[00:00:04] <gisburn> boyd: and there are no resources left to do so.
[00:00:11] * timeless cries
[00:00:11] <timeless> cannot rename 'root_pool/hanoi/1': dataset is busy
[00:00:13] <gisburn> we have a problem
[00:02:13] <gisburn> when is thanksgiving over ?
[00:02:17] <timeless> monday
[00:02:23] <gisburn> oh crap
[00:02:51] <timeless> cool
[00:02:59] <timeless> i think my script worked :)
[00:03:36] <timeless> hrm, well, it /would/ have worked
[00:03:47] <timeless> instead it moved things to hanoi, and then right back to where they started :)
[00:05:24] *** bobbyz has joined #opensolaris
[00:07:32] *** fik has quit IRC
[00:09:27] *** RaD|Tz has joined #opensolaris
[00:09:49] *** dj2 has quit IRC
[00:18:28] *** slowhog has quit IRC
[00:18:54] * boyd grumbles about Thanksgiving. I want to give thanks for my SXCR b53 isos :)
[00:19:13] *** syndrome71 has joined #opensolaris
[00:19:20] <boyd> syndrome71: Hay!
[00:19:35] <Gman> boyd, after the morning i've had, i wouldn't be too thankful yet :/
[00:19:51] <boyd> Gman: Oh really? Damn...
[00:20:02] <Gman> i dunno
[00:20:18] <Gman> i haven't upgraded since b48, and the DRI stuff is probably hurting me
[00:20:37] <boyd> Hey, you said that TJDS is in b53, is that true for the current vermillion builds too?
[00:20:45] <Gman> yeah
[00:20:52] <boyd> 'kay, thanks
[00:21:34] <timeless> what dri stuff?
[00:21:39] * timeless is sitting on b49
[00:24:49] * hali is sitting on b6000
[00:27:01] * gisburn wonders when B666 will be released
[00:28:35] *** bunker has quit IRC
[00:32:55] <timeless> oops, i lost a directory :)
[00:33:17] <boyd> Well, where did you see it last?
[00:33:22] <boyd> Retrace your steps
[00:34:26] <timeless> i don't suppose gman has a way to copy all output from a 50,100 line gnome-terminal? :(
[00:36:04] *** laca has joined #opensolaris
[00:36:13] <timeless> well that's irony for me
[00:36:21] * timeless finds the steps while scrolling
[00:38:21] <boyd> Select all?
[00:39:10] <timeless> yeah, i wish, i can't find that
[00:39:21] <timeless> seems kinda basic, even nt's cmd.exe has it...
[00:41:24] <boyd> Select a little at the end, then shift-click at the other end
[00:42:03] <timeless> thanks :)
[00:42:36] <timeless> ok, i think it was last seen a few lines before one of my favorite error messages
[00:42:56] <timeless> cannot rename 'root_pool/home/svn/lxr-root/lxr-data/world-all.0/world/browser': out of space
[00:43:55] *** trs81 has quit IRC
[00:43:58] <Gman> jmcp, nice mail :)
[00:45:28] *** crash| has quit IRC
[00:47:06] <jmcp> Gman: I thought you'd like it :)
[00:47:16] <jmcp> and on that note .... street breakfast calls
[00:47:19] <jmcp> back later folks
[00:47:21] <Gman> enjoy!
[00:47:24] *** jmcp has quit IRC
[00:49:14] *** bobbyz has quit IRC
[00:49:48] <boyd> jamesd__: Do you understand dennis' latest -discuss mail?
[00:50:01] *** _dreams_ has quit IRC
[00:53:37] *** peteh has quit IRC
[00:56:44] <jamesd__> boyd, the zfs one?
[00:56:49] <boyd> yesh
[00:57:05] <boyd> It's like he's asking the question that you answer... maybe it's Jeopardy
[00:57:28] <jamesd__> yeah.. be was expecting some deep voodo to move a pool like it was when you had SVM or vxvm
[00:57:40] *** freakazoid0223 has joined #opensolaris
[00:58:03] <boyd> Many people are suprised by the lack of voodoo reqd for zfs
[00:58:13] <jamesd__> yeap
[00:58:28] <jamesd__> "it can't be that simple"
[00:59:04] <boyd> I makes one wonder why the hell we've been doing all that voodoo elsewhere
[00:59:44] <jamesd__> because its a corner case that no one bothered with, i guess its easier with  the latest version of vxvm
[01:00:25] <boyd> Well, vxvm has always been similar to zfs in this area. vxdisk -o alldgs list; vxdg import foo
[01:01:12] <jamesd__> i heard it already did that step when you install it on a machine.
[01:01:21] <jamesd__> at least in vxvm 4.x
[01:02:29] <boyd> If the DG is marked as imported on the current host it will, but not if it's been deported of marked for another host
[01:04:29] *** uebayasi has joined #opensolaris
[01:04:43] <delewis> hmm, I don't know what you'd label the AIX LVM.
[01:04:43] *** axisys has quit IRC
[01:04:51] <delewis> it's not quite SVM and it's not quite ZFS or VxVM
[01:05:07] <boyd> LVM?
[01:05:15] <delewis> it's probably closer to VxVM than anything else
[01:05:23] <delewis> physical volumes, logical volumes, volume groups, etc.
[01:05:34] <delewis> which for the most part are 1:1 to VxVM
[01:05:59] <delewis> VxVM has subdisks which doesn't *exactly* fit into the LVM scheme.
[01:06:36] <boyd> I heard that the LVM team went to IBM management and said "OK, we've learned a lot from that implementation... here's how we want to re-do it" mgmt said no, So they left and founded Veritas.
[01:06:47] <timeless> heh
[01:06:49] <boyd> I have no idea how true that is
[01:07:02] <delewis> I don't know, but I've always found the AIX LVM to be way ahead of its time.
[01:07:07] *** axisys has joined #opensolaris
[01:07:12] <delewis> even compared to Veritas, as the the AIX LVM has always been free.
[01:07:23] <delewis> you never, ever see people purchasing Veritas for AIX. :-)
[01:07:38] <boyd> The FS/LVM integration leading to auto-growing volumes can be thought of as nice or awful, depending on your POV
[01:07:38] <delewis> that is unless they have VxVM volumes that want to import
[01:07:49] <delewis> it's actually not that bad
[01:08:01] <delewis> and it leads to far more efficient disk usage
[01:08:08] <delewis> "you only grow your filesystem when you need it"
[01:08:25] <delewis> rather than doing "yeah, we'll *probably* need 20GB for /"
[01:08:40] <boyd> I agree... I mentioned POV :)
[01:09:07] <boyd> I must lead to volume-level fragmentation though
[01:09:18] <boyd> s/I/It/
[01:09:50] <delewis> that's why AIX makes the distinction between inner, outer, middle, etc. layouts on the disk for a volume group.
[01:10:05] <delewis> you can throw all of one volume group on the middle part of each disk if you think that will be accessed a lot
[01:10:12] <delewis> and throw throw the others on the inner or middle
[01:10:21] <delewis> (and there's even an inner middle and outer middle)
[01:13:09] * delewis wishes he would stop receiving stock market spam
[01:17:38] *** fik has joined #opensolaris
[01:17:57] <boyd> You'd prefer herbal v1@grA ?
[01:18:06] <delewis> hehe
[01:18:07] <jbk> or a watch?
[01:18:22] <delewis> or "cheap software"
[01:18:39] <timeless> you're not a fan of teenie?
[01:18:47] <timeless> or japanese or hebrew spam?
[01:19:00] <delewis> not particularly. :-)
[01:19:09] <delewis> and it's impossible to filter for this stuff.
[01:19:22] <boyd> I'm having trouble keeping up with all these inheritances that I've got coming to me. I had no idea I had so many relatives
[01:19:23] *** trs81 has joined #opensolaris
[01:19:29] <timeless> how about new software for you including UFS?
[01:19:32] <timeless> and FreeBSD
[01:19:37] <delewis> I'm even checking for a reverse DNS and all of it seems ineffective.
[01:19:46] <timeless> this has to be the *strangest* spam i've seen in months
[01:19:53] <boyd> Bayesian?
[01:19:57] <delewis> timeless: yes, I've gotten that one, too.
[01:20:18] <delewis> I would be anti-social and setup do Draconian mail filtering
[01:20:35] <timeless> boyd: i'm reading gmail's spam folder
[01:20:43] <delewis> where when someone sends you an email, you send them one, and if they reply back, they're "added to your list of acceptable people to receive mail from"
[01:21:04] <timeless> doesn't work for some mailing lists
[01:21:07] <timeless> like bugzilla-daemon@*
[01:21:12] <delewis> oh, that's true.
[01:21:17] *** RaD|Tz has quit IRC
[01:21:22] * boyd is still appalled that smime hasn't been widely used... it helps to solve so many probs
[01:21:46] <timeless> so, could someone please explain the UFS/FreeBSD/Slackware spam?
[01:21:54] * timeless really doesn't get it
[01:22:01] <delewis> timeless: I haven't been able to figure it out yet, either.
[01:22:12] <delewis> it looks like a concatentation of several FreeBSD mailing list messages.
[01:22:23] <timeless> there's a link to eki-oem.com
[01:22:28] <timeless> maybe that's the key?
[01:22:37] <boyd> If you're stupid enough to buy from spam you'll be stupid enough to buy free SW?
[01:22:40] <delewis> I think understanding the nature of spam is a futile effort.
[01:22:49] <delewis> :-)
[01:23:18] <timeless> i read it because it's entertaining
[01:23:30] <timeless> and lets me test gecko, gmail, and help improve their antiphishing stuff
[01:23:51] <timeless> besides, the japanese stuff has cool japanese ascii-like art
[01:23:52] <TpAway> I like the 419 scam I received from Mrs Deborah Harry (waits to see who is old enough to remember deborah harry)
[01:24:24] * timeless isn't
[01:24:32] <jbk> heh.. a friend of mine had some fun with one of those..
[01:24:37] <TpAway> she fronted the band blondie in teh early 80's
[01:24:46] <jbk> replied calling himself 'michael knight'
[01:24:49] <boyd> TpAway: was it selling blonde hair color?
[01:24:52] <timeless> heh
[01:24:53] <TpAway> :)
[01:24:57] <timeless> michael knight i remember
[01:24:58] *** TpAway is now known as Tpent1
[01:25:03] <richlowe> timeless: junk text that handily dodges filters.
[01:25:14] * richlowe has been wondering if the junk is specific to the lists people's addresses were harvested from.
[01:25:18] <richlowe> in those cases, that is.
[01:25:28] <timeless> i'm not associated w/ fbsd afaik
[01:25:35] <timeless> i'm associated w/ just about everything else
[01:25:46] <richlowe> technical list -> technical text
[01:26:01] *** Tpenta has quit IRC
[01:26:02] <richlowe> I've gotten spam with subjects starting Bug #NNNNN a couple of times.
[01:26:06] *** Tpent1 is now known as Tpenta
[01:26:07] <jbk> http://vitriolic.org/portal.php?article=3&sid=9c5587a4869c74a9c264ecbe398a6f3b <-- if you want to read and are bored
[01:26:08] <timeless> really?
[01:26:13] <timeless> i want to see the bug spam
[01:27:24] <timeless> hrm, i think that's my first sparm for searscard
[01:27:32] *** miffe has joined #opensolaris
[01:27:40] <timeless> why don't spammers learn that spelling correctly would help?
[01:27:49] <timeless> i mean, how hard is it to spell 'restore' correctly?
[01:28:02] <timeless> hrm, maybe they're finnish spammers? :)
[01:28:12] <timeless> (To restoure your Sears credit card features, please follow these steps:)
[01:28:15] <gisburn> timeless: no... hangul!
[01:28:41] *** Tpent1 has joined #opensolaris
[01:28:48] * gisburn waits for the YouTube crash
[01:28:55] *** gisburn has quit IRC
[01:30:12] *** gisburn has joined #opensolaris
[01:30:25] <gisburn> All hail to the crappy plugin implementation in mozilla!
[01:30:33] <gisburn> *grumpf*
[01:31:04] <timeless> at least java doesn't kill you
[01:31:07] * timeless kicks oji
[01:31:12] <richlowe> sure it does.
[01:31:15] <gisburn> heh
[01:31:18] <richlowe> have mozilla load java thingummy, pkill java.
[01:31:19] <richlowe> kaboom.
[01:31:20] <gisburn> richlowe: yeah
[01:31:21] <timeless> oh, i'm not aloud?
[01:31:32] <timeless> erm s/aloud/alone/
[01:31:48] <timeless> this is a new spam
[01:31:55] <richlowe> My other favourite is just how much stuff segvs when X dies.
[01:32:04] <richlowe> X, gdm-binary, firefox, thunderbird, and a couple of other bits I forget.
[01:32:15] <richlowe> (when X crashes, not exits...)
[01:32:20] <timeless> it has USMC, Super Bowl, Aerospace Engineer, Kiefer Sutherland
[01:32:48] <timeless> richlowe: well, i'm still a fan of all the glib based apps that abort()[coring] when they run out of memory
[01:33:11] *** freakazoid0223 has quit IRC
[01:33:53] <boyd> Wow! I always wondered... http://www.drivl.com/posts/view/471
[01:34:38] *** freakazoid0223 has joined #opensolaris
[01:35:23] *** yarihm has quit IRC
[01:37:19] <richlowe> boyd: it ceased being the 80's :)
[01:37:49] <boyd> Hehe... I must say, I always wondered if the technology could allow something other than color-change (other properties) but I guess not.
[01:38:56] *** nwf has quit IRC
[01:40:23] *** Gman is now known as GmanAFK
[01:42:24] *** laca has quit IRC
[01:44:42] *** rodrickbrown has quit IRC
[01:44:43] *** nbkk6fo_ has joined #opensolaris
[01:45:16] *** jmcp has joined #opensolaris
[01:49:26] <gisburn> Tpenta: ping!
[01:49:48] <gisburn> Tpenta: you're aware that your nick made it into our little highscore list, right ?
[01:50:55] *** jamesd__ is now known as jamesd
[01:51:02] *** ChanServ sets mode: +o jamesd
[01:53:28] *** Gadzooks has joined #opensolaris
[01:53:42] <gisburn> does anyone understand termcap/terminfo ?
[01:54:03] *** djgregor has quit IRC
[01:56:22] <Tpenta> gisburn?
[01:56:39] <gisburn> Tpenta: http://svn.genunix.org/repos/on/branches/ksh93/gisburn/scripts/gnaw.ksh
[01:56:49] *** fik has quit IRC
[01:57:08] * gisburn waits for the "ahhggrrllll.... take it out... TAKE IT OUT! OMH PUPPIES ARE DYING!!!!!"
[01:57:18] <gisburn> er
[01:57:21] <gisburn> s/OMH/OMG/
[01:57:48] <Tpenta> ahh, i see you needed some names :)
[01:57:57] <gisburn> Tpenta: the highscores are fake right now... but we needed an attact mode
[01:58:04] <Tpenta> :)
[01:58:07] *** Gadzooks has quit IRC
[01:58:15] <gisburn> Tpenta: well, most the entries are players who use it more than often
[01:58:41] <Tpenta> :)
[01:58:56] <gisburn> Tpenta: if this gets approved by gatekeepers (unlikely) then this will be the minespeeper of solaris... =:-)
[01:59:01] <gisburn> er
[01:59:04] <Tpenta> :)
[01:59:04] <gisburn> minesweeper
[01:59:16] <delewis> I miss /usr/games :-)
[01:59:32] <gisburn> Tpenta: seriously.... what do you think will code review say ? and gatekeepers ?
[01:59:42] *** kleppari has quit IRC
[01:59:47] <boyd> Yeah... I've actually had cause to use factor and morse
[01:59:50] <Tpenta> no idea
[01:59:55] <delewis> boyd: definitely.
[02:00:06] <delewis> trek is always fun, too. :-)
[02:00:11] <boyd> ...although not at the same time :)
[02:00:21] <delewis> nothing like blowing a few Klingons out of an ASCII display.
[02:00:22] <boyd> I liked that ATC thing
[02:00:37] <boyd> (I think that was a /usr/games thing)
[02:00:38] *** simford has joined #OpenSolaris
[02:02:47] <gisburn> $ (TERM=vt52 infocmp -1 | fgrep "cup=")
[02:02:48] <gisburn>         cup=\EY%p1%' '%+%c%p2%' '%+%c,
[02:02:49] <gisburn> # what does this mean ?
[02:03:29] <movement> cat on the keyboard
[02:03:47] <gisburn> ?!
[02:04:06] <gisburn> movement: I mean how does it look like when I insert values there ?
[02:04:15] * delewis thinks gisburn missed the joke
[02:04:47] <rydis> Why not just use "tput cup 42 11"?
[02:05:00] <gisburn> rydis: too slow
[02:05:02] <boyd> Yes, that's what you should do
[02:05:05] <gisburn> rydis: http://svn.genunix.org/repos/on/branches/ksh93/gisburn/scripts/gnaw.ksh
[02:05:37] <gisburn> boyd: this is around 200 times slower than caching the value of infocmp and the use filter+printf with ksh93 builtins
[02:05:40] <gisburn> 200 times
[02:05:41] <gisburn> !!!!
[02:05:42] <boyd> terminfo implements a little stack machine to process those sequences. If you want to avoid it you need to re-implement it. So it's better to pre-cache them.
[02:05:56] <boyd> (with all possible args)
[02:06:00] <gisburn> boyd: I tried that, too
[02:06:10] <gisburn> boyd: 132x25 times
[02:06:13] <gisburn> horrible
[02:06:37] <rydis> I don't seem to have ksh93 installed. Where is it supposed to be on a Solaris 10 machine?
[02:06:39] * boyd shrugs. Anything else is  likely to be an incorrect implementation
[02:06:43] <delewis> rydis: it isn't
[02:06:44] <boyd> rydis: N
[02:06:50] <boyd> not yet integrated
[02:07:03] <boyd> gisburn: is the main man on that project
[02:07:08] <rydis> Ah. /Then/ I understand what all the talking has been about. :)
[02:07:09] <boyd> err rydis ^^
[02:07:10] <gisburn> rydis: http://www.opensolaris.org/os/project/ksh93-integration/
[02:07:59] * jmcp groans @ http://politicsofdancing.blogspot.com/2006/11/as-god-is-my-witness-i-thought-turkeys.html
[02:08:16] <rydis> gisburn: I have found the XTerm escape sequence docs that X{org,Free86} supply to be useful for this sort of stuff.
[02:08:16] <boyd> gisburn: Are you really tellimg me that pre-filling an array with all the results and then looking them up is 200 times slower in ksh93?
[02:08:39] <boyd> rydis: But you can't assume xterm for a quality implementation
[02:09:43] *** djgregor has joined #opensolaris
[02:10:00] <rydis> boyd: I know, but for cup (and a lot of other stuff), it's the same for all "vt100 compatibles".
[02:11:25] <boyd> "vt100 compatibles" ? "all terminals"
[02:11:27] <rydis> But I see know that gisburn doesn't want vt100 compat. Oh, well.
[02:12:15] <gisburn> boyd: I mean ksh93 is FASTER than calling tput all the time
[02:12:36] <gisburn> rydis: I want vt100 and vt52 compatibiltiy
[02:12:42] <boyd> That's what I thought... so pre-cache all the possible outputs of tput with 2 params
[02:12:46] <gisburn> tput is just horrible slow
[02:12:46] <delewis> why vt52?
[02:12:55] * delewis wonders who on earth would have a VT52 terminal still laying around
[02:13:17] <boyd> gisburn: so CACHE the output
[02:13:22] <gisburn> boyd: that are "just" 6600 |fork()|s
[02:13:39] <gisburn> boyd: that needs a minute or worse for startup on my Ultra5
[02:13:44] <boyd> So, less than patchadd then :)
[02:13:52] <gisburn> ha ha ha
[02:13:59] <timeless> i think the windows telnet is closer to vt52 than vt100 :)
[02:14:14] <delewis> timeless: I doubt that's the case. :-)
[02:14:17] <delewis> vt52 is completely non-standard
[02:14:20] <boyd> It's even closer to a bag of sh*t
[02:14:25] <delewis> the escape sequences are entirely different
[02:15:03] <boyd> gisburn: You could make an on-disk cache
[02:15:12] <timeless> i don't suppose there's a spiffy way to ask zfs to find duplicate directories :)
[02:15:23] <boyd> What about memo-ising the lookup. Cache them on first use
[02:15:42] <boyd> timeless: I don't think so. star can do that
[02:15:47] <gisburn> boyd: oh come on, that's getting horrible.
[02:15:52] <timeless> star?
[02:15:52] <boyd> .. and it's POSIX compliant since 5 years
[02:15:55] <delewis> "star can do everything(TM)" :-)
[02:15:58] <gisburn> boyd: that code would become larger than the game logic.
[02:16:04] * jmcp aways to a meeting
[02:16:09] <delewis> star is the answer to the question of life, itself. :-)
[02:16:17] <rydis> From looking at what TERM=vt52 tput cup outputs, it seems it's "ESC Y arg1 arg2", where the args are the cursor position + 32.
[02:16:20] * timeless reads
[02:16:28] <rydis> No extra escapes anywhere.
[02:16:31] *** yusufg has quit IRC
[02:16:34] <boyd> gisburn: No. One function, looks in an array, if the entry is empty calls tput, fills the array, returns the result
[02:17:08] <gisburn> boyd: well, the problem is that the game logic runs in $( ... ) for double-buffering the output
[02:17:12] <timeless> boyd: can it do it w/o having any disk space available? and say a hard limit of 1g of vm
[02:17:15] <gisburn> boyd: caching will not work.
[02:17:54] <dlg> is there an easy way to boot an mp system into a uniprocessor kernel?
[02:18:12] <delewis> dlg: you can just disable the processors after boot
[02:18:25] <gisburn> dlg: without using a shotgun to kill the other cores ?
[02:18:41] <boyd> timeless: Yes, I think so
[02:18:50] <dlg> delewis: what if interrupts are bound to a cpu?
[02:19:11] <delewis> I would assume they'll be unbound
[02:19:32] <boyd> You can disable interrupts on a cpu with psradm... i
[02:19:46] <dlg> hrm, solaris is amazing
[02:20:15] <delewis> let me guess -- Linux screws this up somehow or lacks this functionality completely?
[02:20:18] <boyd> gisburn: What about a shared module (plugable builtin) that gives terminfo access. zsh has one
[02:20:23] <dlg> i dunno, i hardly use linux
[02:23:23] <movement> offlining a cpu won't work if there's threads strongly bound to it, though
[02:23:38] <gisburn> does $ printf "\EY%2\$d %1\$d " 1 1 # move to position 1,1 on vt52 ?
[02:23:58] <movement> and you can't remove the last CPU in a psrset. And finally, offlining doesn't really 'disable' the CPU so much as stop things from running on it
[02:24:21] <boyd> gisburn: IMHO anything less than a full implementation of terminfo is an incorrect one
[02:24:28] <boyd> http://pastebin.com/831750
[02:24:40] <dlg> movement: turns out i can turn the cpu off in the bios
[02:24:49] <gisburn> boyd: I know that.
[02:25:16] <gisburn> boyd: right now I am happy with something which is a) more or less working b) doesn't make the game unplayable on slower machines
[02:25:21] <boyd> movement: What is the functional difference between disabling and not using?
[02:25:46] <movement> not much
[02:25:53] <movement> you can enable it again :)
[02:26:10] <boyd> That's "destroying" :)
[02:26:57] <movement> well it's quite different from cfgadm
[02:27:10] <movement> you can't unplug the cpu and take it somewhere ;)
[02:27:48] <boyd> True. Although even on boxes where you can cfgadm a single CPU (e.g. 25k) you still can't unplug it :)
[02:29:39] <delewis> not without disabling an entire board, anyway.
[02:30:12] <dlg> you cant disable the cpu/board for the boot processor either
[02:30:39] <delewis> yes, you can, given Solaris is "symmetric"
[02:30:52] <gisburn> delewis: SF15k is NUMA
[02:31:02] <delewis> that would be a limitation of asymmetric multi-processing.
[02:31:07] <delewis> gisburn: interesting
[02:31:22] <gisburn> delewis: however the NUMA factor is < 2
[02:31:31] <boyd> There is no specific boot board for a HES domain
[02:31:40] <gisburn> delewis: NUMA usually means something like > 2
[02:31:46] <gisburn> delewis: sometimes 15 or 20
[02:32:04] <gisburn> delewis: which means if memory placement goes wrong performace will really suck
[02:32:17] <gisburn> delewis: this is where IRIX is really great
[02:32:34] *** regx has quit IRC
[02:32:41] <boyd> There is a lot of MPO in solaris for just this reason
[02:33:13] <delewis> heh
[02:33:20] <delewis> acroread in nv52 on SPARC segfaults.
[02:33:23] <delewis> at least with this particular pdf
[02:33:25] <delewis> let me try another
[02:33:29] <gisburn> boyd: not enougth to deal with larger numa factors
[02:33:46] <gisburn> boyd: I bet solaris will crawl if you have a numa factor > 10
[02:34:00] <gisburn> boyd: the OS is not even sheephearding processes
[02:34:25] * dlg would love solaris on an altix
[02:34:27] <gisburn> boyd: e.g. keeping related processes together to avoid that they spam the interconnect to death
[02:34:45] <gisburn> OUCH
[02:34:47] <gisburn> vt52 sucks
[02:34:58] <delewis> gisburn: yes, and that's why there was a vt100.
[02:35:03] <delewis> which became the standard.
[02:35:06] * Tpenta remembers vt52
[02:35:08] <gisburn> they seem to use the ASCII character as position
[02:35:23] <gisburn> OMG pink flying mice
[02:35:30] <gisburn> f*ck
[02:35:33] <delewis> yes, looks like acroread segfaults on any pdf in nv52
[02:35:35] <delewis> :-(
[02:35:42] * delewis doesn't much care for evince
[02:35:43] <timeless> delewis: it could be worse
[02:35:52] <timeless> evince on a device i use causes the device to reboot
[02:35:54] <Tpenta> delewis, change your theme from either of nimbus or blueprint to something else; i saw the same problem
[02:35:54] <boyd> I'm nearly sure that the NUMA factor for a 25k is > 2 Certainly the address interconnect path is ~7 asic longer
[02:36:03] <timeless> on just about any mildly complicated pdf
[02:36:06] <delewis> Tpenta: yeah, I'm running blueprint now :-(
[02:36:19] <Tpenta> i'm using ocean blue and it works
[02:36:20] <delewis> at least gimp works now :-)
[02:36:41] <boyd> syndrome71: Would know better than me
[02:37:12] * boyd lunches
[02:37:15] *** smoco has left #opensolaris
[02:37:25] *** dunc has joined #opensolaris
[02:37:44] * delewis actually likes crux
[02:39:21] <syndrome71> The remote memory latency on a 15K is not much more than 2X a local transaction...
[02:39:48] <syndrome71> I don't remember the actual numbers, but it was mid 200's (ns) for a local, and mid 500'
[02:40:05] <syndrome71> mid 500's for a remote transaction, assuming a CDC hit on the line if the line is remote.
[02:40:20] <syndrome71> if there was a cdc miss on a line that was remote from the home board, it was into the 600's.
[02:40:57] <gisburn> syndrome71: yep
[02:41:23] <gisburn> syndrome71: there is still a away to slow this down, right ?
[02:41:32] <gisburn> (via firmware)
[02:41:44] <syndrome71> a eay to slow it down? Why would you do that?
[02:41:53] <gisburn> syndrome71: I remember someone talking about a firmware option for debugging
[02:42:08] <gisburn> syndrome71: benchmarking/debugging/kernel development
[02:42:13] <syndrome71> perhaps.
[02:42:23] <syndrome71> can't imagine anyone but Sun doing that though.
[02:42:32] <gisburn> syndrome71: e.g. increase numa factor to see where your code starts to suck
[02:42:46] <syndrome71> certainly, many of the ASIC masks can be set to different things, inducing extra wait states at certain levels...
[02:43:29] <syndrome71> Oddly enough, I'm not familiar with the term numa factor... can you tell me more about what it's actually a measure of?
[02:44:34] <gisburn> syndrome71: AFAIK that was the latency min/max/avr for memory access
[02:45:13] <gisburn> syndrome71: (that's not the correct description but gives enougth hints where this goes)
[02:46:26] <gisburn> my god who ever invented vt52 is hopefully dead and buried and eaten by worms and then the worms eaten by komodo dragons
[02:46:47] <syndrome71> hm. I think I understand. Essentially, we have only a few latency groups on the 15K from the OS's NUMA awareness...
[02:46:49] <delewis> gisburn: you're forgetting that vt52 is a 60-year-old terminal.
[02:46:52] <delewis> er 40-year-old*
[02:46:59] *** aijin has joined #opensolaris
[02:47:06] <gisburn> delewis: this is the year 2006
[02:47:18] <gisburn> delewis: not 2026
[02:47:31] <delewis> gisburn: :-P
[02:47:40] <gisburn> the Omarg, the rulers of the universe did not conquer the world
[02:47:42] <gisburn> yet
[02:47:43] <boyd> syndrome71: I believe it's the ratio remote/local mem access time
[02:47:52] <syndrome71> Local memory for the current CPU pair, Other CPU pair's memory on the same board, and remote on a different board.
[02:48:08] <syndrome71> Solaris does not make a distinction for a line of memory that's not 'at home'.
[02:48:15] <boyd> Where remote is "as remote as you can get"
[02:48:32] <syndrome71> So, I'd say Boyd is pretty much spot on with his approximately 2X factor.
[02:48:55] <boyd> Wel.. actually I said > 2 but I'll take that credit :)
[02:48:56] <syndrome71> for the MAX, worst case where a line is not at home, and the CDC has no entry for that line, it might be as much as 3X.
[02:49:04] <syndrome71> :)
[02:49:48] <boyd> Want some things to waste brain cycles over the w/e? http://3quarksdaily.blogs.com/3quarksdaily/2006/11/i_challenge_you.html
[02:51:41] <twincest> will gcc 4.x be integrated into nevada at some point?
[02:53:44] <Tpenta> the people who can answer that are probably sitting back after being stuffed on turkey
[02:54:09] <syndrome71> stuffed with what?
[02:54:20] <gisburn> syndrome71: turkish people
[02:54:27] <gisburn> syndrome71: they eat humans
[02:54:36] <gisburn> syndrome71: different culture in the US
[02:54:40] <syndrome71> mmm... tastes like pork...
[02:54:52] <Tpenta> as ong as it is not the normal method of stuffing an animal...... ummm let's not go there
[02:55:37] <boyd> jmcp: posted a like earlier about some turkeys that were, well, stuffed: http://politicsofdancing.blogspot.com/2006/11/as-god-is-my-witness-i-thought-turkeys.html
[02:55:54] *** galt has joined #opensolaris
[02:56:11] *** g4lt-U60 has quit IRC
[03:07:00] *** dj2 has joined #opensolaris
[03:07:36] <gisburn> Tpenta: erm
[03:07:47] <gisburn> Tpenta: I just realise my joke went wrong.
[03:08:20] * gisburn was thinking "... they eat turkish people..." ... not the "... turkush people.eat ..."
[03:08:23] <gisburn> grumpf
[03:11:23] <Jiko> hmm, docs.sun.com is broken .. again
[03:13:42] <Sporq> hmm
[03:14:11] *** aijin has quit IRC
[03:23:37] *** yongsun has joined #opensolaris
[03:24:38] *** simford has quit IRC
[03:30:55] *** simford has joined #OpenSolaris
[03:33:23] *** GmanAFK is now known as Gman
[03:37:26] <boyd> jmcp: Finally saw that email.... classic
[03:44:24] *** nbkk6fo_ has quit IRC
[03:44:40] *** nbkk6fo_ has joined #opensolaris
[03:46:08] *** regx has joined #opensolaris
[04:20:45] *** kAv_ has joined #opensolaris
[04:21:06] *** yusufg has joined #opensolaris
[04:23:06] *** kAv_ has quit IRC
[04:26:31] *** Plaidrab has quit IRC
[04:27:07] *** Plaidrab has joined #opensolaris
[04:28:00] *** kAv_ has joined #opensolaris
[04:29:07] *** onbot has quit IRC
[04:29:37] *** darkcmd has quit IRC
[04:29:39] *** onbot has joined #opensolaris
[04:30:14] *** onbot has joined #opensolaris
[04:30:23] *** Plaidrab has quit IRC
[04:30:30] <onbot> commit by Govinda Tatti: 6490658 Incorrect Serialid # calculation in pxb_id_props; 6495918 system panics during dr testing in px:px_detach()
[04:30:55] *** Plaidrab has joined #opensolaris
[04:30:58] <movement> hah, it works.
[04:33:00] <twincest> neat
[04:33:13] <boyd> Beat me to it movement, That was my project for next week
[04:33:52] <movement> boyd: I've been waiting for the CIA guy but he's gone AWOL :(
[04:34:03] <movement> stuff's still getting sent there but I can't get its bot to join here
[04:34:34] <boyd> Ah, I didn't know there were such things... I was going to write one myself from scratch :)
[04:34:46] <twincest> i mailed the cia guy weeks ago and never got a reply
[04:35:41] <movement> twincest: yup, he's totally disappeared. I couldn't even find the bot source, so I have a truly vile hack instead
[04:36:20] <boyd> Well done anyway
[04:36:27] <movement> twincest: stuff's still getting added there though: http://cia.navi.cx/stats/project/opensolaris
[04:39:44] *** Plaidrab has quit IRC
[04:40:22] *** Plaidrab has joined #opensolaris
[04:41:40] <movement> boyd: /join #commits for a while ... pretty amazing.. 24/7 commitage
[04:42:02] * boyd joins
[04:55:33] *** KingDillyDilly has joined #opensolaris
[04:59:52] *** knightblader has joined #opensolaris
[05:01:57] <boyd> It's a bit like watching the matrix
[05:03:29] *** yongsun has quit IRC
[05:03:53] <movement> heh
[05:08:19] *** Gman has quit IRC
[05:11:07] <gisburn> my god
[05:11:10] <gisburn> solaris sucks
[05:11:26] * delewis thinks gisburn is drunk
[05:11:27] <gisburn>  /usr/bin/egrep can't handle [[:space::]]
[05:11:37] <gisburn> delewis: unfortunately no.
[05:11:39] <delewis> gisburn: /usr/xpg4/bin/grep can, though
[05:11:44] <gisburn> delewis: yes
[05:12:03] <gisburn> delewis: but seriouly: why wasn't this fixed ?
[05:12:06] <delewis>  /usr/bin/grep and /usr/bin/egrep are horribly, horribly broken when it comes to POSIX compatibility.
[05:12:12] <delewis> gisburn: it should be fixed
[05:12:13] <gisburn> backwards-compatibility ?
[05:12:21] <delewis> gisburn: yes
[05:12:26] <gisburn> f*ck
[05:12:27] <delewis> that's why /usr/bin is so broken.
[05:12:39] * gisburn replaces this with ksh93 builtin tools
[05:12:42] <delewis> I just put /usr/xpg4/bin in the front of my path.
[05:12:43] <gisburn> junk
[05:13:22] * gisburn is so happy that ksh93 can simply do a [[ $var = ~(E)regexp ]] && print "match"
[05:13:36] <gisburn> less trouble
[05:13:38] <gisburn> no fork
[05:13:44] <gisburn> XPG6-compilant
[05:14:10] * gisburn realises why Irek is so angry.
[05:14:39] <delewis> the /usr/bin nonsense is responsible for making Solaris look ancient to freeware developers
[05:15:00] <delewis> it's almost weekly I see some bug report of broken /bin or /usr/bin stuff on the mplayer lists.
[05:15:05] <gisburn> delewis: did you see the discussion about /sbin/sh ?
[05:15:13] <delewis> well, not weekly, but at the very least -- monthly.
[05:15:22] <gisburn> delewis: someone suggested to move the solaris port into a different branch
[05:15:29] <delewis> gisburn: yes
[05:15:31] <delewis> /usr/sun ;-)
[05:15:36] <gisburn> no.
[05:15:45] <gisburn> I mean the application code
[05:15:53] <delewis> oh, I see.
[05:16:02] <delewis> hmm, I don't much care for that idea.
[05:16:08] <gisburn> so the sun "wheenies" can maintain their port and fix that themselves
[05:16:17] <gisburn> seems to be a common idea right now
[05:16:26] <delewis> I think a better solution would be just to move the ancient userland in /usr/sun
[05:16:29] <delewis> as was done with /usr/ucb
[05:16:42] <gisburn> delewis: feel free to get that ARC'ed
[05:17:22] <delewis> I have no idea whatsoever about how to do that, and I'm quite sure that it would be rejected pretty quickly.
[05:17:23] <gisburn> delewis: I fear we will have to feed some managers@sun to komodo dragons before this will happen
[05:17:29] <delewis> yes
[05:17:42] <gisburn> very hungry komodo dragons
[05:17:54] <delewis> I'm just tired of hearing people call Solaris "ancient" because of /usr/bin and friends.
[05:17:55] <gisburn> otherwise I fear the managers will eat the dragons
[05:22:15] *** cheatersrealm has joined #opensolaris
[05:22:16] <kAv_> delewis : so whats "modern" ? for academic interest
[05:22:34] <delewis> kAv_: modern is something that complies with SUSv3
[05:22:45] <cheatersrealm> anyone care to suggest a sata card with 4 ports, be it controller or raid, but it must be flawlessly supported in solaris
[05:23:32] *** jimgris has joined #opensolaris
[05:23:33] <kAv_> cheatersrealm : marvel based chipset ones
[05:23:40] <twincest> what is the solaris version of SIOCGIFADDR?
[05:23:48] <twincest> apparently it's only there for BSD compatibility
[05:23:52] <movement> delewis: actually there's been some discussion of something similar
[05:24:07] <movement> delewis: of course it would default to /usr/bin/ being the old versions.
[05:24:34] <delewis> movement: and some userland script, like netservices, modifies such behaviour to use the newer ones, instead?
[05:24:42] <gisburn> kAv_: modern is something in /usr/bin which doesn't shoot itself into the feet when the locale uses multibyte characters
[05:24:43] <cheatersrealm> kAv_: care to link me?
[05:24:49] <cheatersrealm> kAv_: I can't find it on the sun hcl
[05:25:08] <delewis> I think better default behaviour would be to use the new stuff in /usr/bin and have the user execute such a script to use the old userland.
[05:25:16] <gisburn> kAv_: most tools in /usr/bin/ are not multibyte capable - with horrible results.
[05:25:24] <delewis> just as is being done with netservices limited being the default
[05:25:41] <delewis> (surely, some crazy customers are expecting 'open' by default, but Sun isn't catering to them)
[05:26:06] <gisburn> kAv_: and then Sun runs around and praises their i18n framework and doens't even set /usr/xpg4/bin in front of the $PATH to make sure that users won't get a bad suprise in such locales
[05:26:11] <kAv_> cheatersrealm : my hint is based on the hardware that sun makes - servers that have sata are either the nvidia based or the marvel based chipsets
[05:26:23] <cheatersrealm> k
[05:26:26] <kAv_> so either one should keep on having support and bug fixes
[05:27:24] *** sahafeez has joined #opensolaris
[05:27:57] *** Gman has joined #opensolaris
[05:28:00] <boyd> cheatersrealm: Apparently there is good support for some cards based on siliconimage chipsets
[05:28:05] * gisburn greets jimgris
[05:28:17] <cheatersrealm> boyd: my si3114 chipset thing
[05:28:20] <cheatersrealm> is awful
[05:28:21] <twincest> hmm, defining BSD_COMPAT also declares a struct map at global scope.. that's fun
[05:28:30] <kAv_> gisburn : sounds like you want to get yourself a project :) (making the binaries multibyte support)
[05:28:33] * boyd shrugs... only saying what I read in zfs-discuss
[05:28:46] <cheatersrealm> boyd: I think I have a bad card
[05:28:55] <jimgris> gisburn: hi
[05:28:59] <cheatersrealm> it's a syba, but I can't get it to work in linux (a supported os)
[05:29:26] <boyd> cheatersrealm: Explain that the behaviour is unacceptable and that it will have to sit in the naughty corner if it continues
[05:29:38] <gisburn> kAv_: I already have my little horror project running.
[05:29:48] <cheatersrealm> boyd: I fully plan on lighting it on fire once I get a new (supported) raid card
[05:29:57] <gisburn> kAv_: http://www.opensolaris.org/os/project/ksh93-integration/
[05:30:07] <kAv_> gisburn : checking
[05:30:36] <kAv_> cheatersrealm : get zfs you dont need raid
[05:30:43] <kAv_> *hardware*
[05:30:48] <boyd> Yeah.
[05:31:06] <cheatersrealm> kAv_: I want to run raid-z and zfs
[05:31:33] <boyd> cheatersrealm: If you find a really good one, feel free to tell me
[05:31:36] * gisburn wishes the inventor of VT52 is dead, burried, consumed by worms and the worms eaten by komodo dragons
[05:31:59] <delewis> just forget about vt52 :-)
[05:32:05] <gisburn> delewis: nah.
[05:32:09] <gisburn> delewis: I got it working.
[05:32:15] <delewis> gisburn: who actually has a vt52 terminal laying around?
[05:32:24] <gisburn> delewis: but I learned that there was even a japanese variant of vt52
[05:32:31] <delewis> interesting
[05:32:39] <boyd> gisburn: So you're saying that my wyse60 won't work since you refuse to use terminfo properly
[05:32:45] <gisburn> delewis: don't ask how that's working
[05:32:56] <gisburn> boyd: no, it will work now
[05:32:59] <jamesd> was the japanese version sent to japan on the  enola gay?
[05:33:09] * delewis was using a VT100 on a Netra the other day doing a Jumpstart
[05:33:20] <gisburn> boyd: I have a "whitelist" for xterm/dtterm/sun/sun-color etc.
[05:33:25] <boyd> I have a vt320 somewhere
[05:33:30] <delewis> you configure terminal settings by setting various bits, quite interesting.
[05:33:35] <gisburn> jamesd: Who is Enola Gay ?
[05:33:43] <delewis> my IBM 3153 infowindow has spoiled me :-)
[05:33:51] <delewis> which I stil say is the best. terminal. ever.
[05:33:56] <boyd> gisburn: I think it was the plane that flew over Hiroshima
[05:34:02] <gisburn> boyd: ouch
[05:34:03] <jamesd> gisburn, its the plane that dropped the atomic bomb on hirsoshima
[05:34:06] <boyd> delewis: Ascii
[05:34:07] <boyd> ?
[05:34:22] <delewis> boyd: yes
[05:34:36] * boyd googles
[05:34:37] <jamesd> okay history lesson done, i'm going to bed...  see you all in 8 hours or so.
[05:34:42] <delewis> http://riemann.solnetworks.net/~dlewis/images/pics/infowindow3153/infowindow3153_1.jpg
[05:34:47] <delewis> http://riemann.solnetworks.net/~dlewis/images/pics/infowindow3153/infowindow3153_2.jpg
[05:35:00] <delewis> what's nice about the 3153 is that you use *any* PS2 keyboard
[05:35:07] <delewis> no proprietary keyboard which usually sucks.
[05:35:18] <boyd> Oh *those*... yeah
[05:35:20] <delewis> *cough* Televideo *cough*
[05:35:26] <delewis> *cough* Wyse *cough*
[05:35:31] <boyd> One of the later vts did that too... 340? 540?
[05:35:32] <delewis> god, I hate Wyse keyboards.
[05:35:52] <delewis> why the hell did they have to put ^[ on the right side of the keyboard.
[05:35:58] <twincest> oh that's fun... you don't even need to define anything, just including <net/if.h> gets you 'struct map;
[05:36:13] <boyd> Actually there is something in the architecture that I like about the 3270 system
[05:36:35] <gisburn> delewis: maybe because the left side has the "erase file" key ? =:-)
[05:36:44] <boyd> (local editing at the terminal - then batch sends to the host)
[05:36:47] <gisburn> or was that erase record ?
[05:36:48] <delewis> well the 3153 can do 3270, SCO term, vt100, vt220, and a few more emulations.
[05:36:54] <boyd> delewis: Nice
[05:37:27] <delewis> gisburn: the ^[ on a Wyse is traditionally beside the "Do" key.
[05:37:30] <gisburn> boyd: ah, then the 3270 was responsible for the problem that today's POSIX shells must not have an editor mode enabled
[05:37:38] <kAv_> cheatersrealm: http://www.opensolaris.org/jive/thread.jspa?threadID=16294&tstart=0
[05:38:03] <delewis> gisburn: not necessarily.
[05:38:05] <boyd> gisburn: I doubt it... I'm not sure they were ever used on UNIX boxes to any great extent
[05:38:12] <delewis> I'm using editting mode fine in 3270 mode
[05:38:29] <delewis> boyd: IBM ships a 3153 with every pSeries setup, almost.
[05:38:41] <delewis> and 3270 is default, IIRC.
[05:39:07] <boyd> Well, I did say *were*. and IBM are... somewhat unique with some of their UNIX offerings
[05:39:34] <delewis> quite unique. :-)
[05:39:51] <delewis> IBM also sells a hardware modem with each pSeries setup, as well. :-)
[05:40:03] <delewis> (dial-out and dial-in for IBM support)
[05:40:28] <boyd> Is there a difference between 3153 and 3153 infowindow?
[05:40:51] <delewis> I was using it loosely. My terminal is a 3153 InfoWindow, so I think they're the same.
[05:41:06] <delewis> actually, mine is an InfoWindow II
[05:41:07] <boyd> just wondering
[05:41:08] <KingDillyDilly> Are any of those dudes with the stars next to their names (jamesd, sparc-kly, and Tpenta) Sun employees?
[05:41:09] <delewis> interesting.
[05:41:18] <delewis> I've never been able to find a lot of documentation on it, though.
[05:41:31] <boyd> KingDillyDilly: They have red things on my display :)
[05:41:40] <boyd> KingDillyDilly: Tpenta is
[05:41:53] <KingDillyDilly> Green stars in Gaim.
[05:44:26] *** nbkk6fo_ has quit IRC
[05:44:41] *** nbkk6fo_ has joined #opensolaris
[05:44:56] <gisburn> Ok, vt52, vt400, vt320 work fine.
[05:45:18] <gisburn> and "sun-color" and "sun", too.
[05:45:38] <gisburn> and it works now over a 9600 baud line, too.
[05:45:50] <gisburn> (monocrome, color needs 14400)
[05:46:27] <gisburn> KingDillyDilly: not all people with green stars are sun employees.
[05:46:34] *** gm152 has quit IRC
[05:47:06] <KingDillyDilly> Isn't there a Linux that runs on Sparc? Shouldn't the Sun's Sparc workstations include it?
[05:47:18] <boyd> Yes. No.
[05:47:23] <delewis> KingDillyDilly: Linux on SPARC isn't exactly optimal.
[05:47:47] <gisburn> delewis: well, it's getting better now.
[05:47:50] <delewis> so Sun isn't going out of it's way to support it (though, there was something about Ubuntu a few months back)
[05:47:54] <delewis> gisburn: no, it isn't
[05:47:58] <delewis> gcc code generation still sucks
[05:48:12] <gisburn> delewis: the idiot bugs on my Ultra5 are gone... :-)
[05:48:17] <delewis> and until Sun ports Sun Studio to Linux/SPARC (I don't see this happening), it still won't be optimal.
[05:49:37] *** Aghaster has joined #opensolaris
[05:50:29] <gisburn> WTF is wrong with polaris.genunix.org ?
[05:51:11] <Aghaster> dunno, you are interested in polaris?
[05:51:28] <Aghaster> maybe go in ##mac and ask for Error_404, he's a Polaris developer
[05:51:49] <Error_404> huh?
[05:51:50] <gisburn> Aghaster: erm, I am using the polaris.genunix.org "trac" source browser for the ksh93-integration project
[05:52:12] <Aghaster> ah, well, you're here.
[05:52:18] <Error_404> i've hacked at it a bit trying to figgure it out, sure
[05:52:38] <Aghaster> that's still quite impressive for me
[05:52:41] <Error_404> but i have nothing to do with the repo, and none of my code is in it
[05:53:39] <Aghaster> Error_404, solaris 32bit drivers don't run on 64bit solaris, right?
[05:53:52] <delewis> Aghaster: correct.
[05:53:59] <Aghaster> I think my ethernet adapter only has 32bit drivers...
[05:54:50] <gisburn> Aghaster: sparc or x86 ?
[05:54:53] <Aghaster> it is a yukon marvell gigabit
[05:54:54] <Aghaster> x86
[05:55:07] <boyd> Aghaster: No, there are 64bit ones for that
[05:55:08] <Aghaster> this is an AMD64 system
[05:55:13] <Aghaster> link?
[05:55:16] <gisburn> Aghaster: AFAIK all drivers have been ported to 64bit, too.
[05:55:19] <boyd> Look at syskonnect.de
[05:55:22] <boyd> IIRC
[05:56:38] <Aghaster> hum... looks good, but I have to find my exact model
[05:56:57] <KingDillyDilly> I'm deleting directories containing 300,000 files each and it takes like 30 minutes or more through XP's Windows Explorer. Think it would be faster to delete XFS directories than NTFS ones?
[05:57:19] <KingDillyDilly> (with SOlaris)
[05:57:25] <boyd> Aghaster: Perhaps not... I had to edit my /etc/driver_aliases file
[05:57:36] <twincest> king: XFS?
[05:57:46] <KingDillyDilly> I mean ZFDS
[05:57:51] <KingDillyDilly> I mean ZFS
[05:58:09] <Aghaster> boyd: hum... what do you mean?
[05:58:13] <jbk> could always try it
[05:58:21] <boyd> KingDillyDilly: you mean over the net?
[05:58:33] <KingDillyDilly> No. Locally. No network.
[05:59:21] <boyd> Aghaster: http://blogs.sun.com/jmcp/entry/how_do_i_find_out
[06:00:09] <boyd> Aghaster: The skge driver (the syskonnect one) has inbuild entries in driver_aliases for many nics, but you may have to add new ones to map your nic to the skge driver
[06:00:18] <boyd> Ok, I'm outta here all
[06:00:23] <boyd> Byeeeeeee
[06:00:30] <KingDillyDilly> Byyyyyyyyyye
[06:00:40] * boyd goes offline for the weekend.
[06:01:12] <Aghaster> boyd: thanks for the link!
[06:01:48] <KingDillyDilly> I read "About the only industry left that [Solaris] really has control of is the telecom  industry and software like Asterisk may very well change that finally." Discuss.
[06:02:51] <delewis> what is there to discuss? that's an over-generalization perpetuated by some luser.
[06:03:03] <delewis> you should not waste your time reading such inaccuracies.
[06:03:20] *** bobbyz has joined #opensolaris
[06:03:45] <delewis> and regardless, Solaris was never used in the way that Asterisk is being used nowadays.
[06:03:54] <delewis> most telecoms used Solaris for a dial-in server.
[06:03:58] *** Aghaster has quit IRC
[06:04:18] <delewis> which leads me to conclude that only does the author of that have no clue have Solaris' usages, but the telecom industry, in general, which he claims to know about.
[06:04:25] <twincest> discuss what?  i thought pointless arguments over things like that were restricted to linux channels..
[06:05:55] <KingDillyDilly> Well, I can't concentrate on anything more important with my directories being deleted.
[06:06:32] <KingDillyDilly> I have enough file operations going on without trying to start more.
[06:09:50] *** crash| has joined #opensolaris
[06:10:06] <KingDillyDilly> The quote is from a system's administrator of one of the best web hosting services who has a "heavy Solaris background."
[06:10:51] <delewis> I'd question anyone's mentality who compares an operating system to a userland application, like Asterisk.
[06:11:09] <jbk> and why couldn't it be ported to solaris?
[06:11:13] <jbk> i bet it'd probably run better :)
[06:11:14] <delewis> jbk: precisely.
[06:11:59] <jbk> have to separate the hype from reality
[06:12:36] *** cheatersrealm has left #opensolaris
[06:12:37] <jbk> one of my favorite examples was when oracle was touting 'convert everything to oracle rac on linux on dell' and had this big fancy datacenter they were giving executives tours of to convince them
[06:12:58] <jbk> then what do they do when the need to replace their ERP system?
[06:13:06] <jbk> they run it on some 25ks running solaris
[06:13:17] <Tpenta> actually II thnk they are 12s
[06:13:21] <delewis> I guess elegant datacenters somehow implies better operability and functionality.
[06:13:26] <Tpenta> well they were a few years back
[06:13:37] <delewis> [sic]
[06:13:44] <jbk> but my question was 'why didn't they move it to oracle rac on linux?'
[06:14:24] <delewis> (a) Oracle RAC sucks (b) Linux sucks. sucks + sucks = more suckage.
[06:15:14] <jbk> if we can figure out how to cleanly shut down rac, we might (finally) have figured out everything we need to do to stabalize rac
[06:16:19] <jbk> right now, we can't shut down the cluser without cssd crapping out and rebooting stuff
[06:16:33] <delewis> most of the shops I've seen use VCS or HACMP to HA, rather than RAC in that sense.
[06:16:43] <jbk> this is with vcs
[06:17:18] <jbk> the symantec provided stuff for managing css doesn't seem to work correctly, or we somehow have it configured wrong
[06:18:36] <jbk> but i'm really unimpressed with rac, and think places should look hard at their availability requirements before deploying it
[06:18:40] * jmcp returns
[06:18:43] <jbk> but that's just me
[06:19:05] <jmcp> jbk: not just their availability requirements, but also their application requirements - not all apps are suitable to run with RAC
[06:19:19] <jmcp> of course, your standard salesweasel won't admit that ......
[06:19:27] <jbk> of course
[06:19:50] <jbk> i'm still not sure how they're plannign to use rac with this project, or how well it'll actually handle a node failure
[06:20:01] <jbk> but not my decision
[06:20:05] <twincest> what is the procedure for adding const correctness to old interfaces?  can it be done at all without breaking compatibility?
[06:23:44] <timeless> C?
[06:24:32] <timeless> presumably adding const to returns isn't safe
[06:25:39] <timeless> but for arguments, as long as the calling convention is C, the const isn't in the calling convention, so the method ABI isn't changed
[06:26:11] <timeless> (but i am not a solaris hacker)
[06:29:34] * gisburn wishes he had a better font for http://www.opensolaris.org/os/project/ksh93-integration/screenshots/gnaw_v0_8.png
[06:31:50] <timeless> wow
[06:32:11] <timeless> i think that the center is wrong
[06:32:22] <timeless> it should only  have one open slot at the top, not 3
[06:32:30] <timeless> and it should rarely be open
[06:33:10] <timeless> and can't you use a bullet instead of a period? :)
[06:34:31] <KingDillyDilly> I think a middot would keep it the same size and it would be centered.
[06:35:19] <gisburn> timeless: you mean the monster cave ?
[06:35:24] <timeless> yes
[06:35:56] <gisburn> timeless: it won't work. I am not using the original algoritm and these monsters won't get out fast enougth then.
[06:36:04] <timeless> ah
[06:36:38] <KingDillyDilly> Isn't this just Pacman?
[06:37:39] <gisburn> timeless: the current algoritm is just a follow immediately if there is a straight-line connection between player and monster which starts the homing in that direction. hitting a wall turns homing off. the rest is random movement.
[06:37:47] <gisburn> timeless: primitive but effective.
[06:38:04] <gisburn> timeless: you're quickly cursing the monsters... :-)
[06:38:41] <gisburn> timeless: I had a better one which remembers the turns of a player, up to five times.
[06:38:59] <gisburn> timeless: but the result was devastating since you couldn't get rid of the monsters anymore.
[06:39:51] <gisburn> # Note that this script has been written with the main idea to show
[06:39:53] <gisburn> # many of ksh93's new features (comparing to ksh88/bash) and not
[06:39:54] <gisburn> # as an example of efficient&&clean script code.
[06:40:53] <gisburn> the game fits on 8 DIN-A4 pages and half of the size are the level maps.
[06:41:05] <gisburn> (and comments)
[06:41:22] * timeless frowns
[06:41:27] *** Gman has quit IRC
[06:41:39] <timeless> moving around 8g of data across newly created volumes is really slow
[06:44:41] *** nbkk6fo_ has quit IRC
[06:44:45] *** nbkk6fo__ has joined #opensolaris
[06:53:44] *** sahafeez has left #opensolaris
[06:55:56] <timeless> doh
[06:56:07] <timeless> zfs volumes can't contain +s :)
[06:56:09] *** PosixCompliantNe has quit IRC
[06:58:15] *** nwf has joined #opensolaris
[06:59:08] *** jimgris has left #opensolaris
[07:01:02] *** yongsun has joined #opensolaris
[07:09:19] *** bougie has joined #opensolaris
[07:09:34] *** gisburn has quit IRC
[07:16:24] *** Burana has joined #opensolaris
[07:21:27] <KingDillyDilly> What's with all the "general errors" when I click a company in the results for http://partneradvantage.sun.com/catalog/search/PartnerAdvancedSearch.jsf;jsessionid=189993011c71614d9b9a88e181d11:K98s
[07:21:31] *** Gman has joined #opensolaris
[07:21:41] <Gman> jmcp, ping
[07:21:50] *** Burana has quit IRC
[07:21:51] *** Burana has joined #opensolaris
[07:22:00] <jmcp> Gman: hiya
[07:22:06] <Gman> jmcp, got your mail
[07:22:15] <Gman> i don't have a sun extension given that i work from home
[07:22:28] <Gman> and i purposely don't forward my irish extension to that number given timezone and costs
[07:22:31] <jmcp> you can't redirect a number to your preferred handset?
[07:22:36] *** ShadowHntr has joined #opensolaris
[07:22:39] <jmcp> well yes :)
[07:23:04] <jmcp> btw, Rob was following Havoc's suggestions in http://www.mail-archive.com/gconf-list at gnome dot org/msg00127.html but without success
[07:23:04] <Gman> my number for nz is in namefinder though
[07:23:09] * Gman looks
[07:23:11] * jmcp nods
[07:23:54] <Gman> what was the exact problem?
[07:24:06] <jmcp> I'll forward it to you
[07:24:09] <Gman> ok
[07:25:41] *** karrotx_ has joined #opensolaris
[07:26:06] <timeless> gman: did you get snv_53 happy?
[07:26:17] <Gman> timeless, mostly, still having issues with i915 though
[07:26:45] <timeless> > simple question. run gnome-system-monitor. click the devices tab. click a device. wait 2s. is your selection still present? (control: repeat for process tab)
[07:29:04] <Gman> hrm, yeah, still ok
[07:29:21] <Gman> though i have < 10
[07:29:46] <timeless> i think i have around 300 at this point
[07:29:59] <timeless> ok, so it's fixed in 2.16?
[07:30:18] <Gman> well, it doesn't happen to me in 2.16
[07:30:38] <timeless> it happens for me in 2.14 and for a linux user in 2.12
[07:30:49] <timeless> but if it's fixed in 2.16, then that's another reason for me to upgrade
[07:31:01] * timeless closes the enter_bug window
[07:33:46] *** dj2 has quit IRC
[07:38:19] *** Burana has quit IRC
[07:39:28] <twincest> how do i make vim's :set sm balance <> as well as ()?
[07:40:26] <KingDillyDilly> Do they make editors for Solaris that are more graphical than Vim?
[07:41:04] <KingDillyDilly> Is there a gVim for Solaris like there is for Windows?
[07:42:30] <twincest> gvim, nedit, gedit, kate, (x)emacs, ...
[07:42:59] <KingDillyDilly> Sounds like enough.
[07:43:18] <KingDillyDilly> But I use Wordpad so I shouldn't complain anyway.
[07:43:57] <timeless> gedit, ...
[07:44:28] *** nbkk6fo__ has quit IRC
[07:44:41] *** nbkk6fo__ has joined #opensolaris
[07:46:06] <KingDillyDilly> (x)emacs wasn't around when I last looked for an editor, so I downloaded gvim. Used it for a few days and went back to Wordpad, expecting to return to gvim to learn more but never did.
[07:46:20] <twincest> you last looked for an editor in the 1970's, then?
[07:46:53] <timeless> xemacs isn't in my gnome start menu
[07:47:02] * timeless can never find anything anywhere  else
[07:47:19] <KingDillyDilly> No. Was there an XEmacs then? With an "emphasis is on modern graphical user interface support"?
[07:48:32] <timeless> /opt/SUNWspro/contrib/xemacs-21.4.12
[07:49:12] <twincest> i don't believe that's ever been emacs' or xemacs' emphasis
[07:49:36] <timeless> it has buttons, and a tab
[07:50:07] <timeless> and a hyperlink and an animated start screen w/ a picture
[07:50:32] <timeless> and it takes a while to load/unload since it's loading emacsos :)
[07:51:15] <KingDillyDilly> I like GUIs. I was considering FreeBSD until I saw the special version with a GUI that some guy created. If the real thing doesn't have a GUI, I look elsewhere.
[07:51:30] *** gm152 has joined #opensolaris
[07:51:41] <timeless> you can get kde/gnome for just about anything
[07:51:42] <twincest> you think freebsd doesn't have a gui?
[07:52:23] <KingDillyDilly> There was a need for a version with a better GUI, so I guess the standard one isn't so good.
[07:53:10] *** ShadowHntr has quit IRC
[07:54:00] * KingDillyDilly looks up what he's thinking of
[07:54:11] *** bougie has quit IRC
[07:56:40] *** fik has joined #opensolaris
[07:57:03] *** vaneth has quit IRC
[08:19:05] *** tsoome has quit IRC
[08:19:56] <KingDillyDilly> There's a PCBSD, but I was thinking of http://www.desktopbsd.net/. This stuff should come with the operating system and not have to be sought out in third party software.
[08:21:56] *** vaneth has joined #opensolaris
[08:25:44] <KingDillyDilly> But I don't care whether Solaris has a better GUI or not because I want ZFS.
[08:26:53] <KingDillyDilly> I've been avoiding learning a database for a long time, and maybe with ZFS I can go even longer.
[08:27:13] * timeless isn't really sure how zfs relates to databases
[08:27:18] <timeless> zfs isn't exactly befs
[08:27:26] <twincest> time: let me introduce you to a guy named Mr. Reiser...
[08:27:37] <KingDillyDilly> I know. The murderer.
[08:27:50] <KingDillyDilly> I don't want his buggy file system.
[08:28:56] <KingDillyDilly> I have millions of small files to save and they waste 90% of my HD space with a 4K block size on NTFS.
[08:29:10] <timeless> you aren't using compression w/ NTFS5?
[08:30:29] <KingDillyDilly> No. My backup software does and had no problem. I tried saving without the backup software and without compressing and it's impossible. It crashes XP.
[08:31:20] <KingDillyDilly> Pool file too small on XP, or something, and no way around it. I tried every registry hack I found.
[08:32:25] *** tsoome has joined #opensolaris
[08:34:25] <KingDillyDilly> I wanted to move the files to my USB hard dive and work with them in plain text. There was no room on my internal drive.
[08:40:16] *** sgr1 has joined #opensolaris
[08:40:57] <sgr1> is there a file for avoid the LD_LIBRARY_PATH variable? something like /etc/ld.so.conf in linux but in solaris?
[08:41:15] <asyd> man crle ?
[08:41:39] <razrX> yup, crle is your friend
[08:41:51] <sgr1> oh thanx guys let me see :)
[08:44:27] *** nbkk6fo__ has quit IRC
[08:44:32] *** nbkk6fo__ has joined #opensolaris
[08:47:28] *** gm152 has quit IRC
[08:50:04] *** mazon is now known as Mazon
[08:52:18] <sgr1> and just another simple question.. where can I set the global locales valid for all the users (LC_ALL, LANG, etc...)?
[08:52:28] *** fik has quit IRC
[08:52:47] <jmcp> sgr1: in /etc/default/init
[08:53:24] <sgr1> jmcp: thanx :)
[08:53:51] *** vaneth has quit IRC
[08:55:13] <jmcp> yw
[08:56:47] <Tpenta> jmcp, seen the cricket score at stumps?
[08:59:01] <jmcp> Tpenta: nope, been a bit busy
[08:59:21] <Tpenta> Aus 9/602 decl, Eng 3/53 at stumps
[08:59:27] <jmcp> Tpenta: have you had a chance to think about my psarc questions?
[08:59:30] <jmcp> wtf?!?!?!?!
[08:59:39] <jmcp> did the English bother to turn up?
[08:59:44] <Tpenta> i am going to have to talk to some others about them i think
[08:59:53] <Tpenta> :-D
[09:00:12] <Tpenta> clarke (the bowler) came in in teh tail and hitt wo consecutive sixes
[09:00:24] *** yakov has joined #opensolaris
[09:00:29] <yakov> hi
[09:00:50] <jmcp> ta much
[09:00:51] * jmcp nods
[09:01:02] <jmcp> hi yakov
[09:01:38] <yakov> i'v compiled Qt with SunONE CC compiler and now i cant't compile sample app with g++ and those libraries (made with CC). could you pls explain me - is this behavior correct?
[09:01:52] <yakov> don't ask why i'm compiling with g++...
[09:02:17] <Error_404> Qt is picky about compilers
[09:02:56] <Error_404> you can't even compile Qt with GCC 3.4 and expect something compiled with GCC 4.1 to work
[09:03:11] *** trondn has joined #OpenSolaris
[09:03:20] <jmcp> yakov: the C++ ABI isn't very stable as far as I am aware
[09:03:41] <yakov> guys thank alot!
[09:04:13] <yakov> i thought 'bout ABI.. i hoped that it could possibly compile :)
[09:06:36] <yakov> one more Q about SunONE. I'm using  version 8 with solaris10 on x86. is it reasonable to upgrade to SunONE 10?
[09:06:51] <jmcp> yeeek!
[09:07:02] <jmcp> get yourself a copy of Studio 11 and do yourself a favour
[09:08:43] *** miffe has quit IRC
[09:13:42] <Error_404> holy hell my power bill was stupid this month
[09:13:54] <Error_404> $95 for a small 2 bedroom apartment
[09:20:36] <twincest> morning jm
[09:20:54] <twincest> there is actually a (fairly) stable C++ ABI now
[09:21:14] <twincest> but legacy compilers can't implement it easily because it obviously changes the existing ABI
[09:21:23] <twincest> (gcc just said, 'the abi changed, sorry')
[09:23:17] <yakov> brb
[09:23:22] *** yakov has quit IRC
[09:23:53] <jmcp> twincest: hi
[09:24:02] <jmcp> twincest: typical gcc ....
[09:24:29] *** dunc has quit IRC
[09:24:44] * Tpenta kicks off the builds for the b53 non-debug encumbered bins (sorry i',m getting behind folks)
[09:26:14] *** yakov has joined #opensolaris
[09:27:51] *** Gman is now known as GmanAFK
[09:28:16] *** simfordWFH has joined #OpenSolaris
[09:33:26] *** simford has quit IRC
[09:40:41] *** KingDillyDilly has left #opensolaris
[09:43:37] <Doc> plus you never did send me that email...
[09:44:26] *** nbkk6fo__ has quit IRC
[09:44:41] *** nbkk6fo__ has joined #opensolaris
[09:49:39] *** dunc has joined #opensolaris
[09:49:43] *** kleppari has joined #opensolaris
[10:09:20] *** dlynes_laptop has joined #opensolaris
[10:10:06] *** alanc_away has quit IRC
[10:10:07] *** GmanAFK has quit IRC
[10:10:52] *** alanc_away has joined #opensolaris
[10:11:00] *** GmanAFK has joined #opensolaris
[10:16:42] *** Tpenta has quit IRC
[10:17:36] * timeless sighs
[10:18:03] <timeless> well, i can now move around my huge zfs world using my 2g quota space :), too bad i keep changing my mind about where i want it to go
[10:20:20] <yakov> guys has anyone dual-head OWconfig?
[10:25:52] *** ly___schi has joined #opensolaris
[10:28:14] *** yongsun has quit IRC
[10:29:03] <Berny> morning
[10:30:33] <onbot> commit by George Shepherd: 6433292 Apparent dblk refrence counting issues in TCP/IP
[10:32:35] <Doc> join #solaris
[10:32:42] <Doc> turning it on it trivial - edit one file and reboot
[10:32:43] <Doc> blah
[10:34:00] <jmcp> /me heads
[10:34:24] *** y___schil has quit IRC
[10:34:41] *** jmcp has quit IRC
[10:34:58] *** whaq has joined #opensolaris
[10:35:03] *** Barry has joined #opensolaris
[10:35:08] <whaq> hi ho
[10:35:54] <Berny> .oO(what's the difference between a terrorist and a sysop?)
[10:35:57] *** Burana has joined #opensolaris
[10:36:33] <Berny> .oO(with a terrorist it's possible to negotiate...)
[10:37:14] <whaq>  A sysop doesn't have an agenda.. he has many notes in .txt files.
[10:38:09] <timeless> are there graphical tools that can show me where my zfs space is going?
[10:38:17] <timeless> i feel like i've lost a bunch of pieces
[10:38:48] <Berny> how do you loose pieces of a fs?
[10:39:01] <Barry> Is SPARC pronounced "sparse" or "spark"?
[10:39:03] <whaq> Has anyone had a horror story w/ ZFS?
[10:39:30] <Barry> I guess knowing what the "c" stands for would help.
[10:39:54] * Berny is happily running zfs
[10:39:55] *** ssam has joined #opensolaris
[10:40:22] *** ssam has left #opensolaris
[10:40:33] <Barry> Looked it up. Guess it's spark.
[10:40:49] <delewis> Barry: (S)cable (C)computer (ARC)hitecture.
[10:40:54] <delewis> and yes, it's spar*k*
[10:41:09] <delewis> oh, I forgot the (P)rocessor
[10:41:14] <whaq> Yea, that's how I've been pronouncing it..
[10:41:16] <delewis> that's what staying up till 4AM gets you.
[10:41:19] <whaq> Scalable?
[10:41:28] <timeless> the only problem i have w/ zfs is that if you are silly as i was and overdesign your directory hierarchy and then run out of space on your zpool, it's a pain to fix it
[10:41:31] <delewis> whaq: of course.
[10:41:40] <delewis> that's what SPARC was designed to be -- scalable
[10:41:43] <delewis> in a number of ways
[10:41:50] <delewis> register windows comes to mind
[10:42:00] <Doc> the fact you can get a 144 core machine should prove that...
[10:42:01] <whaq> timeless, what did you mean by overdesign?
[10:42:04] <delewis> a processor vendor can just throw more GP registers onto a SPARC processor
[10:42:11] <delewis> and the compiler does not need to be re-compiled
[10:42:21] <whaq> delewis, right.. I was wonedering what scable meant.
[10:42:44] *** ssam has joined #opensolaris
[10:42:57] *** ssam has left #opensolaris
[10:44:27] *** nbkk6fo__ has quit IRC
[10:44:32] *** nbkk6fo__ has joined #opensolaris
[10:44:33] *** ssam has joined #opensolaris
[10:45:21] *** ssam has joined #opensolaris
[10:45:29] <tomww> scalable doesn't stop at the processor level. the rest of the machine is scalable too.
[10:45:53] <delewis> yes, but that's not the scalability that the designers of SPARC had in mind.
[10:45:54] *** jcea has quit IRC
[10:45:57] <tomww> remember the guy running a 3.5gig gzip for half an hour. parallel bzip did it in 26 seconds
[10:46:04] <delewis> they were more focused on being able to apply the processor to a wide-range of systems.
[10:46:05] <tomww> right
[10:46:12] <delewis> with excellent application compatibility.
[10:46:14] <sgr1> I get a c1::dsk/c1t0d0 when I do a cfgadm -al | grep CD, but when I try to do a mount  -o ro /dev/dsk/c1t0d0s* /cdrom it doesnt works. Any suggestion?
[10:46:38] <tomww> sgr1: volume-manager running?
[10:46:41] *** ssam has left #opensolaris
[10:47:15] <sgr1> tomww: sure ps -ef | grep volm report that it's working
[10:47:46] <sgr1> I mean vold
[10:47:53] <tomww> :-)
[10:48:21] <raph_ael> hello
[10:48:29] <tomww> maybe you have an elephant sitting on your cdrom-device (means: the device is open/locked)
[10:48:41] *** ssam has joined #opensolaris
[10:48:44] <tomww> which OS-release is running?
[10:48:46] *** ssam has left #opensolaris
[10:48:55] <sgr1> tomww: solaris 10
[10:49:22] <tomww> and the vold did not monut the cdrom automatically?
[10:49:39] <sgr1> tomww: no :(
[10:49:42] * delewis wonders when the 64-bit Perl is coming
[10:49:46] <delewis> now that we have a 64-bit Python
[10:49:46] *** yakov has quit IRC
[10:49:57] <tomww> hmm. you might stop vold and retry your manual mount
[10:50:03] <tomww> svcs -av | grep vol
[10:50:26] <sgr1> svcs -av | grep vol reports nothing ?!?!?
[10:50:36] <tomww> svcadm disable <stingforvolummgt>
[10:50:53] *** ssam has joined #opensolaris
[10:50:57] *** ssam has left #opensolaris
[10:51:14] <sgr1> svcs  | grep vol reports nothgin ??
[10:51:41] <tomww> svcs | grep mile
[10:51:42] <tsoome> then it's not running and therefore not mounting cd's ....
[10:53:05] <tomww> i qould sav, you have failed services, and the needed milestone are not reached.
[10:53:27] <sgr1> tomww: and how can change the state of these services?
[10:54:06] <tomww> 1. look at the problems 2. clear maintenance after removing the problems
[10:54:11] <tomww> if you give svcs -x
[10:54:25] <tomww> this gives your services needing your eyes
[10:54:54] *** susisa has joined #opensolaris
[10:55:12] *** lopa has joined #opensolaris
[10:55:29] *** simfordWFH has quit IRC
[10:55:50] <onbot> commit by Gavin Maltby: 6474853 Solaris should be cognizant of the platform while attempting to turn off smi mc polling; 6489753 mc_amd.h not quite ready for revision G; 6491720 DC/IC/BU system infill only capture physical <39:6>; 6495485 i86pc/cpu/scripts do not get clobbered
[10:56:00] *** Doc has quit IRC
[10:56:09] *** Dar_HOME is now known as Dar
[10:56:49] <sgr1> svcs -x reports nothing
[10:56:51] *** Doc has joined #opensolaris
[10:57:33] *** sgr1 is now known as sgnut
[11:00:49] *** lopa has left #opensolaris
[11:05:07] <Barry> Delewis: Ask in #perl6 and tell them Wassercrats sent you.
[11:05:29] <timeless> how do i ask zfs for a list of unmounted filesystems?
[11:07:08] *** nwf has quit IRC
[11:17:01] * timeless settles for zfs get mounted|grep no
[11:22:27] <lasseoe> get mounted=no ?
[11:23:09] <Berny> invalid property 'mounted=no'
[11:23:35] <lasseoe> damn, that would have made sense :)
[11:23:54] <Berny> indeed
[11:23:54] <lasseoe> actually.. a RFE should be made, imho
[11:24:09] <Berny> but not everything in black and whote makes sense
[11:24:16] <lasseoe> nope
[11:24:23] <Berny> .oO(damn i should have some guinness these days)
[11:24:35] <lasseoe> :)
[11:24:47] <timeless> that'd be too much like a database langauge :)
[11:25:18] <timeless> anyway, i now have everything mounted, so i'm happy
[11:25:38] *** axisys has quit IRC
[11:28:50] *** Burana has quit IRC
[11:28:57] *** yakov has joined #opensolaris
[11:30:31] *** frankho has joined #opensolaris
[11:32:45] *** frankho has quit IRC
[11:33:00] *** frankho has joined #opensolaris
[11:41:03] *** susisa has left #opensolaris
[11:44:38] *** freakazoid0223 has quit IRC
[11:48:28] *** susisa has joined #opensolaris
[11:48:33] *** susisa has left #opensolaris
[11:55:04] *** silk has joined #opensolaris
[11:55:31] *** silk has quit IRC
[11:56:19] *** silk has joined #opensolaris
[12:09:08] *** yakov has quit IRC
[12:09:37] <lasseoe> se noai
[12:09:37] <lasseoe> apse ai
[12:10:15] <lasseoe> heh, oops
[12:15:41] <raph_ael> strange language :)
[12:15:57] <trondn> C# ?
[12:16:34] <Kaktus`> :)
[12:16:51] <lasseoe> it's a bug in screen :)
[12:17:18] <lasseoe> when switching windows, and I hit the wrong key combo at the wrong time, then it outputs that
[12:17:23] <quasi> yeah, riiiigt ;)
[12:17:34] <lasseoe> no shit :)
[12:19:37] *** jmcp has joined #opensolaris
[12:23:39] *** karrotx_ has quit IRC
[12:27:44] *** bougie has joined #opensolaris
[12:30:54] * jmcp achieves an open bottle
[12:30:59] <jmcp> for some reason there isn't a corkscrew in this hotel room :(
[12:31:20] <lasseoe> frontdesk is likely to have one
[12:31:53] <lasseoe> with a little charm it's amazing what you can borrow from them
[12:32:13] <jmcp> lasseoe: I called housekeeping, who sent one up ... and once I'd opened the bottle promptly took it back!
[12:32:16] <quasi> jmcp: or grab your sword and see if you can cut the top af the bottle cleanly ;)
[12:32:27] <jmcp> quasi: ah yes ....
[12:32:41] <jmcp> but I'd prefer to avoid the chance of shards of glass in my throat afterwards
[12:32:44] <lasseoe> jmcp: problem solved :)
[12:32:47] <Berny> or get some trained chinese guy to do it ;-)
[12:33:17] <jmcp> lasseoe: for the moment ..... if I acquire a second bottle I'll have to call again
[12:33:19] <jmcp> or go shopping
[12:33:36] <jmcp> Berny: true, true .... but I do like opening my own bottles of vino
[12:33:45] <lasseoe> I'd ask frontdesk/reception, they are usually car nicer
[12:33:48] <jmcp> heh ... saw Limoncello in a shopping centre this evening
[12:33:48] <lasseoe> car=far
[12:33:53] <jmcp> lasseoe: nod
[12:34:12] <jmcp> 98RMB for 200ml though .. a little pricey given the cost of living
[12:34:14] * jmcp shrugs
[12:34:33] <Berny> whats rmb?
[12:34:34] *** Snake007uk has joined #opensolaris
[12:34:45] <lasseoe> rumba?
[12:34:49] <silk> currency
[12:34:50] <jmcp> RenMinBi, the People's Money
[12:34:56] <jmcp> formal name for the Chinese Yuan
[12:35:01] <jmcp> which on the street is the Kuai
[12:35:02] <silk> jmcp, you're in China?
[12:35:06] <jmcp> silk: at the moment
[12:35:11] <jmcp> got another week here in Beijing
[12:35:16] <silk> aah, cool
[12:35:21] <jmcp> yes, very
[12:35:44] <silk> heh, was in Guangzhou in winter once, was chilly
[12:36:07] * jmcp nods
[12:36:40] <jmcp> hm forecast is for rain overnight, and 0-5C http://www.chinadaily.com.cn/weather/en/54511_C_e.html
[12:36:43] * jmcp sobs
[12:37:01] <silk> heh
[12:37:04] <Berny> just like here...
[12:37:10] * Berny wants summer
[12:37:19] * sgnut wants holidays
[12:37:26] <Berny> that as well
[12:37:29] <jmcp> Australia is the land of summer ....
[12:37:42] * dunc is off for 2 weeks after today
[12:37:45] * dunc looks smug
[12:37:54] <sgnut> jmcp: spain also :P
[12:38:13] <Berny> the carribean...
[12:38:34] <Berny> rum, nice girls and all that stuff...
[12:38:43] <jmcp> sgnut: I haven't been to Spain yet .... I will get there one day
[12:39:04] <jmcp> sgnut: last friday night I went to a wine appreciation dinner here ... in a Spanish restaurant
[12:39:06] <sgnut> jmcp: nice weather, nice people, nice girls :P
[12:39:25] <jmcp> excellent!
[12:39:26] <sgnut> jmcp: oh and nice food :)
[12:39:30] <jmcp> even better
[12:39:33] <sgnut> lool
[12:39:43] <jmcp> I think I should develop an appreciation of Spanish wines, too
[12:39:52] <sgnut> yeah
[12:40:00] <sgnut> but I preffer coke loool
[12:40:24] <jmcp> sgnut: sacrilege!
[12:40:25] <jmcp> :)
[12:40:26] <sgnut> loool
[12:46:08] <trygvis> anyon remember what kind of memory that fits in a sun blade 1500?
[12:47:44] * jmcp heads out to get dinner
[12:47:45] <jmcp> back later
[12:47:49] <razrX> sgnut: jmcp: and if you want a little piece of home, lots of Starbucks in Barcelona
[12:47:54] <razrX> about 10 of 'em ;)
[12:48:15] <jmcp> razrX: fortunately, I'm not an American so I know what real coffee tastes like
[12:48:21] <jmcp> :-)
[12:48:24] *** calumb has joined #opensolaris
[12:48:27] <sgnut> razrX: loooooooool sure
[12:48:32] <razrX> jmcp: lol
[12:48:44] <razrX> been to barcelona 2 years back
[12:48:49] <razrX> beautiful city
[12:49:03] <jmcp> I'm from Australia, where we've had the benefit of real European styles since at least 1945
[12:49:07] <jmcp> ok, really dinner time now
[12:49:47] <razrX> enjoy then
[12:50:02] <razrX> just returned from lunch myself
[12:50:18] <razrX> i'm from the netherlands
[12:52:49] <bougie> hello :)
[12:53:04] <quasi> oh, drat - is it lunch time here in .nl? I should probably go find some lunch then
[12:53:44] <razrX> well, you enjoy it when you find some okay? ;)
[12:57:07] *** mlh_ has joined #opensolaris
[12:57:24] <quasi> razrX: this is .nl - even mcd would be better than the company provided white bread with hagelslag ;)
[12:58:33] <Berny> *g*
[12:58:55] <Berny> sounds like the "mensa" in out hospital...
[12:59:13] <Berny> food always looks like the leftovers of some kind of nasty surgery
[12:59:50] <lasseoe> Dutch lunches leave a lot to be desired :)
[13:02:08] *** uebayasi has quit IRC
[13:02:18] *** reg1 has joined #opensolaris
[13:04:29] *** mlh has quit IRC
[13:05:48] <quasi> indeed
[13:14:20] *** trondn has left #OpenSolaris
[13:15:15] <razrX> quasi: i prefer lunch at a Subway restaurant over a 'boterham met hagelslag' anyday ;)
[13:15:21] *** crash| has quit IRC
[13:17:16] <lasseoe> razrX: got Quiznos there yet?
[13:17:32] <razrX> lasseoe: don't know what Quiznos is
[13:17:45] <quasi> razrX: La Place and their cooked to order food is pretty decent
[13:17:48] <lasseoe> razrX: similar to subway, only better
[13:17:53] <razrX> quasi: true
[13:17:57] <razrX> it's decent
[13:18:22] <razrX> lasseoe: hmm, better then subway? (i like my sandwhiches large and fresh so ...)
[13:18:22] *** |denis| has joined #opensolaris
[13:18:44] <lasseoe> razrX: same at Quiznos, they toast it slightly and it just tastes better
[13:19:11] <razrX> lol, sounds like a bk thing
[13:19:18] <lasseoe> anyway, if there isn't a franchise there anyway, it's not like it matters :)
[13:19:27] *** linma` has joined #opensolaris
[13:20:24] <razrX> but i agree with the fact that dutch cuisine isn't actually cuisine at all
[13:20:51] <razrX> i always tend to look for american food joints anyways
[13:21:21] <razrX> ehrm, that's an opinion but it's mine (not a fact)
[13:21:24] <razrX> ;)
[13:21:32] <lasseoe> s close to a fact as it gets :)
[13:21:43] <razrX> :)
[13:21:48] <lasseoe> wonder what wikipedia has to say about dutch food :)
[13:23:35] <razrX> it'll prolly state something about potatoes and 'erwtensoep' (pea soup) or something rather ;)
[13:24:12] <lasseoe> http://en.wikipedia.org/wiki/Dutch_cuisine
[13:24:28] <lasseoe> it mentions frikandel, kroket and other crap :)
[13:24:50] <razrX> i'm on a powerbook so i'm reading it from a dashboard widget
[13:24:59] <razrX> ah yes, the frikandel and kroket
[13:25:05] <lasseoe> I never got the point of Dashboard, it's useless
[13:25:13] <lasseoe> I've removed it
[13:25:38] <razrX> i like it
[13:25:57] *** bunker_work has joined #opensolaris
[13:39:40] *** mlh_ has quit IRC
[13:39:40] *** silk has quit IRC
[13:44:05] *** saif has joined #opensolaris
[13:46:33] <saif> ll
[13:49:18] *** saif has quit IRC
[13:49:40] * jmcp returns
[13:51:42] <razrX> hope you enjoyed your dinner jmcp
[13:51:55] <jmcp> it was decent
[13:52:09] <razrX> sorry to hear that
[13:52:10] <jmcp> but I'm at the point where I wish I could eat dinner with my wife
[13:53:25] <jmcp> razrX: the restaurant at the end of the street is quite acceptable, the food is fine, they have a senior hostess who speaks good English, picture/photo menus, and the prices aren't too bad
[13:53:35] *** saif_indore has joined #opensolaris
[13:53:50] <jmcp> I had an expensive dinner tonight though - a chilli+chicken dish, pancake and a 630ml Yanjing beer for 52RMB
[13:55:22] <razrX> what would be an excellent dinner then jmcp (since you described that as a decent meal)
[13:55:25] <razrX> ?
[13:56:16] <jmcp> well... one I could share with my wife, for a start! :)
[13:56:44] <jmcp> with *really* good quality true organic ingredient
[13:56:47] <jmcp> s
[13:56:58] <jmcp> and beverages to match
[13:57:07] * jmcp is a bit of a food snob at time
[13:57:08] <jmcp> s
[13:57:46] <lasseoe> nothing wrong with that
[13:58:39] <razrX> nope, i agree
[13:59:32] <jmcp> last night I joined some of the office Americans for a Thanksgiving dinner .... I ate turkey with chopsticks
[14:00:17] <razrX> that is something i've not heard of before tbh
[14:00:21] <razrX> turkey with chopsticks
[14:01:00] <jmcp> razrX: when I picked up the knife+fork it just felt wrong
[14:02:05] *** saif_indore has quit IRC
[14:02:10] <razrX> i like eating with chopsticks, i like taking the time for my food  and not rush things when things are good
[14:02:21] <razrX> so i understand
[14:04:59] <jmcp> Robs and I (he's also staying at this hotel, and a good mate) have been getting breakfast from street markets for the last few days too
[14:06:47] *** calumb has quit IRC
[14:07:21] *** gustav3d has quit IRC
[14:08:21] *** Somethingelse has joined #opensolaris
[14:09:34] *** gustav3d has joined #opensolaris
[14:10:03] *** claudiush is now known as cla
[14:10:47] *** cla is now known as claudiush
[14:12:29] <razrX> as lunch goes, i really do hope someone will open up a subway in my home town
[14:12:44] <razrX> now i have to drive for about 45 minutes to Amsterdam to get my lunch
[14:12:54] <razrX> nearest subway restaurant for me
[14:16:13] <Somethingelse> there's no McDonald's near by? :P
[14:16:24] <lasseoe> Gotta be something else nearby
[14:16:38] <Somethingelse> ... i'm in Toronto, Canada...sorry
[14:18:08] <razrX> i hate mcd
[14:19:00] <lasseoe> it's vile
[14:19:39] <jmcp> I'm actually annoyed at how expensive maccas, kfc, starbucks et al are over here ... out of all proportion to the cost of living
[14:20:19] <Somethingelse> yyyep...thas how it is everywhere...thas jus how the fast food industry works
[14:20:33] <razrX> it's not any different over here afaik
[14:20:42] <jmcp> in Australia those three corps seem quite cheap though
[14:20:43] * jmcp shrugs
[14:20:45] <razrX> we don't even have starbucks in the netherlands
[14:20:50] <Doc> wtf would you want to each any of those?
[14:20:59] <jmcp> don' worry, you're not missing outt
[14:21:04] <jmcp> Doc: exactly
[14:21:18] <razrX> well, unofficially we have 1 but it's located inside a Nike building
[14:21:29] <jmcp> razrX: even worse!
[14:21:30] <razrX> jmcp: well, even though i'm european i do like starbucks
[14:21:52] <jmcp> razrX: you've just dropped a level in my estimation.... have you no taste?!?!?!!?
[14:21:58] <razrX> lol
[14:22:03] <Barry> There's a Starbucks inside a building in NY too. Probably more than one.
[14:22:08] <jmcp> Doc: the starbucks next to the office here managed to _burn_ my espresso
[14:23:24] <Barry> Did I break the conversation?
[14:23:43] <razrX> yes you did ;)
[14:24:22] <razrX> but seriously though, the dutch coffee places (and i'm not referring to the dope joints) also leave a lot to be desired IMHO
[14:25:13] <razrX> are the Senseo pads becoming popular outside of holland already?
[14:25:26] <lasseoe> yes, god knows why :-/
[14:25:41] <razrX> does anyone know douwe egberts coffee from holland?
[14:25:46] <lasseoe> yep
[14:25:48] <razrX> lasseoe: you don't like it?
[14:26:01] <lasseoe> razrX: it's an expensive way to make a cup of coffee
[14:26:06] <razrX> very true
[14:26:17] <razrX> for the lazy peeps
[14:26:20] <lasseoe> I'd rather spend the money on a decent espresso maker
[14:26:23] <lasseoe> one of the manual ones
[14:26:34] <Somethingelse> second cup is the best :)
[14:26:47] <lasseoe> second cup, heh
[14:26:53] <lasseoe> it's as bad as Tim Hortons
[14:27:01] <lasseoe> well nearly :)
[14:27:08] * Somethingelse kicks lasseoe
[14:27:12] <Somethingelse> take that back, bish
[14:27:18] <Somethingelse> second cup ownz
[14:27:19] <lasseoe> haha
[14:27:41] <lasseoe> ok ok that was uncalled for, an unfair comparison
[14:28:20] <Somethingelse> i used to like coffee time...only cuz it was the best place to go when i skipped class every day
[14:28:49] <Somethingelse> but their coffee is really bad...about as bad as mcdonald's coffee
[14:29:08] <razrX> that's gotta be the worst coffee in the world
[14:29:36] <Somethingelse> it is...but you'd be surprised how many people go through their drive through every morning just for coffee
[14:29:43] <Somethingelse> i used to work there, i know ;)
[14:29:59] <razrX> they all are f**ked up in my book
[14:30:05] <razrX> terrible shit
[14:30:09] <razrX> pure poison
[14:30:12] <Somethingelse> agreed
[14:30:27] <Somethingelse> not poison...iz just water...and dirt...
[14:30:51] <tomww> if you like coffee, try it in vienna. they know how to and have a good water (makes 50% of the tase IMHO)
[14:30:58] <tomww> s/tase/taste/
[14:31:07] <lasseoe> or Italy!
[14:31:19] <Somethingelse> mmm...i lived in vienna for 2 months, but i didn't drink coffee at that age
[14:31:29] <Somethingelse> was in italy for 6 months...same story
[14:31:33] <tomww> soory, Italy yes!
[14:31:58] <Somethingelse> i was only 8 so...ice cream was more appealing then coffee
[14:32:05] <Somethingelse> and italy had daaaamn good ice cream
[14:32:08] <razrX> Dutch water is renowned for its quality ... terrible they can't put it to good use when making coffee though ;)
[14:32:11] <lasseoe> they still do
[14:32:17] <tomww> my favorite espresso beans are roasted in italy, my espresso machin' is from an italian manufacturer...
[14:32:18] <razrX> italian ice cream ... yummy :p
[14:32:23] <Somethingelse> yea man :)
[14:32:32] <Somethingelse> italy ... yummy
[14:32:35] <Somethingelse> i wanna go back there
[14:32:50] <razrX> hmm, it's been 5 years already since i visited
[14:32:52] <lasseoe> Gelato
[14:32:53] <Somethingelse> had the best time of my life there...don't know why my folks brought me here...cuz it sux
[14:32:54] <lasseoe> Mmm!
[14:33:33] *** jsoft has joined #opensolaris
[14:33:38] <sgnut> for apply a patch is it necesary to have a sun solve plan?
[14:33:48] <sgnut> I mean for download it
[14:33:53] <sgnut> or is there another way?
[14:34:18] <razrX> eek, what's that ... that's not a question about coffee or gelato  ;)
[14:34:25] <razrX> sorry
[14:34:44] <sgnut> razrX: what do u mean?
[14:34:47] <Somethingelse> we're talkin about coffee and ice cream and this guy is asking solaris questions?  pfft, whats this look like to you?
[14:34:51] <Somethingelse> j/k :)
[14:35:13] <razrX> lol
[14:35:20] <sgnut> ok if u anly are intested in OPENsolaris ok.. don't answer me
[14:35:52] <sgnut> :P
[14:35:57] <razrX> for security patches you need a support contract i believe (someone correct me if i'm wrong)
[14:36:10] *** jsoft has quit IRC
[14:36:16] <jmcp> razrX: not that I recall
[14:36:39] <razrX> i'm a little rusty on the patch policy
[14:36:41] <jmcp> as far as I recall, patches for security issues and bits-integrated-into-Solaris are free
[14:36:46] *** gustav3d has joined #opensolaris
[14:36:48] <jmcp> razrX: they keep changing it
[14:36:55] <razrX> that's what i meant
[14:36:58] <jmcp> sgnut: I could be wrong though
[14:37:06] <lasseoe> jmcp: sounds about right
[14:37:10] <jmcp> sgnut: so try downloading the patch, and if it works, great
[14:37:15] <jmcp> otherwise, sorry :(
[14:37:37] <lasseoe> support contracts are cheap though, $120 per socket
[14:37:40] <jmcp> yeah
[14:37:43] <lasseoe> that's not too bad, imho
[14:37:47] * jmcp has been meaning to buy one for ages
[14:38:02] <jmcp> yeah, per-socket and not per-core it's an excellent price
[14:38:24] <sgnut> when I go to download a patch it request a account and says:
[14:38:27] <sgnut> NOTE: Full Access to the SunSolve Knowledgebase and restricted Software Updates (patches) require a valid Sun Service Plan. If your Sun Service Plan information needs to be verified, you will be notified via e-mail in 1 to 3 business days regarding your request.
[14:39:13] <jmcp> bummer
[14:39:19] <jmcp> what is that particular patch for?
[14:39:27] *** lopa has joined #opensolaris
[14:39:33] <lopa> how to do backup in sun server?
[14:39:39] <sgnut> oracle and sap
[14:40:08] <lasseoe> lopa: like most other unixes
[14:40:09] <jmcp> lopa: it depends on what you want and need
[14:40:12] <razrX> i vaguely remember having to use my employers' support contract to register my box to make full use of smpatch(1M)
[14:41:01] <razrX> it's a Solaris 10 system upgraded to 6/06 (originally was a 1/05 i think)
[14:41:14] <lasseoe> razrX: PCA is much nicer than smpatch
[14:41:28] <razrX> lasseoe: need to have a look at it then
[14:41:45] <onbot> commit by Peter Dennis: 6496941 SVM still removes the diskset if the SunCluster nodeid file is missing
[14:47:15] <razrX> just downloaded it
[14:47:24] <razrX> will take a look at PCA when i get home
[14:47:30] <lasseoe> nice and simple
[14:49:30] <razrX> i'm kind of bothered that bug_id 6483014 will only come with the 11/06 release of Solaris 10 (zonecfg misery) and not as a seperate patch set for 6/06 installations
[14:49:51] <razrX> i can't create zones now on my box
[14:51:19] <sgnut> have a nice weekend guys ;)
[14:51:28] <razrX> sgnut: you too
[14:53:57] *** sgnut has quit IRC
[14:56:17] <razrX> i mean the fix to bug_id 6483014 oc ;)
[14:57:16] <jmcp> I really hate it when they put "please see comments"
[14:59:00] <razrX> agreed
[15:00:25] <movement> jmcp: reflexes are hard to break.
[15:00:33] <movement> at one point I got as far as 'sc'
[15:09:31] <jmcp> movement: I know
[15:09:59] <jmcp> movement: it takes a concerted effort ... and a certain amount of physical pain in some cases :)
[15:13:02] <dlynes_laptop> Is int64_t a standard posixly type?
[15:13:30] *** nbkk6fo__ is now known as rodrickbrown
[15:13:42] <movement> dlynes_laptop: yes
[15:14:11] <dlynes_laptop> movement: so, theoretically, solaris should have it defined?
[15:14:52] *** Barry has left #opensolaris
[15:14:52] <jmcp> dlynes_laptop: yes, in /usr/include/sys/types.h
[15:15:24] *** whaq_ has joined #opensolaris
[15:17:27] <dlynes_laptop> jmcp: thanks...it's actually defined in /usr/include/sys/int_types.h
[15:17:45] <jmcp> I got close :)
[15:18:57] <dlynes_laptop> jmcp: it's included from /usr/include/sys/types.h and /usr/include/inttypes.h
[15:19:04] <jmcp> meh
[15:19:05] * jmcp shrugs
[15:19:18] <movement> dlynes_laptop: you should use <inttypes.h> though
[15:19:36] <dlynes_laptop> I must have _KERNEL defined or something though, because int64_t isn't getting defined...hrm...
[15:19:38] <jmcp> dlynes_laptop: what movement said. he knows
[15:20:05] <dlynes_laptop> Yeah...the code I'm using is already including intttypes.h
[15:20:12] <dlynes_laptop> erm s/intttypes/inttypes/
[15:20:43] * jmcp flakes out
[15:20:46] <jmcp> gnite all
[15:20:49] <dlg> nanite jmcp
[15:23:09] <movement> dlynes_laptop: use -E and check the output
[15:23:23] <razrX> nn jmcp
[15:23:24] <dlynes_laptop> movement: ok, thanks
[15:23:34] <dlynes_laptop> movement: Just used to gcc on Linux and Watcom
[15:23:42] *** whaq_ has quit IRC
[15:24:51] *** lopa is now known as marctt
[15:27:16] *** lloy0076 has joined #opensolaris
[15:31:45] *** calumb has joined #opensolaris
[15:32:07] *** qdk has quit IRC
[15:32:28] *** onbot has quit IRC
[15:33:14] *** onbot has joined #opensolaris
[15:33:23] *** onbot has quit IRC
[15:35:32] *** linma` has quit IRC
[15:35:58] *** dj2 has joined #opensolaris
[15:41:38] *** Snake007uk has quit IRC
[15:42:19] *** laca has joined #opensolaris
[15:44:00] *** onbot has joined #opensolaris
[15:44:03] <movement> hey laca
[15:44:31] *** rodrickbrown has quit IRC
[15:44:33] *** nbkk6fo_ has joined #opensolaris
[15:44:45] *** nbkk6fo_ is now known as rodrickbrown
[15:46:14] <laca> hi movement
[15:48:07] *** Mazon is now known as mazon
[15:53:27] *** lloy0076 has left #opensolaris
[15:56:37] *** mrdeviant has joined #opensolaris
[15:57:20] <onbot> commit by sg70180:  6418780 vswitch needs to be able to process updates to its MD node; 6492078 vds does not handle ldc_close failures; 6492423 vSwitch multi-ring code hangs when queue thread not started; 6492705 vsw warning messages should identify device instance number
[15:58:06] <mrdeviant> do we really need a bot monitoring onnv-notify ?
[15:59:51] *** karrotx_ has joined #opensolaris
[16:15:14] *** jcea has joined #opensolaris
[16:16:03] <movement> mrdeviant: yes
[16:16:10] <mrdeviant> why?
[16:16:15] <mrdeviant> can't you just subscribe to onnv-notify?
[16:17:52] <movement> it's a low-impact way of keeping connected with what's happening.
[16:18:27] <mrdeviant> my point is if you care about commits, you can subscribe to a list. but by having a bot, you barrage people who don't care w/ messages.
[16:19:38] *** qdk has joined #opensolaris
[16:22:48] <movement> take it up with the channel admins if it's a real issue for you.
[16:23:24] <mrdeviant> it is. #blastwave did the same thing months ago. it was a stupid move them, it's a stupid move now.
[16:23:44] *** razrX is now known as razrX_afk
[16:26:28] <movement> sigh, pretty easy to fix, isn't it?
[16:26:43] <mrdeviant> yea, it is. just kill the bot.
[16:28:24] <movement> funny how /ignore is topmost in my mind already.
[16:28:58] <mrdeviant> why should every user who doesn't care about commit message have to take action not to receive the messages?
[16:29:11] <mrdeviant> it should be opt-in (i.e., those who care, subscribe to the list), not opt-out
[16:33:09] <movement> so far, only you have complained. you have a very very simple solution to your problem and that's the last I have to say on the topic.
[16:33:24] <mrdeviant> *sigh*
[16:33:31] <mrdeviant> i am so sick of having this argument in every channel
[16:37:54] * quasi finds the onbot a silly idea, but adding an ignore is no problem
[16:38:25] <lasseoe> it is pretty silly, you'll only ever notice the commits announced when you're reading the channel anyway
[16:38:56] <lasseoe> doesn't bother me though, not yet anyway :)
[16:41:21] *** reg1 has left #opensolaris
[16:44:46] *** rodrickbrown has quit IRC
[16:44:49] *** nbkk6fo_ has joined #opensolaris
[16:48:40] <quasi> lasseoe: yeah, same here - I didn't even notice till mrdeviant started bitching ;)
[16:48:45] *** mazon is now known as Mazon
[16:59:58] *** calumb has quit IRC
[17:11:38] *** calumb has joined #opensolaris
[17:13:28] *** calumb is now known as calAFK
[17:24:15] *** eugene has joined #opensolaris
[17:25:06] *** calAFK is now known as calumb
[17:26:59] <PerterB> I also think it's a pretty stoopid idea, but it isn't bothering me either... yet
[17:27:50] *** axisys has joined #opensolaris
[17:37:54] *** bunker_work has quit IRC
[17:40:24] *** tsoome has quit IRC
[17:40:25] *** Dar is now known as Dar_AFK
[17:40:32] *** tsoome has joined #opensolaris
[17:45:26] *** frankho has quit IRC
[17:46:43] *** calumb has quit IRC
[17:52:17] *** calumb has joined #opensolaris
[17:57:26] *** onbot has quit IRC
[17:57:31] *** onbot has joined #opensolaris
[18:12:13] *** raph_ael has quit IRC
[18:19:04] <leal> dlg: it's working now... :))
[18:19:11] <leal> dlg: the SAN boot.
[18:24:31] *** GmanAFK is now known as Gman
[18:26:17] <tomww> leal: on X86?
[18:27:31] <leal> tomww: yeap.
[18:28:59] *** dclarke has joined #opensolaris
[18:29:09] <jamesd> hi dennisa
[18:29:24] <dclarke> hello
[18:30:01] <dclarke> do you recall offhand the amazing command to get Solaris to spill its guts about memory usage ?
[18:30:10] <dclarke> it was something with adb -k
[18:30:15] <dclarke> or thereabouts
[18:31:14] <dclarke> sorry .. mdb
[18:31:32] <PerterB> ::memstat
[18:31:58] <dclarke> okay .. yep .. thats it
[18:32:15] <dclarke> I think I need mdb -k /dev/kernel something something  ::memstat
[18:32:30] <jamesd> echo "::memstat"  | mdb -k
[18:33:48] <dclarke> thats close ... to what I was looking for
[18:36:24] <B|nTaRa> Some cool info: Tibeten monks, after twenty years or so of practise in the Himalaya, control their brain stem - they can control their heart beat, blood pressure etc.  After thirty years they can connect to the internet purely by meditation, setting TCP stacks in their neurons and stuff.  Right now I am chatting with a monk who is sitting naked in an ice storm on his towel, his only possesion.  He's using ipv6.
[18:38:58] <dclarke> http://www.rafb.net/paste/results/Kytmwo93.html
[18:39:41] <dclarke> B|nTaRa: further proof that Douglas Adams was right .. you really do NEED a towel
[18:45:54] <jamesd> dclarke, a solaris 10 box with 65 MB of ram?
[18:47:21] *** Tzoa has joined #opensolaris
[18:47:46] <dclarke> jamesd : yes
[18:48:00] <dclarke> jamesd: less .. only 64MB of RAM
[18:48:09] <dclarke> and every meg count
[18:48:13] <dclarke> counts
[18:48:23] <jamesd> i didn't even know that  solaris 10 would even boot in such a box.
[18:48:24] <dclarke> the trick is to get rid of all non-required crud
[18:48:39] <dclarke> guess what .. it ain't easy .. but it does
[18:49:13] <jamesd> i guess you can bring that box to its knees, if you fired up java, or use  zfs on it..
[18:50:45] <dclarke> just about anything would be the kiss of death
[18:50:56] *** Tzoa has left #opensolaris
[18:50:59] <dclarke> Netra t1 (UltraSPARC-IIi 440MHz), No Keyboard
[18:50:59] <dclarke> OpenBoot 3.10.25 ME, 64 MB memory installed, Serial #14240682.
[18:50:59] <dclarke> Ethernet address 8:0:20:d9:4b:aa, Host ID: 80d94baa.
[18:55:15] *** silk has joined #opensolaris
[18:58:23] *** indrajit has joined #opensolaris
[19:00:11] *** laca has quit IRC
[19:03:15] *** claudiush is now known as cla|
[19:03:30] <prstat> has anyone else been getting the 'Sun and VMware virtualization' mail?
[19:03:47] <prstat> I'm trying to decide if its worth attending..
[19:03:59] <prstat> We use zones for virtual servers now...
[19:04:26] <prstat> I cant imagine there is a worthwhile way of virtualizing x86 on sun hardware..
[19:05:03] <prstat> but there is not enought info in the spam to determine if they have a real product behind this one.
[19:05:46] <dclarke> prstat : there is really very little for virtualization on UltraSparc
[19:05:58] <dclarke> prstat: you can *try* QEMU
[19:06:07] <dclarke> prstat : its not pretty but it works
[19:07:02] *** Gman has quit IRC
[19:07:10] *** Yamazaki-kun has joined #opensolaris
[19:09:02] *** mega has joined #opensolaris
[19:09:15] <jamesd> dclarke, thanks for the idea for a blog entry  http://uadmin.blogspot.com/2006/11/thanks-for-memories.html
[19:10:31] <dclarke> $ prtconf -v | grep Memory\ size
[19:10:31] <dclarke> Memory size: 32704 Megabytes
[19:10:38] <dclarke> thats the other extreme end
[19:10:52] <jamesd> is that in solaris 10?
[19:11:03] <dclarke> $ uname -a
[19:11:03] <dclarke> SunOS saturn 5.10 Generic_118855-19 i86pc i386 i86pc
[19:11:09] <jamesd> nice
[19:11:14] <dclarke> yeah ..
[19:11:17] <dclarke> its a BIG box
[19:11:24] <Error_404> *yawn* morning all
[19:11:30] <jamesd> but some how i would perfer  solaris 8  on a box what that little amount of ram
[19:11:49] *** indrajit has quit IRC
[19:11:56] <dclarke> true ... but where is the challenge
[19:13:06] *** gustav3d has joined #opensolaris
[19:13:23] <Error_404> which box?
[19:14:41] <jamesd> dclarke, i guess i'm getting old.. i don't like challenge as much as i used to.. i used to love running X on a 5MB ram box.. and 24 hour linux kernel builds, but not so much these days.
[19:15:00] <dclarke> ha ha .. you're a riot
[19:17:30] *** reflect_ has joined #opensolaris
[19:17:31] *** reflect has quit IRC
[19:17:42] <reflect_> re
[19:17:49] <reflect_> did anyone see what I asked?
[19:18:12] <jamesd> no?
[19:18:52] <reflect_> I need a favor from someone with open/solaris. I need the output of prtconf, prtdiag and scanpci.. if you have both sparc and x86, all the better. if you can, add a few switches (-v, etc)..
[19:19:01] *** darkcmd has joined #opensolaris
[19:19:10] <darkcmd> hi everyone
[19:19:12] <dclarke> .me busy here
[19:19:20] <darkcmd> anyone ever mess with MO disks?
[19:19:22] * dclarke busy
[19:19:36] <onbot> commit by pj149606:  6495324 SUNWroutr package doesn't have a copyright file (fix build)
[19:19:36] <onbot> commit by Artem Kachitchkin:  6490549 Tamarack fails when a sparc system is bfu'd with nightly archives built with gcc; 6495813 rmmount should allow foldcase and nofoldcase for pcfs; 6495819 eject/rmmount -l output not quite correct
[19:20:07] *** darkcmd has left #opensolaris
[19:20:52] <axisys> my /etc/TIMEZONE shows TZ=US/Eastern but the date shows 1 hr slow, 12:20pm
[19:20:55] <reflect_> I'm trying to build a tool to gather better HCL data..  but, as I don't have access to a solaris box, I need a bit of helf
[19:21:00] <reflect_> help, even
[19:21:05] <axisys> why so can anyone explain?
[19:24:00] *** dlynes_laptop has quit IRC
[19:25:51] *** klocze1 has quit IRC
[19:26:15] <reflect_> paste to output to pastebin or perhaps into a file and then place it on a http capable machine, if possible
[19:27:14] *** kloczek has joined #opensolaris
[19:27:30] *** mlh has joined #opensolaris
[19:28:33] <reflect_> I've put up a lil bit of info regarding the project here > http://www.acggbg.org/~reflect/opensolaris/ should you be interested in knowing more about the hcl idea
[19:30:02] *** reflect_ is now known as reflect
[19:36:14] *** Somethingelse has quit IRC
[19:37:28] *** richloo has joined #opensolaris
[19:39:14] *** calumb has quit IRC
[19:39:31] *** onbot has quit IRC
[19:44:32] *** richlowe has quit IRC
[19:44:34] *** richloo is now known as richlowe
[19:45:26] *** onbot has joined #opensolaris
[19:47:08] *** Somethingelse has joined #opensolaris
[19:50:10] *** dunc has quit IRC
[19:50:34] *** Fish- has joined #opensolaris
[19:52:41] <leal> the zfs default is raidz right?
[19:52:54] <richlowe> the default is a plain stripe.
[19:53:17] <leal> raid0?
[19:53:20] <elektronkind> yes
[19:53:41] <elektronkind> zpool create <device> <device> <device> ....
[19:53:48] <elektronkind> that will create only a stripe
[19:54:05] <elektronkind> err, zpool create <pool name> <device> <device> <device> ....
[19:54:18] <leal> Ok, i will create a pool of EMC discs, i want the full capacity. I have redundancy on the EMC already..
[19:54:51] <elektronkind> that's perfectly fine
[19:55:14] <Fish-> hello
[19:55:21] <elektronkind> hi Fish
[19:56:21] <leal> I have two sata discs on that machine.. So, i will need to configure raidz to have some redundancy.. right?
[19:56:34] <silk> 2 disks, you might as well create a mirror
[19:56:36] <leal> After create the pool, there is a option to set?
[19:56:40] <elektronkind> raidz requires 3 or more devices.
[19:56:46] <richlowe> not quite true.
[19:56:53] <richlowe> it works with only 2 devices, it's just really really silly to do that.
[19:57:01] <richlowe> (it's basically a very very expensive mirror in that case)
[19:57:17] <elektronkind> ah well, then *essentially* requires :)
[19:57:33] <elektronkind> never tried it with two for the reasons you mention
[19:57:39] <silk> leal, you set teh raid level when you create teh pool
[19:58:20] <leal> silly? i mean, maybe i want 160GB (for sure), and no 320GB (in doubt)..
[19:58:35] <twincest> leal: raid-z is like raid-5.  raid-5 for two devices makes no sense
[19:58:46] <twincest> leal: if you want redundance, use raid-1 (mirror)
[19:59:14] <leal> ok. that is really sily. sorry.
[19:59:40] <richlowe> twincest: so, you should add priocntl to your wad of error message fixes :)
[19:59:42] *** LordKing has joined #opensolaris
[19:59:47] <twincest> rich: i heard ;-)
[20:00:00] <twincest> i'm waiting for gisburn before i can commit this stuff though
[20:00:01] <richlowe> twincest: yeah, well I typoed a taskid again, and it refreshed my memory :)
[20:00:17] <leal> raid1. raidz is raid5 (three or more discs) .
[20:03:02] *** postwait has quit IRC
[20:03:39] <richlowe> twincest: how come it's waiting on gisburn?
[20:04:32] <twincest> rich: apparently it conflicts with ksh93-integration (!)
[20:04:41] <richlowe> How?
[20:04:48] <twincest> they want to modify some of cmd/ to use their stuff
[20:04:51] *** axxl has joined #opensolaris
[20:04:59] <richlowe> twincest: well, quite, but they want to do that *after*
[20:05:35] <twincest> still, there's no point doing it now if they're going to do it later anyway.  but i'm waiting for something a bit less vague
[20:05:45] <reflect> I'm building a HCL tool, and I need a favor from someone with open/solaris. I need the output of prtconf, prtdiag and scanpci.. if you have both sparc and x86, all the better. if you can, add a few switches (-v, etc)..
[20:08:35] <richlowe> twincest: I still think a desire to change something at some indeterminate time in the future isn't blocking.
[20:10:05] *** Burana has joined #opensolaris
[20:10:24] *** postwait has joined #opensolaris
[20:10:41] <postwait> If I wanted to compile a kernel module available in opensolaris for Solaris 10
[20:10:47] <postwait> How would I go about building it?
[20:11:11] <twincest> rich: well one of them did ask me to wait.  i was expecting a quicker response :)
[20:11:22] <twincest> if they don't reply i might just submit it anyway
[20:12:24] <whaq> Say.. if opensolaris were a woman, how do you visualize her? What kind of a woman is she?
[20:16:51] <jamesd> not one woman, many women... (zones)  ...  with different colored wigs ( brandz) with a memory like a dozen elephants (zfs), and constantly able to keep 1000's of tasks running (smf)  and can communicate like know other ( fireengine)
[20:17:17] *** Gman has joined #opensolaris
[20:19:02] <richlowe> mornin' Gman
[20:19:37] <dclarke> if OpenSolaris were a woman?  She is older but still wants to act like she is twenty two.  She had a long long relationship with a real rich man and now can not figure out how to live cheaply.  She keeps lots of secrets while telling you that its an "open" relationship.  When pissed off she can fight like a marine and is more than able to take carte of herself.
[20:21:43] <Gman> hey richlowe
[20:22:20] <reflect> is there any way to probe the pci bus on a sparc?
[20:22:29] <reflect> I mean the entire pci bus, also the devices on the motherboard
[20:22:34] <elektronkind> to list devices?
[20:22:42] <reflect> yes
[20:23:07] <elektronkind>  /usr/X11/bin/scanpci is the most straight-forward and human-readable way.
[20:23:27] *** Fish- is now known as Fish
[20:23:32] <reflect> and you've tried that on a sparc?
[20:24:17] <elektronkind> can't recall if I've done that on sparc or not. I have lots of systems.
[20:24:28] <elektronkind> doesn't hurt to try though
[20:24:37] <elektronkind> there are other commands you can use, too
[20:24:45] <reflect> I'm listening
[20:25:01] <elektronkind> prtconf  -vp
[20:25:15] <reflect> I got scanpci, prtconf and prtdiag
[20:25:45] <elektronkind> prtpicl -v
[20:25:48] <elektronkind> is another
[20:26:29] <reflect> excellent
[20:27:51] <richlowe> and there's HAL, in onnv_51 and above.
[20:28:04] <reflect> and that is what...  a command?
[20:28:33] <elektronkind> sometimes I miss the simplicity of IRIX's 'hinv' command :)
[20:28:44] <reflect> building a HCL tool, I do too!
[20:29:09] <reflect> hell, even linux's "lspci" seems to have more coherency and simplicity than this
[20:29:56] <elektronkind> prtpicl is actually pretty good, it just give a TON of detail in a not very human-readable method of displaying it
[20:30:21] <reflect> that doesn't matter, as long as the output format is the same no matter what hardware is underneath
[20:30:32] <elektronkind> http://blogs.sun.com/dmick/entry/prtpci_digest_and_display_prtconf
[20:30:36] <elektronkind> check that out
[20:30:39] <reflect> is seems like prtdiag changes the output format if you sneeze
[20:30:45] <elektronkind> prtpci is something akin to lspci in linux
[20:31:20] <elektronkind> it takes prtconf output and puts it into a format easily parsed by humans
[20:31:21] <reflect> yes!!
[20:31:24] <reflect> that's exactly what I want
[20:31:54] <reflect> actually, if it only gave me the vendor id, I'd be happy.. but this works too
[20:31:55] <elektronkind> it uses the PCI ID database to present sane names for the devices
[20:32:34] <reflect> sane names for a hcl gathering tool..  for that part, I don't give a crap. however, searching the hcl database, then you need sane names
[20:32:49] <reflect> (and a great interface, alot better than what exists today)
[20:33:59] <reflect> the hcl today, if you'll pardon my french, is horrible
[20:34:11] <elektronkind> this HAL thing richlowe speaks of is new to me, though
[20:34:27] <elektronkind> the HCL just isn't detailed enough
[20:34:33] <richlowe> HAL as in http://hal.freedesktop.org/wiki/Software/hal
[20:34:46] <richlowe> the stuff recent gnome uses for it's hardware-ish stuff.
[20:34:55] <richlowe> and I believe how Ubuntu's HCL-ish app functions.
[20:35:09] <richlowe> (there was some discussion of that app 6 or 7 months ago I think, and I believe the Nexenta folks have it functioning on there)
[20:35:15] <reflect> thanks, I'll take a peek at it
[20:35:34] <reflect> elektronkind: details isn't the real problem
[20:35:37] <reflect> it's the focus
[20:36:04] <lasseoe> does prtpci work on SPARC ?
[20:36:19] *** Somethingelse has quit IRC
[20:36:42] <reflect> if a certain chip is supported, it works.. no matter if it's onboard, hooked up to the pci bus, or connected through a PCI-X card
[20:36:58] <lasseoe> oh never mind, of course it does.
[20:37:11] <lasseoe> it parses prtconf output
[20:37:29] <reflect> it's a nightmare getting the current hcl to list all devices capable of talking to a SATA drive, for instance
[20:37:40] <reflect> you can substitute SATA for just about anything
[20:38:00] <lasseoe> well.. I don't have a single x86/x64 box so I don't really use it
[20:38:09] <reflect> and I mean listing them in a "human readable format" so that you many decide which hardware you want to buy
[20:38:16] <reflect> lasseoe: que?
[20:38:52] <elektronkind> I'd also like to see things like "The xyz driver in Solaris supports the following hardware and revisions..."
[20:39:11] <elektronkind> so if you have, say, a
[20:39:12] <reflect> elektronkind: ah, nice
[20:39:15] <reflect> good idea.
[20:40:02] <elektronkind> so if you have, say, a card with a "FooComm Super4332 Rev B" chip on it, you can see if the "fcom" driver in Solaris will speak with it
[20:40:03] * reflect steals it
[20:40:48] <elektronkind> at the very least it would cut down on lost time and RMA'ing hardware that you though would work :)
[20:41:08] <lasseoe> reflect: servers: SPARC, desktop: G4 Macs
[20:41:29] <reflect> lasseoe: still not sure what you're on about
[20:41:50] <reflect> can't a sparc server make use of a SATA drive?
[20:41:59] <elektronkind> sure it could
[20:42:06] *** postwait has quit IRC
[20:42:21] <reflect> so, he's probably not talking about that :)
[20:43:52] <lasseoe> reflect: just saying I don't use the HCL :)
[20:43:58] <reflect> ah
[20:44:24] <reflect> well, I wanted to try out zfs
[20:44:35] <reflect> after an hour wrestling with the hcl, I gave up
[20:44:59] <richlowe> Hm, there's the java thingy that's intended to automate some of that.
[20:45:13] *** Gman has quit IRC
[20:45:23] <reflect> richlowe: http://www.acggbg.org/~reflect/opensolaris/
[20:46:51] *** mega has quit IRC
[20:47:38] <leal> hello, i did install a solaris 10, and the
[20:47:50] <richlowe> http://www.sun.com/bigadmin/hcl/hcts/device_detect.html
[20:47:51] <richlowe> that thingy.
[20:48:21] <richlowe> though, and who knows why, it refuses to work if you're already running Nevada.
[20:48:28] <richlowe> so I can't actually see how good it is :)
[20:48:29] <lasseoe> heh
[20:48:30] <leal> X starts but the monitor (server switch), does not works ( out of sync)... i can't login
[20:49:07] <leal> That is a fresh install, i can't login via ssh, because there is no user.
[20:49:15] <reflect> if anyone would like to help out, I'm all for it
[20:49:20] <leal> What the better approuch to fix it?
[20:49:30] <reflect> wether it's suggestions or coding time, or database design
[20:50:09] <leal> How can i boot the machine without start the X?
[20:50:22] <reflect> boot it into single user
[20:51:07] <leal> reflect: ok, i did that.. but cant create a user. permission denied to "passwd" utility.
[20:51:28] <reflect> then you're not root?
[20:52:04] *** deather_ is now known as deather
[20:52:49] <leal> reflect: i did a boot in single user, then i try to create a user, and the passwd utility did not work.
[20:53:05] <Yamazaki-kun> Is the root volume mounted read-only?
[20:53:12] <leal> reflect: i will try disable dtlogin (kdmconfig right?)
[20:53:21] <reflect> then something is very wrong, type 'mount' to check wether your root is rw or ro
[20:53:26] <leal> no i did a "remount" rw.
[20:53:47] <reflect> leal: to be honest, I haven't been using a solaris machine for some 3-4 years ;)
[20:53:51] <leal> is rw, because i can create the user. i could not change the passwd.
[20:54:21] <leal> reflect:ok, thanks.
[20:54:42] <reflect> truss the command?
[20:55:34] <lasseoe> problem is /home is managed by the volume manager
[20:55:39] <lasseoe> create your homedir in /export/home
[20:55:47] <reflect> autofs, eh
[20:55:48] <reflect> ofcourse
[20:56:06] <leal> no, autofs and passwd?
[20:56:18] <leal> i guess not.
[20:56:24] <reflect> autofs is on at /home default
[20:56:25] <reflect> iirc
[20:56:54] <Error_404> damned shame that sun's Startup program is for USians only
[20:57:00] <leal> reflect: it's in single user. there is no autofs.
[20:57:08] <lasseoe> Error_404: yes :-/
[20:58:14] <lasseoe> but then again, there really aren't any other countries outside of US borders ;)
[20:58:21] <reflect> single user and no passwd.. sorry, you're gonna have to run it by the truss command
[20:59:20] <Error_404> it would be incredibly cheap to run a datacenter where I am right now
[20:59:33] <leal> reflect: dtconfig -d
[20:59:34] <Error_404> no cooling required, just vent the air from outside
[20:59:35] <leal> thanks all
[21:11:15] *** reflect_ has joined #opensolaris
[21:11:16] *** reflect has quit IRC
[21:14:25] *** Gman has joined #opensolaris
[21:16:20] *** nplt has joined #opensolaris
[21:18:12] <axisys> what is the fastest way to convert 1000 1M gzip'd files to zip format?
[21:19:09] <delewis> axisys: for-loop
[21:20:12] <delewis> for i in *.gz; do gunzip $i; zip `echo $i | sed s/\.zip//`; done
[21:20:23] <delewis> nasty, as it's just something I came up with off the top of my head.
[21:20:48] <axisys> delewis: so there is no other tool like convert file.gz file.zip ?
[21:20:59] <reflect_> it'll have to be slightly modified, I think
[21:20:59] <delewis> axisys: not that I know of
[21:21:15] <axisys> ofcousre covert could be your wrapper but nothing like a app ?!
[21:21:18] <delewis> and why should there be when you can do a short, one-line command that acheives the same?
[21:21:30] <delewis> this is the UNIX philosophy.
[21:21:41] <axisys> delewis: ofcourse..
[21:21:46] <lasseoe> I can't see what a specialised app could do that's better or faster
[21:22:17] <axisys> delewis: i guess i could mv a file in /tmp and do the work and move it back to gain speed
[21:22:18] <delewis> and you'll want to quote that regex, too
[21:22:23] <lasseoe> if you have multiple cpus or cores I'd make a few loops
[21:22:37] <delewis> yes
[21:22:42] <delewis> but the concept is still the same
[21:22:57] <axisys> lasseoe: well the originial gzip files are in nfs server and i will do the process in my ultra 20
[21:23:07] <lasseoe> moving it into tmp won't speed it up I don't think ,it's only 1MB files
[21:23:11] <delewis> pick all the gz's, gunzip them, remove the gz extension, and zip them up
[21:23:25] <reflect_> for i in *.gz; do gunzip $i; zip `echo $i | sed 's/\.gz//'`; done  <-- something like that ought to be right
[21:23:26] <axisys> i have 2G memory and it is a quiet system .. it is my workstation
[21:23:27] <delewis> and that should be s/\.gz//, obviously
[21:23:36] <delewis> reflect_: yes
[21:23:51] <lasseoe> actually..it's only 1G in total, it won't take that long
[21:23:52] <reflect_> nicely spotted
[21:24:05] *** reflect_ is now known as reflect
[21:24:25] <axisys> cool.. thnx guys
[21:24:44] <delewis> hmm, I think the battery on this IO board is bad.
[21:24:53] <delewis> I keep having to do copy-clock-tod-to-io-board
[21:36:52] *** nplt has left #opensolaris
[21:40:17] *** LordKing has quit IRC
[22:00:11] *** mrdeviant has quit IRC
[22:01:31] *** sahafeez has joined #opensolaris
[22:03:18] *** sahafeez has quit IRC
[22:03:27] *** sahafeez has joined #opensolaris
[22:04:58] *** Gman is now known as GmanAFK
[22:09:56] <Burana> Anybody defined root as an rbac role ? Is login via console still possible?
[22:10:04] <twincest> no, you have to su
[22:10:19] <tsoome> roles can't do login
[22:10:46] <Burana> i would like to have only a couple of users that can su - to root. these users are in ldap. what, if ldap is not reachable?
[22:11:01] <Burana> then I can't fix it...
[22:11:13] <Burana> ...if the problem is on the host.
[22:11:42] <trygvis> I always have backup accounts in /etc/passwd
[22:12:01] <Burana> I would like to have no password for root...
[22:12:54] <Burana> where there is no password, no policy can require to change it :-)
[22:12:59] <trygvis> just set it to blank?
[22:13:06] *** leal has quit IRC
[22:13:36] <Burana> I know, but I need to somehow do an emergency login if users allowed to su - are not around...
[22:13:51] <Burana> and that would be through the console...
[22:15:22] *** deedaw has joined #opensolaris
[22:16:46] <andersmo> I'd suggest using pfexec or similar instead? Give the accounts you want to be "root-like" the "Primary Administrator" profile, and they can use pfexec and friends to run commands (including shells) as uid 0.
[22:17:28] <axisys> delewis, reflect .. i guess zip needs two arguments like tar and unlike gzip
[22:17:29] <andersmo> Then give root a proper password, which you write down on a note and store in the safe.
[22:20:58] <andersmo> A tad more strict: instead of granting the accounts the "Primary Administrator" profile directly, give a "rootlike" role that profile, and give the specified users access to that role. The root account remains unused.
[22:26:41] *** GmanAFK has quit IRC
[22:35:02] <lasseoe> axisys: man zip
[22:36:35] *** _william_ has joined #opensolaris
[22:36:41] <Burana> I really prefer to have only root, application and user accounts
[22:37:12] <Burana> the only part that keeps me from defining root as a role, is the lack to not being able to login as root directly on the console
[22:37:24] <_william_> hi all
[22:39:19] <axisys> lasseoe: yep
[22:45:20] *** nplt has joined #opensolaris
[22:56:03] *** MikeE has joined #opensolaris
[22:56:24] *** bougie has quit IRC
[22:59:13] <reflect> Burana: would using sudo be a problem?
[23:02:40] *** Gman has joined #opensolaris
[23:02:41] *** hell` has quit IRC
[23:12:50] *** MikeE has quit IRC
[23:19:00] *** deather_ has joined #opensolaris
[23:23:31] *** slowhog has joined #opensolaris
[23:24:25] <reflect> how many of you are hardware driver developers?
[23:25:05] <jamesd> not i
[23:25:11] *** dj2 has quit IRC
[23:25:52] <PerterB> nor I, not since sunos 4 days...
[23:27:49] <reflect> ok.. it's possible that I may be looking for a hw developer
[23:28:07] <reflect> ie, a paid job
[23:28:07] <PerterB> did you check behind the sofa?
[23:28:58] <reflect> if you're not interested, then just don't reply
[23:29:12] <PerterB> depending what kind of hardware, it might be worth asking again more during aussie hours or after the yanks are done with thanksgiving
[23:29:30] <PerterB> and don't get bossy, it's not very endearing
[23:29:46] <reflect> maybe someone would put it in the topic so I don't have to repeat myself
[23:29:57] <delewis> PerterB: surely if someone throws around cash in #opensolaris they can walk on top of you! :-)
[23:30:07] <PerterB> oh right, silly me :)
[23:30:07] <reflect> it's not a ton of money, but it's something
[23:30:17] <delewis> it's amazing how many feature requests we get on the MPlayer mailing list at the lure of money
[23:30:23] <delewis> "implement this! and I'll send you cash!"
[23:30:36] <delewis> reflect: we're just joking.
[23:30:45] <reflect> well, I'm not.
[23:31:16] * delewis is currently looking for a job
[23:31:21] *** gm152 has joined #opensolaris
[23:31:23] <delewis> but alas, I'm not much of a hardware developer.
[23:31:39] <reflect> I'm just looking for a driver developer
[23:31:40] <delewis> UNIX/SAN administration? sure.
[23:31:51] <delewis> and maybe some C development
[23:32:41] *** lisppaste3 has quit IRC
[23:34:06] <lasseoe> reflect: what do you want a driver for ?
[23:34:39] <reflect> I have connections with a company making SATA controllers
[23:35:18] *** marctt has left #opensolaris
[23:35:25] <reflect> so, I want drivers for them
[23:35:33] *** deather has quit IRC
[23:36:26] <lasseoe> right on
[23:37:05] *** delewis has quit IRC
[23:37:48] *** lisppaste3 has joined #opensolaris
[23:39:01] <reflect> the low-end hardware..   and especially controllers..  is important to the spreading of solaris
[23:41:14] <dclarke> hello .. anyone have a suggestion for fibre based storage that is rock solid dependable but not really really expensive ?
[23:41:27] <dclarke> no SATA or IDE solutions please ...
[23:41:32] <jamesd> a5100
[23:41:40] <dclarke> you mean the old A5200
[23:41:57] <lasseoe> dclarcke: T3B ?
[23:42:02] <dclarke> thats getting really long in the tooth and where can you get them in top notch shape
[23:42:11] <Auralis> a5200 cheap ass these days and good
[23:42:14] <dclarke> the T3 .. hrmmm
[23:42:19] <reflect> we have really bad experience with the t3
[23:42:22] <dclarke> the A5200 does rock
[23:42:35] <dclarke> and I have heard nothing but bad stuff about the T3 also
[23:42:39] <lasseoe> yes 5200's are nice
[23:42:54] <dclarke> but .. where does one get A5200's loaded with disks ?
[23:42:59] <lasseoe> ebay
[23:43:01] <dclarke> ebay .. I guess
[23:43:05] <dclarke> yep ..
[23:43:10] <dclarke> well hell .. let me look
[23:43:15] <lasseoe> :)
[23:43:16] *** deedaw has quit IRC
[23:43:29] <dclarke> I need fibre hubs for the A5200 however
[23:43:33] <lasseoe> you'll want some FC100 SBus cards for them
[23:43:45] <dclarke> like .. if I need to connect three or four systems to them
[23:43:49] <lasseoe> or whatever they are called as PCI, not that common though
[23:44:18] <dclarke> I don't see that I can do SAN stuff with the mat all unless I have expensive Veritas software
[23:44:39] <dclarke> I will have to check Ben Rockwoods site also .. he has some great veritas resources there
[23:44:56] <lasseoe> The A5200 gives you a free VxVM license
[23:45:07] <lasseoe> nice but not all that useful now that we have ZFS
[23:45:09] <jbk> i don't know that's necessairly the case you need veritas
[23:45:39] <dclarke> ZFS is only good if you have Solaris 10 everywhere
[23:45:48] <dclarke> and you can't share it out via NFS
[23:46:28] <lasseoe> very true
[23:46:37] <tomww> dclarke really?
[23:46:48] <dclarke> yes .. really
[23:46:51] <dclarke> you *can
[23:46:57] <dclarke> but it sucks
[23:47:00] <richlowe> What now?
[23:47:01] <dclarke> really really bad
[23:47:19] <dclarke> richlowe: just the usual
[23:47:33] <dclarke> richlowe: if you set a ZFS filesystem to be NFS shared
[23:47:35] <dclarke> it sucks
[23:47:46] <richlowe> Which I have, many of them in fact...
[23:47:47] <Stric> performance issues
[23:47:54] *** gm152 has quit IRC
[23:47:55] <richlowe> that latter bit however, I don't particularly see.
[23:48:02] <richlowe> I assume you're meaning the bits in the thread on zfs-discuss.
[23:48:03] *** reflect has quit IRC
[23:48:08] <dclarke> go check the zfs discuss maillist
[23:48:08] <richlowe> in which case you may want to tack on "for some workloads" :)
[23:48:15] <dclarke> its all there
[23:48:20] <dclarke> gotta go eat
[23:48:23] <dclarke> be back in a bit
[23:48:34] *** reflect has joined #opensolaris
[23:48:56] <richlowe> also, did you read Roch's replies in that thread?
[23:49:04] <Stric> richlowe: writes on 10u2 sucks bigtime.. each write causes a sync of the entire pool
[23:49:07] <PerterB> well, that's a pretty precise reference.... "go read all of zfs-discuss to see what I mean"
[23:49:38] <tomww> not everyone shares his opinion
[23:49:49] <dclarke> well here .. this is more precise : "ZFS shared out via NFS sucks"
[23:49:58] <dclarke> is that precise enough for you ?
[23:50:13] <dclarke> now then .. I'll be back in a bit
[23:50:19] <tomww> no
[23:50:30] <tomww> :-)
[23:54:59] <qdk> dclarke: its the problem related specific to ZFS and NFS working together, or is it just NFS that sucks?
[23:55:16] <Stric> qdk: the first at least
[23:55:32] <richlowe> qdk: zfs is more strict with honouring certain NFS requests.
[23:55:37] <richlowe> it actually flushes them to disk.
[23:56:48] <qdk> ok, so if i were to use a Solaris cluster as frontend to a iSCSI storage and provide NFS access to it to my webfarm, what should i use then?
[23:57:22] *** Mazon is now known as mazon
[23:59:30] <richlowe> (where UFS may write to disk, and assume the data is on disk when the write returns, zfs flushes the write cache and makes sure)





top