other question
[15:56:41] <dclarke> not well
[15:56:46] <dclarke> but it would work
[15:56:51] <dclarke> not at 29fps
[15:56:53] <icon> what are the extra usb ports on there?
[15:56:58] <dclarke> maybe 15fps
[15:57:02] *** calumb has joined #opensolaris
[15:57:09] <dclarke> yes
[15:57:14] <dclarke> those are USB ports
[15:57:19] <dclarke> :-/
[15:57:20] <icon> let me rephrase :)
[15:57:27] <icon> why are they there?
[15:57:32] <icon> does it support locally attached devices?
[15:57:38] * dclarke desperately reaches for coffee
[15:57:39] <icon> (other than keyboard/mouse)
[15:57:44] <dclarke> duh
[15:57:47] <dclarke> yes
[15:57:57] <dclarke> like a CD or DVD burner
[15:58:03] <icon> hm
[15:58:06] <dclarke> or a USB memory stick
[15:58:11] <icon> thats useful
[15:58:31] <icon> last time i used a thin client was in the xterm heyday
[15:58:36] <dclarke> or a USB-serial converter that allows a remote doctor to use a serial based digital instrument to collect data
[15:58:54] <icon> do they come with a keyboard and mouse?
[15:58:56] <dclarke> xterm ?
[15:59:03] <dclarke> I _live_ with Xterm !
[15:59:04] <icon> haha
[15:59:05] <dclarke> love it
[15:59:07] <dclarke> live it
[15:59:08] <dclarke> use it
[15:59:13] <dclarke> Xterm .. the last terminal
[15:59:18] <dclarke> you will ever need !
[15:59:20] <icon> i meant the actual physical boxes :)
[15:59:24] <dclarke> oh
[15:59:33] * dclarke needs more coffee
[15:59:49] *** MattAFC is now known as MattMan
[15:59:54] <icon> *really* tempted to grab a ray
[16:00:03] <dclarke> get one
[16:00:05] <dclarke> use it
[16:00:12] <icon> throw it away :)
[16:00:13] <dclarke> you will love it and never look back
[16:00:16] <icon> you think?
[16:00:23] <icon> do they ship with keyboards and mice?
[16:00:27] <icon> cant really tell from the page
[16:00:32] <dclarke> when you see other people with power eating moron boxes under there desks .. you will giggle at them
[16:00:44] <dclarke> yes .. mouse and mice included
[16:00:48] <icon> type 7?
[16:01:00] <dclarke> or .. get one off eBay and then attach any decent USB keyboard and mouse
[16:01:10] <dclarke> type 6 I think
[16:01:11] <icon> im a bit partial to sun keyboards
[16:01:14] <icon> i miss my old type 5
[16:01:17] <dclarke> me too
[16:01:22] <dclarke> I have a type 5c here now
[16:01:28] <dclarke> and I'm not letting go
[16:01:32] <dclarke> :-)
[16:01:38] <icon> oh nice
[16:01:41] <icon> well
[16:01:55] <icon> this may sound silly, but is the 2fs more performant?
[16:02:12] <jteo> ...
[16:02:22] <dclarke> it looks nicer
[16:02:29] <dclarke> in every measurable way
[16:02:32] <icon> well im curious... it certainly has more hardware
[16:02:51] <icon> 100fx, higher resolutions, 2 dvi, 2 extra usb ports
[16:02:52] <dclarke> the real trick is the back end
[16:02:56] <icon> of course
[16:03:08] <icon> thats where you see most of the difference
[16:03:17] <dclarke> the original SunRay 1 .. which I also have here .. was capable of 100fx
[16:03:18] *** Dar is now known as Dar_HOME
[16:03:30] <dclarke> but the graphics was not as sweet at all
[16:03:38] <dclarke> I have to do something that looks like work now
[16:03:49] <dclarke> so .. I'll wander off to the coffee pot ..
[16:03:56] <dclarke> but .. I'll be around
[16:09:08] <jengelh> icon: 320x240 would be more accurate
[16:10:48] <damienc> http://www.opensolaris.org/jive/thread.jspa?threadID=15785&tstart=0
[16:10:54] <damienc> JDS svn repository is live!!
[16:11:44] <jteo> damienc: :)
[16:12:05] <jteo> damienc: uh..b.o.o. tracks RFEs and bugs for JDS as well?
[16:12:25] *** LordKing has joined #opensolaris
[16:12:32] <damienc> I don't understand the question.
[16:13:01] <damienc> Our internal svn server for JDS spec files is dead. We're actively committing to the external one now.
[16:14:14] *** LordKing has quit IRC
[16:14:25] <jteo> damienc: then where is the bug database? or do we go upstream to GNOME?
[16:14:31] <asyd> opensolaris is supposed to switch to mercurial, isn't it ?
[16:15:24] <oxygene> the ON consolidation
[16:15:37] <damienc> asyd: yeah, ON, not JDS.
[16:15:52] <asyd> ok
[16:15:53] *** LordKing has joined #opensolaris
[16:16:01] <damienc> jteo: we'd prefer if you filed bugs upstream
[16:16:23] <damienc> for spec file bugs then you can submit a patch
[16:16:38] <damienc> See the last item in the desktop-discuss post:
[16:16:43] <damienc> * Write access will be granted after your first approved code
[16:16:44] <damienc> change, as per the Code Review process being deployed:
[16:16:44] <damienc> http://www.opensolaris.org/jive/thread.jspa?threadID=14977
[16:17:21] <jteo> damienc: thanks. :)
[16:17:29] <damienc> np.
[16:18:14] <damienc> I'm already working on 2.17. When GNOME 2.16 goes into Solaris Express then we'll branch the spec files and I'll commit my work.
[16:20:16] <jteo> damienc: basically this repo is for specfiles + patches to make GNOME work on Solaris.
[16:22:17] <damienc> exactly.
[16:22:26] <jteo> damienc: cool.
[16:22:27] *** pjlv has quit IRC
[16:29:03] *** mage2 has joined #opensolaris
[16:30:00] *** mrdeviant has joined #opensolaris
[16:32:11] *** DataStream has quit IRC
[16:32:16] *** halton has left #opensolaris
[16:32:33] <mage2> anyone had enough coffee to answer a question?
[16:32:45] <dclarke> not yet
[16:32:54] <dclarke> but fire away
[16:33:07] <dclarke> and I'll think about it as I trot off to the coffee pot of life
[16:33:21] <mage2> good. alright so i have a odd question
[16:33:36] <dclarke> excellent .. I'm an odd person
[16:33:36] <mage2> I have a solaris box with broadcom drivers
[16:33:40] *** svoboda has joined #opensolaris
[16:33:40] <dclarke> me too
[16:33:58] <mage2> I have updated the drivers to the bcme drivers and configured software to bond to it and all is happy.
[16:34:07] <dclarke> uh huh ....
[16:34:07] *** DataStream has joined #opensolaris
[16:34:23] <DataStream> Hmm Dont know what happend there
[16:34:30] <mage2> now if i install a patch from SUN would it possible overwrite those drivers back to the bge drivers and break my install?
[16:34:43] <mage2> and i mean install a patch in the future.
[16:35:18] <dclarke> patch installs don't break your config
[16:35:50] <dclarke> I have a machine with open source drivers for AMD PCNet network cards because the Sun supplie pcn drivers suck all day
[16:36:04] <dclarke> I apply patches ..
[16:36:08] <dclarke> and I am still fine
[16:36:32] <elektronkind> patches only affect the packages they're patching
[16:36:35] <mage2> I am new to solaris . I just dont want my production server going down.
[16:36:43] <dclarke> $ ifconfig -a
[16:36:44] <dclarke> lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
[16:36:44] <dclarke>         inet 127.0.0.1 netmask ff000000
[16:36:44] <dclarke> ae0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
[16:36:45] <dclarke>         inet 192.168.35.42 netmask ffffff00 broadcast 192.168.35.255
[16:36:56] <dclarke> $ modinfo | grep ae
[16:36:56] <dclarke> 106 fea68ae4   16f4  25   1  ptc (tty pseudo driver control 'ptc')
[16:36:56] <dclarke> 124 f9e26000   b244 203   1  ae (pcnet driver v2.2.0)
[16:36:56] <dclarke> 131 f9dd3624    ae4  24   1  pts (Slave Stream Pseudo Terminal dr)
[16:37:04] <dclarke> see that driver there ?  the ae one ?
[16:37:07] <mage2> yes
[16:37:10] <mage2> pcnet
[16:37:12] <dclarke> its not from Sun
[16:37:15] <dclarke> it works fine
[16:37:28] <dclarke> same goes for hardware from other vendors
[16:37:36] <dclarke> they supply the drivers and when you patch
[16:37:39] <dclarke> nothing breaks
[16:37:49] <DataStream> Out of interest, do all patches come with the backout option, or is it juts theclusters that have that optio, ( i have only patched sol 9, so its been a while)
[16:38:02] <mage2> so it should not overwrite the driver or the interface names right?
[16:38:07] *** PosixCompliant has quit IRC
[16:38:08] <dclarke> I can't give you a long term promise .. but .. I can not recall installing a sun patch that broke my config
[16:38:24] <dclarke> and I have been at this since the 80's
[16:38:33] <dclarke> you're safe man ..
[16:38:47] <DataStream> fAMOUSE LAST WORDS :p
[16:38:50] <dclarke> on that note .. I need to trot off to that coffee pot I mentioned earlier
[16:39:01] <dclarke> DataStream: :-P
[16:39:09] <dclarke> DataStream: play nice
[16:39:10] <DataStream> :)
[16:39:17] <mage2> any scarred!
[16:39:20] <mage2> haha
[16:39:35] <mage2> err ant
[16:39:36] <mage2> bah
[16:39:52] <dclarke> be back shortly
[16:40:02] <DataStream> Bye
[16:40:03] <mage2> ok is there a way to trick solaris or change the nic interface name?
[16:40:17] <PerterB> 108993 :P  (at some revision around -23 they backported the sol9/10 LDAP client which broke the host profiles of all my servers... It _was_ documented in the README but the implications weren't really spelled out)
[16:40:18] <mage2> say i wanted bcme0 to be bge0
[16:40:29] <DataStream> Why do that?
[16:41:36] <jteo> mage2: clearview will support that I believe.
[16:42:08] <elektronkind> mage2: no, the interface name is registered by the driver itself, although the Crossbow/Clearview projects on opensolaris.org has one of its feature being the ability o arbitrarily name a interface.
[16:42:09] *** Symmetria has joined #opensolaris
[16:42:10] <Symmetria> hey all
[16:42:49] <Stric> Anyone have a clue how large Nexenta is for mirroring?
[16:42:59] <Cyrille> about this large: <---->
[16:43:03] <Cyrille> sorry.
[16:43:03] <Symmetria> what is nexenta if I can ask a stupid question :p
[16:43:11] <Stric> Symmetria: An OpenSolaris dist
[16:43:14] <DataStream> a opensolris dist
[16:43:15] <Symmetria> since if its available for mirroring I need to mirror it :p
[16:43:16] <DataStream> kinda
[16:43:18] <Cyrille> GNU flavoured distribution of OpenSolaris
[16:43:25] <trygvis> it's a opensolaris-based distro with ubuntu coating
[16:43:26] <Symmetria> heh, will add it to my list of mirrors, where do I get it :p
[16:43:30] <DataStream> a solaris kernel, with ubunto userland
[16:43:37] * icon grumbles
[16:43:53] * Symmetria laughs, Im busy moving my mirrors onto a solaris platform at the moment
[16:43:53] <DataStream> Stric: no, never used ot for raid'ing
[16:44:13] <Symmetria> moving everything from one san to another over gigE to move away from the linux file systems
[16:44:15] <icon> although i do find it quite ironic... after all, Gnu's not Unix
[16:44:16] <Symmetria> gonna take a LONG time
[16:44:22] <Symmetria> apool                  3.8T    55G   3.8T     2%    /mirror/iscsi-san-1
[16:44:33] <Symmetria> heh 55G of 2.8 terabytes moved, and thats just SAN-1 :(
[16:44:45] <mage2> now thats fun
[16:44:47] <mage2> :)
[16:44:48] <DataStream> elektronkind: and it looks like your gonna have to find out the orifinal name for the card, i VERY much dought it iwll let you use another one
[16:44:49] <Stric> DataStream: kinda not what I meant :P .. becoming a download mirror..
[16:44:57] <Symmetria> heh mage what is fun?
[16:45:05] <Cyrille> icon: I guess it's part of the long tradition of "put linux everywhere", such as an xbox, a toaster, or even OpenSolaris ;-)
[16:45:14] <mage2> 55g of 2.8 terra
[16:45:24] <Symmetria> apool                  3.8T    58G   3.8T     2%    /mirror/iscsi-san-1
[16:45:26] <Symmetria> :p 58 now
[16:45:29] <Symmetria> <3 gigE
[16:45:46] <Symmetria> heh I still gotta move san's 2 and 3 as well, and between them thats another 5 terabytes
[16:45:49] <mage2> its moving pretty quick
[16:45:50] <Symmetria> but one at a time :)
[16:46:17] <mage2> has anyone worked with this "clearview" thing before?
[16:46:19] <Symmetria> well, solaris is handling the san's a HELL of a lot better than linux
[16:46:29] <DataStream> Stric: tell me about it, last saunday i had toe back up some this it dvd,m it too all fscking day, and i  only got 12 done ://///////
[16:47:01] <DataStream> Stric: tell me about it, last saunday i had toe back up some  dvd,s it too all fscking day, and i  only got 12 done ://///////
[16:47:07] <mrdeviant> Symmetria, by SAN do you mean FC or iscsi ?
[16:47:17] <DataStream> (Hows its shout look)
[16:47:19] <DataStream>  :P
[16:47:24] <Symmetria> mrdeviant iSCSI SAN on a QLogic 4050C HBA
[16:47:39] *** stevel has joined #opensolaris
[16:47:40] *** ChanServ sets mode: +o stevel
[16:47:45] <Symmetria> well, 3 x iSCSI SAN's on 3 x 4050C HBA's
[16:47:58] <DataStream> Whos stevel
[16:48:01] <Symmetria> I was running software iSCSI but it chews CPU at high speeds
[16:48:09] * stevel is stevel
[16:48:10] <DataStream> karouac?
[16:48:59] <icon> Steve Lau
[16:50:58] <jteo> stevel: mercurial mirror. wooot.
[16:51:07] <stevel> jteo: :-)
[16:51:52] <jteo> though i'm still waiting for hg clone to finish.
[16:51:54] <jteo> ;)
[16:52:36] <Symmetria> if someone can give me a list of opensource solaris stuff I can mirror, or mirrorable content
[16:52:40] <Symmetria> it would be appreciated
[16:52:55] <Symmetria> :) I run the largest mirror server on the african continent, and it sux that I've got all the linux stuff and all the bsd stuff but almost no solaris stuff
[16:53:13] <mrdeviant> do you have blastwave and sun freeware?
[16:53:26] <Symmetria> I've got blastwave now
[16:53:29] <Symmetria> not sun freewhere
[16:53:31] <Symmetria> freeware I mean
[16:53:32] <Symmetria> :p
[16:53:39] <Symmetria> blastwave was relatively small, only about 16gigs
[16:53:42] *** Mazon is now known as mazon
[16:54:51] <Symmetria> mrdeviant where do I grab sun freeware?
[16:55:00] <tsoome> Symmetria: and this linux stuff is same software packaged over and over and over for almost all thinkable linux distributions? ;)
[16:55:08] <coffman> www.sunfreeware.com
[16:55:09] <mrdeviant> http://www.sunfreeware.com/
[16:55:38] <Symmetria> tsoome heh *shrug* there is a lot of duplication yes, linux I carry gentoo, ubuntu, debian, slackware, centos, knoppix I think
[16:56:02] <Symmetria> tsoome but thats the nature of running a large download mirror, gotta have everything as the systems will expect it for updates etc and its not like Im short on disk space
[16:56:07] <mrdeviant> Symmetria, what about fink and darwinports for os x folk
[16:56:20] <Symmetria> mrdeviant ohh where do I grab that stuff :)
[16:56:41] * Symmetria laughs, I have 7 TB free
[16:57:45] <mrdeviant> http://fink.sourceforge.net/ and http://www.macports.org/
[16:58:37] <Symmetria> hrm, if fink is on sourceforge I already have it, cause while IM not officially a sourceforge mirror, I mirror downloads.sourceforge.net once every 2 days (will do a daily sync once sourceforge wakes up and makes me an official mirror, but getting holda anyone at sf is virtually impossible)
[16:59:27] <tsoome> yeah, keeping up good mirrors is not that easy at all...
[16:59:32] *** LordKing has quit IRC
[16:59:32] <Symmetria> holy crap zfs is a shitload faster than anything I've used under linux
[17:00:16] <Symmetria> tsoome heh, we do it because its cost effective, the price we pay for international bandwidth in .za is so insanely sick that the amount of bandwidth I save in peak hours by having that mirror pays for the mirror hardware many times over
[17:00:27] <Symmetria> (our bandwidth runs dead at night since all the universities bugger off home)
[17:00:48] <mrdeviant> Symmetria, do you have CPAN mirrored
[17:00:57] *** bengtf_ has joined #opensolaris
[17:01:39] <Symmetria> mrdeviant its on my list for as soon as the system is migrated properly
[17:01:40] <tsoome> Symmetria: I can understand that, I have worked in univ myself and mastered mirrors.....
[17:01:55] <Symmetria> tsoome *nod* I run an NREN
[17:02:16] <Symmetria> (I'm CTO of TENET, the south african national research and education network, heh, got 90 universities behidn my backbone)
[17:05:03] *** lloy0076 has quit IRC
[17:07:24] *** gnu2it2 has joined #opensolaris
[17:07:30] *** schily_ has joined #opensolaris
[17:08:05] <coffman> Symmetria: do u proxy that?
[17:11:05] <stevel> so has anyone successfully pulled/cloned the onnv-gate repository yet?
[17:12:10] <jteo> stevel: i'm stilling pulling/cloning.
[17:12:15] <jteo> *still.
[17:12:34] <movement> stevel: yep
[17:13:04] <stevel> jteo: how long has it taken?
[17:13:19] <stevel> movement: how long did it take you? (and were you pulling onto SWAN via the proxies? or onto an external machine)
[17:13:47] <movement> externally, sorry, I forgot to time it
[17:14:11] <stevel> no worries. as long as it wasn't heinously long
[17:14:47] <jteo> stevel: been pulling for the past 30 mins.
[17:14:51] <jteo> stevel: :(
[17:15:02] <stevel> jteo: where are you (geographically)?
[17:15:10] <jteo> stevel: .sg
[17:15:15] <stevel> ah
[17:15:27] <stevel> we should setup some international mirrours
[17:15:53] *** bengtf__ has joined #opensolaris
[17:16:09] *** bengtf has quit IRC
[17:16:14] *** bougie has joined #opensolaris
[17:16:17] *** bengtf__ is now known as bengtf
[17:16:32] <bougie> hello
[17:17:15] <janus_man> Does anyone know of a way of turning on some level of kmem debugging? I think we have some code either doing a double kmem_free() or a kmem_free() with a bad address?
[17:17:21] <janus_man> Solaris 10 btw
[17:18:59] *** tsoome has quit IRC
[17:20:15] *** trs81 has joined #opensolaris
[17:21:13] <quasi> janus_man: dtrace would be an obvious answer
[17:24:07] *** Symmetria has quit IRC
[17:24:15] *** Symmetria has joined #opensolaris
[17:24:39] *** schily has quit IRC
[17:24:49] <mrdeviant> janus_man, http://docs.sun.com/app/docs/doc/816-5041/6mb7ae3l7?a=view
[17:26:24] *** Disorganized has joined #OpenSolaris
[17:27:14] <janus_man> thanks
[17:27:22] <Disorganized> hi all
[17:28:52] *** eboutilier has joined #opensolaris
[17:29:16] * elektronkind wonders if there will be a SUNWhg package
[17:29:33] *** bengtf_ has quit IRC
[17:29:35] <jengelh> I wonder why it could not just be called hg
[17:30:24] <movement> elektronkind: there already is SUNWmercurial
[17:31:36] *** DataStream has quit IRC
[17:32:17] * stevel needs a volunteer to help him debug something
[17:32:25] <stevel> can someone affiliate themselves to the onnv project?
[17:32:31] <stevel> http://opensolaris.org/os/project/onnv
[17:32:36] <dclarke> what ?
[17:32:41] <dclarke> sure ..
[17:32:44] <dclarke> one sec
[17:32:47] <mrdeviant> stevel, ok. done
[17:33:00] <dclarke> cool .. I don't have to then :-)
[17:33:43] <mrdeviant> btw, the hg clone took me about 15 min.
[17:33:45] <stevel> mrdeviant: okay cool. thanks, you've just helped me verify a somewhat annoying bug in the affiliation code for the website
[17:34:34] *** slowhog has joined #opensolaris
[17:36:07] *** jcea has joined #opensolaris
[17:36:24] <jteo> stevel: my hg clone completed. about 1 hr.
[17:36:34] <stevel> yikes.
[17:36:43] <stevel> well keep it around so you don't have to do full clones every time :)
[17:36:49] <movement> hrm.
[17:38:53] <quasi> anyone got an idea why a thumper isn't accepting input on the serial interface? I saw lots of boot info and ended with: SUNSPXXXXXXXXXX login:
[17:43:55] *** tsoome has joined #opensolaris
[17:45:21] <oxygene> waht kind of history does the repository contain, only the releases or every checkin since 2005-06-14?
[17:45:44] *** tsoome has quit IRC
[17:45:53] <stevel> every checkin since 2005-06-14
[17:46:04] *** sommerfeld has joined #opensolaris
[17:46:04] *** ChanServ sets mode: +o sommerfeld
[17:46:42] <oxygene> nice
[17:46:43] *** tsoome has joined #opensolaris
[17:47:41] <quasi> gah, where is the x4500 hiding on sunsolve? I can't see it in the handbook
[17:49:03] <g4lt-mordant> in the "you haven't paid nearly enough to us to see the manuals" section
[17:50:19] <quasi> nope, I'm already looking at the registered content
[17:50:59] *** loke has joined #opensolaris
[17:51:49] <sommerfeld> quasi: X4500 hasn't been added to the internal Sun System Handbook either.
[17:52:13] *** Uraeus has quit IRC
[17:54:11] <quasi> sommerfeld: ok, thanks - I suppose I'll just have to do the honorable thing and start guessing ;)
[17:58:33] *** calumb is now known as calMTG
[18:00:24] *** dduvall has quit IRC
[18:03:59] *** solarisjon has joined #opensolaris
[18:03:59] *** loke_ has joined #opensolaris
[18:05:08] *** loke has quit IRC
[18:06:46] *** gisburn has quit IRC
[18:07:35] *** dduvall has joined #opensolaris
[18:10:32] *** mikaeld_ has joined #opensolaris
[18:14:40] * steleman hopes an AMD64 guru is around
[18:16:08] <dclarke> hello
[18:16:13] <dclarke> not a guru
[18:16:27] <steleman> ld.so.1: xmlwf: fatal: relocation error: R_AMD64_PC32: file /usr/gnu/lib/amd64/libexpat.so.1: symbol errno: value 0x280012fb459 does not fit
[18:16:44] <dclarke> ick
[18:16:47] <dclarke> gack
[18:16:55] <Cyrille> you forgot urk.
[18:17:06] <dclarke> I was getting there
[18:17:09] <Cyrille> ah good.
[18:17:14] <dclarke> urk
[18:17:14] *** andrei has joined #opensolaris
[18:17:30] <stevel> 'morning andrei
[18:17:33] <steleman> and the reason for this (i suspect) is: the  relocation for errno is defined as R_AMD64_COPY                0x41b260          0  .SUNW_reloc    errno
[18:17:46] <andrei> stevel: morning... congrats on mercurial
[18:17:51] <stevel> thanks :)
[18:18:01] <dclarke> I can only assume that you are using Studio 11   ?
[18:18:04] <andrei> stevel: I haven't tried it myself yet though
[18:18:05] <steleman> but the library defines it as R_AMD64_PC32                  0xcacf 0xfffffffffffffff8  .SUNW_reloc
[18:18:27] <steleman> yes it's studio 11 and i'm using -xmodel=medium and all the other incantations
[18:18:30] <dclarke> I'll need coffee for this
[18:18:41] <dclarke> brb
[18:18:46] <andrei> R.I.P. vold?
[18:18:47] <dclarke> I'm interested however
[18:18:54] <dclarke> brb ..
[18:18:55] <stevel> andrei: yup. long live tamarack
[18:19:03] * dclarke trots off in search of coffee
[18:19:10] <andrei> did that putback include removing the -V option from pkill?
[18:19:37] * stevel didn't know pkill had a -V option
[18:19:52] <andrei> ha! it's undocumented :-)
[18:20:06] <andrei> it's a shortcut for "pkill vold" :-)
[18:20:13] <stevel> seriously?
[18:20:20] <andrei> go check the source :-)
[18:20:59] *** besonen_mobile__ has quit IRC
[18:21:00] <steleman> hi stevel great job with mercurial :_)
[18:21:17] <stevel> thanks steleman
[18:21:44] <stevel> andrei: where's the source for pkill?
[18:21:52] <stevel> i would guess usr/src/cmd/pkill
[18:21:53] * steleman writes aliases to make hg look like cvs :-P
[18:21:57] <nightswim> isnt it called pgrep
[18:22:33] <ProfMikey> http://www.bash.org/?700015
[18:23:21] <nightswim> p* are the same binary
[18:23:22] <stevel> ROFL
[18:23:25] <stevel> [stevel@donuthole:onnv_external] 551$ hg diff -r 2947 -r 2948 | grep PKILL_OPTS
[18:23:25] <stevel> +static const char PKILL_OPTS[] = "fnovxc:D:u:U:G:P:g:s:t:z:J:T:";
[18:23:26] <andrei> stevel: yep.. artem remove it :-)
[18:23:46] <stevel> that's hilarious
[18:23:47] <andrei> stevel: usr/src/cmd/pgrep/pgrep.c look at the last delta
[18:23:59] *** Cyrille has quit IRC
[18:24:10] <andrei> stevel: it doesn't do anything now
[18:24:51] *** eboutilier has quit IRC
[18:25:01] <andrei> stevel: are there any plans to backport tamarack to S10?
[18:25:28] <stevel> andrei: i'm not sure
[18:25:41] <stevel> i don't see why not. we're backporting everything else and the kitchen sink
[18:25:51] *** andersmo has quit IRC
[18:25:57] <oxygene> mostly the parts that were promised for s10, I think
[18:26:21] <andrei> that would be great... there are a couple of vold-related bugs against vmware tools for solaris
[18:27:45] <loke_> wait a second... are you seriously _removing_ options from a command?
[18:28:14] <andrei> no, the option is still there
[18:28:22] <andrei> relax :-)
[18:28:59] *** andersmo has joined #opensolaris
[18:29:06] <loke_> andrei: haha. :-)  I'm just qurious, because a few years ago it was a long, LONG uphipp struggle to make any change that could even coniveably create any form of regression at a customer site
[18:29:39] *** loke_ is now known as loke
[18:30:28] <mrdeviant> of course, wasn't it an undocumented option ?
[18:30:44] <loke> mrdeviant: hmm... good point
[18:30:59] <loke> by the way, what's the latest justification of /usr/bin/cd ?
[18:31:20] <alanc> POSIX stupidity
[18:31:39] <loke> alanc: finally, a real answer :-)  So how does POSIX justify it?
[18:31:42] <stevel> we're undecided about replacing it with scd (schily-cd) or scd93
[18:31:45] <stevel> cd93 rather
[18:32:03] <oxygene> how about gnome-cd?
[18:32:15] <loke> haha
[18:32:17] <icon> posix has to justify??
[18:32:19] <andrei> java-enterprise-cd
[18:32:20] <oxygene> not to forget sun java change directory
[18:32:21] <icon> thats new :D
[18:32:41] <stevel> oxygen: it'll use a bunch of zenity dialogs ;-)
[18:32:48] <alanc> I don't remember exactly - I think it's so you can run "/usr/bin/cd somedir" and check the exit code to see if the directory exists and you have permissions to change to it
[18:32:55] <icon> yup
[18:32:58] <oxygene> stevel: hint: gnome-cd exists already ;)
[18:33:00] <icon> it doesnt mess with cwd
[18:33:11] <alanc> it's been explained on comp.unix.solaris a couple of times, should be able to find it on groups.google.com
[18:33:27] <icon> http://bama.ua.edu/cgi-bin/man-cgi?cd
[18:34:42] <loke> there is one more of those crazynesses I'd like to hear a justification for (I tried to induce a change back many years ago, but failed just as much as with the /usr/bin/cd thing)... What is the justification for "cp -i" to default to yes if stdin is not a terminal? (it started when a friend of mine wanted to copy missing files to a directory containing newer versions and ended up losing a lot of files when he typed: yes n | cp -i /foo/*
[18:34:43] <loke> /bar
[18:35:35] <icon> its usually in the name of backwards compatibility
[18:35:57] <loke> there is specific code to check for stdin being a tty which disabled the -i option, and I just can't understand why they did it
[18:36:06] <icon> oft times it seems (at least to me) posix tends to operate off of precedence
[18:37:31] <oxygene> "it sucks, but at least it sucks everywhere" - isn't that the credo of most standard bodies?
[18:39:01] <icon> haha
[18:41:16] <loke> oxygene: good point
[18:41:32] <loke> oxygene: but is the -i on non-tty-devices behaviour mandated by POSIX?
[18:41:56] *** damienc has quit IRC
[18:42:33] <oxygene> no idea.. honestly, my intuition would propose an exit with error in this case.. "interactive" with non-interactive shell feels so.. "undefined"
[18:44:03] <loke> oxygene: true, and worse, the more dangerous behaviour (and completely contrary to the -i option itself) is introduuced when on in a terminal
[18:44:11] <loke> it's so counter-intuitive
[18:44:31] <loke> you'd expect the option to do what the manual says it should do
[18:44:43] <loke> (I don't believe it was documented back in 2.5.1)
[18:46:53] <loke> anyway, bedtime
[18:46:56] *** loke has quit IRC
[18:48:01] *** nwf has joined #opensolaris
[18:55:07] <AbeFroman> i just found a bug in xorg... i have a type 7 usb keyboard, and if numlock is on, ctrl+backspace will kill it
[18:55:19] <timeless> that's well known
[18:55:24] <AbeFroman> ok
[18:55:31] <timeless> the recommendation is to disable the feature entirely
[18:55:36] <timeless> it drives me nuts, but ...
[19:01:27] <Symmetria> wheee
[19:01:30] <Symmetria> 190gigs transferred
[19:01:38] <Symmetria> and my gawd I love zfs and solaris's file system
[19:01:59] <Symmetria> Im pushing data to that file system at like, 200+mbit, millions upon millions of tiny files
[19:02:11] <Symmetria> and it still performs like Im not doing anything at all when reading off it
[19:02:17] <Symmetria> because of its caching etc
[19:03:06] <sommerfeld> the i/o deadline scheduling is helping there.
[19:03:13] <sommerfeld> prioritizes reads over writes
[19:03:57] <Symmetria> heh well I can definately say that it WAY outperforms linux or fbsd doing this type of thing
[19:04:28] *** mjf has joined #opensolaris
[19:04:53] *** alanc is now known as alanc_away
[19:05:06] <mjf> hey guys, what's the purpose of the map layer in UFS? (e.g common/fs/ufs/lufs_map.c)
[19:05:17] *** bondolo has joined #opensolaris
[19:05:58] <sommerfeld> the lufs_* files generally relate to UFS logging
[19:06:21] <mjf> yeah, i meant waht purpose do they have for logging support
[19:07:27] <timeless> mjf: hanging my system for hours at boot :)
[19:07:40] *** andrei has left #opensolaris
[19:07:44] *** Trident has joined #opensolaris
[19:07:47] <mjf> timeless: haha, doesnt that defeat the purpose of a journalled FS? ;)
[19:08:05] <mjf> well, one of
[19:08:09] <timeless> mjf: well, if it finished doing its job, i'm sure it'd be ok
[19:08:29] <timeless> i just don't have however many millenia it was going to take to recover from one stupid user error on my part :)
[19:08:44] <mjf> hmm?
[19:08:53] <mjf> what user error?
[19:08:56] <timeless> oh, i managed to get it into a very specifically bad spot
[19:09:08] <timeless> where it decided to spin arond calculating a certain sizing
[19:09:25] *** bougie_ has joined #opensolaris
[19:09:44] <timeless> for reference: you don't want to try to make an 800g file in / as root, when / is a 4g fs that has ~3.1g used
[19:09:58] <mjf> ouch :P
[19:10:17] <timeless> but hey, i learned kernel debugging :)
[19:11:58] <mjf> :D
[19:15:59] *** bougie has quit IRC
[19:18:34] *** mazon is now known as Mazon
[19:19:49] *** Disorganized has quit IRC
[19:19:54] *** Disorganized has joined #OpenSolaris
[19:21:20] *** MattMan has quit IRC
[19:24:27] <PerterB> bugger... I have a server in a state where Sun Cluster still has a record of an SVM set in its device groups, but due to idiocy the set no longer exists...
[19:24:52] <PerterB> So now I can't create a new set with the same name, nor find any way to delete the existing set from sun cluster
[19:28:52] *** yjsong has joined #opensolaris
[19:29:22] *** jacotton has quit IRC
[19:32:17] *** dunc has quit IRC
[19:32:49] *** alanc-away is now known as alanc
[19:35:42] *** jerome__ has left #opensolaris
[19:35:58] *** salamanders has quit IRC
[19:37:58] *** deather_ has joined #opensolaris
[19:42:52] *** deather__ has joined #opensolaris
[19:51:54] *** mjf has left #opensolaris
[19:53:12] *** bengtf has quit IRC
[19:54:11] *** deather has quit IRC
[19:54:29] *** slowhog has left #opensolaris
[19:54:35] *** deather_ has quit IRC
[19:54:51] *** sparc-kly|WORK has quit IRC
[19:55:02] *** neoxed has quit IRC
[20:01:48] *** calMTG has quit IRC
[20:01:53] *** neoxed has joined #OpenSolaris
[20:02:35] *** mage2 has quit IRC
[20:03:42] *** mkhl has joined #opensolaris
[20:06:04] *** besonen_mobile has joined #opensolaris
[20:10:24] <gdamore> hi *
[20:12:05] *** eboutilier has joined #opensolaris
[20:12:16] <jbalint> hey
[20:15:25] *** Teknix has joined #opensolaris
[20:19:12] <timeless> swift% prstat
[20:19:12] <timeless> zsh: fork failed: not enough space
[20:19:15] <timeless> ouch?
[20:19:32] <alanc> move the knives and spoons then to make more room
[20:19:50] <gdamore> exec prstat
[20:19:51] <timeless> "System Monitor" has quit unexpectedly
[20:19:59] <alanc> (sorry, too close to lunchtime - brain already out to lunch)
[20:20:09] <gdamore> mmm.... lunch.  hungry....
[20:20:43] <timeless> what's cheaper than prstat?
[20:20:48] <timeless> i seem to be out of swap
[20:20:55] <delewis> increase swap space :-)
[20:21:03] <tsoome> add some memory;)
[20:21:10] <gdamore> exec it. :-)  or better yet, exec swapadd ... :-)
[20:21:10] <timeless> i'd like to know how much i'm using first
[20:21:22] <timeless> what would exec do?
[20:21:33] <timeless> ok, that killed the shell
[20:21:34] <gdamore> exec trades your shell for the program.
[20:21:39] <timeless> i don't think tha'ts a good trade
[20:21:39] <delewis> replaces whatever you're running with the image of prstat
[20:21:50] <delewis> timeless: sure it is, if you're really, really wanting to use prstat
[20:21:58] <timeless> but it didn't work!
[20:22:05] <timeless> i lost the entire tab
[20:22:08] <gdamore> it probably did, and then the window closed. :-)
[20:22:30] <gdamore> exec pkill firefox is always a good choice.
[20:22:32] <tsoome> svcadm disable -t sendmail or some other disposable service;)
[20:22:52] <gdamore> you will have a hard time running svcadm unless you exec, if you're out of swap
[20:22:57] *** yippi has joined #opensolaris
[20:23:45] <timeless> i think i just lost my window manager
[20:23:51] <timeless> at least, i have no window decorations
[20:24:05] * gdamore curses the idjits that write web-based training software that _requires_ internet exploder.
[20:24:08] <delewis> yes, Solaris gets very angry when there's nothing to allocate :-)
[20:24:09] <tsoome> kill X and use serial terminal;)
[20:24:28] <delewis> of course, something nastier like your process table could be full
[20:24:37] <timeless> nah
[20:24:42] <timeless> i had a cross reference running
[20:24:47] * gdamore curses the idjits at the company that write policies that require i take said web training course, regardless of the fact that I don't have a windows PC.
[20:24:51] <timeless> it only had to chew through 600mb of data
[20:25:01] <delewis> gdamore: classic. :-)
[20:25:18] <delewis> and I guess you don't have an x86/x86_64 system laying around, anyway?
[20:25:33] <gdamore> oh, i have those.  just no windows install media.
[20:25:38] <timeless> wine
[20:25:42] <delewis> gdamore: time for IE for SPARC :-)
[20:25:47] <delewis> Solaris/SPARC, that is.
[20:25:49] <gdamore> ewww.... i've been there.
[20:26:02] <gdamore> no fun.  the website requires active X.
[20:26:17] <alanc> IE for SPARC was painful the first time you started it and it took 15 minutes to scan all your fonts
[20:26:35] <timeless> it worked
[20:26:41] <gdamore> actually, they've shipped me a disk, loaded with the company's "standard windows install", that i'm supposed to boot up in my talin.  what a PITA.
[20:26:49] <delewis> of course, you were emulating TCP/IP via winsock
[20:26:55] <delewis> (IIRC)
[20:26:59] <gdamore> (talin = our 15" laptop loaded with JDS 2.)
[20:27:23] <delewis> yeah, the Talin is the x86 portable isn't it?
[20:27:30] *** dunc has joined #opensolaris
[20:27:52] <delewis> alright, dumb question. :-)
[20:27:56] <timeless> ok, gdamore: since you're familiar w/ gds
[20:28:06] <timeless> how do i kill individual windows when they have no borders?
[20:28:13] <timeless> and i can't seem to focus anything
[20:28:14] <gdamore> xkill.
[20:28:35] <delewis> ugh, creating the boot environment is taking forever.
[20:28:37] <gdamore> you could always exec twm or something. :-)
[20:28:44] <timeless> i can't focus a terminal!
[20:29:04] <gdamore> what app has focus?
[20:29:15] <timeless> not really sure
[20:29:17] <gdamore> and can you come in over the network?
[20:29:18] <delewis> probably some Java app :-)
[20:29:23] <timeless> hrm
[20:29:47] <timeless> i can't kill the java app
[20:29:50] <timeless> it's running as root
[20:29:54] <delewis> I was being sarcastic :-)
[20:30:21] <timeless> hrm
[20:30:27] <timeless> INIT: Couldn't write p
[20:30:31] <timeless> that's probably not so good
[20:30:36] <gdamore> i once wrote for qualcomm a program called "erbd" (emergency reboot daemon).  It listens on a named port waiting for the root password, and then does a uadmin() to reboot the system.  It does this without forking or allocating any memory (though it did preallocate some at startup.)
[20:30:52] <gdamore> it was nice for recovering from scenarios like this.  you could "cleanly" restart your system remotely.
[20:30:58] *** Seawolf99 has joined #opensolaris
[20:30:59] *** besonen_mobile has quit IRC
[20:31:33] <timeless> ok
[20:31:51] <timeless> INIT: Couldn't write persistent state file '/etc/svc/volatile/init.state'.
[20:31:52] <gdamore> i used it to reset FTP servers when they got horked at a remote ISP.  better than hitting the power switch.
[20:31:55] * timeless wonders what that means
[20:32:13] <gdamore> that's bad.  not enough memory for tmpfs?
[20:32:20] <timeless> ok
[20:32:28] <timeless> glimpseindex 2.5GiB resident
[20:32:41] <timeless> i'm hoping that was the nexenta zone and not my real init
[20:32:47] *** axisys has joined #opensolaris
[20:32:57] <timeless> is there a semi safe way to shut down my zone?
[20:33:22] *** nachox has quit IRC
[20:33:24] <axisys> timeless: zlogin zone init 5
[20:33:25] <timeless> the only thing i have that's talking to me is something like task manager
[20:33:32] <timeless> i still don't have a console!
[20:33:54] <axisys> zlogin -C zone
[20:34:11] <gdamore> well, if you have a task manager, start killing off windows.
[20:34:17] * timeless kills evolution
[20:34:37] * gdamore expects timeless to log off any minute now as he kills his irc client.
[20:34:40] <timeless> wah, i can't kill sendmail, i'm not root
[20:34:48] <timeless> irc client is running on a remote computer
[20:35:02] <timeless> it's ssh/screen/epic, so i switched to my laptop to talk :)
[20:35:10] <gdamore> clean up any files you have in /tmp
[20:36:26] *** Gman has joined #opensolaris
[20:36:42] <gdamore> at this point, as soon as you can get a shell running, or run a command, i'd just uadmin it off.
[20:37:24] <gdamore> uadmin 1 1 would be a good idea.
[20:37:56] * timeless sighs
[20:38:38] <timeless> oh wow
[20:38:46] <timeless> that 2.5gb process went away
[20:39:11] * timeless logged out and ran to console w/o x11
[20:39:14] <gdamore> it probably couldn't malloc. :-)
[20:39:36] <timeless> ok, how do i check how much swap i have/had?
[20:40:12] <gdamore> swap -a
[20:40:24] <gdamore> swap -l i mean
[20:41:00] *** Yamaraj has joined #opensolaris
[20:42:35] * timeless sighs
[20:42:36] *** alanc is now known as alanc-away
[20:42:47] * timeless goes to reread the man pages about swap
[20:43:40] *** Disorganized has quit IRC
[20:44:00] *** stevel has quit IRC
[20:46:37] <axisys> svcadm enable ipfilter failed to start ipfilter.. saying pfil not plumbed on any network interfaces
[20:47:19] <tsoome> svcadm enable pfil ?
[20:47:45] *** slowhog has joined #opensolaris
[20:47:53] <timeless> is there a safe way for me to verify that my /etc/vfstab is correct? :)
[20:47:58] <axisys> tsoome: it is .. reran it
[20:48:06] <tsoome> or add pfil with ifconfig or just reboot....
[20:48:16] <Seawolf99> NEED HELP. I am trying to update snv_46 to BFU-archive-20061016 and when I run the bfu command it aborts saying "BIND 8 has been removed from ON; BIND 9 is available from SFW. Run /ws/onnv-gate/public/bin/migrate_bind9 to migrate to BIND 9". Can some tell me where can I find this script I searched almost everywhere that I could think of and even googled it but it only showed scripts that has reference to this script
[20:49:11] *** tsoome has quit IRC
[20:49:36] <axisys> tsoome: rebooting
[20:49:45] <astinus> Seawolf99:  find / -name 'migrate_bind9' -print
[20:49:58] <Seawolf99> thanks but already did that
[20:50:36] <twincest> /ws/onnv-gate is a sun internal path, you shouldn't see that
[20:50:38] <Seawolf99> I have been yelling since yesterday but you are the first person who atleast said something. Thanks
[20:50:56] <twincest> i'd suggest mailing one of the lists
[20:51:07] <timeless> seawolf99: i was under the impression that richlowe started poking someone in response to your inquiry yesterday
[20:51:10] <axisys> i have to do this to make ipfilter working on solaris 10
[20:51:12] <axisys> http://learningsolaris.com/archives/2005/05/19/ipfilter-on-solaris-10-primer/
[20:52:13] <timeless> ok, other than sendmail, what other fun things should i disable? :)
[20:52:42] <astinus> timeless:  I'd say anything you aren't planning on using.
[20:52:50] * timeless wonders how to disable that java updater service
[20:53:01] <timeless> hrm
[20:53:09] <timeless> doesn't dbus require x11 to be useful?
[20:53:15] <astinus> lord no
[20:53:57] <astinus> (most stuff using dbus = X11 though)
[20:55:37] * timeless wonders what network/rpc/cde-calendar-manager is
[20:55:58] <astinus> one would guess, something to do with calendars in CDE :p
[20:56:17] * timeless waves goodbye to it
[20:56:39] <astinus> feel free to yell, scream and shout at me - I'm a newb  .. converting to Solaris after 5-6 years in Linux
[20:56:48] * astinus grins
[20:58:03] <Seawolf99> timeless: I didn't see anything from richlowe yesterday. If I missed it then I am sorry
[20:58:16] <timeless> he didn't address you directly
[20:58:31] <timeless> it just seemed like he was responding to your problem by trying to deal w/ the core issue
[20:58:35] <timeless> i could be wrong
[20:58:52] <Seawolf99> I have been using linux since last 7 yrs, trying to learn solaris to better evaluate what's good for what
[20:58:52] *** betzi has joined #opensolaris
[20:58:56] <timeless> personally, i'd just disable or uninstall bind8
[20:59:01] *** Kernel86|Laptop has quit IRC
[20:59:02] <timeless> after backing up the bind8 settings
[20:59:13] <timeless> and then i'd pray that the updater doesn't complain
[20:59:26] <timeless> well, that's if i were feeling enduserish
[20:59:40] <timeless> if i was feeling hackerish, i'd go read the migration script and probably hack it instead
[20:59:43] <betzi> hlo. I tried to install solaris express on an external usb hdd. Install went fine, but on boot it stops: saying
[20:59:54] <betzi> scsi transport failed reason timeout
[21:00:14] * timeless wonders if grub is ok w/ betzi's config
[21:00:14] <betzi> and vn_rdwr failed with error code 0X503
[21:00:33] <axisys> /etc/ipf/pfil.ap I do not see nge driver here for x2100
[21:00:39] <betzi> well grub is OK, it loads the kernel etc
[21:00:53] <axisys> i need to uncomment it for ipfilter to work.. but i do not see it here
[21:01:30] <timeless> nge not bge?
[21:01:46] <timeless> presumablye you can just add nge -1 0 pfil
[21:01:48] <betzi> and another msg: something like kubj_load_module mc_amd read failed or smthg
[21:01:53] <timeless> the file is just a convenience
[21:01:56] <timeless> at least from a quick read
[21:02:16] <timeless> note that the only thing i've managed to do w/ ipf was panic my system by pinging the global zone from the local zone
[21:02:19] <timeless> but :)
[21:02:36] <timeless> besides, that's fixed in ON already :|
[21:03:23] <timeless> you could probably file a bug asking them to add nge to the template, heck, you could include a patch and try to become a contributor ;-)
[21:03:43] <axisys> timeless: nge yes
[21:03:54] <timeless> anyway, the file's header clearly says it's just a format
[21:04:10] <timeless> it seems they listed the common drivers to make people's life easier
[21:04:29] <axisys> timeless: rebooting again
[21:04:30] *** Kernel86|Laptop has joined #OpenSolaris
[21:04:47] <timeless> you shouldn't need to reboot!
[21:04:53] <Seawolf99> well I don't think I have Bind 8 installed, from the message it looks like its was part of ON in snv_46 and now its not. I tried pkginfo |grep -I "name" but it could find bind
[21:04:56] <timeless> the docs explain what you can do
[21:04:59] <axisys> that did it
[21:05:16] <timeless> it helps that i fought this stuff yesterday :)
[21:05:26] <axisys> timeless:just wanted to see if does not break ut
[21:05:27] <axisys> it
[21:05:40] <timeless> seawolf99: well, in that case, it's probably time to hack the migration script
[21:05:46] <axisys> timeless: sure did :-)
[21:06:12] <timeless> seawolf99: your choices are grep/truss basically
[21:06:22] <timeless> i've found truss works wonders
[21:06:56] <timeless> well, you might be able to search the cvs.opensolaris.org server for the migration reference
[21:06:57] <timeless> one sec
[21:07:36] <timeless> ok, it's in bfu.sh
[21:08:14] <timeless> there's a -f floating around if you really want it
[21:08:18] <timeless> but instead of doing that
[21:08:26] * timeless reads http://cvs.opensolaris.org/source/xref/on/usr/src/tools/scripts/bfu.sh#2105
[21:09:02] <timeless> grep libdns /var/sadm/install/contents
[21:09:06] <timeless> what does it say?
[21:09:37] * timeless tries to remember how negations existence, ands, and suches work after 10pm
[21:10:18] <timeless> seawolf99: i presume you're comfortable w/ patch?
[21:10:26] <Seawolf99> timeless: /usr/lib/gnome-vfs-2.0/modules/libdns-sd.so f none 0755 root bin 10436 19051 1160506743 SUNWgnome-vfs
[21:10:38] <Seawolf99> there is no libdns.so in /usr/lib/dns
[21:11:08] <Seawolf99> yeah
[21:11:25] <Seawolf99> patch on linux, have no clue on solaris
[21:11:25] *** _william_ has joined #opensolaris
[21:11:34] <Seawolf99> but I could figure out
[21:12:06] <_william_> hi all
[21:12:11] <timeless> there's gpatch on my solaris
[21:12:33] *** nwf has quit IRC
[21:12:53] <timeless> http://www.webwizardry.net/~timeless/solaris/seawolf99.bfu.46.patch
[21:13:19] <timeless> no warranty
[21:13:33] <timeless> but that should get you past it
[21:13:42] <Seawolf99> timeless: Thanks and I am used to no warranty after working 7 yrs with linux
[21:14:03] <timeless> unfortunately, given that the script in question is closed, i can't say what it's doing
[21:14:36] *** Yamaraj has quit IRC
[21:15:03] <timeless> i think that if you actually cared about bind9 you would get it from SFW
[21:16:09] <timeless> bfu is interesting, kinda a super checksetup.pl (bugzilla thing)
[21:16:22] <timeless> but having read a few bits of it, i'm pretty sure i can't try it
[21:17:58] *** tsoome has joined #opensolaris
[21:18:45] * timeless goes to grab elinks for nexenta(zone)
[21:19:46] <Seawolf99> I don't really care about installing bind8-or-9, but bfu did. I am trying to learn ON upgrade. I also compiled the latest src for that and took almost 6hr 45 mins on dell optiplex (P4, 512MB)
[21:20:18] <timeless> hey, it could be worse
[21:20:32] <timeless> i had a process that took 7hrs :( and it wasn't building anything useful either :(
[21:20:38] *** bengtf__ has joined #opensolaris
[21:20:42] *** bengtf__ is now known as bengtf
[21:21:05] <Seawolf99> I wasted 5 hrs the day before not knowing that 1.5GB would not be enough and it ran out of disk space
[21:23:26] *** delewis has quit IRC
[21:25:46] <Error_404> 1.5 gigs is nothing
[21:26:09] <Error_404> like, for anything
[21:26:47] <timeless> nexenta fits in <400mb :)
[21:27:07] <timeless> although, i'm not sure if that counts the real solaris kernel that lives in the global zone
[21:27:45] <Error_404> perhaps, but i don't ever think "how much space do I need?" and come up with a number "smaller than a DVD"
[21:27:58] <astinus> especially with disks so damn cheap these days
[21:28:08] <Error_404> if i need to think about how much space I need, chances are I'm thinking in gigabytes
[21:28:14] 
[21:28:40] * timeless tries to remember the name of the computer store here
[21:29:19] <Error_404> i need to pick up another drive actually
[21:29:33] <Error_404> to mirror my zfs drive, and to put my root on SATA
[21:29:43] <Error_404> unsupported ATA controller = painful
[21:29:56] <Seawolf99> well, I know now and will buy disks, but hey I tried nexenta to start with before SX
[21:30:13] <timeless> i'm not advertising nexenta
[21:30:16] <astinus> Any of you lot UK based and know where I could obtain something like a B1000 for not much money?
[21:30:35] <timeless> i'm only using it as a zone to do some automated apt-get'ing of an unrelated system
[21:31:06] <jamesd_> ebay.co.uk
[21:31:28] <astinus> james_:  things seem overpriced on there lately, esp. Sun gear
[21:32:40] <jamesd_> i don't think you are going to do any better... last i looked  ebay.co.uk had even better prices than  ebay.com
[21:33:13] <jamesd_> esp on sun gear
[21:33:16] <mrdeviant> did you factor in the fact that it's not in usd
[21:33:29] <jamesd_> yes
[21:34:38] <jamesd_> http://search.ebay.co.uk/search/search.dll?from=R40&satitle=sun+1000   139 pounds for a b1k?  sounds damm  decent to me
[21:34:42] 
[21:35:59] <Seawolf99> timeless: thanks once again, it installed fine
[21:38:45] * timeless chuckles
[21:38:50] <timeless> /usr/lib/0 at 0 dot so.1 rocks
[21:39:46] <elektronkind> huh? what's that?
[21:40:06] <elektronkind> libmanwithglassesandabignose.so ?
[21:40:23] <timeless> it's in man ld.so.1
[21:41:46] *** zarathustra has joined #opensolaris
[21:46:32] * timeless considers disabling fingderd
[21:47:09] <Gman> Hoooray!!
[21:47:13] <Gman> JDS is LIVE! :)
[21:47:38] <Gman> http://mail.opensolaris.org/pipermail/desktop-discuss/2006-October/001987.html
[21:47:39] <timeless> so solaris ships w/ telnet enabled?
[21:48:04] <sommerfeld> timeless: that was fixed recently
[21:48:23] <tsoome> timeless and is it bad?
[21:49:06] * timeless shrugs
[21:49:09] <timeless> just interesting
[21:49:46] *** zarathustra is now known as sopor_aeternus
[21:50:08] <timeless> rsh is also enabled?
[21:50:14] <tsoome> I would rather use telnet than blindinlgy believing the ssh (or some other fancy new solution) is safe
[21:50:47] <icon> heh
[21:50:57] <icon> im not sure how to react to that ;)
[21:51:07] *** sparc-kly__ has joined #opensolaris
[21:51:16] <Triskelios> calling ssh a "fancy new solution" is pretty silly
[21:51:35] <tsoome> maybe
[21:51:36] <timeless> ssh has its share of securiity flaws
[21:51:39] <tsoome> maybe not
[21:51:44] <icon> well, i would take 2kbit public key encryption over a raw socket any day
[21:51:45] * timeless shrugs
[21:51:54] <icon> timeless: which are handled quite promptly
[21:52:00] <timeless> icon: sure
[21:52:05] <timeless> that doesn't help me though
[21:52:08] <icon> remember the berkely telnet flaw in 99?
[21:52:15] <timeless> no :)
[21:52:20] <icon> it was more than just nasty
[21:52:21] <coffman> svccfg apply /var/svc/profile/generic_limited_net.xml
[21:52:37] * timeless goes to less /var/svc/profile/generic_limited_net.xml
[21:52:45] <coffman> thats what i recommend
[21:53:07] <tsoome> my point is - know your "enemy" and use when needed and as needed without hurting yourself.
[21:53:20] <timeless> man, that could have saved me so much time
[21:53:34] <Error_404> if you're throwing a default install on the internet without any sort of security at all, you're either dumb or inconsequential anyways
[21:53:45] <tsoome> true:D
[21:53:46] <timeless> coffman: is there a convenient way to compare that to my current config?
[21:53:47] <icon> tsoome: honestly, there is no reason to use telnet whatsoever these days unless you lock yourself out of sshd
[21:54:19] <coffman> uh
[21:54:21] <coffman> dunno
[21:54:51] <icon> im not talking about limited devices, thats a different story
[21:55:04] <icon> this is #opensolaris, not #embedded :P
[21:55:16] *** solarisjon has left #opensolaris
[21:55:22] <timeless> hrm, i was hoping this was #embedded
[21:55:29] <tsoome> if I have trusted network, why I should?
[21:55:37] <icon> tsoome: never trust the network
[21:55:54] *** sahafeez has joined #opensolaris
[21:55:58] *** slowhog has left #opensolaris
[21:56:21] <coffman> tsoome: u trust ure network?
[21:56:34] <tsoome> not any network;)
[21:57:01] <icon> ive worked some places that wouldnt even trust transport security
[21:57:02] <tsoome> but - I know I can trust my cluster interconnect for instance;)
[21:57:05] <timeless> coffman: i trusted loopback, but then something like it panic'd me :)
[21:57:13] <icon> haha
[21:57:18] *** Symmetria has quit IRC
[21:57:25] <Error_404> icon: ssh is slow
[21:57:44] <timeless> is that w/ -1, -3, or -9 compression?
[21:57:51] <Error_404> if my system's under a *lot* of I/O load, sometimes it takes over a minute for login w/ ssh...
[21:57:52] <Triskelios> Error_404: there are remedies for that if you need to dump a lot of data
[21:57:57] <Error_404> telnet works all the time
[21:58:01] <coffman> security is something for the aplication level - not for the network
[21:59:22] <coffman> *level - layer even
[22:00:05] *** sparc-kly_ has quit IRC
[22:13:44] <betzi> no one has solaris express/opensolaris on an external usb drive?
[22:14:01] *** sparc-kly__ has quit IRC
[22:14:32] *** lisppaste3 has joined #opensolaris
[22:14:34] *** bougie_ has quit IRC
[22:16:33] *** yippi has quit IRC
[22:21:37] *** deather__ is now known as deather
[22:23:20] <timeless> how can a  java service that uses gnome be running when x11 isn't?
[22:26:38] <Stric> not easily
[22:27:05] <trygvis> sure, there are lots of x11 emulation libs
[22:27:15] <Stric> Xvfb, Xvnc, ...
[22:27:28] <jamesd_> it could also be ran remotely...  no X neeeded on the server
[22:27:54] <trygvis> timeless: did you ever get proper NAT set up on your zones?
[22:28:07] <timeless> trygivs: no, i setup ssh forwarding
[22:28:14] <timeless> lemme see if i posted it
[22:28:21] <trygvis> slacker! now I need the configuration! :)
[22:28:23] <timeless> looks like i didn't
[22:28:33] <timeless> do you need ssh and stuff
[22:28:37] <timeless> or is http good enough?
[22:28:48] <timeless> basically i setup apache as a local proxy
[22:29:02] <trygvis> yah, we're doing that aswell, but I need the zones to be able to get out on the internet
[22:29:08] <timeless> actually, i have two proxies floating around
[22:29:31] <timeless> in my case, my box itself can't go out on the internet
[22:29:36] <timeless> so that's not much of an issue
[22:29:46] <timeless> if out is ssh friendly, just setup keys for ssh tunneling
[22:29:51] <timeless> which is what i mostly did
[22:30:08] <timeless> the only thing i can do in the intranet is talk to an http(ftp/https) proxy
[22:30:22] <trygvis> right
[22:30:27] <timeless> so ssh -L 8000:real_proxy:8080 localhost
[22:30:48] <timeless> http_proxy=http://localhost:8000/ sh
[22:31:03] <timeless> i'm not quite sure what i did wrong
[22:31:20] <timeless> i did find out that the panic i got from ping'ing localhost was already fixed in ON
[22:31:22] <trygvis> is it the /dev/rdsk/c5t0d0s2 or the /dev/dsk/c5t0d0s2 device I want to copy if I want a block copy of the drive?
[22:32:10] *** nwf has joined #opensolaris
[22:32:42] *** dunc has quit IRC
[22:32:45] <timeless> /dev/zpool/dsk/root_pool/swap
[22:32:50] <timeless> is something i can use as swap
[22:33:21] * timeless thinks
[22:33:36] *** dunc has joined #opensolaris
[22:33:42] <timeless> yeah
[22:34:02] <timeless> http://docs.sun.com/app/docs/doc/816-5177/6mbbc4gcr?a=view
[22:34:35] <timeless> that seems like a reasonable explanation
[22:36:26] * timeless waits as ssh crawls
[22:36:43] *** bondolo is now known as bondolo_
[22:45:16] *** phretor has joined #opensolaris
[22:45:19] <phretor> hi there
[22:45:47] <phretor> suppose that I've a symlink from /bin/foo to /usr/bin/foo - what does the kernel do when I execute /bin/foo? Does it execute /usr/bin/foo? Does it execve('/usr/bin/foo')?
[22:46:11] <phretor> looking at some BSM audit files I've found many duplicated PIDs and they belong to symlinked executables: that's because I'm asking
[22:47:27] <elektronkind> symlink resolution happens on the fs level
[22:47:49] <phretor> elektronkind: even in Solaris 2.5.1? :)
[22:47:57] <elektronkind> even in 2.5.1.
[22:48:26] <phretor> elektronkind: thanks
[22:48:42] <phretor> elektronkind: may I read about this topic somewhere?
[22:48:43] <axisys> i have a user wanted to know why LD_LIBRARY_PATH is bad and where can he find more about -L and -R
[22:48:48] <timeless> phretor: consider
[22:48:51] <timeless> #!/bin/sh
[22:48:54] <timeless> echo $0 $*
[22:48:59] <timeless> ^Z
[22:49:02] <elektronkind> from wikipedia: Symbolic links operate transparently, which means that their implementation remains invisible to applications. When a program opens, reads, or writes a symbolic link, the operating system will automatically redirect the relevant action to the target of the symlink. However, functions exist to detect symbolic links, so applications may find and manipulate them if needed.
[22:49:06] <timeless> ln -s that.sh foop.sh
[22:49:07] <elektronkind> http://en.wikipedia.org/wiki/Symbolic_link
[22:49:08] <tsoome> axisys: docs.sun.com and linker and libraries guide
[22:49:13] <timeless> chmod +x that.sh
[22:49:17] <timeless> ./foop.sh
[22:49:20] *** mrdeviant has quit IRC
[22:49:26] <axisys> thnx guys
[22:50:06] <elektronkind> man ld
[22:50:10] <elektronkind> axisys
[22:50:13] <timeless> although each os is free to have its own strange answer for process names
[22:50:29] <timeless> there are some great rants in portability libraries about that :)
[22:50:32] <tsoome> and LD_LIBRARY_PATH is not bad if used in [wrapper] scripts and if you really know why;)
[22:50:56] *** betzi has quit IRC
[22:54:31] <tsoome> but why it's bad - suppose you have gnome/gtk libs in /lib, /opt/csw and /opt/sfw, how you are going to set this LD_LIBRARY_PATH for uses and making all apps to behave in a correct way;)
[22:55:57] <axisys> http://www.blastwave.org/userguide/ also has some comments on not using LD_LIBARRY_PATH..
[22:55:59] <timeless> ~/bin/{foopy} scripts :)
[22:56:29] <axisys> i know why it is bad.. but i needed to hand something to this user for his own good
[22:56:53] <icon> err
[22:56:58] <icon> why not link with -L and -R set?
[22:57:19] <icon> setting LD_LIBRARY_PATH is bad kharma
[22:57:48] <tsoome> gnu configure/makefiles are quite often broken regarding to -R :(
[22:58:34] <icon> oh?
[22:58:35] <icon> ive never had issues
[22:59:41] <tsoome> I have seen quite many times.... ok, perhaps this has been improved lately...
[23:00:08] <bengtf> check anything that uses libfontconfig , that is not compiled correctly on blastwave or opensolaris, had to specify LD_LIBRARY_PATH for that
[23:00:36] <icon> nothing you can do for that :/
[23:00:56] <icon> ldconfig helps a bit, but well, you know how that is ;)
[23:01:19] <bengtf> especially if you use alot of languages on your machine, I use most european and asian , and that seems to brake anything that uses fonts ...
[23:01:31] <bengtf> brake/break
[23:01:34] *** besonen_mobile has joined #opensolaris
[23:02:55] <richlowe> you can use clri, rather than LD_LIBRARY_PATH
[23:02:58] <richlowe> no, not that.
[23:02:59] <richlowe> clre
[23:03:09] <richlowe> crle, even.
[23:03:11] <tsoome> its as bad imo:)
[23:03:12] * richlowe gives up
[23:03:38] <icon> ive heard bad bits about crle, ive never really messed with it though
[23:03:50] <icon> i generally dont have an issue with ldconfig
[23:03:52] <tsoome> except env is taking little less space...
[23:04:15] <icon> its convienient when you need to isolate lib trees
[23:04:29] <icon> ie: /opt/csw/lib vs. /opt/sfw/lib etc.
[23:05:06] <icon> and setting -L and -R in build scripts is a bit inconvenient when you want/need to move things around
[23:05:40] <timeless> richlowe: clri was what saved me yesterday :) most people really shouldn't be using it ;-)
[23:06:00] <bengtf> I wouldnt care if they fixed the bugs in libfontconfig ;)
[23:06:01] <icon> ie: tomorrow dclarke wakes up and decides change roots to /opt/foo
[23:07:49] * timeless ponders
[23:08:04] <timeless> does solaris have any provisions for getting rid of zombies? :)
[23:08:13] <icon> dont make them :D
[23:08:15] <bengtf> machineguns ?
[23:08:17] <icon> wait() is your friend
[23:08:31] <timeless> what am i wait()ing for? :)
[23:08:37] <icon> haha
[23:08:44] <icon> fork() and conquer!
[23:09:28] *** Kernel86|Laptop has quit IRC
[23:09:31] <timeless> seriously though
[23:09:36] <jamesd_> timeless, your very select()ive
[23:09:37] *** Kernel86|Laptop has joined #OpenSolaris
[23:10:03] <timeless> /bin/true
[23:10:06] <icon> timeless: are you creating zombie processes?
[23:10:10] <icon> or do you just want to kill them?
[23:10:33] <timeless> well, if you mean did i cause them to be zombies by trying to kill whatever they were? yes
[23:10:46] <timeless> if you mean did i just randomly spawn(zombies) no
[23:10:55] <icon> ah okay
[23:10:58] <bengtf> kill the zombies parents and hope that process one waits them
[23:11:05] <icon> there you have it :D
[23:11:14] <timeless> that leads me to the next question
[23:11:35] <timeless> without installing pstree, how do i figure out what the parent is? :)
[23:11:39] <bengtf> that was how it was done ages ago, I hope it is still the same ;)
[23:11:54] <icon> bengtf: its how i always did it ;)
[23:11:55] <bengtf> ps -ef  ofcourse
[23:12:22] <timeless> ps -ef|grep defunct|wc -l
[23:12:23] <timeless> 144
[23:12:27] <timeless> that's a good start
[23:12:36] <icon> haha
[23:12:40] <icon> what the hell are you running?!
[23:12:50] <timeless> care to be more specific?
[23:13:02] <icon> ps -ef|grep defunct |xargs kill -9
[23:13:04] <bengtf> check check 3rd column that should be one ?
[23:13:05] <timeless> root|timeless {randompid} 1
[23:13:26] <icon> (dont run that, i was kidding :P)
[23:14:01] <bengtf> would HUP 1 work for this or does the machine go down ?
[23:14:04] <timeless> all of the defuncts are ppid 1
[23:14:12] <icon> oh ouch
[23:14:15] <axisys> preap kills zombie
[23:14:24] <icon> HUP on init doesnt down the system
[23:14:41] <bengtf> it used to rewait
[23:14:46] <icon> thats sort of cool actually... init spawning defucnt proc's
[23:14:53] <bengtf> p_reap ? or preap ?
[23:15:04] <timeless> preap(1)
[23:15:26] <bengtf> that would be the new command to do it then ;)
[23:15:27] <icon> what were you doing timeless ?
[23:15:46] <timeless> icon: well, this is still the same system that ran out of memory a while ago
[23:15:53] <timeless> it didn't fall over very dead
[23:15:57] <timeless> just kinda/sorta dead
[23:16:01] *** laca2 has joined #opensolaris
[23:16:23] <timeless> i could probably run mdb or check /proc if you want me to try to figure out what things i had run
[23:16:33] <timeless> but basically most of x11 fell apart :)
[23:16:54] <icon> ahh
[23:17:10] <icon> why not just init 6?
[23:17:13] <timeless> i had a bunch of gnome-terminals, a gnome-edit, firefox, a bunch of sshs(dozens in all likelyhood),
[23:17:31] <timeless> i kinda like my system and would rather learn how to maintain it than just knock it over
[23:17:45] <icon> good enough
[23:17:53] <icon> pride over pragmatism :)
[23:18:29] <icon> well, reap the zombies and i would check to make sure that your smf services are still intact
[23:18:32] *** laca2 has quit IRC
[23:18:36] <icon> no panic which is good
[23:18:48] <timeless> preap: Failed to reap {foo}: the only non-defunct ancestor is 'init'
[23:19:02] <icon> might drop back into single user, and back into your current init
[23:19:13] <timeless> ?
[23:19:37] <icon> icon: you might want to reinit your current runlevel in case you had some services blow out on you
[23:20:01] <timeless> svcadm didn't complain about anything
[23:20:08] <icon> did you try preap -F
[23:20:12] <timeless> not yet
[23:20:15] <timeless> i'm rereading preap
[23:20:18] <timeless> (1)
[23:20:21] <timeless> i'm confused
[23:20:30] <timeless> it claims init(1m) periodically reaps
[23:20:39] <icon> it has a safety check to not reap decendants of init
[23:20:40] <timeless> what's the definition of periodically? once a year?
[23:21:18] <icon> its saying init *may* periodically reap
[23:21:24] <icon> inow, read the init man page ;)
[23:21:57] <timeless> it does not say may
[23:22:15] <icon> ah my mistake
[23:22:31] <icon> may was on the line above, i think i had one of my dyslexia moments ;)
[23:22:36] <timeless> and offhand, init(1m) doesn't mention reap or defunct
[23:22:58] <icon> regardless, i would preap -F those pids, then issue init s
[23:23:05] <icon> and then init back to your main runlevel
[23:23:18] <tsoome> it defunct's real parent is gone, init will be set to be parent, this does not mean init is creating them
[23:23:37] <timeless> i would never accuse init of creating defunct processes
[23:24:13] <icon> most likely:
[23:25:21] <icon> preap -F pids && init s && init 3
[23:25:56] <icon> then again, im anal - i prefer to reinit when a box starts throwing sharts
[23:26:17] <timeless> i prefer to watch as much fireworks as possible
[23:26:22] <timeless> they're generally rare and quite amusing
[23:26:29] <icon> haha
[23:26:37] <icon> personal box i hope then :)
[23:27:43] <timeless> actually, it's theoretically running a webservice which i want people to use heavily
[23:27:52] <timeless> (the updater for that service is what knocked over the box)
[23:28:02] <timeless> but in practice, it's 12.30am and everyone is at home
[23:28:05] <icon> dont have it jailed or under limits?
[23:28:06] <timeless> hopefully sleeping
[23:28:30] <timeless> i was planning on eventually doing the update in a zone
[23:28:31] <tsoome> zone it and set cpu, memory and other limits.....
[23:28:40] <icon> yup
[23:28:41] <timeless> but for basic experimenting, i hadn't gotten that far
[23:28:42] <icon> zones are your friend :)
[23:28:49] <icon> especially if its mem hungry like that
[23:28:54] <timeless> zones panic'd my box last night!
[23:29:05] <icon> if its java, set your -Xmx and -Xms
[23:29:08] <timeless> (repeatedly, reproducably, for buggyness)
[23:29:13] <timeless> it's perl
[23:29:17] <icon> ouch
[23:29:18] <timeless> or mostly perl
[23:29:55] <tsoome> then create project and use limits with this project....
[23:30:23] * timeless will have to man that next
[23:30:31] <timeless> first i need to finish reading init(1m)
[23:30:35] *** mkhl has quit IRC
[23:30:52] <Gman> ooh, first onnv-notify mail that i've received
[23:30:53] *** alanc-away is now known as alanc
[23:31:01] * icon cheers
[23:31:21] <richlowe> Gman: heh.
[23:31:23] <Gman> the format is puke.
[23:31:24] <richlowe> damnit.
[23:32:25] *** bondolo_ is now known as bondolo
[23:32:47] *** christof has joined #opensolaris
[23:32:53] <timeless> richlowe: was i correct earlier in concluding that your comments indicated you were concerned about bfu and /sw/ internal bits?
[23:33:12] *** christof has left #opensolaris
[23:33:55] <richlowe> timeless: when I was talking with Alan?
[23:34:01] <timeless> probably
[23:34:15] <richlowe> Part of my concern is dealing with cross-consolidation flag-days better now that the source is available in realtime.
[23:34:18] <timeless> i haven't quite memorized everyone's name yet, and i've also lost a few sleep cycles
[23:34:25] <richlowe> part of it was dealing with in-consolidation flagdays regarding the closed bins.
[23:34:46] <richlowe> though, as Alan said, there's only 4 closed bins actually needed to build, there is still the possibility of open flagging against closed.
[23:35:12] <timeless> so what's the story of this bind8->9 migrate thing?
[23:36:16] <richlowe> I don't know, I've never actually seen that one.
[23:36:22] <richlowe> or I have, but so long ago I forgot.
[23:36:33] <timeless> http://cvs.opensolaris.org/source/xref/on/usr/src/tools/scripts/bfu.sh#2105
[23:36:53] * Tpenta yawns
[23:37:18] <richlowe> Yeah, but I don't think i've ever been in a position where bfu chose to attempt to run that script.
[23:37:26] <richlowe> that *is* the kind of thing we were talking about however, yes.
[23:37:35] <timeless> good
[23:38:05] <timeless> hrm
[23:38:14] <timeless> is it bad that i have *log* files that are 1G? :)
[23:38:29] <timeless> my xref is 261 for a 600m repository
[23:38:31] <bengtf> well yes
[23:39:16] <timeless> hrm
[23:39:23] <timeless> "name too long: ........."
[23:39:28] <timeless> that's a great error message
[23:40:07] <timeless> only 4 1/3 *million* instances of that error message
[23:40:38] <timeless> ok, so, i read init(1m)
[23:40:42] <timeless> and my xref built
[23:41:50] * timeless goes to google ENAMETOOLONG
[23:42:00] * icon &
[23:42:54] * timeless wonders what one is supposed to do upon encountering that error
[23:43:46] <bengtf> change your name to timel ?
[23:44:13] <bengtf> but, what application is logging that error ... ?
[23:44:41] <timeless> glimpse i'd expect
[23:44:56] <timeless> maybe find
[23:45:00] * timeless isn't really sure
[23:46:15] <bengtf> somtimes it is the path that is too long for some stupid reason,
[23:46:28] <timeless> like a symlink?
[23:46:48] <timeless> /................/yip -> ../../../../blah/....
[23:46:57] <bengtf> hmm, recursive links ...
[23:48:18] <timeless> /udev-0.084/upstream/tarballs/udev-084/test/sys/class/tty/ttyUSB0
[23:48:21] <timeless> device ->
[23:48:44] *** Kernel86|Laptop has quit IRC
[23:48:58] *** Kernel86|Laptop has joined #OpenSolaris
[23:48:59] <timeless> it's kinda strange
[23:49:22] <timeless> device is 1000:1000, everything else is more normally owned
[23:52:14] <timeless> ok, none of the other directories that are siblings of that one have 'devices'
[23:53:19] <timeless> err 'device'
[23:56:26] *** ShadowHntr has joined #opensolaris
[23:56:44] <timeless> so, assuming i don't have bad entries in my zfs file system
[23:59:11] *** gtc has joined #opensolaris





top