[00:00:11] *** estibi_ has joined #opensolaris
[00:00:55] *** estibi has quit IRC
[00:01:00] *** estibi_ has quit IRC
[00:01:15] *** estibi has joined #opensolaris
[00:01:48] *** derchris has quit IRC
[00:01:56] *** derchris has joined #opensolaris
[00:04:22] <FunkyELF> whats up guys, I tried booting into opensolaris but I keep getting this error.....
[00:05:03] <FunkyELF> Warning: init(1M) exited on fatal signal 9: restarting automatically
[00:05:05] * jmcp caffeinates
[00:07:37] *** nostoi has quit IRC
[00:09:46] *** fedorared has joined #opensolaris
[00:10:50] *** phus has quit IRC
[00:11:33] <Chipdancer> jmcp: ping, do you want real caffeine?
[00:11:48] *** Murmuria has quit IRC
[00:12:50] <cla> 23
[00:12:58] <jamesd> 24
[00:13:31] <cla> sorry, it was supposed to be /23
[00:13:37] <cla> :)
[00:13:44] <FunkyELF> so I did some googling, aparently that error is common and because my processor doesn't support SSE.  Anyone know where I could get a live cd built for i386....you know, the lowest common denominator?
[00:14:35] <jamesd> FunkyELF, how old is this system? i think even p2's supported  sse.
[00:14:50] <e^ipi> I thought SSE was an addition to P3
[00:14:56] <FunkyELF> yeah....SSE was p3
[00:14:59] <delewis> P2s had MMX, but no SSE.
[00:15:00] <FunkyELF> this is a Thunderbird
[00:15:15] <flyingparchment> Funky: what build are you using?
[00:15:20] <FunkyELF> It was better than the pentiums at the time....just didn't have that instruction set
[00:15:22] <flyingparchment> i thought SSE bug was fixed
[00:15:39] *** SYS64738 has joined #opensolaris
[00:15:41] <flyingparchment> it doesn't actually require SSE, the linker incorrectly adds that dependency, or so
[00:15:44] <FunkyELF> flyingparchment, lemmie check the date of my iso file
[00:15:59] <FunkyELF> 2007-08-16
[00:16:05] <FunkyELF> thats the day I downloaded it
[00:16:39] <Triskelios> FunkyELF: the filename has the build
[00:16:49] <Triskelios> SXCE, right?
[00:17:00] <FunkyELF> Triskelios, filename was lost ;-)  Thats what happens when you need to unzip / concatonate files
[00:17:05] <FunkyELF> lemmie check firefox download manager
[00:17:44] <FunkyELF> sol-nv-b69-x86-dvd-iso
[00:18:03] <Triskelios> yeah, you want b70
[00:18:08] <FunkyELF> of course
[00:18:13] <FunkyELF> always the case ;-)
[00:18:30] <FunkyELF> "its fixed in SVN"
[00:18:37] <FunkyELF> i love that stuff
[00:18:54] <Triskelios> no, b70 is the current release
[00:18:59] <FunkyELF> Triskelios, should I get the DVD or the CD?
[00:19:10] <e^ipi> what's wrong with things being fixed in SVN?
[00:19:14] <Triskelios> DVD, that many CDs are a pain in the ass
[00:19:23] <Triskelios> e^ipi: having to build
[00:19:26] <e^ipi> you're running experimental code.. it's got bugs that get fixed in later builds
[00:19:31] <e^ipi> deal with it
[00:19:32] *** Trident has joined #opensolaris
[00:19:39] <FunkyELF> e^ipi, nothing.  I just find it funny to see that everywhere
[00:19:57] <e^ipi> so run S10
[00:20:00] <flyingparchment> b70 isn't "fixed in svn", it was released a while ago
[00:20:45] <FunkyELF> i know, I was just saying, that is a common thing to see in the open source community.  I said I got b69, you say I need b70
[00:21:07] <flyingparchment> when the code is changing all the time, stuff is likely to be broken and fixed quickly
[00:21:14] <flyingparchment> it's not really that surprising
[00:21:24] <flyingparchment> (we tend to break things less frequently than linux :)
[00:21:25] <e^ipi> beats the pants off "it's a known issue... we might fix it later"
[00:21:28] <FunkyELF> i know its not....I'm not flaming or anything
[00:21:41] <e^ipi> at least now if the issue is important enough, you can fix it yourself
[00:22:19] <Triskelios> it's sort of surprising if the installer worked even though this regression happened, though
[00:22:36] <FunkyELF> Triskelios, this was trying to boot the live cd
[00:23:03] <flyingparchment> careful with b70, it has the craptastic new installer
[00:23:08] <FunkyELF> if it is a live cd....I don't know, I haven't got that far ;-)
[00:23:13] <flyingparchment> which has a pretty UI and all the features were removed
[00:23:55] <Triskelios> FunkyELF: it's not supposed to be a "live CD", just an environment for the installer and/or for recovery
[00:24:11] <FunkyELF> good, thats what I'm used to....I run Gentoo now
[00:24:40] *** Fish has quit IRC
[00:24:55] <FunkyELF> I have a box I want to use as a fileserver / webserver.  It has 4 IDE controllers so I can have 8 hard drives.
[00:25:12] <FunkyELF> I will probably only get 4 since if one goes down the other one on the same cable will probably too
[00:25:28] <FunkyELF> but I wanted to try out running raidz on zfs
[00:25:59] <Chipdancer> FunkyELF: once you pop, you can't stop
[00:26:11] <Chipdancer> FunkyELF: (zfs is teh awesome)
[00:26:27] <FunkyELF> can solaris boot ZFS?
[00:26:36] <Triskelios> as flyingparchment notes, you probably want to avoid the new installer if you want options like partitioning; in b70 selecting SXCE rather than the default SXDE option will get you the old installer
[00:26:43] <flyingparchment> yes, but the installer doesn't support it, so it's difficult(er) to set up
[00:27:16] <Chipdancer> FunkyELF: yes, see blogs.sun.com/timf/
[00:27:30] <Chipdancer> flyingparchment: it's not actually very hard.. especially  when you use Tim's script :)
[00:27:37] <flyingparchment> yeah
[00:27:42] <FunkyELF> flyingparchment, well...I could do a software RAID 1 (or whichever RAID the copying one is) for the system and then ZFS for the rest of each hard drive
[00:27:58] <flyingparchment> Funky: that's how it's traditionally done
[00:28:08] <Chipdancer> just don't try it at home if you're using snv66+xen as there is a bug in grub where both the zfs bootloader and the xen hypervisor try to use the same section of memory and end up overwriting each other
[00:28:19] <flyingparchment> only problem is, say you have 4 250G drives, and a 20G root, you "lose" 20G on the rest of the drives
[00:28:29] <flyingparchment> (because a raid set is limited to the size of the smallest drive)
[00:28:39] <Triskelios> Chipdancer: that's a great bug
[00:28:48] <Chipdancer> Triskelios: it really caught me out!
[00:28:57] <FunkyELF> or what would you guys recommend if I got four 500Gb hard drives?  20Gb for OpenSolaris on each drive in RAID 1 then the remaining 4x480Gb in raid-z?
[00:29:10] <flyingparchment> a four-way mirror is excessive
[00:29:14] *** onlinebacon has joined #opensolaris
[00:29:21] <flyingparchment> i'd probably do 20G root, 20G scratch and zfs the rest
[00:29:40] <Chipdancer> flyingparchment: both in raid1?
[00:29:50] <Triskelios> FunkyELF: I'd just RAID-Z everything not used for software
[00:29:57] <flyingparchment> Chipdancer: yes
[00:30:05] <e^ipi> caveat: you can't boot off a raidz yet
[00:30:14] <flyingparchment> could use the scratch slice for LU or whatever too
[00:30:19] *** WickedWicky has quit IRC
[00:30:20] <e^ipi> just singles & mirrors
[00:30:39] <Triskelios> FunkyELF: although really you might as well use a smaller disk for the OS and give ZFS the whole drives
[00:30:43] <FunkyELF> flyingparchment, what do you mean scratch?  SWAP?
[00:31:03] <flyingparchment> could be swap if you like - 20G is rather a lot though
[00:31:12] <flyingparchment> LU is probably the most useful application
[00:31:13] <FunkyELF> Triskelios, it will be hard enough trying to get 4 hard drives in this case, I don't wanna try to get 5
[00:31:16] <flyingparchment> (online upgrade between releases)
[00:31:25] <FunkyELF> LU?
[00:31:42] <jmcp> live upgrade
[00:31:53] <Triskelios> (as opposed to "dead upgrade")
[00:32:03] <flyingparchment> LU is like "apt-get dist-upgrade" but it upgrades into a new partition
[00:32:16] <flyingparchment> if the upgrade fucks up, you can reboot back into the old environment
[00:32:18] * FunkyELF knows nothing about apt-get
[00:32:24] <flyingparchment> "emerge world"? :)
[00:32:27] <Chipdancer> FunkyELF: http://blogs.sun.com/timf/entry/zfs_bootable_datasets_happily_rumbling
[00:33:41] <FunkyELF> with RAID all 4 drives would need to be identical, what about ZFS?  That is why I thought I'd need to lose 20Gb on each disk
[00:34:06] <flyingparchment> Funky: same
[00:34:14] <FunkyELF> They split the DVD iso into 3 zip files and only let you download 2 things at once so I can't walk away from the damn computer
[00:34:31] <richlowe> Only 2?
[00:34:38] <richlowe> I thought it was 3.
[00:34:40] <richlowe> (and thus, handy)
[00:34:41] <flyingparchment> Funky: you mirror 20G on disks 1+2 for /+swap+whatever, then create another mirror on the 3+4 and mount it at /aux0 or something, and use that for live upgrade
[00:35:25] <FunkyELF> right....so I have 2 different RAID 1 mirrors, one for the working system and one for messing around with or upgrading
[00:35:30] <flyingparchment> yes
[00:36:35] <FunkyELF> so RAID-Z with the remaining 480 on each disk would give me 1,440Gb.....3/4 for data and 1/4 for parity
[00:37:10] <FunkyELF> nice
[00:37:58] <Chipdancer> FunkyELF: just like a floppy :)
[00:38:08] <flyingparchment> or use a hardware raid controller and don't lose any space :)
[00:38:33] <FunkyELF> I'm glad this ZFS came along.  I have a music collection that is starting to get skips in the mp3s where there weren't before.
[00:38:56] <FunkyELF> I was looking into RAID-5 but this seems nicer with built in compression and all
[00:39:05] <FunkyELF> just wish I could do encryption
[00:39:25] <jbk> that should be happening soon enough
[00:40:04] <FunkyELF> how have the upgrades and new features been going with ZFS?  Can you get the new features or do you have to start with a new filesystem?
[00:40:12] <FunkyELF> can you upgrade I mean
[00:41:09] <flyingparchment> there is a "zpool upgrade" when the on-disk format changes, which isn't very often
[00:42:54] *** jamesd_ has joined #opensolaris
[00:43:20] <FunkyELF> these downloads couldn't be more annoying
[00:43:34] <flyingparchment> untrue.  you could've been downloading it a few months ago.
[00:43:38] <flyingparchment> it had 5 parts then :)
[00:43:47] <FunkyELF> first there are 3 parts to a DVD and I can only download 2 parts at a time
[00:43:51] <FunkyELF> then they're zipped
[00:43:56] <FunkyELF> then I need to concatonate them
[00:44:07] <FunkyELF> so you need 3 times the space for a DVD to get it
[00:44:32] <FunkyELF> unless you unzip, concat, delete one at a time as you go
[00:45:52] <FunkyELF> zipping doesn't get you much anyway, I'd imagine that the livedvd image is compressed already
[00:48:22] <FunkyELF> So with ZFS, will compression kill my 1.33GHz AMD Athlon Thunderbird?
[00:48:38] <flyingparchment> no
[00:48:43] <flyingparchment> it'll probably be faster, less disk access
[00:48:57] <FunkyELF> cool
[00:49:16] <flyingparchment> the default compression uses a very quick algorithm.. zlib is available too
[00:49:49] <FunkyELF> does it take ZFS a long time to delete big files?
[00:49:58] <flyingparchment> no
[00:50:14] <Chipdancer> FunkyELF: surprisingly quick
[00:50:15] <FunkyELF> the only comparison I've read on different file systems was for when I was setting my my MythTV box and that seemed to be the big issue
[00:50:23] <flyingparchment> although iirc it's not quite as quick as QFS :)
[00:50:46] <flyingparchment> (QFS removes the inode and returns immediately, then deletes it in the background)
[00:50:47] <FunkyELF> so I wound up setting up jfs or some bullshit, then every time the power went out I'd have to fsck.jfs
[00:51:12] *** sparc-kly has joined #opensolaris
[00:51:13] *** ChanServ sets mode: +o sparc-kly
[00:52:13] <FunkyELF> it'll be interesting to give this opensolaris a test drive to see if it can replace my current fileserver
[00:52:36] <onlinebacon> FunkyELF, why didnt you use xfs?
[00:52:53] <FunkyELF> onlinebacon, don't remember....but I remember that I was between jfs and xfs
[00:53:07] <onlinebacon> xfs would have been better if you get powercuts a lot
[00:53:15] <FunkyELF> I just need Samba, and Gallery (requires Apache, MySQL, PHP)
[00:53:29] *** jamesd has quit IRC
[00:53:52] <flyingparchment> you'll probably want coolstack for that (sun's apache/mysql/php package)
[00:54:53] <onlinebacon> what's gallery?
[00:55:01] <flyingparchment> a web-based image gallery
[00:55:08] <onlinebacon> oh that makes sense :P
[00:56:17] <FunkyELF> oh, I need imagemagick too
[01:02:53] <Doc> bawhaw..  Skype reads /etc/passwd - it's obviously evil (according to slashdot anyway)
[01:05:41] <purserj> Why would skype need to read /etc/passwd?
[01:05:57] <delewis> getpwnam() or something would be my guess.
[01:05:59] <hali> good question, probably to send skype information about you
[01:06:14] <delewis> no, its probably grabbing your gecos.
[01:06:15] <hali> delewis: good old fopen()
[01:06:23] <flyingparchment> it's stealing your credit card info
[01:06:30] <hali> someone posted strace output
[01:06:35] <delewis> in any case, the Linux ls(1) does the same.
[01:06:37] <hali> on skypes forum
[01:06:38] <delewis> is ls eveeeel, too!?
[01:06:57] <hali> and it goes through your firefox profile and some other stuff as well
[01:07:01] <hali> big evil strace
[01:07:14] <delewis> probably seeing if you have the Firefox Skype plugin installed.
[01:07:29] <flyingparchment> stealing your form passwords
[01:07:52] <delewis> those should be encrypted.
[01:07:53] <Doc> or looking for proxy servers
[01:08:11] <Doc> are credit card numbers stored in /etc/passwd or /etc/shadow ?
[01:09:29] <delewis> hali: strace output can no longer be trusted.
[01:09:50] <delewis> the same goes for anything that interpositions syscalls.
[01:11:01] <delewis> it seems to me that if Skype were truly doing something malicious, the developers would've exploited that vulnerability.
[01:11:23] <delewis> that, or there as dumb as the people that think interpositioning system calls is somehow 'secure'
[01:11:27] <delewis> s/there/they're/
[01:12:51] *** karrotx has quit IRC
[01:14:18] *** CIA-16 has quit IRC
[01:14:55] *** jamesd__ has joined #opensolaris
[01:14:59] *** karrotx has joined #opensolaris
[01:20:28] *** Red_Cloud has joined #opensolaris
[01:22:09] *** Red_Cloud has quit IRC
[01:23:14] *** CIA-16 has joined #opensolaris
[01:23:17] <Pietro_S> I don't think that skype wants to do something malicious on linux, I would suggest them more on win platform, btw. collapse of skype network caused by windows auto-updates, made me laugh ;-)
[01:23:33] *** onlinebacon has quit IRC
[01:25:22] <purserj> followed by the crash of the Windows Genuine Advantage system
[01:26:41] *** jpdrawneek has joined #opensolaris
[01:30:55] *** jamesd_ has quit IRC
[01:35:50] *** Red_Cloud has joined #opensolaris
[01:36:00] *** pg_user has joined #opensolaris
[01:39:52] *** pg_user has quit IRC
[01:44:07] <Red_Cloud> g4lt-sb100 I have now installed firefox v2.0.0.6 from source and it is working wonderfully.
[01:48:35] *** freakazoid0223 has quit IRC
[01:52:24] *** jhawk has joined #opensolaris
[01:52:26] *** Red_Cloud has quit IRC
[01:54:46] *** uebayasi has joined #opensolaris
[01:57:04] *** jhawk has quit IRC
[01:58:43] *** Tpenta has joined #opensolaris
[02:04:51] *** jpdrawneek has quit IRC
[02:10:27] <Tempt> Doc: You're brilliant.
[02:10:42] <Tempt> Doc: If users keep their credit card details in /etc/passwd, we can bill them every time they annoy us.
[02:10:50] <Tempt> Doc: Or add some fields to /etc/shadow perhaps.
[02:11:05] <Tempt> I'd like a pam module which handles the billing on login.
[02:11:13] *** Pietro_S has quit IRC
[02:11:32] *** JSRJ has joined #opensolaris
[02:17:34] *** Red_Cloud has joined #opensolaris
[02:19:22] *** cajetanus has quit IRC
[02:20:45] *** Red_Cloud has quit IRC
[02:25:09] <bda> http://www.yankodesign.com/index.php/2007/08/24/backing-up-your-trash/ # ...
[02:26:23] <g4lt-sb100> you have got to be shitting me
[02:29:01] <Tempt> That's insane.
[02:29:09] <Tempt> How many layers of stupidity protection can one have?
[02:29:19] <dlg> not enough it seems
[02:29:49] * jmcp yawns
[02:30:18] <dlg> ola jmcp
[02:30:48] <jmcp> ola
[02:31:25] <bda> It's not even the "Extra Extra Trash" thing.
[02:31:28] <bda> It's *Bluetooth*.
[02:31:41] <Triskelios> wait, does that thing actually have an LCD outside showing how full it is?
[02:32:27] <Tempt> Hmm, bluetooth storage.
[02:32:33] <Tempt> That'd be speedy *cough, cough*
[02:32:35] <jmcp> dlg: time for me to log a p1 bug against the u20's bios update
[02:32:36] <g4lt-sb100> tryesw, yes it does
[02:32:51] <Tempt> jmcp: Hmm? Ultra20 problem?
[02:35:32] <jmcp> Tempt: I updated the bios in my u20 using the current version from sun.com/downloads ... now the bloody thing won't even POST
[02:35:39] <jmcp> Tempt: so I need a new motherboard, really
[02:36:20] <g4lt-sb100> jmcp, now you se why they basically forced you into a support contract for one ;P
[02:36:47] <jmcp> g4lt-sb100: this is work's u20 and it's still under warranty
[02:36:52] <Tempt> jmcp: That's a bit crap.
[02:36:57] <jmcp> yeah
[02:37:01] <Tempt> jmcp: What prompted the BIOS update?
[02:37:04] <Triskelios> there's no backup ROM in the U20s?
[02:37:05] <dlg> jmcp: any chance to recover?
[02:37:19] <jmcp> Tempt: It is/was about 5 revs behind current
[02:37:23] <jmcp> Trident: apparently not
[02:37:27] <jmcp> dlg: not that I can see
[02:37:33] <Tempt> jmcp: Man, this isn't like an OBP upgrade.
[02:37:37] <dlg> pwned
[02:37:42] <Tempt> jmcp: You don't flash BIOS unless there's a specific problem.
[02:37:45] <jmcp> afaik even if there's a backup rom option, you still need to be able to get past POST first
[02:37:57] <jmcp> Tempt: I know, and I thought I had one
[02:38:00] <dlg> hot plug the proms :)
[02:38:03] <Tempt> jmcp: Nah, sometimes the backup boot option requires twiddling a jumper.
[02:38:10] <jmcp> ie, the 2gb of ram in this box were showing up as 1gb
[02:38:16] <jmcp> I checked - I've got 2x1gb sticks
[02:38:21] <Tempt> That doesn't sound like a BIOS problem.
[02:38:27] <jmcp> dlg: not possible - soldered onto the board
[02:38:35] <jmcp> Tempt: quite possibly
[02:38:42] * jmcp shrugs
[02:38:53] <jmcp> this is the mobo with the dodgy pin in one of the dimm slots anyway
[02:38:54] <dlg> jmcp: you'll need side cutters then
[02:38:59] <jmcp> dlg: muahahahahaah no
[02:39:23] <Tempt> dodgy pin in a dimm socket and you thought the RAM not detecting was a BIOS bug?
[02:39:25] <dlg> at least its not your main machine
[02:39:34] * Tempt boggles
[02:39:36] <jmcp> Tempt: calm down
[02:39:37] <jmcp> dlg: true
[02:39:49] <jmcp> Tempt: in answer to your question *NO* I did not think that
[02:39:54] <dlg> pity it happened though
[02:39:58] <Tempt> Hey, at least you've got a contract on it and you can get a replacement board.
[02:40:01] <jmcp> true
[02:40:02] <dlg> end of august draws ever nearer
[02:40:06] <jmcp> once I've logged the bug
[02:40:11] <jmcp> dlg: day by day, indeed
[02:40:16] <Tempt> dlg: Is that a reference to no new Sol10?
[02:40:22] <dlg> no
[02:40:31] <dlg> its a reference to objections to an rfe i submitted
[02:40:38] <Tempt> Aah.
[02:41:23] <jmcp> Tempt: re the dodgy pin - read this http://blogs.sun.com/jmcp/date/20060609
[02:42:43] <Tempt> aah, that joy of FMA.
[02:42:47] <jmcp> yes
[02:43:01] <Tempt> I find it especially handy that it doesn't log some faults to syslog, only FMA.
[02:44:02] <Tempt> Otherwise, FMA is nifty.
[02:47:37] *** NikolaVeber has quit IRC
[02:51:29] *** ChanServ sets mode: +o Tpenta
[02:51:49] <jmcp> hi Tpenta
[02:51:56] <Tpenta> james
[02:52:05] <jmcp> Tempt: it's under continual improvement, fortunately
[02:52:50] <Tempt> jmcp: Mm. To be honest, I haven't had any failures on my Sun machines to generate any activity to really evaluate it. I guess that's a big win in itself.
[02:53:00] <jmcp> I totally agree
[02:53:03] <dlg> is it hard to write fma providers?
[02:53:04] <jbk> i've seen one
[02:53:15] <jbk> which itself was kinda amusing
[02:53:39] <jbk> since monitoring was handled by a separate group (who were largely uncooperative) fma events were not monitored
[02:53:52] <jbk> and in january, it had offlined a stick of ram in a v890
[02:54:00] <jbk> didn't notice it until april or may
[02:54:22] <jmcp> dlg: not sure, haven't had to write one
[02:54:53] <dlg> does fma make sense for things like raid controllers?
[02:55:04] <jmcp> yes
[02:55:11] <richlowe> Yes.
[02:55:19] *** _hsilva has joined #opensolaris
[02:55:22] <dlg> k
[02:55:40] <dlg> but mgmt needs a separate tool?
[02:56:12] <jmcp> perhaps
[02:56:13] <jamesd__> jbk, dont feel bad,  i did a scan of  prtdiag, and found  4 boxes with problems including  2 bad cpus and  6 dimms disabled....  and no one noticed for god knows how long.
[02:56:24] <jmcp> if you could tie in fma with management, that would be very handy
[02:56:32] *** simford has joined #opensolaris
[02:56:41] <g4lt-sb100> "does fault managment architecture need neeed a seperate amagement tool"?!
[02:56:43] *** kFuQ has quit IRC
[02:57:07] <jmcp> g4lt-sb100: I don't think that's what dlg was asking
[02:57:31] <dlg> creating volumes via fma doesnt make much sense to me
[02:57:35] <g4lt-sb100> jmcp, yes, he was.  just expanding the acronyms to illustrate the question
[02:57:38] *** kFuQ has joined #opensolaris
[02:58:07] <g4lt-sb100> dlg NOT creating volumes via fma makes a LOT of sense
[02:58:08] <jmcp> g4lt-sb100: ah
[02:58:27] <dlg> ah
[02:58:30] *** hohum has quit IRC
[02:58:35] <dlg> jmcp knows what i was asking though
[02:58:43] <g4lt-sb100> "create a volume using c0t0d0s0" "c0t0d0s0 ius broken, are you sure"?
[03:03:48] <g4lt-sb100> in fact, one day, when I have a LOT of round tuit, I may very well try to play with that.  poll fma for a NIC that boned up, if it does, don't send anyting vital on it
[03:05:09] <dlg> urgh
[03:05:20] <g4lt-sb100> fma events prolly aren't going to be as useful as system messages for complete failures, but impending failures can and should still be handled nicely
[03:05:26] <dlg> i have a natural aversion to any process that has polling as part of its design
[03:05:45] <dlg> nics already have mechanisms for dealing with failure
[03:05:57] <dlg> usually a watchdog picks up if they dont wrk and brings them down
[03:06:04] <dlg> the stack copes and tries to route around it
[03:06:05] <jamesd__> dlg, so you dislike all shells that sit there waiting for input (polling)
[03:06:15] <g4lt-sb100> well, yeah, espewcially since this particular program is pretty much speed sensitive.  I've never gotten it to handle GBe every well
[03:06:22] <dlg> the poll syscall is a bit of a misnomer
[03:07:22] *** hsilva has quit IRC
[03:07:27] *** _hsilva is now known as hsilva
[03:07:27] <g4lt-sb100> dlg a lot of my program bypasses a lot of extraneous junk from the kernel in an effort to speed things up
[03:07:49] <dlg> ...
[03:08:38] <g4lt-sb100> think of a filtering ethernet bridge and you'll be on the right track
[03:08:48] *** cajetanus has joined #opensolaris
[03:08:58] <dlg> yeah...
[03:09:02] * dlg still like the kernel
[03:10:50] <g4lt-sb100> no disagreement, but the moment that the data become available to userland, I need to take them, bnecause I have about 5-10k instructions to apply to them and then send them out on the other nic. do this millions of times a second and see where I live
[03:11:06] <jmcp> g4lt-sb100: you live in hell
[03:11:15] <g4lt-sb100> jmcp, yes, yes I do
[03:11:43] <dlg> cant you do it in the kernel?
[03:11:53] <dlg> you avoid the context switches and the copyin copyout stuff
[03:12:02] <Tempt> The bloat the kernel up.
[03:12:29] <dlg> so?
[03:12:36] <dlg> solaris has kernel modules
[03:12:37] <Triskelios> that's probably not an important concern
[03:12:44] <dlg> fi you're not using it you dont load it
[03:12:52] <g4lt-sb100> if I could just better manage kernelspace-userspace transitions, I 'd shave about half the instructions off :(
[03:13:04] <Tempt> If it screws up in userspace it dumps core; if it screws up in the kernel it dumps a much bigger core.
[03:13:44] <Triskelios> Tempt: if everyone thought like you, everyone would be using Mach
[03:13:51] <Tempt> Heh.
[03:13:57] <dlg> Tempt: it is a balancing act
[03:14:02] <Tempt> I just shudder at the linux extreme of putting webservers in the kernel.
[03:14:05] <dlg> all i can say is "so dont screw up"
[03:14:06] <Tempt> It is a balancing act indeed.
[03:15:19] <g4lt-sb100> dlg, Ted Sturgeon for you on line one....
[03:15:37] <dlg> who is ted sturgeon?
[03:15:55] <g4lt-sb100> Sturgeon's law: 90% of everything is crap
[03:16:17] <dlg> he's an optimist
[03:16:59] *** JSRJ has quit IRC
[03:17:26] <jbk> yet, for systemtap, weren't they trying to figure out some super-duper ultra-high speed kernel->user data transfer mechanism to feed the probe data out of the kernel, because processing it at the source of the probe (like dtrace does) would be 'bloat' ?
[03:18:33] <jamesd__> Tempt, last i checked there is at least one kernel http server, and maybe more now
[03:18:37] <jmcp> jbk: probably, and because their definition of "bloat" is fscked in the head
[03:18:51] <g4lt-sb100> I'd hope not, the kernel->user data mechanism in linux is horrid.  it's based on the x86 ring model, which could uyse a lot of help
[03:18:59] <Tempt> Huh? Kernel httpd in Solaris? Since when?
[03:19:15] <dlg> Tempt: no, lunix
[03:19:26] <Tempt> Oh, lunix has had two kernel httpds.
[03:19:27] <dlg> though im sure ive heard of solaris implementing a telnet server as a streams module
[03:19:32] <Tempt> One of which was renowned for panic'ing.
[03:19:40] <Tempt> Sorry, not panic, AIEEEE or whatever.
[03:19:47] <dlg> oops
[03:19:48] <fedorared> Linux did have a kernel httpd at one point.
[03:20:11] <jmcp> dlg: the idea of saying "oops" rather than "panic" to me indicates a certain lack of professionalism
[03:20:24] <dlg> jmcp: well, look at where it came from
[03:20:27] <Tempt> I wouldn't be surprised if the linux guys find a way to move GNOME into the kernel.
[03:20:28] <dlg> history has inertia
[03:20:31] <moazamraja> jmcp: oops.
[03:20:43] <jmcp> moazamraja: :-P
[03:20:46] <jbk> i like how the fact that null pointer dereferences don't kill things (in all cases) and it's considered a 'feature'
[03:20:46] <jmcp> dlg: exactly
[03:21:06] <dlg> jbk: on linux?
[03:21:11] <jbk> yes
[03:21:18] <dlg> yeah...
[03:21:22] <dlg> thats a tough one
[03:21:57] <g4lt-sb100> unless you're trying for a wiffleunix, which universally traps errors so you don't have to
[03:22:02] <jmcp> "I'm sorry, the device you are looking for disappeared, so your pointer is now null. I'll just keep on reading from what I think you want anyway...."
[03:22:44] <dlg> jmcp: bad code like that can exist anywhere :)
[03:22:49] <jmcp> yeah
[03:22:51] <Tempt> No, the bit that gets me is Linus having an aversion to crash dumps because it "makes it too easy".
[03:23:02] <jmcp> arrogance of the highest order
[03:23:13] <dlg> can we end this discussion?
[03:23:16] <dlg> im getting angry
[03:25:57] <Tempt> Okay.
[03:26:04] <jmcp> dlg: no stomach for aggro and adrenaline first thing in the morning?
[03:26:07] <jmcp> softy :-)
[03:26:12] <dlg> no
[03:26:13] *** xuewei has joined #opensolaris
[03:26:20] <dlg> im in my c++ lecture
[03:26:25] <dlg> i can only take so much pain
[03:26:28] <Tempt> To be honest, nothing they do to linux could make me angry, only make me laugh.
[03:26:33] <jbk> :)
[03:26:33] <jmcp> dlg: muahahahahahaha
[03:26:52] <Tempt> jmcp: They make him use Visual C++ ;)
[03:27:02] * jmcp hears echoes
[03:27:59] <dlg> Tempt: SHUT UP SHUT UP
[03:28:01] <g4lt-sb100> dlg s/cout/printf/ for much fun ;P
[03:29:05] <jmcp> dlg: don't forget to use c++ in the kernel
[03:29:15] <dlg> ha
[03:29:20] <dlg> linus isnt even that stupid
[03:29:38] <g4lt-sb100> and use /*comments*/
[03:29:39] * jmcp whistles while reading sc3.x manual set .....
[03:29:57] <dlg> jmcp: if you want a good reason not to use c++ in the kernel go read the ehci driver in macos
[03:30:04] * jmcp rotfl
[03:30:49] <g4lt-sb100> dlg make it a game to put as much ANSI C in your C++ programs as physically possible just to piss off the grader ;P
[03:30:58] <jmcp> dlg: I don't need any more of a reason than "wesolows is totally against it"
[03:31:11] <dlg> g4lt-sb100: its automated marking, they barely look at teh code
[03:31:21] *** neoxed has quit IRC
[03:31:22] <Tempt> Hah.
[03:31:24] <Tempt> That's all class.
[03:31:27] <dlg> i could probably embed lua in it and noone would notice
[03:31:58] <Tempt> Write it in pascal and use p2c
[03:32:07] <dlg> now we're just getting silly
[03:32:24] <g4lt-sb100> dlg this would be a perfect place to use #macros and see if you can't break the grader
[03:32:33] <dlg> the problem with this subject is it breaks my brain
[03:32:53] <dlg> it takes me a while to move back into working on kernel code
[03:33:00] * g4lt-sb100 will award many extra points to someone who's C can brak an automatic grader
[03:33:44] <jbk> #define BEGIN {
[03:34:02] <jmcp> jbk: nice
[03:34:24] * g4lt-sb100 will award many MORE extra points to someone who's C can break a human grader
[03:35:08] <Tempt> Write the entire thing in macrospeak and use #macros to make it work
[03:35:10] <Triskelios> I was just talking to someone about int main(x, y, z) { z = ... }
[03:35:21] <Tempt> nah
[03:35:36] <Tempt> int CAN(HAS, A, CHEEZBURGER)
[03:35:48] *** neoxed has joined #OpenSolaris
[03:36:00] <fedorared> Bah, just do it in lolcode
[03:36:18] <Tempt> Sadly, there are no viable lolcode implementations yet.
[03:36:32] <g4lt-sb100> for an automatic grader, I wonder if you can't carefully set up the output such that it does everything that's required and still hits a race condition or a buffer overflow in th grader
[03:37:07] <Tempt> Or just write the whole thing in a mess of inline SPARC assembly.
[03:38:17] <dlg> lecture over
[03:38:18] <dlg> bbl
[03:38:58] <jbk> apl
[03:39:05] <Tempt> Got rejected from a job I interviewed for a couple of weeks back. How rude.
[03:39:08] <Tempt> "Overqualified".
[03:39:09] <Tempt> Bah.
[03:39:26] <Tempt> In other words they want to pay peanuts and employ monkeys.
[03:39:39] <jmcp> Tempt: you wouldn't have liked working there then
[03:40:02] <Tempt> I know, but still, I'm not used to being told "no" in that situation. I can't remember when I last didn't get an offer from an interview.
[03:40:02] *** purserj has quit IRC
[03:40:11] * jmcp nods
[03:40:32] <Tempt> I probably would have declined the offer anyway because I'm not in dire need of a new job and I've got something much better in the pipeline
[03:41:07] <jbk> i'm still debating about this job i'm at now
[03:41:50] <jbk> they wanted a senior guy, but so far, all the work as been entry-to-mid level type stuff with one minor exception
[03:41:54] <Tempt> The Melbourne UNIX admin market has ramped the pay scale from an average of $70k-$85k two years ago to $90k+++ these days.
[03:42:20] <jbk> so i'm rather bored
[03:42:29] <Tempt> jbk: The place I'm at now hired me as a senior guy and I'm spending half my time staring at stack traces and doing basic deployment stuff.
[03:42:35] <jmcp> Tempt: what about for people who have skillz ?
[03:42:46] <jbk> that's better than what i'm doing
[03:43:01] <Tempt> jmcp: People with skillz are contracting at $700-$900 a day.
[03:43:11] <Tempt> jmcp: Hence I'm contracting at the moment :)
[03:43:11] <jbk> it seems 90% of the work is: cleaning up file systems when they get too full, and growing file systems
[03:43:19] <jbk> of course, on solaris, they don't use leadville
[03:43:24] <jbk> so it means rebooting
[03:43:34] <jmcp> Tempt: hmm
[03:43:47] <jbk> so they also do it on linux (not sure if it can handle additional san luns w/o rebooting or not)
[03:44:00] <Tempt> jmcp: You're contracting, right?
[03:44:05] <jmcp> yup
[03:44:14] <Tempt> The thing is, while the market is paying well, the work is dull
[03:44:19] <Tempt> The big money is from IBM
[03:44:27] <Tempt> and the work is dull, dull, dull.
[03:44:38] * jmcp shudders
[03:45:20] <Tempt> According to one pimp I was talking to recently, there's just thousands of linux weenies throwing their CVs in trying to get some lucre, and a major shortfall of Solaris skills.
[03:45:34] * jmcp smiles broadly
[03:45:45] <e^ipi> hooray!
[03:45:50] *** linma has quit IRC
[03:46:11] <Tempt> It's just like I predicted. The obsession with loonix decimated the UNIX admin pool, driving our salaries up.
[03:46:35] <e^ipi> like windows did a decade previous
[03:47:01] <Tempt> A lot of UNIX people have left the market - retired, moved onto other things, opened cafes, generally vanished.
[03:47:21] <Tempt> And there's a perception even with recruiters now that linux experience only = no experience.
[03:48:19] *** neoxed has quit IRC
[03:48:31] <Tempt> There was a shortfall of AIX admins a few years back and it drove AIX salaries to 25% more than any other UNIX.
[03:49:08] <e^ipi> and I own a pseries...
[03:49:09] *** whaq is now known as ysss
[03:49:10] <e^ipi> hooray!
[03:49:23] <jbk> e^ipi: yeah, but would you want to work on aix? :)
[03:49:34] <e^ipi> depends how much they're willing to pay me for it
[03:49:37] <Tempt> Why not? AIX 6 is looking feature-rich.
[03:49:52] <e^ipi> i unfortunately don't own a system that can run aix6
[03:49:57] <moazamraja> AIX 6 was looking feature-rich many years ago.
[03:50:16] <Tempt> I know that when this IBM deal stops hiring locally the rates for Solaris-only is going to slide quite a bit, so having platform diversity maximises earning potential.
[03:50:34] <Tempt> e^ipi: This is why I'm buying a p510
[03:51:03] <e^ipi> I don't have that sort of money, unfortunately
[03:51:22] <e^ipi> the only reason i've got what i've got is because  i picked it up from a sucker getting rid of it for $50
[03:51:40] <moazamraja> Thank the Lord most of american big operations run on Solaris.
[03:51:47] *** hile_ has joined #opensolaris
[03:52:05] <Tempt> I thought there was reasonable AIX penetration in the US?
[03:52:19] <Tempt> I know Australia's defence dept has a lot of AIX out there.
[03:52:21] <jbk> depends on the industry
[03:52:27] <moazamraja> finance is mostly solaris
[03:52:34] <moazamraja> web2.0 is linux
[03:52:35] <Tempt> Telco is always Solaris, finance is usually Solaris.
[03:52:40] <Tempt> Web2.0 might as well run on IIS.
[03:52:44] <moazamraja> web...linux/solaris
[03:52:49] <moazamraja> yeah, iis also
[03:53:03] <hile_> finance (at least from my point of view working for a large frim of that type) is split AIX/Solaris
[03:53:09] <hile_> Healthcare has as large AIX presence
[03:53:19] <hile_> defence has AIX and solaris
[03:53:21] <jbk> someone tell that to the developers I have to deal with :)
[03:53:22] <Tempt> Defence has AIX.
[03:53:35] <Tempt> A lot of shops moved from mainframes / as400s to AIX.
[03:54:14] <jbk> my old job had a little bit of aix
[03:54:14] <Tempt> A lot of SAP deployments, for example, land on AIX.
[03:54:20] <jbk> well at least until our old cio
[03:54:33] <Tempt> Every time I see someone wearing an "Armani Exchange" shirt I have to laugh.
[03:54:39] <Tempt> Walking around town with a giant AIX logo.
[03:55:13] <jbk> his 'IBM is CMMI level 5, we are only 3, therefore they are better' mentality led him to had IBM redesign the entire public website (which all went on AIX)
[03:55:31] <Tempt> Heh.
[03:56:07] *** matsutsu has joined #opensolaris
[03:56:35] <jbk> it still sucks :)
[03:56:48] <jbk> and was constantly having issues
[03:57:07] *** Lusitanian has quit IRC
[03:59:10] <Tempt> Don't blame the OS for that.
[04:00:30] *** matsutsu has quit IRC
[04:01:55] <jbk> well in a number of cases it was problems with how aix dealt with load
[04:02:02] <jbk> though often it was just crappy code
[04:02:42] *** neoxed has joined #OpenSolaris
[04:02:42] <Tempt> Hmm, AIX usually does quite well under load. Surprising.
[04:03:24] <Tempt> anyway, lunch
[04:03:25] <jamesd__> Tempt, its written by people that do mainframes... why wouldn't it handle loads...  aix boxes run  1000's of apps at once.
[04:08:23] *** kalou has quit IRC
[04:09:32] *** sehat_tan has joined #opensolaris
[04:09:58] <sehat_tan> hi. everybody in forum
[04:10:18] <sehat_tan> i am a newbie here want to ask some questions..
[04:10:34] *** xinkeT has quit IRC
[04:10:58] *** xinkeT has joined #opensolaris
[04:12:34] <jmcp> sehat_tan: please ask
[04:13:05] <jmcp> sehat_tan: we might ask you questions before answering - just to make sure that we know what you are really asking
[04:13:10] <jmcp> we don't like to confuse people :-)
[04:13:21] <sehat_tan> i just received 2 set of DVD Opensolaris
[04:13:42] <sehat_tan> when i read the manual and found that Opensolarisbased on Linux "Ubuntu"
[04:13:50] <jmcp> nope, that's not correct
[04:13:57] <sehat_tan> okay...please.
[04:14:13] <jmcp> the OpenSolaris DVD kit has several distributions of OpenSolaris on it
[04:14:46] <jmcp> one is Belenix (a GNU userland, I think), another is Nexenta (like Debian), then there's Schillix (just different) and Solaris Express (which is what Sun produces)
[04:15:01] <sehat_tan> yeah. i meant the Nexenta one
[04:15:09] <jmcp> right
[04:15:25] <jmcp> I haven't used Nexenta
[04:15:31] <sehat_tan> may i go through installatin (not from fresh install)
[04:15:44] <sehat_tan> because i have Ubuntu on my machine
[04:16:01] <sehat_tan> i think going installation on the Ubuntu maybe okay?
[04:16:21] <jmcp> you'll need a separate DOS primary partition to install into
[04:16:43] <sehat_tan> you meant make a new partition even on Ubuntu
[04:16:48] <jmcp> yes
[04:16:58] <sehat_tan> resize the disk and the rest going for nexenta?
[04:17:10] <jmcp> yes
[04:17:11] <jmcp> I would do that
[04:17:24] <jmcp> you need to use the "grub" which comes with your OpenSolaris distribution
[04:17:25] <sehat_tan> okay. thanks.
[04:17:29] *** blindfish has quit IRC
[04:17:50] <sehat_tan> and I also received SUN Solaris Developer Express from SUN itself
[04:17:59] <jmcp> yeah
[04:18:08] <jmcp> that will have the Sun Studio Compilers on it
[04:18:10] <sehat_tan> what's the difference between Opensolaris and Solaris Developer edition?
[04:18:20] <jmcp> OpenSolaris - technically - is just a bunch of code
[04:18:37] <jmcp> Solaris Express (either Developer or Community Editions) is an actual distribution
[04:18:38] <sehat_tan> can i find SUN compiler in Opensolaris?
[04:18:41] <jmcp> yes
[04:18:43] <jmcp> well ...
[04:18:54] <jmcp> it's in the Developer Edition dvd
[04:19:08] <jmcp> we usually refer to the developer edition as "SXDE"
[04:19:16] <jmcp> and community edition as "SXVE"
[04:19:19] <jmcp> sorry, "SXCE"
[04:19:53] <sehat_tan> the difference between them is one that packaged with the tools for developer, right?
[04:20:16] <jmcp> I think that is the case
[04:20:17] <e^ipi> the difference is one is older
[04:20:22] <jmcp> that too
[04:20:27] <jmcp> SXCE comes out more often
[04:20:30] <e^ipi> SXCE comes with studio too, does it not?
[04:20:46] <jmcp> dunno
[04:21:15] <jmcp> sehat_tan: sorry, I'm not giving you very good answers :(
[04:21:20] <jbk> if not, it's a free download
[04:21:45] <jmcp> sehat_tan: you'll probably find this blog posting useful to get started with some of the terminology:: http://whacked.net/what_is_opensolaris
[04:21:48] *** linma has joined #opensolaris
[04:22:35] <sehat_tan> do u mean that prefer to install a SXCE than the SXDE because the SXCE has more new features on it?
[04:22:53] <jmcp> I would go with SXCE because it is more up to date
[04:23:00] <jmcp> SXDE only comes out every few months
[04:23:09] <jmcp> sehat_tan: I have to go out for a bit, but there are plenty of people here who can help you
[04:23:18] <jmcp> sehat_tan: thankyou for joining us here
[04:24:28] <sehat_tan> okay thanks for your annswers
[04:24:34] <sehat_tan> and nice to chat with you..
[04:24:55] <sehat_tan> hope can learn and share more 'bout opensolaris with you further..
[04:25:49] *** sparc-kly_ has joined #opensolaris
[04:26:21] *** sehat_tan has quit IRC
[04:27:06] <boyd> Gah! Can somebody tell me what exactly *is* on-topic for opensolaris?
[04:27:15] <boyd> err... -discuss
[04:27:24] <jbk> haha
[04:28:52] <boyd> It seems that every community need some forum for general, loosely on-topic, chit-chat. It further seems that there is nowhere more like that place in this community than opensolaris-discuss.
[04:29:17] <palowoda> Got to have one trash barrel I guess.
[04:30:04] <boyd> I think that expecting big parts of the OpenSolaris community to not have a conversation about SUNW->JAVA is unreasonable
[04:30:55] <palowoda> Than suggest a opensolaris-garbage-bin-disscuss list.
[04:31:08] <richlowe> we have one, opensolaris-discuss
[04:31:16] <palowoda> Like I said.
[04:31:20] <richlowe> just change it's name, and we're set.
[04:31:31] <palowoda> Actually that is a good idead.
[04:31:36] <palowoda> idea
[04:32:14] <palowoda> Next topic for steve to bring up at OBG.
[04:34:48] <palowoda> sorry ogb.  I get it backwards sometimes.
[04:35:01] <boyd> bgo?
[04:35:11] <palowoda> ogbyn?
[04:36:23] *** tamr has joined #opensolaris
[04:39:27] *** sparc-kly has quit IRC
[04:42:54] <delewis> new reason to stop using acroread -- dtrace -n 'syscall:::entry /execname == "acroread"/ { @[probefunc] = count();'
[04:43:15] <delewis> in a period of 4 or 5 seconds acroread, while idle, was responsible for ~ 12000 syscalls.
[04:43:32] <hile_> holy shit..
[04:43:38] <ysss> sounds like it's working hard..
[04:43:51] <delewis> while it should be doing nothing.
[04:43:54] <jbk> but doing what?
[04:44:07] <ysss> to kill himself and the host
[04:44:27] <hile_> hey delewis
[04:44:31] <delewis> http://pastebin.com/m75a42bb6
[04:44:32] <hile_> check out /msg :)
[04:45:55] <axisys> delewis: what is it doing? try @[ustack()]
[04:48:00] *** sparc-kly_ has quit IRC
[04:48:32] <delewis> http://pastebin.com/m2c7d091f
[04:48:36] <delewis> there's some rwsnoop output
[04:48:50] *** sparc-kly_ has joined #opensolaris
[04:49:56] <axisys> delewis: do u have a pdf file open at the moment?
[04:49:59] <jbk> so looks like perhaps some poorly designed x interaction?
[04:50:17] <delewis> axisys: several.
[04:50:51] <axisys> delewis: i guess jbk said it is paranoid..
[04:50:52] <delewis> its doing about 3000-4000 syscalls/second, though.
[04:51:16] <delewis> fortunately, not a lot of xcalls.
[04:51:24] <axisys> delewis: some paranoid app .. isn't it?
[04:51:43] <axisys> is my file open x 4000 .. hehe
[04:52:00] <delewis> and why the fuck is it doing getuid() over and over.
[04:52:06] <delewis> I'll do a ustack() there.
[04:52:53] <axisys> u can also do ustack(),stack() to see both on same aggr
[04:53:47] <jbk> of course, even if you figure out the problem, think adobe would actually fix it?
[04:54:12] <axisys> jbk: prob'ly not.
[04:54:32] <axisys> i usually use xpdf or evince.. seems a lot faster
[04:55:00] <delewis> looks like its all glib's fault.
[04:55:11] <delewis> its generating all the pollsys's.
[04:55:12] <hile_> well we know glib/gtk suck :)
[04:55:20] <richlowe> seems more likely acrocread is provoking glib to do it
[04:55:43] <delewis> libX11 is responsible for the read()s and write()s
[04:56:22] <delewis> http://pastebin.com/m3250aa28
[04:58:09] <delewis> they must be doing something funky with glib, as I've never seen another application using glib doing that many syscalls/second.
[04:58:18] * boyd loves acroread... it's a perfect butt of jokes on dtrace courses
[04:58:55] <delewis> boyd: really? first I've heard about acroread doing interesting things. I was just looking at mpstat(1) output and noticed a load of syscalls.
[04:58:58] *** jumpi has joined #opensolaris
[04:58:59] <jbk> it looks like some sort of callback
[04:59:27] <delewis> did a dtrace -n 'syscall:::entry { @[execname] = count(); }' and *poof* there was acroread in first place.
[04:59:41] <boyd> delewis: It seems always to be near the top for me whan I run soemthing like that
[04:59:53] <delewis> xmms scored fairly high as well.
[05:00:04] <delewis> but I suspect it might be doing sound-related polling.
[05:00:16] *** rottle has quit IRC
[05:00:45] <delewis> in any case, the read()s and write()s acroread is doing aren't causing any mjf's which is good, too.
[05:01:08] <delewis> still bugs me, though.
[05:03:22] <axisys> delewis: how many acroread instance are u running?
[05:03:33] <axisys> @[execname,pid] that is
[05:03:48] <axisys> and then may be use pid$PID provider to get info
[05:04:00] <axisys> more info that is
[05:04:22] <delewis> axisys: 2, I think. one real instance, and one that was spawned by Firefox via the acroread plugin.
[05:04:41] <axisys> what does @[execname,pid] tell u? 2 ?
[05:05:14] <delewis> http://pastebin.com/m442a0968
[05:05:20] <delewis> there's something funky its doing with getuid().
[05:05:22] <delewis> axisys: one sec.
[05:05:49] *** linma has quit IRC
[05:06:16] <delewis> http://pastebin.com/mc607cc4
[05:06:43] <delewis> the above rwsnoop I pastebin'd was run on both pids
[05:06:46] <boyd> delewis: It's in case it got r00t so it can install the backdoors
[05:06:51] <delewis> one was doing some sort of X11 polling
[05:07:04] <delewis> and the other pid was read()ing and write()ing to <unknown>
[05:07:12] <delewis> anonymous memory perhaps?
[05:07:37] <axisys> well pid18531:::entry give u anything interesting?
[05:07:40] <delewis> in any case, acroread is doing whatever it's doing when its not supposed to be doing anything (minimized)
[05:09:04] <delewis> it's taking dtrace awhile to instrument acroread :-)
[05:10:07] <axisys> delewis: with pid? yes because there are probably anout 7000 probes :-(
[05:10:51] *** linma has joined #opensolaris
[05:10:52] <delewis> dtrace: description 'pid18531:::entry ' matched 242416 probes
[05:10:53] <delewis> :-)
[05:11:33] <axisys> hehe
[05:12:02] <axisys> u can change that may be by adding one one of the function u found on ustack
[05:12:17] <axisys> pid18531::FUNCTION:entry
[05:12:29] <axisys> s/change/reduce/
[05:14:36] <axisys> pid18531::UnixAppMain:entry
[05:15:37] *** sponix has joined #opensolaris
[05:16:28] <delewis> http://pastebin.com/m30813131
[05:16:43] <delewis> the frame starting at line 41 is particularly interesting
[05:18:12] * delewis has a new fondness for gtk/glib applications
[05:18:28] *** jumpi has quit IRC
[05:19:19] <Tempt>   java                                                           5110
[05:19:22] <Tempt> one second ...
[05:19:24] <Tempt> geez.
[05:21:43] *** sartek has joined #opensolaris
[05:24:13] *** karrotx has quit IRC
[05:27:20] *** rootard has joined #opensolaris
[05:30:25] *** neoxed has quit IRC
[05:32:10] *** sarah has quit IRC
[05:33:46] *** tamr has quit IRC
[05:34:58] *** phus has joined #opensolaris
[05:35:32] *** uebayasi has quit IRC
[05:41:26] *** neoxed has joined #OpenSolaris
[05:51:08] *** sartek has quit IRC
[05:51:44] *** coraline has joined #OpenSolaris
[06:00:59] *** phus has quit IRC
[06:01:56] *** karrotx has joined #opensolaris
[06:07:52] *** libkeiser has joined #opensolaris
[06:08:49] <dlg> richlowe?
[06:19:12] *** ashner has joined #opensolaris
[06:22:09] *** Gman has quit IRC
[06:27:11] *** Gman has joined #opensolaris
[06:32:50] <jbk> wb
[06:36:04] <Gman> hrm, my keyboard has started to wedge itself into block letters until i thump it again :(
[06:36:16] <Tempt> Hmm, want a Lifebook U1010.
[06:36:20] <Tempt> That is one nice looking toy.
[06:39:17] <Chipdancer> Tempt: wouldn't be afraid of breaking it?
[06:44:48] *** daninj has joined #opensolaris
[06:44:49] *** nostoi has joined #opensolaris
[06:45:15] <daninj> anyone know of some sun hardware refurbishers in .au ? point me in the right direction!
[06:45:22] *** daninj is now known as andyshack
[06:51:54] <Tempt> Chipdancer: Naah, I only trash phones
[06:52:28] <Chipdancer> andyshack: unlikely - Sun Ripened Kernels was put out of business by SUN via legal action if I understand correctly
[06:52:33] <Chipdancer> andyshack: your best bet is ebay
[06:53:33] <Tempt> SRK was *not* put out of business by Sun.
[06:53:52] <Tempt> SRK went out of business because Craig couldn't find *anyone* who wanted to pay $1000 for a SPARCstation in 2004.
[06:54:12] <Tempt> andyshack: Is this for commercial or personal use?
[06:54:22] <andyshack> Tempt : both
[06:54:39] <Tempt> andyshack: Interactive (www.interactive.com.au) sell service contracts for older Sun hardware and have a parts inventory.
[06:54:44] <Tempt> andyshack: But they charge business rates.
[06:55:03] <andyshack> great party thanks.
[06:56:09] <jmcp> Tempt: I'm fairly sure there was at least threatened legal action
[06:56:18] <jmcp> but I don't know whether it ever made it to court
[06:56:20] <jmcp> not that I care, really
[06:56:41] <Tempt> There was threatened legal action but they didn't put him out of business.
[06:57:04] <Tempt> I think his behaviour had a role in making SunSolve handbook access contract-only.
[06:57:29] <Tempt> andyshack: Otherwise, look on ebay or ask here if you're looking for anything in particular
[06:57:31] <andyshack> after todays run around with both sun and some of their local resellers i am now questioning the viability of sticking with proprietary hardware. It doesnt appear worth it unless im dumping over 6 digits a year into a datacenter.
[06:57:49] <Tempt> andyshack: Are you dealing with anyone other than frontline.com.au?
[06:58:17] <jmcp> andyshack: "proprietary" is defined for you as "not a PC" ?
[06:58:29] <Tempt> s/PC/whitebox PC/
[06:58:30] <Tempt> heh.
[06:58:47] <jmcp> Tempt: yeah, "see also: crap"
[06:58:56] <Tempt> Australia has one Sun reseller worthy of your coin, and that's frontline.
[06:59:02] <Tempt> I wouldn't do business with anyone else.
[06:59:05] <andyshack> ill speak with them thanks.
[07:02:30] *** uebayasi has joined #opensolaris
[07:02:41] <LeftWing> jmcp: That would appear to be an unfortunately common definition.
[07:03:08] <Tempt> White Box Yum Cha PC = Crap PC.
[07:03:13] <Tempt> Yep, that's a reasonable choice.
[07:04:22] <LeftWing> Well, the only reseller experience I've had personally was buying Sun Ray 2s... and that was pretty painless.
[07:04:44] <LeftWing> Thought just about everyone with a blog has a horror story to tell these days.
[07:04:50] <LeftWing> *Though
[07:05:51] <e^ipi> hey hey
[07:06:04] <Tempt> LeftWing: Which reseller?
[07:06:16] <LeftWing> Tempt: Arion. Local to Newcastle as far as I'm aware.
[07:06:31] <Tempt> Aah, k.
[07:06:32] <andyshack> tempt : you are correct. theyre helpful
[07:06:49] <Tempt> andyshack: Which state are you in?
[07:07:03] <andyshack> i sit in wa
[07:07:08] <Tempt> Aah, k.
[07:07:29] <Tempt> So they're taking care of you? Well, you just need to wait for the hardware to arrive and eat the chocolate they ship with every machine
[07:07:52] <andyshack> oh shit, Theyre the people i head you mention ship the chocolate
[07:08:09] <Tempt> If I had a dollar for everyone I sent to FrontLine that was happy with the service, I'd have ... 37 dollars.
[07:08:25] <Tempt> If I had 1% of those sales, I'd have ... lots and lots of dollars.
[07:08:34] <Tempt> andyshack: Yep, these are the candy people.
[07:08:37] <jmcp> :)
[07:10:07] <Tempt> caaaaaaaaaaaaandies.
[07:10:14] <andyshack> i had a survey done many years ago to look into the viability of sending beer with orders. surprisingly people weren't wanting free beer. ;/
[07:11:18] <Tempt> That's a bit odd.
[07:11:29] <Tempt> Although I think people prefer a rep to come out and buy beers at the pub.
[07:15:02] *** BatonT has joined #opensolaris
[07:16:41] *** Me[-_-]gaf has joined #OpenSolaris
[07:18:55] <e^ipi> Tempt: got a second ?
[07:23:29] <Gman> mmm, anil's doing some nice things with the usbdump stuff
[07:30:21] *** estibi has quit IRC
[07:30:38] *** monzie has joined #opensolaris
[07:30:53] <Tempt> e^ipi: Sure.
[07:30:56] <Tempt> e^ipi: What's up?
[07:32:03] <e^ipi> what could be the cause of this... I've got 20 second shots, but the first 2 seconds or so are gross and thick black, like dirty oil coming from a car
[07:32:13] <e^ipi> after then they look good, and the end product is okay
[07:33:06] <Tempt> Hmmm
[07:33:09] <Tempt> Let's have a look.
[07:33:26] <Tempt> You shouldn't be seeing any output for the first 3-4 seconds.
[07:33:40] <e^ipi> no, i mean when the shots start pouring
[07:33:51] <Tempt> Aah, okay, so your total shot time is around 25 seconds or so?
[07:34:01] <e^ipi> no, it's 20 seconds-ish
[07:34:15] <Tempt> I'm talking from pump-on to end of the line
[07:34:29] <e^ipi> the first 2 seconds is nothing, then another couple seconds of gross tarry blackness, then a proper looking shot
[07:34:57] <Tempt> Are you dosing "on the line" or above?
[07:35:27] <e^ipi> I dose it to the top of the basket
[07:35:59] <Tempt> Okay.
[07:36:08] <Tempt> Try taking your grind a little finer and dosing a little less.
[07:36:21] <Tempt> It could be related to too much heat from the shower screen and not enough room for the grounds to expand.
[07:36:42] *** Me[-_-]gaf has quit IRC
[07:42:30] <Tempt> e^ipi: Well, did that help?
[07:47:25] <e^ipi> yeah, a little better
[07:47:36] <e^ipi> I'll keep tweaking, I think you're right
[07:47:41] <e^ipi> thanks :)
[07:48:13] * steleman_away is back.
[07:48:18] *** steleman_away is now known as steleman
[07:49:51] <Tempt> When you updose too much you risk burning the top of the puck before the water can penetrate down through the rest of the coffee.
[07:50:27] <Tempt> Which can lead to turbulent flow through the puck causing micro-channelling. Hence getting some sump oil dripping at the start followed by a stream.
[07:51:08] <e^ipi> *nod*
[07:52:23] *** JSRJ has joined #opensolaris
[07:56:05] <e^ipi> espresso and geeks are a perfect match....
[07:57:37] <trochej> tru
[07:57:42] <Tempt> Espresso is the perfect blend of art and science.
[07:57:43] <trochej> To a very scary extent
[07:57:53] <Tempt> And there's no end of stuff to geek on about with espresso.
[07:58:12] <Tempt> I've still got a PID sitting on my desk from when I had a Silvia.
[08:05:22] *** phimic has joined #opensolaris
[08:08:26] *** cajetanus has quit IRC
[08:11:15] *** cajetanus has joined #opensolaris
[08:15:36] *** timeless has joined #opensolaris
[08:16:32] *** tsoome has quit IRC
[08:16:46] *** _d3 has quit IRC
[08:17:58] *** nostoi has quit IRC
[08:24:13] *** micken has joined #opensolaris
[08:25:25] *** JSRJ has quit IRC
[08:28:52] *** mbalmer has quit IRC
[08:34:21] *** Dar has joined #opensolaris
[08:36:14] *** cajetanus has quit IRC
[08:36:52] *** cajetanus has joined #opensolaris
[08:40:00] <sponix> Who makes the Best Drives ? Seagate, Hitachi, Samsung, Fujitsu ?
[08:41:31] <coraline> why don't you go to each manufacturer's website and look up their specs and MTBF ratings?
[08:41:38] <coraline> and do an actual comparison.
[08:42:26] <coraline> instead of asking for....someone's completely uneducated opinion.
[08:43:04] <jmcp> sponix: it often seems to come back to personal biases, assuming that you're looking at the same class of disk (eg, enterprise FC/SAS, or enterprise SATA, or consumer SATA)
[08:43:15] *** nostoi has joined #opensolaris
[08:43:35] <sponix> well its a question that prompts for personal opinion/bias
[08:43:52] <coraline> when I worked at Sun, I replaced a lot of Seagates.  But that's because Sun used a lot of Seagates.
[08:43:53] <sponix> I am looking at Cheap Ass IDE ATA100
[08:44:04] <coraline> THe only drives I've had fail on me personally at home, are Western Digital.
[08:44:12] <jmcp> if I'm paying out of my own pocket, I tend to go for Seagate. If I'm getting work to pay .... Hitachi or Fujitsu depending on connection type
[08:44:21] <jmcp> coraline: Maxtor have failed for me at home
[08:44:40] <sponix> I have went through more maxtor drives than I can shake a stick at
[08:44:46] <sponix> and most of them didn't last long either
[08:45:04] *** JSRJ has joined #opensolaris
[08:45:14] *** cajetanus has quit IRC
[08:48:07] <sponix> Right now I have 2x Seagate and 1x Hitachi that are going into the software raid, 2x Highpoint Rocket100 ATA100 Cards
[08:49:47] *** neoxed has quit IRC
[08:49:57] <palowoda> The Highpoint Rocket works with Opensolaris?
[08:50:00] *** neoxed has joined #OpenSolaris
[08:50:28] *** tg has quit IRC
[08:50:44] *** tg has joined #opensolaris
[08:51:04] *** dlynes_laptop has joined #opensolaris
[08:51:48] <sponix> not 100% sure _yet_
[08:52:09] <sponix> adaptec was 4x higher and out of stock
[08:53:28] <dlg> get an areca!
[08:54:13] <sponix> wtf is that
[08:54:29] <dlg> www.areca.com.tw
[08:54:31] <palowoda> A nice controller for Solaris.
[08:55:15] <trochej> I got me nice 200 GB 2,5" hdd
[08:55:19] <trochej> yummy
[08:55:25] <trochej> 7,2 krpm
[08:55:36] <sponix> came out with 250G 2.5 recently
[08:55:39] <palowoda> Hey how much was that trochej
[08:56:22] <jmcp> I really wish that 2.5" disks were the standard rather than 2.5"
[08:56:47] <palowoda> you mean 3.5's?
[08:57:56] <jmcp> yeah
[08:58:00] <jmcp> thinko
[08:58:57] *** tsoome has joined #opensolaris
[08:59:06] <palowoda> But the standard your talking about is tied to a volume where OEM force the footprint on you.  Maybe too much price dependent right now.
[08:59:23] <sponix> dlg: 2 issues, 1) Don't see anything ATA100 at newegg 2) I only shop at newegg
[08:59:35] <dlg> ata100?
[08:59:40] * dlg look confused
[08:59:43] <palowoda> Old stuff you know.
[09:00:06] <dlg> i didnt read enough history
[09:00:24] <dlg> id try and get an ata100 controller with the SiI680 chip on it
[09:00:32] <dlg> but any generic ata controller should just work
[09:00:36] <sponix> I have 3-5 ATA/PATA100 (EIDE) drives already
[09:00:48] <trochej> palowoda: I don't know, it's birthday gift from my company. :)
[09:00:54] <dlg> are you looking for a raid controller?
[09:00:57] <dlg> or just ata?
[09:01:04] <palowoda> EIDE is kind of dieing out right now.
[09:01:22] <sponix> Just ATA, doing software raid, drive brand/type is mismatched
[09:01:48] <dlg> sii680 then
[09:01:58] <dlg> but it really doesnt matter
[09:02:08] <sponix> palowoda: I know, but tossing 400-$500 worth of drives seems silly
[09:02:10] <BadKarma> trochej: a b-days gift... darn my cmpny never gave my any gift out
[09:02:53] <palowoda> sponix: You mean your considering tossing the controller not the drives.
[09:03:13] <trochej> BadKarma: :)
[09:03:22] <trochej> BadKarma: And to think I want to leave them :)
[09:03:30] <sponix> I have at least 5x ATA100 500G drives right now
[09:03:53] <BadKarma> I see, nothing further
[09:03:58] <sponix> Several of which are in external shells... I want to make better use of them by putting them into a software raid
[09:03:59] <Tempt> Heya, has anyone found any wierd behaviour using -ctime +X with find to grab files old than X days?
[09:04:00] <palowoda> And ATA only has 4 channels.
[09:04:07] <Tempt> doesn't seem to work properly with more than 7
[09:04:25] <sponix> 4 channels ? ... Per Controller ... riiight ?
[09:04:36] <palowoda> Oh your talking about the Hotpoint controller that works on opensolaris right?
[09:06:11] <palowoda> What model of Adaptec ATA controller did you mention before?
[09:06:29] *** Erwann has joined #opensolaris
[09:06:45] <boyd> Tempt: more than 7 days?
[09:07:08] <Tempt> yeah, more than 7 days
[09:07:13] <sponix> palowoda: there is only one at newegg that isn't raid .. about 65 dollars
[09:07:27] <Tempt> Perhaps I'm tried and my brain isn't working properly and I'm looking at a line from one of our DBAs...
[09:07:31] <palowoda> URL?
[09:07:38] <Tempt> But ... wierdness abounds.
[09:07:39] <sponix> palowoda: one sec
[09:08:02] <sponix> http://www.newegg.com/product/product.asp?item=N82E16816115009
[09:08:11] <sponix> that is highpoint rocket100
[09:08:16] <sponix> brb with adaptec
[09:08:23] <sponix> only other option is promise
[09:09:29] <sponix> http://www.newegg.com/Product/Product.aspx?Item=N82E16816103023
[09:09:39] <palowoda> I don't think Promises ATA controllers work with opensolaris either.
[09:09:58] *** mikefut has joined #opensolaris
[09:10:14] <palowoda> Err that Adaptec only has 4 ports.
[09:11:40] <palowoda> I'd check the Adaptec site see if they have drivers for Opensolaris in that case.
[09:12:57] <palowoda> Adaptec, Promise, Hotpoint they all are kind of into closed drivers specially for the old stuff.
[09:14:45] *** cydork has joined #opensolaris
[09:15:50] <palowoda> Sad the good old days of ATA is dieing a quicker death.  But rejoice we have cheap SATA. :)
[09:19:46] <palowoda> I guess the world would have been better off if Richard Staleman gave out free hardware rather than software. :)
[09:22:37] <palowoda> By the way Sponix just use a combination of ATA and SATA.  I'm sure you can find a controller based on some cheap 3124 Silicon Image chipset to compliment your systems storeage in the future.
[09:22:59] <sponix> I like that SATA is cheap, and that it offers a faster interface, but from whave I've seen in benchmarks current drives don't take advantage of it
[09:23:20] <palowoda> What benchmarks are you talking about?
[09:23:26] <sponix> PATA && SATA drives are maxing out at 80-85MB/s reads roughly
[09:23:56] <palowoda> I have three SATA drives with ZFS and I'm getting in the 120Meg/sec writes.
[09:24:08] <sponix> palowoda: you would ask me that after I just closed it 15min ago :(
[09:24:22] <palowoda> And that is with raidz ZFS.
[09:25:00] <sponix> what is pretty sweet... what system specs && controller(s) ?
[09:25:24] <sponix> what do you get for reads ?
[09:25:29] <palowoda> http://www.fiver.net/misc/solarisamdbox.html  via chipset.
[09:26:05] <renihs> you get 125mb/s ....with that?
[09:26:11] <palowoda> Reads are typically buffered too high.  Why are you concerned?
[09:26:12] <renihs> skeptical me is
[09:26:13] <sponix> honestly, if I can write at 10MB/s and read at 20MB/s consistantly, I'll be very happy
[09:26:36] <palowoda> God Lord are you outdated.
[09:27:20] <palowoda> It is the year 2007.
[09:28:03] <sponix> how do you get zfs raidz out of 2x 500G maxtors ?
[09:28:14] <palowoda> 3x500G
[09:28:21] <sponix> ha, wrong box :P
[09:28:24] <sponix> didn't read down...
[09:28:45] <palowoda> Friggen cheap for the performance too.
[09:29:54] <palowoda> You want better on SATA just get the onboard cache up or get the 1T 32meg Hitachi drives.
[09:30:16] *** lisppaste3 has quit IRC
[09:30:25] <sponix> On Board Cache ... On the Motherboard ?
[09:30:36] <palowoda> On board cache on the controller
[09:30:50] <sponix> 500G Seagates for $99 is good, paying $119 for them at newegg
[09:31:21] <palowoda> Oh wow now where talking about 20.00.  Wait a few days.
[09:31:26] <sponix> you like AMD chips ?
[09:31:42] <palowoda> I hate all hardware because I have to pay for it.
[09:31:43] <quasi> mmmm, amd
[09:32:01] <Triskelios> palowoda: I hear you, brother
[09:32:40] <e^ipi> palowoda: go to the office part of town and start jumping in dumpsters
[09:32:55] <e^ipi> occasionally you stumble across some great equipment
[09:33:30] <palowoda> e^ipi: I'm too old for that now.
[09:35:27] <palowoda> If you want some universal opinion I'm sure we all could agree that power costs are the root of all evil now.
[09:35:29] *** lisppaste3 has joined #opensolaris
[09:36:05] <Tempt> e^ipi: How did your coffee go?
[09:36:35] <andyshack> god i hate paying the power bill
[09:37:07] <e^ipi> Tempt: well, thank you :)
[09:37:27] <e^ipi> real challenge is to see if I can't replicate it tomorrow
[09:37:40] <andyshack> the power bill is the single restriction that prevents me from playing with a 4 acre farm of retro stuff.
[09:37:57] <Tempt> andyshack: How bad is your power bill?
[09:38:36] <andyshack> dev room that i use as my personal teach yourself phd is setting me back about 500au pm running at about 30%
[09:38:53] <Tempt> Jesus
[09:38:54] <andyshack> i cant really justify booting it all up at once
[09:38:56] <Tempt> And the rest of your house?
[09:39:08] <andyshack> nfi
[09:39:17] <Tempt> My power bill is $AU900/quarter
[09:39:22] <Tempt> and that includes a V880!
[09:39:37] <palowoda> Tempt: That is a little outragoues isn't it?
[09:39:44] <andyshack> residential rates ?
[09:39:49] <andyshack> that seems cheap
[09:40:07] <andyshack> im still trying to work out whats eating the juice although i dont really ahve the time.
[09:40:30] <Tempt> palowoda: Not really.
[09:40:37] <Tempt> palowoda: No gas in this place.
[09:40:42] <palowoda> Tempt: Oh wait that was per quarter.
[09:41:06] <andyshack> my interim solution is to just leave what im not playing with switched off. id love to pick up a few 4800's or other fridge like devices, yet i cant justify it as i have really no idea what im doing :)
[09:41:07] <Tempt> Yes, per quarter
[09:41:26] *** WickedWicky has joined #opensolaris
[09:41:28] <Tempt> includes running my 3.5kW coffee machine for at least an hour a day during the week and most of the weekend
[09:42:04] <palowoda> Tempt: It's your refigerator or freezer that gets you.
[09:42:39] <sponix> palowoda: 7600GS ... that isn't supported by the nvidia drivers on Solaris/OpenSolaris is it ?
[09:42:54] *** bunker has joined #opensolaris
[09:43:01] <Tempt> andyshack: So what machines are costing you five hundred a month in power bills? You got some 11/780s in there?
[09:43:45] <palowoda> sponix: Hey I'm using them on three opensolaris machines.  Enjoy several streams of dvd video on 24inch LCD's.  What does support mean.
[09:43:56] *** UGH has joined #opensolaris
[09:44:15] <sponix> But you are using standard X11 nv to support them?
[09:44:38] *** Ljuvnix has joined #opensolaris
[09:44:45] <palowoda> I use Nvidia OpenGL drivers.  Why would I use nv drivers?
[09:45:00] <renihs> security reasons only if at all
[09:45:13] <renihs> :p
[09:45:51] <andyshack> i think im mainly getting stung from the emc distro boards. 5 racks, 2 in each sucking from the 3 phase. i currently have nothing more modern than a 280r, although there are lots of disks that like to spin. im also guessing they eat juice as scsi doesnt spin down unless its off.
[09:46:41] <UGH> and your cooling
[09:46:42] <Tempt> You're running 3 phase in to power up machines lighter than 280R?
[09:46:59] <Tempt> Distro boards should not consume power anyway.
[09:47:17] <andyshack> for power load balancing to distribute over each phase.
[09:47:41] <Tempt> you need to get some power measuring PDUs
[09:47:49] <andyshack> the fat assed loaded 450's seem to like to suck it down.
[09:48:02] <andyshack> ohh yes i need some bad. its on the list of things to do.
[09:48:33] <UGH> Those Cyclade powerswitches we have in the DC show us exactly which rack uses how much Amperes
[09:48:42] <andyshack> i cant run 20 - 30 machines on a single phase though.
[09:48:53] *** WickedWicky has quit IRC
[09:48:56] *** UGH is now known as WickedWicky
[09:50:12] <Tempt> The fact that you have 20-30 Sun machines indicates you have a hardware habit even more dangerous than mine
[09:50:30] <sponix> palowoda: just looked at the HCL for the nvidia drivers the other day, nvidia site only has Quadro listed for Solaris, no 7xxx 8xxx or 9xxx series cards
[09:50:53] <sponix> palowoda: thought that was odd, on Linux and FreeBSD it lists them
[09:51:05] <Auralis> sponix: thaat hcl is crap, the solaris driver is identical in function and suport as the linux drivers
[09:51:14] <Tempt> anyway, time to knock off for the day, enough office time
[09:51:16] <renihs> shouldnt matter since that is one unified huge binary blob
[09:51:17] <Tempt> later all
[09:51:22] <WickedWicky> see ya later
[09:51:23] <andyshack> oh yeah. the main idea of it is to be able to say "NO YOURE WRONG!" to the next sysadmin that fsks with me.
[09:51:35] <sponix> Auralis: they should be smacked for that then, I almost bought an expensive card for no reason :P
[09:52:08] <Auralis> they should yes, i guess its because sun only sells quadros
[09:52:41] <sponix> time for me to crash out...
[09:52:54] <oxygene> also, in the beginning, the solaris driver really only registered the quadro pci-ids
[09:52:55] *** jamesd__ has quit IRC
[09:53:02] <oxygene> but that changed _long_ ago
[09:53:11] <palowoda> Ah the good old days.
[09:53:24] *** Ljuvefreya_ has joined #opensolaris
[09:54:57] *** cmihai has joined #OpenSolaris
[09:58:50] *** Ljuvefreya has quit IRC
[09:59:43] *** Ljuvsolaris has quit IRC
[10:00:05] *** dmarker has quit IRC
[10:00:18] *** dmarker has joined #opensolaris
[10:06:59] *** tombhadAC has quit IRC
[10:08:23] *** andyshack has quit IRC
[10:10:40] <palowoda> I sometimes wonder if AMD/ATI could learn something from the OEM relationship that Sun has with Nvidia graphics support in a mixture of open source and prorietary graphic drivers.  The platforms are so far apart.
[10:15:25] <dlg> blah
[10:15:40] <palowoda> Reality bites.
[10:17:28] *** sponix has quit IRC
[10:17:47] *** Fish has joined #opensolaris
[10:18:02] *** ptribble has joined #opensolaris
[10:18:23] <Fish> hello
[10:23:13] <rootard> Hi Fish.
[10:24:29] <renihs> fish go moo
[10:27:19] <Tempt> moooo
[10:29:25] <dlg> oooow
[10:29:27] <dlg> oh wait
[10:29:38] <dlg> you're the right way up relative to me
[10:34:33] *** MattMan has joined #opensolaris
[10:34:43] *** migi has joined #opensolaris
[10:37:04] *** NikolaVeber has joined #opensolaris
[10:42:56] *** nostoi has quit IRC
[10:49:34] *** FBdev_ has joined #opensolaris
[10:50:36] *** _dreams_ has joined #opensolaris
[10:50:45] <_dreams_> hello
[10:50:58] <renihs> hello mammal
[10:51:10] <g4lt-sb100> lies
[10:55:28] *** cmihai has quit IRC
[10:55:49] *** _setuid_H has joined #opensolaris
[10:56:21] <_setuid_H> May I use anew onbld tools to bfu older archives?
[10:57:11] <_setuid_H> new onbld tools sorry
[10:57:14] *** _setuid_H has left #opensolaris
[10:57:19] *** _setuid_H has joined #opensolaris
[10:58:50] *** FBdev has quit IRC
[11:00:09] *** _setuid_H has quit IRC
[11:20:25] *** seanmcg has joined #opensolaris
[11:20:26] *** timsf has joined #opensolaris
[11:21:18] *** cmihai has joined #OpenSolaris
[11:22:05] <palowoda> _setuid_H: You never expected to BFU to older achives to begin with.
[11:24:29] *** pizdec has quit IRC
[11:24:40] *** pizdec has joined #opensolaris
[11:28:22] *** Pietro_S has joined #opensolaris
[11:29:20] <Pietro_S> hello, how can I discover arguments of running proocces? man prstat didn't help me :-(
[11:29:30] *** bnitz has joined #opensolaris
[11:30:10] <palowoda> /usr/ucb/ps -auxxx
[11:30:31] <palowoda> Wide terminal that is.
[11:30:41] <kjetilho> palowoda: pargs
[11:30:51] <kjetilho> sorry
[11:30:52] <palowoda> Oh ok two ways.
[11:30:53] <kjetilho> Pietro_S: pargs
[11:31:10] <palowoda> At least he has a choice.
[11:31:50] <kjetilho> both pargs and ucb ps will display what the process has stored in argv[], this can be corrupt
[11:32:16] <kjetilho> ps -o args contains a kernel copy of the original arguments, but only the first 128 bytes or thereabouts
[11:32:21] <palowoda> Hey an idea.  How about removing ucb ps if it corrupts?
[11:32:29] <Pietro_S> thanks
[11:32:32] <kjetilho> no, the *process* corrupts it
[11:32:49] <kjetilho> e.g. sendmail changes it to "listening for connections" and stuff like that
[11:33:15] <palowoda> kjetilho: And the solution is?
[11:35:30] <quasi> never use sendmail ;)
[11:35:52] <palowoda> Thank God for Postfix.
[11:37:44] <Triskelios> Wietse Venema is God?
[11:38:15] <Triskelios> hm, that would explain a few things...
[11:38:58] <palowoda> After you do some custom confguration on sendmail and went to hell you can consider postfix, exim etc as heaven.
[11:39:31] <dlg> mmmsendmail
[11:41:12] <palowoda> Although sometimes I enjoy the good old ways of beating ourself with a big stick over our heads with the traditional way. :)
[11:41:21] <Tempt> I found a command to make sendmail more secure and more reliable
[11:41:35] <renihs> hah :p
[11:41:50] <dlg> Tempt: is it rm?
[11:41:55] <renihs> giggles :)
[11:42:00] <Tempt> pkgrm
[11:46:37] <palowoda> Tempt: I was thinking about your power bill.  That works out to about 300 USD per month right?
[11:47:05] <palowoda> What is the killo-watt hour cost?
[11:47:17] <Tempt> $AU300/month
[11:47:30] <Tempt> I'd have to get up and get my bill for that :)
[11:47:54] <palowoda> I was using 2.5 months for 900AUS.
[11:48:01] <Tempt> Next bill will be smaller, because I've changed power companies having found the competition will go 25% lower.
[11:48:04] *** mbalmer has joined #opensolaris
[11:48:08] <mbalmer> hya
[11:48:15] <palowoda> Or is it three months?
[11:48:23] <Tempt> Three months.
[11:48:30] <Tempt> My power usage is considered quite high
[11:48:33] <dlg> mbalmer: now i look dumb
[11:48:40] <mbalmer> ;-)
[11:48:55] <Tempt> Although I know quite a few suburban family units that burn $600/quarter in *gas* bills during winter,.
[11:48:58] <Tempt> which is quite insane.
[11:49:14] <Tempt> Last place I lived had gas hot water/cooking and the bill was around $15/month
[11:49:33] <palowoda> I'm about 230 USD for about 6 machines turned on in the house on a regular basis.
[11:49:59] <Tempt> for what time period?
[11:50:06] <palowoda> A month
[11:50:15] <Tempt> 'kay. What sort of machines?
[11:50:48] <palowoda> All of them AMD 65W machines now Some 90W machines.  The drives are what is killing me.
[11:51:47] <Tempt> My total household machine count: Sun Fire V880 (8 CPU, thirsty!), Sun Blade 1000, Sun Ultra 20, Sun Netra T105, hackjob Sempron PC, two laptops and my gf's linux workstation.
[11:51:54] <Tempt> plus a coupla 'rays
[11:52:45] <Tempt> and a tape library
[11:52:56] <Tempt> and a coupla cisco switches and a cisco router
[11:55:31] <palowoda> I have a couple of older U2's and 80's but that reall cranks up the power.  6 laptops in the house but rarely are they turned on a a 24/7 baises.
[11:56:44] <Tempt> Oh, I've got a pair of U2s, a classic, three spare netras, a v210, an e3000, two storage arrays de-powered.
[11:56:51] <Tempt> and pair of e250s, de-powered
[11:57:00] <dlg> mmmnetra
[11:57:13] <Tempt> dlg: Dude, come and get it.
[11:57:30] <dlg> its not like you're next door
[11:57:38] <dlg> i'll figure something out one day though
[11:58:07] <Tempt> I must find something to plug into the LSI card in my Ultra 20 too.
[11:58:24] *** alobbs has joined #opensolaris
[11:58:25] <Tempt> Seems a waste to have a dual-channel card sitting there doing nothing but slowing my boot process
[11:58:26] <dlg> which lsi card?
[11:58:35] <Tempt> dlg: Can't remember the model, dual channel LVD320
[11:58:42] <Tempt> sun branded.
[11:58:45] <dlg> ah
[11:58:46] <dlg> yeah
[11:58:58] <dlg> they all have the same pci id
[11:59:03] * dlg calls them the 1030
[11:59:24] <Tempt> I should plug all my assorted SCSI devices in
[11:59:33] <Tempt> MO drive, tape drives, old 4Gb external drives...
[11:59:36] <dlg> heh
[11:59:45] <palowoda> Thats when the power goes up.
[11:59:54] <Tempt> I was thinking about getting a scanner, I might buy an old HP, that'll use a SCSI port.
[12:00:33] <palowoda> Heh can you buy SCSI scanners anymore?
[12:00:34] <Tempt> palowoda: I know, that's why I don't have any of this stuff plugged in. I don't need to read old QIC tapes very often either.
[12:00:49] <Tempt> palowoda: Sure, all the fancy scanners are still SCSI (unless they're ethernet)
[12:01:35] <palowoda> Tempt: most of them are usb.  I have an Epson 4800dpi usb scanner and it shuts the power down when I'm not using it.
[12:01:52] *** simford has quit IRC
[12:01:53] *** derchris has quit IRC
[12:02:02] <Tempt> The fancy film scanners and stuff are still being released in SCSI
[12:02:09] *** derchris has joined #opensolaris
[12:02:34] *** calumb has joined #opensolaris
[12:02:46] <palowoda> Tempt: Your getting old.  USB 2.0 scanners are the way.
[12:03:05] <Tempt> You see, I was happy enough with SCSI. I don't need this new-fangled USB
[12:03:06] <palowoda> Mine is a film scanner.
[12:03:33] <Tempt> Well, I'll get myself an old SCSI scanner and scan things while rockin' in my rocking chair.
[12:03:52] <palowoda> Not that I use the film scanning feature (or related hardware that is in it).
[12:05:58] <dlg> Tempt: newfangled usb uses scsi inside
[12:06:22] <Tempt> SCSI over USB
[12:06:23] <Tempt> sad.
[12:06:32] <dlg> heh
[12:06:46] <dlg> its good
[12:07:03] <palowoda> The market isn't changing either.
[12:07:28] <palowoda> For some reason usb is winning out.
[12:08:13] <palowoda> For scanners, cams etc.
[12:08:55] <palowoda> Just think they put scsi in the device and charge you 4 times the cost.
[12:09:16] <palowoda> Bridge it and it's 4 times cheaper.
[12:09:24] <Tempt> USB
[12:09:25] <Tempt> cheap
[12:09:36] <renihs> but nice cables :p
[12:09:39] <Tempt> market perception that SCSI is a premium product
[12:09:58] <palowoda> And the asshole manufactures that do straight usb without scsi bridges are making out like bandits.
[12:10:09] <Tempt> Like vendors charging more for LCDs with DVI than analog
[12:10:20] <Tempt> analog requires converting crap inside, DVI doesn't, DVI is more expensive.
[12:10:25] <dlg> theres a difference between the protocol and the hardware
[12:10:42] <dlg> usb has been using the scsi protocol for years
[12:10:45] <Tempt> Yeah. Never seen a SCSI cupwarmer/xmas tree/dildo/whatever
[12:11:07] <dlg> actually talking to scsi hw is hard
[12:11:25] <Tempt> How much current can you pull on TERMPWR? Perhaps we *can* have a SCSI cupwarmer.
[12:12:07] <palowoda> LandRover has SCSI cupwarmers.
[12:17:01] <kjetilho> and a SCSI winch
[12:17:15] *** Jondice_ver_2 has quit IRC
[12:17:20] <palowoda> and SCSI headlights.
[12:17:43] <Tempt> SCSI towball
[12:18:13] <WickedWicky> SCSI girlfriends
[12:19:22] *** UGH has joined #opensolaris
[12:19:32] <UGH> that's it.
[12:19:39] <palowoda> You drive a LandRover in the middle of the desert what do you expect.
[12:19:42] <UGH> this D-link wireless crap router is going out of my house, today
[12:20:15] <Triskelios> UGH: yeah, they universally suck
[12:20:25] *** prg3__ has quit IRC
[12:20:36] *** WickedWicky has quit IRC
[12:20:39] *** UGH is now known as WickedWicky
[12:20:41] *** uebayasi has quit IRC
[12:21:35] <WickedWicky> so as I was saying, since SCSI is hard to talk to, I am gonna ask my date next time when she's being difficult, why she has to be so SCSI
[12:22:16] <palowoda> I thought you have a U5. How can you have a date?
[12:22:29] <renihs> haha :p
[12:22:30] <WickedWicky> correction, I have a SS20
[12:22:33] <WickedWicky> and a U5 soon
[12:22:45] <palowoda> Oh lord.
[12:22:46] <WickedWicky> palowoda: I am a geek, not a nerd, geeks have sex and dates
[12:22:54] <WickedWicky> it's for free, how can I say no?
[12:23:00] <WickedWicky> (the U5)
[12:24:39] <palowoda> Check for deseases.
[12:24:51] <WickedWicky> always
[12:25:07] <WickedWicky> the U5 can run Opensolaris though
[12:25:11] <WickedWicky> should make you happy
[12:25:42] <Triskelios> hey my U5 has facilitated sex more than once
[12:25:48] <palowoda> Err are you asking me if you want me to be happy?
[12:25:56] <WickedWicky> I want everybody to be happy
[12:26:03] <Tempt> Triskelios: Propping your bed up?
[12:26:04] <WickedWicky> life is cheerfull and peachy
[12:26:31] <palowoda> Ok I'd use the U5 as a fishing sinker.
[12:26:49] <WickedWicky> just cause it doesnt do 6 terra on bogomips?
[12:27:13] <palowoda> Hey now bogo stuff is lugnuts terms.
[12:27:26] <WickedWicky> you get my point nevertheless
[12:27:37] <palowoda> Your sinker.
[12:27:51] <WickedWicky> if it works it works. if it does what its expected to do, even better
[12:28:11] <WickedWicky> I dont have Blade 2000 money
[12:28:20] <Tempt> Yes you do.
[12:28:30] <Tempt> You just spent it on wine, women, song and colombian marching powder.
[12:28:31] <WickedWicky> no I dont, any idea what the permit costs for canada are?
[12:28:58] <Tempt> permit?
[12:29:07] <WickedWicky> to live there, permanent residentship
[12:29:16] <palowoda> Colombian marching powder.  Now that one I've never heard.
[12:29:19] <Tempt> moving to canadia?
[12:29:28] <WickedWicky> I dont know, maybe
[12:29:51] <Gman> http://www.gnome.org/~gman/yahoo-finance.png
[12:30:19] <renihs> heh lol
[12:30:20] <palowoda> You can't import U5 into Canada anylonger.
[12:30:22] <WickedWicky> gnome runs wonderfull on 64MB of memory btw
[12:30:39] <WickedWicky> you cant say cunt in canada either
[12:30:40] <renihs> wonderfull is taste :p
[12:30:53] <Gman> they didn't migrate the history either! ):
[12:30:56] <Tempt> Ha, the Sun ad looks very wrong next to that message.
[12:31:13] <Tempt> It should Say JAVA microsystems and Javalaris
[12:31:42] <palowoda> No Jalaris sounds better.
[12:32:16] *** jamesd has joined #opensolaris
[12:32:17] *** ChanServ sets mode: +o jamesd
[12:33:44] <palowoda> We should give marketing any ideas.
[12:33:53] <palowoda> shouldn't
[12:34:36] *** snaud has quit IRC
[12:34:48] <palowoda> Oh man I'm I off base.  Marketing doesn't have any ideas.
[12:35:02] <Tempt> We should encourage them.
[12:35:36] <Tempt> I for one welcome JAVA java Javalaris with Java JavaContainers and Javatrace and the new JavaEngine JCP/JP stack
[12:35:48] <palowoda> Oh yeah lets call the next java version Solava.
[12:36:08] *** calumb is now known as calAFK
[12:36:57] <palowoda> Actually Sava might have been a better name.
[12:37:48] <Tempt> Palava
[12:38:04] <palowoda> Church of Bob.
[12:38:28] <Tempt> Anyway, I'm going to have a cup of Java and a Javarette
[12:38:50] * WickedWicky is gonna buy a new wireless router
[12:40:06] <palowoda> WickedWicky: At least get an 802.11n version so we don't know what your talking about.
[12:40:39] <WickedWicky> you dont do that now either so I might as well stick with g
[12:41:56] *** mlh_ has quit IRC
[12:43:56] <Tempt> Cisco 877
[12:44:01] <Tempt> Wireless + ADSL2+
[12:46:29] *** mlh has joined #opensolaris
[12:54:36] *** Gman is now known as GmanAFK
[12:56:27] *** calAFK is now known as calumb
[12:56:54] <Tempt> Hey, if you think your machines have cruft in them, check this production host http://www.purplecow.org/?p=26
[13:02:01] *** Chihan has joined #OpenSolaris
[13:06:22] *** deather_ has joined #opensolaris
[13:08:23] *** nostoi has joined #opensolaris
[13:09:25] <quasi> Tempt: I've seen older in production - but weren't allowed to take pics
[13:09:45] <Tempt> Probably *shouldn't* have taken that pic, but meh.
[13:09:52] *** derchris has quit IRC
[13:10:22] <Tempt> There's a whole series of Eclipses from around the place here.
[13:10:24] *** derchris has joined #opensolaris
[13:11:32] *** migi has quit IRC
[13:12:39] *** derchris has quit IRC
[13:12:41] *** migi has joined #opensolaris
[13:13:33] *** derchris has joined #opensolaris
[13:13:39] *** deather has quit IRC
[13:13:41] *** deather_ is now known as deather
[13:30:28] *** JSRJ has quit IRC
[13:32:39] *** cajetanus has joined #opensolaris
[13:35:35] *** JSRJ has joined #opensolaris
[13:35:36] *** cajetanus has quit IRC
[13:36:53] *** cajetanus has joined #opensolaris
[13:36:53] *** JSRJ has quit IRC
[13:39:16] *** cajetanus has left #opensolaris
[13:40:24] *** nivox has joined #opensolaris
[13:46:17] <WickedWicky> so palowoda, they didnt have the 802.11n router.. but they told me that when I'd rant about gucci the effect would be more or less the same
[13:50:45] *** WickedWicky has quit IRC
[13:54:04] *** Chihan has quit IRC
[13:59:19] *** fedorared has quit IRC
[14:01:14] *** WickedWicky has joined #opensolaris
[14:01:22] <WickedWicky> let's see if things are more stable now
[14:03:09] *** tsoome has quit IRC
[14:03:19] *** calumb is now known as calLNCH
[14:11:26] *** master_o1_master has joined #opensolaris
[14:23:54] *** master_of_master has quit IRC
[14:24:11] *** JSRJ has joined #opensolaris
[14:25:25] <Gekkko[PDA]> boo
[14:26:20] <jmcp> Gekkko[PDA]: generally expanded
[14:26:36] <Gekkko[PDA]> lol
[14:26:43] *** JSRJ has quit IRC
[14:29:55] *** laca has joined #opensolaris
[14:31:55] *** bowdengl has joined #opensolaris
[14:32:13] <Tempt> !seen e^ipi
[14:32:17] <Tempt> e^ipi: ping
[14:32:32] <Drone> e^ipi is currently online in #opensolaris and last spoke on Mon 27 Aug 2007 07:37 GMT, saying 'real challenge is to see if I can't replicate it tomorrow'.
[14:37:17] *** umdstu_ has joined #opensolaris
[14:37:44] <umdstu_> morning
[14:38:42] <Gekkko[PDA]> hi
[14:39:38] *** mbalmer has quit IRC
[14:39:46] *** bowdengl has quit IRC
[14:39:49] *** bowdengl_mac has joined #opensolaris
[14:40:25] *** ashner has quit IRC
[14:41:43] <umdstu_> you have any experience with mail servers
[14:42:21] <Tempt> I eat mail servers for breakfast.
[14:42:26] <Tempt> They tend to get stuck in the teeth.
[14:43:51] <umdstu_> lol
[14:44:00] <umdstu_> i'm having a heck of a time getting something going
[14:44:13] <umdstu_> i just need a simple mail server for about 24 computers to email each other
[14:44:19] <umdstu_> all in the same room, no outside network
[14:44:20] <Tempt> use exim
[14:44:25] <umdstu_> on solaris sparc
[14:44:30] <Tempt> Use exim.
[14:44:31] <umdstu_> i havn't heard of it
[14:44:37] <umdstu_> i'll look at it
[14:44:38] <Tempt> You have now.
[14:44:47] <umdstu_> i was trying to use communications suite
[14:44:51] <umdstu_> which is huge
[14:44:54] <Tempt> Honestly, it's quicker to get exim rolling nicely than to read enough sendmail docs to get there.
[14:44:59] <umdstu_> yea
[14:45:10] <umdstu_> i spent a couple weeks reading over sendmail and qmail
[14:45:18] <umdstu_> and it made me hate mail servers
[14:45:23] <Tempt> Seriously, exim is pretty easy.
[14:45:26] <umdstu_> i'll take a look at exim, thanks
[14:45:30] <umdstu_> sweet
[14:45:31] <Tempt> No worries.
[14:45:33] <umdstu_> brb
[14:45:37] <Tempt> The documentation for exim is sweet.
[14:46:55] <jmcp> umdstu_: http://www.exim.org
[14:46:57] <jmcp> it's *very* good
[14:50:16] <Berny> yikes
[14:50:44] <Berny> who broke my local patchserver and smpatch again?
[14:50:44] *** cmihai has quit IRC
[14:51:37] <Tempt> jmcp: You an exim man too?
[14:51:42] <jmcp> Tempt: yeah
[14:51:58] *** jamesd has quit IRC
[14:51:59] <Tempt> jmcp: Good to see more people with taste around here.
[14:52:07] * jmcp read's JS's latest blog entry ... 364 comments!
[14:52:10] <Berny> smpatch get just gives an "Failed to load main class: com.sun.patchpro.cli.PatchProSetup"
[14:52:59] <umdstu_> haha alright guys i'm back, i'll start reading about exim
[14:53:26] <jmcp> umdstu_: it's very easy to setup and maintain, and the docs are well written
[14:55:35] *** vmlemon has joined #opensolaris
[14:55:46] <umdstu_> that's good news
[14:56:00] <umdstu_> let me read up on it and then i'll pester you with any questions
[14:56:03] <umdstu_> is that ok
[14:56:22] <jmcp> you'd  probably get better answers from a #exim somewhere
[14:57:06] <jmcp> and there's a #exim on freenode, so you're in luck
[14:57:15] <jmcp> that's not to say that we won't try to help
[14:57:37] <Tempt> Exim on Solaris is the same as Exim on BSD is the same as Exim on AIX is the same as Exim on IRIX is the same as Exim on HPUX is the same as Exim on ...
[14:57:42] <Gekkko[PDA]> lies.
[14:57:47] <Tempt> You get the picture, there's nothing particularly Solaris specific.
[14:58:11] <Gekkko[PDA]> Exim on Linux? lol
[14:58:16] *** Me[-_-]gaf has joined #OpenSolaris
[14:58:19] <Tempt> I don't know anything about Linux.
[14:58:25] <Tempt> Other than the fact that it sucks.
[14:58:28] <jmcp> Gekkko[PDA]: doesn't that run inside the kernel?
[14:58:32] <Gekkko[PDA]> lol.
[14:58:42] <Gekkko[PDA]> jmcp: no idea.
[15:00:31] <umdstu_> oh alright good deal
[15:00:57] <umdstu_> well first thing i notice is that the first line says its intended for unix systems connected to the internet, i won't be on the internet, does that matter
[15:01:05] <Tempt> Anyway, when I did my first exim install I didn't even read the docs. Just read through the configuration file and take care of everything step by step and make sure your DNS is correct.
[15:01:22] <Tempt> Doesn't matter, no.
[15:01:34] <umdstu_> lol ok
[15:01:35] <umdstu_> good
[15:02:32] *** karrotx has quit IRC
[15:02:47] *** calLNCH has quit IRC
[15:08:17] *** _dreams_ has quit IRC
[15:14:01] *** ichigo has joined #opensolaris
[15:17:33] *** vmlemon has quit IRC
[15:18:32] *** vmlemon has joined #opensolaris
[15:21:21] *** cmihai has joined #OpenSolaris
[15:25:11] *** vmlemon has quit IRC
[15:26:05] *** vmlemon has joined #opensolaris
[15:28:42] *** cajetanus has joined #opensolaris
[15:34:04] *** cajetanus has quit IRC
[15:34:34] *** EchoBinary has quit IRC
[15:34:42] *** nostoi has quit IRC
[15:37:27] *** monzie has quit IRC
[15:38:37] *** cajetanus has joined #opensolaris
[15:45:32] *** tsoome has joined #opensolaris
[15:47:30] *** nachox has joined #opensolaris
[15:47:43] <nachox> morning
[15:49:41] *** vmlemon has quit IRC
[15:49:59] *** Teknix has joined #opensolaris
[15:50:19] *** vmlemon has joined #opensolaris
[15:54:31] <Berny> .oO(am i blind or stupid or both or isn't the local patch server package for sol10 not on sdlc anymore?)
[15:57:17] <Tempt> It is my belief that Sun is dedicated to preventing people from patching their hosts.
[15:57:29] <Stric> Sounds plausible.
[15:57:39] <cmihai> Tempt, oh please, you should see how HP handles this.
[15:57:49] * WickedWicky is patching pebbles as we speak
[15:57:52] <Tempt> Do HP even have an OS anymore?
[15:57:57] <cmihai> 3 even.
[15:58:00] <Stric> Tempt: Sure they do.. WinXP ;)
[15:58:04] <WickedWicky> HPUX
[15:58:04] <cmihai> HP-UX, Tru64 UNIX (dead) and OpenVMS.
[15:58:10] * Tempt thinks WickedWicky will ne up for a late one if he's watching that patch.
[15:58:21] <WickedWicky> ah yes, of course.. OpenVMS, they aquired Digital
[15:58:21] <Tempt> HP is in the process of killing OpenVMS too.
[15:58:26] <Tempt> And setting fire to NonStop
[15:58:30] <SYS64738> what the right term to use in screen ? I don't reach to edit a file
[15:58:38] <cmihai> NonStop OS hasn't gotten updates since 1984
[15:58:39] <Stric> Tempt: so NonStopBurning? :)
[15:58:44] <cmihai> Did you see the documentation for NonStop?
[15:58:51] <cmihai> It's scanned paper mate
[15:58:53] <Tempt> Never worked with it.
[15:58:54] <cmihai> (c) 1984
[15:58:59] <WickedWicky> SYS64738: TERM=vt100 && export TERM
[15:59:01] <cmihai> Had to once
[15:59:04] <cmihai> OMg :-)
[15:59:05] <Tempt> Seen a few boxes around machinerooms though.
[15:59:15] <Tempt> Tandem era, not HP crap.
[15:59:15] <cmihai> It's ancient and garbage.
[15:59:18] <cmihai> Yeah
[15:59:21] <cmihai> HP bought Tandem
[15:59:29] <WickedWicky> Tandem?
[15:59:37] <Tempt> You mean they *trashed* tandem for the brand.
[15:59:40] <WickedWicky> I used to have a 286 by that make
[15:59:46] <Tempt> Which they then decided they didn't like.
[15:59:55] *** smallfoot- has joined #opensolaris
[15:59:59] <WickedWicky> 4mb on simms
[16:00:02] <smallfoot-> when will sun make opensolaris under GPL?
[16:00:03] * Tempt goes off to cackle at "serviceguard"
[16:00:08] <Tempt> smallfoot-: Hopefully never.
[16:00:17] <smallfoot-> sun microsystems owns!! SOALRIS!! OPEN SOLARIS! ZFS!! DTRACE! OPEN SOURCE!!!!! ROX
[16:00:26] <smallfoot-> Tempt, why? GPL rocks, its greatest thing ever
[16:00:34] <jmcp> smallfoot-: no, it's not
[16:00:37] <Tempt> smallfoot-: No, it isn't.
[16:00:40] <smallfoot-> sun + google = 2 coolest companies ever
[16:00:41] <jmcp> it's a particular solution to a particular problem
[16:00:43] <smallfoot-> oh yes it is
[16:00:54] <jmcp> smallfoot-: you seem to be a troll
[16:00:55] <Auralis> it never was
[16:00:56] <jmcp> we don't like trolls here
[16:01:03] * quasi remembers a cow-orker who wanted to prove that the tandem could hotplug cpus - he just shouldn't have when the other cpu was out for repair ;)
[16:01:12] <WickedWicky> AHAHAHA
[16:01:15] <cmihai> Hell, NotStop documentations consists on tips  on how to use google ffs ;_)
[16:01:19] <WickedWicky> see? hot offlined!
[16:01:25] <jmcp> quasi: that's "courageous", in the best Yes Minister sense of the term
[16:01:39] <Tempt> jmcp: A banana skin, no less
[16:01:47] <jmcp> indeed
[16:02:03] *** smallfoot- has left #opensolaris
[16:02:11] <Tempt> Geez, that was a short troll.
[16:02:16] <WickedWicky> courageous being totaly random and stupid in this context?
[16:02:19] <quasi> jmcp: same guy who busted a $20k ibm memory board because he didn't want to wait for a tech
[16:02:24] <Tempt> [00:02] CTCP VERSION reply from smallfoot-: mIRC v6.21 Khaled Mardam-Bey
[16:02:27] <Tempt> haha.
[16:02:35] <jmcp> quasi: ouch
[16:02:39] <jmcp> quasi: *big* ouch
[16:02:44] <jmcp> Tempt: sigh
[16:02:45] <WickedWicky> expensive ouch too
[16:02:52] <jmcp> Tempt: yabt
[16:02:53] <quasi> aye
[16:03:04] <Tempt> Haha, reminds me of a guy who fscked an Alcatel OmniCore switch by slamming a blade in.
[16:03:10] <Tempt> wouldn't seat first time so he used BFI
[16:03:16] <Tempt> pity he didn't remove the plastic cover first.
[16:03:22] <WickedWicky> kik
[16:03:28] <Berny> lol
[16:03:28] <quasi> although ibm ate it as DOA because they owed us bigtime
[16:03:52] <nachox> morning all
[16:03:53] <WickedWicky> that's pure luck then
[16:05:02] <jmcp> gnite all
[16:05:10] <WickedWicky> nite
[16:05:37] <Tempt> Some guy was sent to remove the faulty HBA from a machine
[16:05:37] <Tempt> He didn't know which one so he pulled the one with the flashing orange light.
[16:05:37] <Tempt> Same guy was given some boxes to build
[16:05:37] <Tempt> Pity he didn't attach the second half of the mirror when doing the disksuite
[16:05:37] <Tempt> Same guy was sent to fix a two disk machine with a disk out.
[16:05:39] <Tempt> Pity he rebooted it
[16:05:42] <Tempt> ....
[16:06:04] *** JSRJ has joined #opensolaris
[16:06:05] <Tempt> Man, I've got a hundred stories about that one guy.
[16:06:45] <nachox> Tempt: and he didnt get fired?
[16:06:52] <Tempt> naaah
[16:06:56] *** Me[-_-]gaf has quit IRC
[16:07:05] <Tempt> Nobody ever gets fired for fucking up or being useless in my experience.
[16:07:08] *** cajetanus has quit IRC
[16:07:13] <Tempt> They only get fired for knowing too much and making managers feel small.
[16:07:27] <Tempt> (err, and being a smartarse)
[16:07:31] <ofu> how do I send a break on ilom?
[16:07:37] <nachox> Tempt: depends on what you screw up imho
[16:07:45] <Tempt> ofu: With a hammer?
[16:07:53] <quasi> ofu: which sort of box?
[16:08:00] *** cmihai has quit IRC
[16:08:19] <Tempt> nachox: Well, at this particular place it appeared the only way it was possible to get fired was sexual harassment.
[16:08:22] <ofu> cant really tell
[16:08:38] <Tempt> nachox: I remember at one point I decided I'd do zero work for a week to see what happened.
[16:08:47] <Tempt> nachox: Absolutely nothing, except a mountain of work piled up.
[16:09:01] <quasi> ofu: then the answer has to be "can't really tell" ;)
[16:09:41] *** cypromis has joined #opensolaris
[16:10:03] <ofu> quasi: T51x0
[16:10:43] *** pfa3rh has joined #opensolaris
[16:11:21] *** Jondice has joined #opensolaris
[16:12:37] <nachox> Tempt: oh, so he was knowingly doing nothing just to test how much he could pull till the string breaks
[16:12:50] *** RaD|Tz has joined #opensolaris
[16:13:21] <SYS64738> where is the cc compiler in the full oem ?
[16:14:17] <WickedWicky> /opt/SUNWspro/bin/cc
[16:14:37] <nachox> SYS64738: it is not, you need to install sun studio
[16:14:40] *** bowdengl_mac is now known as bowdengl
[16:14:46] <SYS64738> ah ok
[16:14:47] <SYS64738> thanks
[16:15:00] <WickedWicky> oh, I assumed he used opensolaris
[16:15:05] <SYS64738> thank I needed to install solaris express developper edition ?
[16:15:10] <nachox> SYS64738: there is gcc though in /usr/sfw/bin
[16:16:16] <nachox> WickedWicky: even then sun studio is not installed in the full oem version, you need to install the development crap which even leads to a different installer
[16:16:17] *** sehat_tan has joined #opensolaris
[16:17:05] <SYS64738> nachox, how can I say to the cc wrapper in /usr/ucb that I want to use /usr/sfw/bin/gcc ?
[16:17:25] <kjetilho> you do NOT want to use /usr/ucb/cc
[16:17:34] <WickedWicky> CC=gcc;  export CC
[16:17:45] <nachox> you dont want to use /usr/bin/cc
[16:17:46] <kjetilho> hardly any software is easier to compile with that anymore
[16:17:59] *** cajetanus has joined #opensolaris
[16:17:59] *** pmh has quit IRC
[16:18:04] <kjetilho> (it changes the semantics of signal and some other BSD stuff)
[16:18:06] *** pmh has joined #OpenSolaris
[16:18:10] *** JSRJ has quit IRC
[16:18:21] <nachox> it is there for backwards compatibility reasons and i think there is an RFE to remove it
[16:18:25] <kohju>  /usr/ucb/cc is trap.
[16:18:40] <Tempt> It's a trap. Get an axe.
[16:18:47] <kjetilho> there was one or two programs I used to compile with it back in Solaris 2.5
[16:19:25] <SYS64738> my make continue to search cc also after the export CC=gcc
[16:19:54] <WickedWicky> you'll have to rerun ./configure
[16:20:27] <Pietro_S> AR=gar MAKE=gmake ./configure
[16:20:43] <Pietro_S> and AR=gar MAKE=gmake gmake
[16:20:50] <Berny> yikes
[16:21:04] <Berny> seems like the latest java patch broke the patchserver :-\
[16:21:05] <SYS64738> I am tryng to compile qmail, there isn't a ./configure
[16:21:18] <kjetilho> SYS64738: it's a sign.  Qmail is a trap.
[16:21:25] *** tsoome has quit IRC
[16:21:39] <quasi> kjetilho: indeed
[16:22:02] <Triskelios> qmail lets you specify the compiler, check the build instructions
[16:22:07] *** mega has joined #opensolaris
[16:27:31] <Shinden> http://www.joemonster.org/filmy/2554/Numer-z-martwym-Jeleniem to jest fajne
[16:27:52] *** ptribble has quit IRC
[16:30:40] <SYS64738> I think I'll use postfix
[16:31:18] <bda> Wise move.
[16:31:46] <Tempt> Naah
[16:31:46] <Tempt> exim
[16:31:58] <paul> hmm, 'ERROR: Cannot copy miniroot from media /net/...../Solaris_11/Tools/Boot' why is luupgrade giving me this error? nothing immediately obvious from truss either.
[16:32:18] <paul> ok nm
[16:32:40] <Pietro_S> laca: this spec file http://home.zcu.cz/~sobotkap/SFEwesnoth.spec , won't pkgtool (pkgbuils ends with NO ERROR) - I probably have there some mistake, but in logs there is no clue for it - looks like pkgtool bug or at least missing some detailed error message
[16:34:16] <laca> Pietro_S: I suspect something goes wrong in the packaing phase?
[16:34:22] <laca> can you send me the build log?
[16:36:35] <nachox> postfix is really nice, i always had problems with exim :(
[16:36:36] *** oninoshiko has joined #opensolaris
[16:36:53] <Triskelios> exim confuses me more than sendmail
[16:37:13] <Pietro_S> sure, but you need to wait a bit, need to compile it (I faced this problem yesterday and /tmp/SFEwesnoth.log was destroyed by shutdown)
[16:40:59] <Tempt> Triskelios: huh? How?
[16:41:06] <Tempt> Triskelios: I'm boggled.
[16:41:12] <bda> Its existance.
[16:41:47] <Tempt> It works, it is simple, it rocks.
[16:41:54] <Tempt> The MTA to end all MTAs.
[16:41:58] *** AtomicPunk has joined #opensolaris
[16:42:26] <nachox> Triskelios: have you tried configuring sendmail without the m4's? it cannot get harder than that
[16:48:17] *** vmlemon has quit IRC
[16:49:51] *** vmlemon has joined #opensolaris
[16:51:30] <bda> One of our sites lacks a Solaris box and it's driving me up the goddamn wall.
[16:51:45] <bda> "Oh, you want to set up dedicated egress MXes for per-list stuff? Sure, that's no probl--- goddamnit!"
[16:52:10] <bda> Other two sites, it takes two minutes to bring a new zone up, especially one that trivial.
[16:52:13] <bda> sigh
[16:54:25] *** ichigo_ has joined #opensolaris
[16:54:34] <Tempt> Time to order some hardware.
[16:54:41] *** Snake007uk has joined #opensolaris
[16:55:03] <bda> Shit, I have plenty of hardware from a site I just tore down. Getting it out there is the problem. :)
[16:55:06] *** Snake007uk has quit IRC
[16:55:13] <bda> What's hilarious is we have two /26s out there.
[16:55:18] <bda> And like.. ten machines.
[16:55:27] *** Snake007uk has joined #opensolaris
[16:55:32] *** tsoome has joined #opensolaris
[16:57:13] <Tempt> Hey, my site has a class B and four frontfacing hosts.
[16:57:30] <bda> Legacy addressing from when we had T1s through UUNET back in the day.
[16:57:48] <Tempt> Wasting an entire class B is almost criminal though.
[16:57:51] <bda> Weird they didn't just give us /24s, which is what I'd have expected.
[16:57:52] <bda> Yeah.
[16:57:54] <bda> It is.
[16:57:57] <bda> Fired into the sun!
[16:58:14] <Tempt> I used to have most of a class C at home
[16:58:21] <Tempt> had to share it with the guy that owned it from way back
[16:58:30] <Tempt> I think my frontfacing stuff gobbled up 30 IPs :)
[16:59:42] <bda> My second day at Cisco, years and years ago, the boss was like "So what do you think you'll need for this network you're building for us? A /16, a /24?" "uh. A /29, probably..." "No, really."
[17:00:05] <Tempt> Yeah.
[17:00:08] <Tempt> I know what you mean.
[17:00:13] <bda> Those were salad days.
[17:00:14] <Tempt> And I have to fight to get /28s for my colo machines.
[17:00:17] <bda> Mostly because I was so green.
[17:00:23] <bda> Same.
[17:00:32] <bda> They're charging us $100 for allocs now.
[17:00:37] <Tempt> fark
[17:00:42] <Tempt> For how many addresses?
[17:00:45] <bda> Doesn't matter.
[17:00:53] <bda> Place just got bought by PAETEC.
[17:00:58] <Tempt> aah, so it encourages ambit grabs of as many as possible.
[17:01:03] <bda> They've restructured their power pricing, allocs..
[17:01:31] <bda> Yeah, I've asked for a /27, which should see us through next year.
[17:01:41] <Tempt> Might as well bid for a /24
[17:01:46] <Tempt> Leave some room for other things.
[17:01:52] *** Tpent1 has joined #opensolaris
[17:01:55] <bda> The other place I just built out, and realy really like, you only have to pay for /24 and up, which is apparently (?) ICANN policy now.
[17:02:43] <Tempt> Mmm
[17:02:50] <bda> Or so I was told, anyway. shrug.
[17:03:00] <Tempt> I can see why IPV6 hasn't gone anywhere.
[17:03:09] <bda> Heh.
[17:03:10] <Tempt> It'd be destroying a cheery business built on IP scarcity.
[17:03:29] <bda> I was watching a Google Tech Talk the other day by Van Jacobson on rethinking networking.
[17:03:32] <bda> It was pretty great.
[17:03:42] <Tempt> Nifty.
[17:03:46] <Tempt> What were the main ideas?
[17:04:01] <bda> Disseminated networking. Information as named objects.
[17:04:23] <bda> It's become a pretty common idea, or anyway I remember reading papers on it several years ago.
[17:04:34] *** charlesnw has left #opensolaris
[17:04:35] <Tempt> No real implementation yet, though.
[17:04:42] <bda> But he's an older guy, and been in the game since the 60s, and apparently a total ninja.
[17:04:46] *** EchoBinary has joined #opensolaris
[17:04:48] <bda> It was a good talk.
[17:05:02] <Tempt> Man, you've never heard of Van Jacobson header compression?
[17:05:05] <bda> I'm sure there are lots of PhDs. ;)
[17:05:14] *** EchoBinary has quit IRC
[17:05:31] <bda> I recognized it when the guy introduced him, but.. my memory is awful.
[17:05:49] <Tempt> Yeah, good to see some of those guys are still working on stuff.
[17:05:50] <bda> If I don't use a piece of information regularly, or make a concerted effort to remember it, it's sent into the cornfield.
[17:05:53] <kjetilho> he came up with the hack to forge hop counts to make traceroute :)
[17:06:11] <bda> I believe it. :)
[17:06:26] <kjetilho> anyway, it was a good visionary talk.  it made me feel it's within grasp
[17:06:43] <bda> "We could totally do this now if people didn't suck."
[17:06:52] <bda> (my take)
[17:06:56] *** jafari has joined #opensolaris
[17:07:15] <Tempt> Aah, good speaker then.
[17:07:38] <Tempt> And when you got out and settled down and the inspiration wore off, it was depressing because great ideas aren't smiled upon these days.
[17:07:45] <Tempt> Only Web 2.0 mashups or something.
[17:08:19] <oninoshiko> most people cant even get web 1.0 right *rolls eyes*
[17:08:43] <bda> There's just so much weight behind things, it seems impossible (compared to the 40s-60s jumps) to make basic paradigm shifts.
[17:08:48] <bda> But probably that's because we're living in it.
[17:09:00] <Tempt> It isn't a research network anymore.
[17:09:18] <Tempt> It's all profit-tastic and you can't change anything without ensuring backwards compatibility.
[17:09:27] <Tempt> Unless, of course, you mean stitching shit together with javascript.
[17:09:31] <bda> Well, people have already built new networks on top of the Internet, like they did with the phone system. ;-)
[17:09:56] <bda> Pretty soon it'll be like goddamn BBSes all over again.
[17:10:26] *** ichigo has quit IRC
[17:10:35] <Tempt> heh, BBSes still exist.
[17:10:51] <bda> Yeah, but, c'mon. :P
[17:11:03] <Tempt> Hey, I want the ANSI art again, right?
[17:11:08] *** worklez has joined #opensolaris
[17:11:39] <bda> RIP art!
[17:12:00] <oninoshiko> the way to change from ipv4 to ipv6 is easy. take away all the ipv4 address presently being used by companies to give to users, this is a group that can be easily transitioned to ipv6, and those with servers will be accessible to both.
[17:12:54] <oninoshiko> its just too bad ICANN doesn't have the chutzpah to do it
[17:13:07] *** vmlemon has quit IRC
[17:13:23] <bda> Put NSI in charge. They'll just figure out a way to wildcard route everything through themselves.
[17:13:33] *** worklez has quit IRC
[17:13:37] <bda> If the addr block isn't allocated, you get sent to a :*'d virtual host...
[17:13:52] *** vmlemon has joined #opensolaris
[17:14:51] *** vmlemon has quit IRC
[17:15:21] *** vmlemon has joined #opensolaris
[17:15:41] *** Snake007uk has quit IRC
[17:16:41] <Pietro_S> laca: here is log http://home.zcu.cz/~sobotkap/SFEwesnoth.log
[17:17:00] *** worklez has joined #opensolaris
[17:18:54] *** cajetanus has left #opensolaris
[17:19:30] *** worklezz has joined #opensolaris
[17:22:13] <laca> Pietro_S: it looks like pkgmk dies or something
[17:22:19] <laca> can you upload the prototype file too
[17:22:28] *** stevel has joined #opensolaris
[17:22:28] *** ChanServ sets mode: +o stevel
[17:22:35] <laca> .../packages/PKGMAPS/proto/SFEwesnoth.proto
[17:22:43] <laca> morning stevel
[17:22:49] <stevel> hey laca
[17:27:17] *** eboutilier_ has joined #opensolaris
[17:28:09] *** mega has quit IRC
[17:28:54] <Pietro_S> laca: here is prototype: http://home.zcu.cz/~sobotkap/SFEwesnoth.proto
[17:30:09] <laca> hmm...
[17:30:24] <laca> Pietro_S: what happens if you try to run pkgmk with this proto manually?
[17:30:30] * laca looks up the exact command line
[17:32:02] <laca> pkgmk -o -d /tmp -f /export/home/sobotkap/packages/PKGMAPS/proto/SFEwesnoth.proto -r /var/tmp/pkgbuild-sobotkap/SFEwesnoth-1.3.7-build
[17:32:34] *** bowdengl has quit IRC
[17:32:46] *** nachox has quit IRC
[17:33:12] * stevel can't believe that "An Open Letter..." thread on osol-discuss is still going on
[17:33:45] <Pietro_S> wow: ## Building pkgmap from package prototype file.
[17:33:46] <Pietro_S> Segmentation Fault (core dumped)
[17:33:57] <laca> there
[17:34:12] <laca> can you upload pstack core to you site pls
[17:34:25] <laca> what build of nevada, btw?
[17:34:37] <laca> which
[17:35:03] *** laca has quit IRC
[17:35:50] *** laca has joined #opensolaris
[17:36:03] <Pietro_S> this is still 65b, but 69b (notebook)  has same problem
[17:37:06] *** mlh has quit IRC
[17:37:06] *** Erwann has quit IRC
[17:37:07] *** ysss has quit IRC
[17:37:17] <laca> grrr... there's some magic kbd combination that i keep pressing instead of Ctrl-Alt-Right that kills the xserver like Ctrl-Alt-BackSpace  :/
[17:37:24] *** worklez has quit IRC
[17:37:48] <laca> not always though, so i can't reproduce it if I'm trying to
[17:37:56] <laca> Pietro_S: did you upload the pstack output?
[17:38:20] <axisys> hmm.. 4 more days before sol 10 u4 should release?
[17:38:43] <laca> Pietro_S: i will need the core file itself and the prototype for a bug report
[17:38:50] <Pietro_S> laca: http://home.zcu.cz/~sobotkap/stack.txt
[17:39:23] <Pietro_S> laca: np, I'll upload it too
[17:39:49] <laca> Pietro_S: huh? is that all?
[17:40:28] <Pietro_S> yes
[17:40:37] <Pietro_S> quite strange
[17:41:26] *** vmlemon has quit IRC
[17:41:48] <Pietro_S> the core is written to current dir, right?
[17:41:56] <laca> yes
[17:42:03] <laca> normally
[17:42:04] *** vmlemon has joined #opensolaris
[17:42:35] *** dlynes_laptop has quit IRC
[17:45:17] *** worklezz has quit IRC
[17:45:24] *** mbalmer has joined #opensolaris
[17:48:46] <sickness> http://www.cmyos.com/index.php
[17:50:40] *** bunker has quit IRC
[17:54:34] *** bondolo has joined #opensolaris
[17:54:58] *** charlesn1 has joined #opensolaris
[17:55:59] *** charlesn1 has quit IRC
[17:56:11] *** charlesn1 has joined #opensolaris
[17:57:42] *** vmlemon has quit IRC
[17:58:22] *** vmlemon has joined #opensolaris
[17:58:53] *** sehat_tan has quit IRC
[17:59:48] *** vmlemon has quit IRC
[18:00:27] *** vmlemon has joined #opensolaris
[18:00:43] *** ichigo has joined #opensolaris
[18:01:40] *** ichigo has left #opensolaris
[18:06:05] <oninoshiko> thats interesting, sickness
[18:08:08] *** mikaeld has quit IRC
[18:09:06] *** andyshack has joined #opensolaris
[18:09:10] *** slowhog has joined #opensolaris
[18:09:10] *** vmlemon has quit IRC
[18:10:31] <oxygene> sickness: looks like http://eyeos.org/
[18:11:45] *** endikos has left #opensolaris
[18:11:51] <axisys> we are planning to use 10 x4100 servers as sniffers in our network.. some people want to go with redhat.. but i want to use sol10.. looking for some suggestions
[18:12:08] <axisys> mainly be using etherreal
[18:12:14] <wesolows> suggestions for what?  I agree you should use Solaris.
[18:12:37] <axisys> wesolows: why solaris over redhat is what I am looking for suggestions on
[18:13:38] <wesolows> um, because Red Hat is a pile of shit?
[18:13:55] <wesolows> Seriously, the advantages of Solaris over GNU/Linux are well known and basically plastered all over sun.com.
[18:14:03] <oninoshiko> oxygene: according to wiki-peadia they are related
[18:14:04] <wesolows> If you want something independent, try Usenet.
[18:14:44] *** vmlemon has joined #opensolaris
[18:14:58] <wesolows> stability, reliability, more features, more innovation, in this particular case a better networking stack although you won't benefit much if you're just sniffing.  But you've gotta store that data somewhere, eh?  How about ZFS?
[18:15:06] <axisys> wesolows: i can't agree more.. but i am looking for why solaris is better .. looking for some help getting some strong points
[18:15:30] * oninoshiko hugs ZFS
[18:16:15] <wesolows> how about the fact that we aren't going to break your sniffer in a patch (at least not so long as it's written using public interfaces)?
[18:16:23] <kjetilho> axisys: I don't see a good reason to prefer Solaris over Linux for this application
[18:16:31] *** MattMan has quit IRC
[18:16:42] <wesolows> how about the fact that we actually test our NIC drivers?
[18:16:46] <kjetilho> snort and other tools are better known and supplied with the OS
[18:17:10] * oninoshiko doesn't see a reason to use Linux over solaris in this application
[18:17:13] <wesolows> really what you want for this is an appliance anyway
[18:17:40] <wesolows> you're not going to get much out of your OS either way; the best differentiator here is probably ZFS because you have to store the data
[18:18:00] <axisys> they will use them as sniffers on our voice traffic to troubleshoot from the collected data
[18:18:05] *** RaD|Tz has quit IRC
[18:18:27] <axisys> wesolows: how about the tcp stack .. how does the two os compares?
[18:18:35] <wesolows> if you're doing crazy math on the data to look for problems, you'd likely benefit from Solaris's greater CPU scalability
[18:18:49] <axisys> wesolows: good point
[18:18:59] <trochej> And better scheduler?
[18:19:07] <wesolows> how about http://www.sun.com/bigadmin/xperts/sessions/11_fireengine/
[18:19:15] <axisys> trochej: i was gonna ask about that ;-)
[18:21:20] <axisys> i am also looking for an indepdent study on tcp/ip on diff os
[18:21:27] *** TBCOOL has quit IRC
[18:21:32] *** ichigo_ has quit IRC
[18:21:33] *** TBCOOL has joined #opensolaris
[18:21:56] <wesolows> axisys: good luck finding that.  Most independent testing labs are really expensive, and the ones who test and publish for free rarely do a good job
[18:22:24] <wesolows> why don't you get some evaluation systems and start benchmarking, if performance is a big deal for you?
[18:22:41] <wesolows> snooping is hardly a typical task; you'll find specweb numbers but that's nothing like what you're doing
[18:23:53] <axisys> wesolows: thanks for the suggestions.. running two os and do the perf test would probably be an excellent way to find the answers..
[18:24:03] <sickness> oxygene: eheh
[18:24:08] <wesolows> a revelation, I'm sure
[18:25:22] <wesolows> if you want to be an effective advocate, you need more than "this guy in irc said it's better at X"; research is hard work and takes time.  The more you personally know about the problem and understand the characteristics of the candidate solutions, the more effective an advocate you'll be.
[18:25:54] <wesolows> get it in the lab, understand the software, test, test, test
[18:25:59] <umdstu_> wesolows: if 'this guy' happens to be myself
[18:26:04] <umdstu_> then thats all the argument you need
[18:26:16] <umdstu_> i'm a bit of a genius
[18:26:21] <wesolows> I'm sure you are
[18:26:28] <umdstu_> lol
[18:27:09] <wesolows> appeals to authority may or may not work on stupid IT directors, but even so it's better to understand the problem yourself
[18:27:22] *** tg has quit IRC
[18:27:29] *** tg_ has joined #opensolaris
[18:27:42] *** tg_ is now known as tg
[18:27:54] <umdstu_> i agree, i was only kidding
[18:27:59] <umdstu_> i work in a lab
[18:28:04] <umdstu_> all i do all day is test test test
[18:28:17] <umdstu_> that and break things for fun
[18:28:38] <umdstu_> so exim channnel seems a little slow.  anyone in here use Exim MTA?
[18:29:14] *** prg3 has quit IRC
[18:32:32] <andyshack> is a v480 half an 880 (ish?)
[18:36:35] <kjetilho> umdstu_: sure
[18:37:12] <umdstu_> kjetilho: oh hey
[18:37:20] <mustang> andyshack: yes, if you ignore the disk slots.
[18:37:29] <umdstu_> so i tried to install it but am getting stumped at the 'make'
[18:37:32] <andyshack> cool thanks
[18:37:54] <umdstu_> I followed the instructions and copied the /src/EDITME file to /Local
[18:37:55] <kjetilho> umdstu_: eh.  sorry, I don't have patience to help you with building
[18:37:58] <umdstu_> after editing it ofcourse
[18:38:01] <umdstu_> haha ok
[18:38:14] <kjetilho> off to the pub anyways :-)
[18:38:23] <umdstu_> enjoy :)
[18:38:28] <kjetilho> good luck :)
[18:38:34] <umdstu_> thanks
[18:38:55] *** comay has joined #opensolaris
[18:44:07] <richlowe> mornin' comay
[18:44:30] *** JSRJ has joined #opensolaris
[18:46:05] <comay> hey rich
[18:54:20] <umdstu_> jmcp: you there?
[18:57:01] <quasi> probably the middle of the night in his part of the world
[18:57:28] <andyshack> 3am
[19:00:26] <CIA-16> tsien: 6550889 The cpumem de does not include the chassis-id in the FRU FMRI (fix ELF runtime)
[19:00:27] <CIA-16> mmusante: 6573276 spare device inuse check problem
[19:00:28] <CIA-16> rie: PSARC/2007/474 new ldd(1) -w option, 6595139 various applications should export yy* global variables for libl
[19:04:58] *** estibi has joined #opensolaris
[19:06:46] *** nostoi has joined #opensolaris
[19:08:47] *** mikefut has quit IRC
[19:10:02] *** cajetanus has joined #opensolaris
[19:11:15] *** migi has quit IRC
[19:13:23] *** bnitz has left #opensolaris
[19:14:37] *** timsf has quit IRC
[19:16:51] *** freakazoid0223 has joined #opensolaris
[19:18:28] *** SunTzuTech has quit IRC
[19:18:31] *** JSRJ has quit IRC
[19:21:24] *** Saltsa has quit IRC
[19:21:24] *** reflect has quit IRC
[19:21:24] *** jwit has quit IRC
[19:21:24] *** phips has quit IRC
[19:21:24] *** BatonT has quit IRC
[19:21:24] *** PerterB has quit IRC
[19:21:24] *** tg has quit IRC
[19:21:24] *** dlg has quit IRC
[19:21:25] *** razrX has quit IRC
[19:21:28] *** andyshack has quit IRC
[19:21:28] *** Pietro_S has quit IRC
[19:21:28] *** hsilva has quit IRC
[19:21:29] *** CIA-16 has quit IRC
[19:21:29] *** Trident has quit IRC
[19:21:29] *** sparvu has quit IRC
[19:21:29] *** madhatter has quit IRC
[19:21:29] *** LeftWing has quit IRC
[19:21:30] *** Triskelios has quit IRC
[19:21:30] *** jmcp has quit IRC
[19:21:30] *** Kush- has quit IRC
[19:21:30] *** logic_ has quit IRC
[19:21:31] *** tomww has quit IRC
[19:21:31] *** polk__ has quit IRC
[19:21:31] *** Jiko has quit IRC
[19:21:32] *** hallu_ has quit IRC
[19:21:32] *** kito has quit IRC
[19:21:32] *** elektronkind has quit IRC
[19:21:32] *** phalenor has quit IRC
[19:21:33] *** cormac has quit IRC
[19:21:34] *** fluffle has quit IRC
[19:21:34] *** hali has quit IRC
[19:21:35] *** schily has quit IRC
[19:21:35] *** FrostCS has quit IRC
[19:21:35] *** dosiu has quit IRC
[19:21:35] *** iMax has quit IRC
[19:22:19] *** tg has joined #opensolaris
[19:22:19] *** andyshack has joined #opensolaris
[19:22:19] *** Pietro_S has joined #opensolaris
[19:22:19] *** BatonT has joined #opensolaris
[19:22:19] *** hsilva has joined #opensolaris
[19:22:19] *** CIA-16 has joined #opensolaris
[19:22:19] *** Trident has joined #opensolaris
[19:22:19] *** sparvu has joined #opensolaris
[19:22:19] *** madhatter has joined #opensolaris
[19:22:19] *** LeftWing has joined #opensolaris
[19:22:19] *** dlg has joined #opensolaris
[19:22:19] *** Saltsa has joined #opensolaris
[19:22:19] *** Triskelios has joined #opensolaris
[19:22:19] *** jmcp has joined #opensolaris
[19:22:19] *** Kush- has joined #opensolaris
[19:22:19] *** logic_ has joined #opensolaris
[19:22:19] *** razrX has joined #opensolaris
[19:22:20] *** tomww has joined #opensolaris
[19:22:20] *** reflect has joined #opensolaris
[19:22:20] *** jwit has joined #opensolaris
[19:22:20] *** polk__ has joined #opensolaris
[19:22:20] *** dosiu has joined #opensolaris
[19:22:20] *** kito has joined #opensolaris
[19:22:20] *** FrostCS has joined #opensolaris
[19:22:20] *** schily has joined #opensolaris
[19:22:20] *** iMax has joined #opensolaris
[19:22:20] *** phips has joined #opensolaris
[19:22:20] *** PerterB has joined #opensolaris
[19:22:20] *** elektronkind has joined #opensolaris
[19:22:20] *** Jiko has joined #opensolaris
[19:22:20] *** cormac has joined #opensolaris
[19:22:20] *** phalenor has joined #opensolaris
[19:22:20] *** fluffle has joined #opensolaris
[19:22:20] *** hallu_ has joined #opensolaris
[19:22:20] *** hali has joined #opensolaris
[19:34:18] *** andyshack has quit IRC
[19:36:32] *** Phenox has quit IRC
[19:37:24] *** alanc-away is now known as alanc
[19:39:47] <tsp> what am I doing wrong? I'm typing "stty -ixon" at my shell, but if I hit ^s it still stops the output
[19:41:50] *** freakazoid0223 has quit IRC
[19:42:15] *** freakazoid0223 has joined #opensolaris
[19:44:28] *** bubbva has joined #opensolaris
[19:46:00] <tsp> ah, screen
[19:48:12] *** onlinebacon has joined #opensolaris
[19:48:59] *** onlinebacon has quit IRC
[19:53:46] *** MacMarc has joined #opensolaris
[19:55:18] *** cydork has quit IRC
[19:56:42] <mbalmer> hey macmarc...
[19:57:12] <MacMarc> hi mbalmer
[20:00:00] *** MacMarc has left #opensolaris
[20:13:53] *** UGH has joined #opensolaris
[20:14:12] *** WickedWicky has quit IRC
[20:16:21] *** freakazoid0223 has quit IRC
[20:17:55] *** UGH is now known as WickedWicky
[20:19:01] *** cajetanus has quit IRC
[20:21:09] *** Murmuria has joined #opensolaris
[20:27:07] *** freakazoid0223 has joined #opensolaris
[20:33:04] *** Ljuvnix is now known as Solfreya
[20:34:32] *** RaD|Tz has joined #opensolaris
[20:34:57] * oninoshiko stabs intel hardware with a screwdriver "1t w45 n07 1337 3n0ugh"
[20:35:32] <e^ipi> o_O
[20:36:05] <Solfreya> a student at our course centre managed to pour nail polish into the computer keyboard in one of our classes today
[20:36:32] <oninoshiko> sorry, there tech support people have managed to piss twice in as many days
[20:36:40] <oninoshiko> piss me off*
[20:37:10] <e^ipi> Solfreya: all the more reason to require students to buy their own computer equipment
[20:37:18] <oninoshiko> "we are sorry, because you have opensolaris ON THE FREAKING NETWORK we will not support out hardware
[20:37:26] <oninoshiko> our*
[20:37:49] <e^ipi> oninoshiko: you were calling about consumer goods, or server gear?
[20:37:52] <Solfreya> oninoshiko: tell them you suspect there might be a Solaris box somewhere in their county. They should shut down operations now before its too late!
[20:38:42] <oninoshiko> e^ipi: Pro/1000 PT Dual-port server card
[20:39:17] <e^ipi> and they don't support this on solaris?
[20:39:18] <e^ipi> that's odd
[20:39:19] <Solfreya> its allergic to solaris, police academy 6 and Scary Movie 4
[20:39:37] <WickedWicky> the latter two I can opt for
[20:39:59] <oninoshiko> whats odd is that im not trying to use it on a solaris machine... im trying to use it to connect two a solaris machine...
[20:40:07] <oninoshiko> to*
[20:40:24] <Solfreya> tell them you try to connect to a Windows Vista machine
[20:40:54] <oninoshiko> the twit saw "OpenSolaris" in the email and sent me back a canned reply
[20:40:57] *** micken_ has joined #opensolaris
[20:41:18] <micken_> oh this was a slow grps line ..:(
[20:41:23] <micken_> gprs
[20:41:27] <e^ipi> but still... a server board not supporting Solaris?
[20:41:37] <e^ipi> seems like an odd business strategy to me
[20:41:49] <umdstu_> any exim users out there
[20:41:59] <WickedWicky> try asking HP for help when you wanna run their SNMP agents on linux, same thing
[20:42:01] <e^ipi> like a digital camera or something where the company doesn't support windows
[20:42:15] <WickedWicky> dutch HP support stalls as soon as you come with a non windows product
[20:42:26] *** delewis has quit IRC
[20:42:26] *** micken_ has quit IRC
[20:42:32] <e^ipi> WickedWicky: try it with HPUX or VMS
[20:42:39] <oninoshiko> EVER saying "we wont support you" to a paying costumer seems like an odd business strategy to me
[20:43:01] <WickedWicky> well, last time I told them that we might as well pull the plug for the support contract then
[20:43:02] <wesolows> depends, if supporting them would cost more than the profit on their business, it may be wise
[20:43:15] <WickedWicky> which made their opinion about what's supported and not turn 160 degrees
[20:43:24] <WickedWicky> 180 too
[20:43:46] <Solfreya> I used HP support text chat to ask for help with an HP 2200 printer on Windows Server 2003. He kept insisting I had to buy a new printer.
[20:44:12] <Solfreya> then I told the support guy "oh, wait, I managed to get it working by using another printer driver" and he went insane ;) "No no no! No supported! Stop stop" ;)
[20:44:13] <WickedWicky> HP 2200 is quite old, init? :P
[20:44:23] <oninoshiko> wesolows: ahh but that usually depends on the amount of the product the client is going to buy if support is offered.... if this works, i plan on getting about 200 of them
[20:44:44] <e^ipi> i used to do tech desk for HP... I got one support call for !windows the whole time i was there
[20:45:11] <wesolows> oninoshiko: yes...the business decision has to be considered carefully.  But hardware margins are small and people are expensive
[20:45:14] <e^ipi> *shrug* I helped the guy out, but he's lucky he got someone that used linux @ home at the time
[20:45:59] <oninoshiko> bad PR is not cheap either
[20:48:05] <Solfreya> it's cheaper than you might think :-p
[20:50:18] <oninoshiko> i would have even been willing to accept "well we need 1k for a support contract to help you"
[20:51:39] <Solfreya> our company is an MS Premium Gold Business Partner and whatnot. I asked them for help with sysprep and WIM (their new Vista-esque ghost replacer), and they never got back to me despite reminding them :p
[20:53:15] *** Mdx4 has joined #opensolaris
[20:53:47] <oninoshiko> well is it any wonder we encourage clients to use things other then windows?
[20:54:11] <Solfreya> and HP? ;)
[20:54:59] <madhatter> Hello. Can somebody help me here: I am planning to install opensolaris on my x86 and am looking for the right hardware. So I downloaded the check-cd for installation and run it with a tp-link wlan pci card installed.
[20:55:38] <madhatter> I got the message that no drivers were available.
[20:56:12] <madhatter> So I bought a used D-Link dwl-510 card because I read on the page for that driver that it is possible to use it.
[20:56:31] <madhatter> Now I run the check again with the d-link card installed. No drivers available.
[20:57:05] <madhatter> Is that cd not up-to-date? Does it not know about all drivers available?
[20:57:34] <oninoshiko> do you know which version of the CD you have?
[20:58:00] <madhatter> Hm, no. I downloaded it a couple of weeks ago.
[20:58:56] <oninoshiko> most of my experience with OS has been using wired Ethernet, so i cant speak much about 802.11 cards
[20:59:25] <madhatter> The router is in another room, so I really need to use wireless network access.
[20:59:47] <SYS64738> does exist a pkg for postgrey (postfix greylisting) ?
[20:59:53] <e^ipi> my fixed machines are all wired...
[21:00:04] <e^ipi> my laptops have always been atheros
[21:00:08] <madhatter> But what I am wondering about is that the first card was with an Atheros chip, which I was told would work well with most unix and linux
[21:00:35] <oninoshiko> madhatter: is this a desktop or a laptop?
[21:01:03] <madhatter> e^ipi: I had an atheros card installed. Vendor was tp-link, but that did not work either. Or better: the check said that it would not work.
[21:01:12] <madhatter> oninoshiko: It's a desktop
[21:01:33] <WickedWicky> I have a dlink 650 and it works fine, it uses the atheros chipset
[21:01:37] <e^ipi> why not just run some cat5 ?
[21:02:08] <oninoshiko> hrm... alot of the desktop adaptors (at least used to) use a pci-pcmcia bridge... i wonder if that might be the issue
[21:02:24] <madhatter> e^ipi: I live in the rooms. So putting wires on the floor a) looks bad and b) forces discussions with the miss
[21:02:38] <e^ipi> so put it above the roof
[21:02:49] <Solfreya> oninoshiko: the problem with Mac, IBM Aix and Sun Solaris and other systems is that there aren't enough problems with them to sustain an entire market of troubleshooters like us ;)
[21:02:51] *** jafari has quit IRC
[21:02:52] <e^ipi> or under the flooring
[21:03:07] <madhatter> e^ipi: No flooring
[21:03:13] <Solfreya> or as large a market anyhow
[21:03:17] <e^ipi> or clip it to the corners (ceiling or floor) with cable clips, it looks fine
[21:03:30] <CIA-16> rs200217: 6597645 SUNWdsdr is missing copyright file
[21:03:36] <madhatter> I will look up that driver's page
[21:04:52] <oninoshiko> you can alway just try it and see... i havent tried the support-check disk, so i cant testify to its accuracy
[21:04:56] <madhatter> http://tinyurl.com/355fec
[21:05:21] <oninoshiko> i think there are a couple of live-cd solaris based systems
[21:05:38] <madhatter> oninoshiko: Hehe, I wanted to be safe. So I tried to figure out which hardware setup will work best first.
[21:05:54] <RElling> you'll want to use the checker for Nevada 7/07, whenever it is available... it is usually faster to download the latest NV build and take a look
[21:06:06] *** dlg has quit IRC
[21:06:10] <oninoshiko> madhatter: Hence trying a live cd ^_^
[21:06:11] <madhatter> oninoshiko: I tried the b*something distiro live-cd. It crashed when starting x-window system
[21:06:23] <oninoshiko> oh :(
[21:07:57] *** cypromis has quit IRC
[21:08:29] *** cypromis has joined #opensolaris
[21:09:27] <e^ipi> how accurate is that check CD anyways?
[21:09:32] <e^ipi> I've never bothered with it
[21:10:32] <madhatter> e^ipi: That's what I am wondering about. When you look at the page I posted, there is the card listed. Exactly that card.
[21:10:33] *** RaD|Tz has quit IRC
[21:11:03] <e^ipi> install nevada... if it works, great, if it doesn't oh well
[21:11:07] *** Dar is now known as Dar_HOME
[21:12:42] <madhatter> e^ipi: Which of the both has the name nevada? Developer or Community?
[21:13:02] <e^ipi> CE.. i'm still not certain why DE exists
[21:13:22] <madhatter> I thought DE is the more tested/ stable one
[21:13:36] <RElling> DE exists so that someone can purchase a "support" contract
[21:13:49] <oninoshiko> in theory, be we all know the developers are perfect
[21:13:49] <richlowe> where the quotes around "Support" are very important.
[21:14:19] <richlowe> the kind of support where you give them money, and more than likely, they don't actually help you. :)
[21:14:38] <madhatter> So there is nothing really more stable about DE?
[21:15:36] <oninoshiko> they provided source, what more do you want?
[21:15:57] <RElling> that can only be answered on a case-by-case basis... DE 7/07 will be more stable than b70, since it is built on b70a or b70b,pending testing
[21:16:01] <oninoshiko> its better then another company that has made my %#@ list
[21:18:01] <madhatter> RElling: Those namings for versions still make my head hurt
[21:19:28] <richlowe> integer is the build number, characters are the number of times changes pulled back into that build.
[21:19:41] <richlowe> or in some cases, integrated into the snapshot first, because DE makes people lose all sense of bloody decorum.
[21:19:41] <RElling> yeah, it looks kinda silly when b70a arrives after b71... IMHO b72 has enough changes to make it compelling, at least for my hardware
[21:20:11] <RElling> BTW, a CE release includes the DE bits, so you don't really lose anything
[21:20:13] <richlowe> In any practical case, DE is more trouble than it's worth for everyone involved.
[21:20:28] *** shirish has joined #opensolaris
[21:20:29] <e^ipi> madhatter: when DE comes out, the build it's based on gets another week or two of testing
[21:20:59] <richlowe> when DE comes out, the build is delayed while people do stupid things for the benefit of DE
[21:21:02] <e^ipi> but ultimately it's no more tested than any other old version of CE
[21:21:03] <richlowe> *then* it gets more testing.
[21:21:10] <richlowe> let's not sugar coat how stupid this entire thing is.
[21:21:11] *** shirish has left #opensolaris
[21:22:29] *** dlg has joined #opensolaris
[21:22:34] <oninoshiko> *IM* use CE, so you sould too :p
[21:22:37] <madhatter> richlowe: So, the first character, 'b', is fix?
[21:22:48] <richlowe> No, characters after the digit.
[21:23:00] <richlowe> snv_70 == build 70, snv_70b == build 70 + two spins of extra fixes
[21:23:06] <e^ipi> they mean something went horribly, horribly wrong with that build
[21:23:15] <richlowe> except in the case of DE.
[21:23:16] <wesolows> or not so horribly, depending
[21:23:23] <richlowe> where someone could have broken wind at the wrong juncture.
[21:23:45] <richlowe> as I said, the safest thing to do with DE is denigrate it as often as possible in the hopes people get the right idea
[21:23:51] <wesolows> the whole thing would make a lot more sense if Sun's product-specific actions were hidden behind an iron wall
[21:24:04] <richlowe> wesolows: not really, it'd just piss me off less.
[21:24:13] <richlowe> it's still stupid, it would just not have a downside I care about
[21:24:23] <madhatter> richlowe: Because you wrote b70a up there. So that is snv_70a?
[21:24:34] <wesolows> richlowe: well, let me put it this way, then: what you would see would make a lot more sense if all the SXDE stuff were hidden
[21:25:09] <e^ipi> then there'd just be delayed builds for no reason whatsoever
[21:25:16] <richlowe> No, because they'd do things properly.
[21:25:22] <e^ipi> which, given... also happens
[21:25:23] <richlowe> and not delay the snapshot for some retarded pseudo-product.
[21:25:50] <richlowe> which may actually be happening, given what happened with 70
[21:26:04] <richlowe> it was late, but it wasn't "let's wait for the (still not done) respins of 70" late.
[21:30:00] *** nostoi has quit IRC
[21:34:32] *** sahafeez has joined #opensolaris
[21:35:15] *** micken has quit IRC
[21:39:00] *** mbalmer has quit IRC
[21:41:08] *** sparc-kly_ has quit IRC
[21:41:29] *** sparc-kly_ has joined #opensolaris
[21:44:02] *** sparc-kly_ has quit IRC
[21:44:31] *** sparc-kly_ has joined #opensolaris
[21:44:49] <Stric> Not opensolaris, but has anyone seen that on sol10 when patching (with pca at least) that you in addition to /var/sadm/pkg/<pkg>/save/<patch> get /var/sadm/pkg/<pkg>/save/pspool/<pkg>/save/<patch> which seem to contain the same info..
[21:45:00] *** oninoshiko has quit IRC
[21:45:07] <Stric> so the pspool directory is just a copy..
[21:45:08] *** sahafeez has quit IRC
[21:45:38] *** sahafeez has joined #opensolaris
[21:46:19] *** sahafeez has joined #opensolaris
[21:46:20] <Stric> hm. google seem to say it's some zone related stuff..
[21:46:46] *** sahafeez has quit IRC
[21:48:02] *** Murmuria has left #opensolaris
[21:49:03] *** tulcod has joined #opensolaris
[21:49:12] <tulcod> hello
[21:49:19] <tulcod> I have a little question
[21:49:24] <tulcod> no
[21:49:25] <tulcod> wait
[21:49:26] <tulcod> I have 2
[21:50:04] <tulcod> 1. are the overall configuration files stored in /etc ? are other configuration files stored in ${HOME}/.somedir/somefile.conf ?
[21:50:30] <alanc> system configuration in /etc, user-specific configuration in $HOME/.whatever
[21:50:35] <tulcod> k
[21:50:36] <tulcod> thx
[21:54:13] *** tulcod has left #opensolaris
[21:57:55] <elektronkind> you should have said "In the Solaris Registry file. See the man page for /usr/sbin/regedit"
[21:58:08] <sickness> ghgh
[22:00:44] <alanc> it's a "Service Repository", not a "Registry"  8-P
[22:01:14] *** vmlemon has quit IRC
[22:01:22] <richlowe> /usr/sbin/regedit should clearly be a link to the vpanels bits.
[22:01:24] <Solfreya> lol /sbin/regedit ;)
[22:01:51] <Solfreya> smc is kind of like mmc though ;)
[22:01:57] *** sahafeez has joined #opensolaris
[22:02:17] <sickness> mmc is much more lightweight, tho :|
[22:02:42] <elektronkind> yes, and it's stored on the 15th cylinder of slice 1, and using swap(1m) on days with even numbers might, just might, accidentally clobber all or parts of the registry stored at that location. And there's no backup of it anywhere else.
[22:02:45] <Solfreya> mmc is pretty heavy
[22:02:47] <Solfreya> oops
[22:02:51] <Solfreya> in version 3
[22:03:01] <Solfreya> the one in Vista and Windows Server 2008
[22:03:27] <Solfreya> someone shoot me, I held a course in Windows Server 2003 today and I manage to get all the references to Windows 2000 Server and Windows Server 2003 right ;)
[22:03:35] *** Murmuria has joined #opensolaris
[22:03:42] *** Murmuria has left #opensolaris
[22:03:54] <Solfreya> Microsoft are EXTREMELY adamant about getting "2000 Server" and "Server 2003" right ;)
[22:04:18] * elektronkind never realized there was a difference
[22:04:50] *** helpme_ has joined #opensolaris
[22:05:03] <Solfreya> there are, Microsoft switched to an MWPG naming syntax after Windows 2000 which had an MWGP syntax
[22:05:56] <Solfreya> at least that's what they told us at the Ms TTT course ;)
[22:06:43] <Solfreya> there's two things that you need to know to train Microsoft classes - their acronym rules and that they take security seriously ;)
[22:06:45] * Solfreya giggles
[22:09:14] *** switch_ has quit IRC
[22:09:22] <holcomb> can there be more than one name_service in a sysidcfg?
[22:12:15] <coffman> hm
[22:12:21] <coffman> flash craps at me
[22:12:24] *** cajetanus has joined #opensolaris
[22:15:26] *** Jerome71 has joined #opensolaris
[22:19:46] *** migi has joined #opensolaris
[22:21:02] <duri> is sol10u4 coming out today ?
[22:21:52] <elektronkind> one could wish
[22:22:33] *** delewis has joined #opensolaris
[22:23:15] <elektronkind> no mention of it on blogs.sun
[22:23:42] *** JSRJ has joined #opensolaris
[22:23:51] *** cajetanus has quit IRC
[22:24:28] <quasi> duri: the last I heard was pushback to early september, but still keeping the 8/07 name
[22:25:10] <duri> just saw on article on techtarget.com mentioning that it was coming out today ... thanks
[22:25:22] <elektronkind> they should just rename it "10/07" so it would come out early/on time for a change ;)
[22:25:34] <delewis> not on SDLC, yet.
[22:25:48] *** JSRJ has quit IRC
[22:26:05] * quasi kicks elektronkind - bloody pessimist ... if it gets postponed even furter, maybe I should blame you ;)
[22:26:10] *** cajetanus has joined #opensolaris
[22:26:25] * delewis wonders why Sun can't set a date and meet it
[22:26:50] <palowoda> All the good stuff is in opensolaris anyways.
[22:26:51] <Tpent1> maybe because we care about shipping something that is correct rather than on time?
[22:27:05] *** Jondice_ver_2 has joined #opensolaris
[22:27:10] <delewis> Tpent1: do both?
[22:27:12] <Tpent1> and no, I don't have any information on release date
[22:27:15] <bubbva> delewis: because sometimes quality is more important than a date.
[22:27:18] <alanc> you can have s10u4 bits ship in August, or you can have s10u4 bits that don't corrupt your data - your choice
[22:27:30] <alanc> Sun chose "don't corrupt data"
[22:27:31] <quasi> delewis: they've refocused by doing s/SUNW/JAVA/ - getting that substitued all ower the code probably led to a few regressions ;)
[22:27:34] <delewis> bubbva: I suppose patches released on SunSolve don't adhere to that?
[22:27:37] <bubbva> (I don't agree with naming our releases for dates, but as long as they don't predict out too far in advance)
[22:27:58] <delewis> (let's not get into the kernel patch that renders v880/v480 systems unbootable if you don't take a chance to select the right dependency patch to install)
[22:28:06] <bubbva> delewis: actually they do. often we badpatch patches once they get into our patch test pipeline before they go onto sunsolve.
[22:28:17] <bubbva> even if it is a critical patch that "everyone" is waiting for.
[22:28:20] <alanc> and changing the name to 9/07 at this point would add a lot more delay that no one wants
[22:28:21] <elektronkind> I've seen tons of blogs that mention "x feature will be in u4, y update will be in u4" ... I wonder if this is a case where the features/fixes list got a bit too ambitious
[22:28:24] <bubbva> mistakes happen, but the patches are tested.
[22:29:33] <bubbva> elektronkind: I'd be interested in those blogs - we aren't supposed to promise features in a specific release until it's on it's way out the door.
[22:29:51] <elektronkind> just search blogs.sun.com for "update 4" :)
[22:29:52] <bubbva> techleads & gatekeeper reserve the right to back things out that are broken
[22:30:07] <sahafeez> any of the old 1u's work on a 2 post telco rack?
[22:31:03] <sahafeez> something that i could get on ebay
[22:31:35] <bubbva> thanks elektronkind.  blogs.sun.com is a weird spot, but I'll mention it to folks.
[22:31:37] <elektronkind> and I imagine that what shows up there is only a part of what the complete list is. I'm not complaining, just miffed that bug fix patches for existing installations of s10 seem to be held up.
[22:32:35] <elektronkind> since the first thing I dive into is the ExtraPatches directory on the update's ISO.
[22:32:45] <alanc> if you remember the discussion earlier today about adding letters after build names, and notice the fixed in build on http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6564677, you can see why it was delayed a little   (not much, due to the broken "see comments" description, but the synopsis is telling enough)
[22:34:25] <bubbva> elektronkind: updates and patches do share release trains. it is annoying.
[22:35:09] <bubbva> changes were made in the U4 timeframe to make it less crappy, but the end of the release always brings delays when things tighten up to maintain quality
[22:35:49] *** estibi_ has joined #opensolaris
[22:35:59] <Tpent1> rubbish like "they're all concentrating on some media shit that happened in the last few days" are comments spoken out of pure ignorance
[22:36:02] <RElling> sahafeez: worst case, put them on a shelf
[22:36:19] <elektronkind> I can understand the reason for that. I guess I'm being kind of selfish in my miffed-ness since, being that I work for a .edu, our patching windows come August/September number in the slim to none :)
[22:36:32] <sahafeez> hum. that would be ugly.
[22:36:34] *** cajetanus has left #opensolaris
[22:37:13] <bubbva> elektronkind: don't you get another window in November? also, doesn't your school do a braek in October? or we just spoiled in the midwest?
[22:37:50] <alanc> break in October?   never heard of that...
[22:37:53] <elektronkind> the next break for us is thanksgiving. We don't even get Columbus Day off.
[22:38:04] <delewis> alanc: "fall break"
[22:38:20] <alanc> never had one
[22:38:22] <bubbva> also, until just recently, you should've been getting patches. We've maintained a better "split gate" model in U4 for ON, keeping "not quite ready" features out of tehre - so more patches could get out with quality
[22:38:51] <bubbva> alanc: yeah, we had "October break" at Purdue. They still haev it, because I went there in October a ocuple years back and the campus was deserted
[22:39:01] <Tpent1> the unfortunate part of that is that many things that probably should have gone into teh sustaining gate for patches, didnt, so we got a lot of "fixed in update 4" type issues
[22:39:20] <alanc> since Berkeley ran on semester system, classes ran unbroken (except for a long weekend or two) from late August until Thanksgiving
[22:39:30] <elektronkind> bubbva: a break in october sounds good. A chance to study/blow off steam before mid-terms I suppose
[22:39:58] <delewis> we only get 2 days off here in the fall for our break.
[22:40:01] <elektronkind> and for the sysadmins to reboot core systems ;)
[22:40:36] <bubbva> alanc: purdue was on semesters.
[22:40:53] <bubbva> we also got MLK day & Presidents day in the spring (along with spring break)
[22:41:06] <elektronkind> anyhow, I can probably fenagle a window after drop/add is over
[22:41:09] <alanc> the other California schools I knew of just had classes from mid-September to Thanksgiving with no fall break
[22:41:12] <bubbva> delewis: October break was like a 4 day weekend.
[22:41:25] <delewis> bubbva: that's what our's is here.
[22:41:40] <bubbva> just enough time to go home & do 2 months worht of laundry ;-)
[22:42:46] *** switch has joined #opensolaris
[22:44:34] <delewis> never, ever pkgrm your screen package while you've got a screen open doing several things, like pkgadd'ing packages.
[22:44:52] <elektronkind> doh
[22:46:10] <e^ipi> heh
[22:47:05] <Tpent1> delewis: that sounds close to the faux par of trussing your X server ;)
[22:47:51] <delewis> I was trying to clean up this particular a system a bit by uninstalling several of my own packages and installing packages off the CCD, instead. When I detached that screen, the package database was locked, and thus I couldn't re-install screen. I finally extracted the screen binary from my screen package and copied it somewhere temporarily to execute it and finish installing the packages I was installing to unlock the package d
[22:48:09] <alanc> I've often pkgrm'ed my Xserver from in an xterm running it - usually works okay if you've got enough RAM that everything X needs is paged in
[22:48:38] <delewis> Tpent1: trussing every pid on your system is even worse. :-)
[22:48:49] *** estibi has quit IRC
[22:49:03] <alanc> pmap Xsun was the one that always used to deadlock me before Bart fixed it to buffer the output until after it's let go of the target process
[22:49:52] <Tpent1> to use a quote from many of bmc's talks "Have you been trussing your X server again?"\
[22:50:23] * alanc put dtrace probes in his X server so he wouldn't have to truss it...
[22:50:37] <richlowe> alanc: I've had fairly evil things happen gcore'ing Xorg
[22:50:38] <delewis> someone should suggest to bmc that he take a look at acroread next time he is bored and wanting something to look at with DTrace.
[22:50:43] <richlowe> but I've never been able to do it reliably.
[22:50:52] * delewis saw an acroread last night doing 12000 syscalls/second
[22:50:56] <Tpent1> well the latest was trash-applet
[22:51:01] <richlowe> delewis: I don't think bmc is in a position to get bored.
[22:51:25] <alanc> delewis: but what would he do afterwards?   not like we could fix acroread's problems once they're found
[22:52:04] <alanc> richlowe: I get enough Xorg cores on my own without having to use gcore to make more thank you...
[22:52:28] <delewis> does Sun not have source to acroread or does Adobe just hand you a binary to ship? I thought you guys might've had a situation like you do with RealPlayer.
[22:52:40] <delewis> of course you probably can't comment if you do or don't. :-)
[22:53:21] <richlowe> maybe it's in the same room as DPS.
[22:53:54] * bubbva is back (gone 03:12:36)
[22:54:46] *** dunc has joined #opensolaris
[22:56:20] *** Mdx4 has quit IRC
[22:56:33] <alanc> delewis: we download the binary from adobe.com
[22:56:53] <alanc> if we had source, don't you think we'd have built on x86 by now?
[22:57:08] <delewis> alanc: not if Adobe won't let you. :-)
[22:58:29] <alanc> I don't think Sun would sign a contract that dictated which of our platforms we can build sources on
[22:59:18] <Stric> alanc: I think I've gotten fscked by that pmap/X thing.. ;)
[23:00:07] <elektronkind> what does pmap do to X (haven't tried it, not so sure if he should ;)
[23:00:08] *** GmanAFK has quit IRC
[23:00:15] <alanc> shouldn't be a problem in S10 - on older releases you just need to remember to do "pmap > /tmp/x ; less /tmp/x" so it's not trying to draw the output at the same time it's locked the server process to get the data
[23:00:33] <Stric> |dd bs=1024k   :)
[23:00:54] <Tpent1> any prog that tries to do output to the Xserver, while it has stopped it is probably going to cause you grief (eg truss)
[23:01:15] <elektronkind> ah, I see
[23:01:20] <Stric> elektronkind: pmap freezes the process to take a reading.. then wants to output data to that frozen X server
[23:01:34] <Stric> instant deadlock :)
[23:01:47] <alanc> it deadlocks with pmap waiting for xterm to finish reading from stdout, xterm waiting for the Xserver to clear it's command buffer, and the Xserver waiting for pmap to let go so it can run again
[23:02:59] *** baijiutong has joined #opensolaris
[23:07:16] <Tpent1> the good thing is that yo can kill it from another screen
[23:07:22] *** RElling has quit IRC
[23:08:55] <trygvis> when I'm setting a property with svccfg, how do I "commit" it? if I get it with svcprop I'm getting the old value
[23:09:39] *** RElling has joined #opensolaris
[23:12:36] *** Jondice has quit IRC
[23:13:10] *** Jondice_ver_2 has quit IRC
[23:13:37] <kjetilho> alanc: how did you fix it in S10?  setbuf to something larger?
[23:14:13] <richlowe> make pmap wait until it's released the process to start output.
[23:14:14] <trygvis> the output from svccfg -s .. postgresql/data is different from svcprop -p
[23:14:32] <kjetilho> richlowe: ok thanks, was just bringing up the code browser :)
[23:17:13] <richlowe> huh, though I maybe wrong.
[23:18:11] <alanc> I thought that's what bart did
[23:21:56] <trygvis> aha, I need to run svcadm refresh!
[23:26:45] *** NeZetiC_ has joined #opensolaris
[23:29:14] *** Jondice has joined #opensolaris
[23:29:20] *** Jondice_ver_2 has joined #opensolaris
[23:29:37] *** luosm_ has joined #opensolaris
[23:34:57] *** NeZetiC has quit IRC
[23:35:16] *** NeZetiC_ is now known as NeZetiC
[23:38:03] *** ahl has joined #opensolaris
[23:38:51] *** Gman has joined #opensolaris
[23:42:34] *** migi has quit IRC
[23:44:56] *** tamr has joined #opensolaris
[23:48:04] *** nachox has joined #opensolaris
[23:48:30] <nachox> evening
[23:48:55] <tamr> 'afternoon
[23:49:40] *** cajetanus has joined #opensolaris
[23:50:59] *** Fetch has joined #opensolaris
[23:53:14] *** cajetanus has quit IRC
[23:53:29] *** cajetanus has joined #opensolaris
[23:54:55] *** tamr has quit IRC





top