November 7, 2006 
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30
November 7, 2006 [00:00:08] <alanc> you don't want to imagine the size of Sun's internet traffic bills [00:00:31] <gisburn> alanc: Do they account each package seperately on paper ? [00:00:52] <gisburn> s/package/packet/ [00:02:24] <gisburn> bugs.opensolaris.org is slow, too. [00:02:55] <richlowe> so it is. [00:03:02] <richlowe> that's been happening on and off for a few days. [00:03:11] *** lloy0076 has left #opensolaris [00:03:26] <alanc> can't be overloaded from the minimal amount of data it publishes [00:04:19] <richlowe> stevel: see, alanc bitches too. [00:04:38] <richlowe> since various people seem to have the impression only I complain :) [00:05:28] <gisburn> "Your request timed out. Please try connecting to bugs.opensolaris.org later or contact the administrators." [00:05:55] * gisburn looks after the room where the admins hide. [00:06:14] * gisburn hears that the door gets locked and barred from inside. [00:06:42] <Triskelios> speaking of slow, I've been downloading b51 for 7 hours now [00:07:02] <alanc> at that rate you should be done just in time for b52 release [00:07:14] <gisburn> heh [00:08:52] <gisburn> Is the new Futurama episode out ? [00:09:06] *** bunker has quit IRC [00:09:18] * gisburn remembers the day when the 144MBit uplink from his university was saturaed. [00:09:40] <gisburn> more than 80% of the traffic was the download of the last futurama episode. [00:10:02] <gisburn> Maybe sun empoyees are doing the same today ? [00:10:41] <Error_404> stealing TV shows? [00:11:00] <gisburn> Error_404: nah [00:11:01] <alanc> we wouldn't do that at work - wouldn't want sun.com to show up on the RIAA/MPAA trackers [00:11:18] <gisburn> Error_404: the german law has a hole there. [00:11:41] <gisburn> Error_404: as long as the "crime" was commited whle noone licensed it in germany it's legal. [00:13:29] <gisburn> Error_404: more interestingly those contracts are not "public", however to prove a violation such a contract would have to be made public in front of the court. However some cases have been dropped or rejected because the license holders were not willing to open the contract terms. [00:13:51] <jamesb> while we're griping about sun.com... has anyone noticed that googling site:sun.com <guery> is always faster and provides better results than using the sun search box? Perplexing... [00:13:56] <gisburn> (note: I am no lawyer so I don't use the correct terms) [00:14:33] <boyd> jamesb: It's quite common across the net unfortunately [00:15:47] <jamesb> I suppose google do have pagerank to help them, but still... Navigating the java class hierarchy is a paritcular gripe of mine [00:20:17] *** bobbyz has quit IRC [00:22:31] <Gman> http://www.gnome.org/~gman/blog/08112006 [00:22:38] <Gman> i don't suppose it'll help, but it's worth a try :) [00:25:03] *** lloy0076 has joined #opensolaris [00:25:28] <lloy0076> I just BFU'ed to 20061103 and I think the initial terminal process is core dumping. [00:25:43] <lloy0076> I get the classic kernel core dump, see the file syncing message and then get summarily rebooted [00:26:02] <lloy0076> ...and now I'm stuck. [00:26:37] *** rodrickbrown has joined #opensolaris [00:29:23] <alanc> intel graphics? could be http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6487609 [00:30:56] <lloy0076> alanc: Might be that - is there a way I can freeze the dump on the screen rather than have it go past me really quickly? [00:31:15] <alanc> I don't know - maybe one of the kernel guys does [00:32:05] *** sparc-kly_ has joined #opensolaris [00:32:24] <lloy0076> Is there any way to "downgrade" the system... [00:32:46] <lloy0076> (without reinstalling the whole lot and keeping what I want) [00:33:29] <gisburn> lloy0076: ".... sorry, ths is SolarisNT, we suggest to reinstall from scratch..." [00:33:57] <Triskelios> lloy0076: boot with -kd? [00:34:24] <lloy0076> I could boot with -kd -- where do I put that in the boot menu for the kernel? [00:34:34] <lloy0076> <-- isn't familiar with passing Solaris kernels extra parameters [00:35:12] <Triskelios> is this sparc or x86? x86 just uses grub... [00:35:47] <stevel> richlowe: phew. no other discrepencies [00:36:13] <lloy0076> Triskelios: Sorry, I'm using an AMD XP, 32bit x86. It's using grub. [00:36:17] *** estibi has quit IRC [00:36:40] <Triskelios> lloy0076: press e or whatever to edit the parameters [00:36:43] * gisburn wonders whether the string "hacked by h0Xtor, die, opensolaris, die" in the HTML code of src.opensolaris.org is intentional... [00:37:04] <lloy0076> Triskelios: *hmm* I know how to edit the parameters - I could just add -kd "somewhere". [00:37:11] * gisburn watches stevel getting a heart attack [00:37:15] <Triskelios> at the end is fine [00:37:23] <lloy0076> Ok, I'll try that. [00:37:58] *** lloy0076 has quit IRC [00:38:04] * gisburn is unhappy because stevel isn't running screaming around after reading gisburn's joke. [00:38:05] <Error_404> what in the... Run /ws/onnv-gate/public/bin/update_dbus to update D-Bus. [00:38:05] <Error_404> bfu aborting [00:38:12] <Error_404> where in blazes is /ws ? [00:38:28] <gisburn> WorkSpace ? [00:38:40] <gisburn> WombatSplatter [00:38:47] <Error_404> there's no /onnv-gate in my workspace [00:38:51] <gisburn> WildS*x [00:38:54] <Error_404> because i'm trying to bfu [00:39:01] <alanc> it's in the eng.sun.com domain automount maps [00:39:09] *** unixconsole has joined #opensolaris [00:39:14] <alanc> ^domain^NIS domain [00:39:22] <Error_404> alanc: so.... where's /ws ? [00:39:23] *** sparc-kly has quit IRC [00:39:28] <Gman> Error_404, you could grab the latest dbus from dlc.sun.com/osol/jds/download/ [00:39:43] <alanc> on the wrong side of the firewall to you [00:39:53] <unixconsole> quick question.. when will cd playback work again in JDS? [00:40:02] <gisburn> Gman: erm, does that mean bfu now depends on JDS ? [00:40:03] <alanc> complain to dduvall about ON gatekeepers allowing internal paths in bfu [00:40:09] * gisburn kicks Gman preventively [00:40:12] <Gman> gisburn, no, sound slike a bug [00:40:15] <alanc> not bfu, ON [00:40:29] <alanc> bfu is just installing the ON bits that depend on JDS [00:40:36] <gisburn> alanc: UMPF [00:40:57] <dduvall> Well, bfu is complaining that you haven't done so. [00:41:05] <richlowe> alanc: eh? [00:41:08] <dduvall> If you follow the flag-days, you won't get those messages. [00:41:15] <richlowe> alanc: the alternative is *nobody* passing flag days easily :) [00:41:20] <alanc> richlowe: wha? [00:41:29] <richlowe> alanc: internal paths in bfu... [00:41:44] * gisburn sees dduvall [00:41:47] * gisburn attacks [00:41:47] <Error_404> I only need jds-vermillion_52x.tar , correct? [00:41:48] * dduvall hides [00:41:51] <gisburn> dduvall: Hi! [00:41:53] <Error_404> i mean *need* [00:41:58] <alanc> sorry, should have said "complain about them providing only internal paths without external equivalents" [00:42:00] <dduvall> 'Zup. [00:42:07] <Error_404> it's a server, i don't care if the windowing system is out of date [00:42:15] <Gman> Error_404, hrm, i thought there was a separate package you could download [00:42:22] <richlowe> Most of the stuff that deals in /ws can be told not to, and does not if you follow the devref. [00:42:23] <Error_404> nope [00:42:34] * Gman looks [00:42:37] <richlowe> the only thing (I can think of, currently) that makes you force it not to by hand would be ws(1) [00:43:05] <unixconsole> so why is cd audio playback broken in JDS? [00:43:06] <richlowe> stevel: nice. [00:44:14] <alanc> unixconsole: because you're not using a version with HAL support yet? [00:44:17] <richlowe> Error_404: the short answer, which I don't see anyone actually telling you, is you need to be on snv_46 or greater (or have equivalent dbus bits) to build. [00:44:19] <Gman> Error_404, i fear it's in the gnome tarball :/ [00:44:36] <Gman> lacaAFK could confirm [00:44:44] <Error_404> :-/ [00:44:47] <Error_404> okiedokie [00:45:03] <Gman> Error_404, i can create some packages quickly if you like [00:45:25] <richlowe> alanc: and I have complained about them doing that. [00:45:27] <Error_404> it's on its way over, ETA about 2 minutes [00:45:30] <richlowe> alanc: you can see how effective it was. :) [00:45:34] <Gman> Error_404, ok, coool [00:45:37] <unixconsole> alanc: what is the status of HAL support? I have sx:cr build 49 running on my sparc and x86 boxes. [00:45:39] <Error_404> thanks though [00:45:50] <Gman> unixconsole, it's integrated [00:45:57] <Gman> unixconsole, should be in b49 [00:46:01] <Gman> i think [00:46:12] <unixconsole> gman: doesn't seem to work. [00:46:13] <alanc> don't you still need vermillion-devel for the JDS side? [00:46:38] <alanc> and I thought Tamarack didn't integrate until 51... [00:50:44] *** |tsoome| has quit IRC [00:50:45] *** rachel has quit IRC [00:50:45] *** yarihm has quit IRC [00:50:45] *** jamesb has quit IRC [00:50:46] *** miffe has quit IRC [00:50:46] *** slowhog has quit IRC [00:50:46] *** sickness has quit IRC [00:50:47] *** Kernel86 has quit IRC [00:50:48] *** boyd has quit IRC [00:50:49] *** Saltsa has quit IRC [00:50:49] *** richlowe has quit IRC [00:50:50] *** DataStream has quit IRC [00:50:50] *** claudiush has quit IRC [00:50:50] *** elektronkind has quit IRC [00:50:50] *** libkeiser has quit IRC [00:50:51] *** heffnerd has quit IRC [00:50:51] *** Trident has quit IRC [00:50:51] *** Drone has quit IRC [00:50:52] *** sparc-kly_ has quit IRC [00:50:52] *** rodrickbrown has quit IRC [00:50:53] *** LeftWing has quit IRC [00:50:53] *** dvorak has quit IRC [00:50:53] *** deather_ has quit IRC [00:50:55] *** icon has quit IRC [00:50:55] *** udos has quit IRC [00:50:55] *** mazon has quit IRC [00:50:55] *** TBCOOL has quit IRC [00:50:56] *** Kernel86_ has quit IRC [00:50:56] *** dduvall has quit IRC [00:50:56] *** moazamraja has quit IRC [00:50:56] *** Odin- has quit IRC [00:50:57] *** Cyl has quit IRC [00:50:57] *** Berny has quit IRC [00:50:58] *** hali has quit IRC [00:50:58] *** doownek has quit IRC [00:50:58] *** noyb has quit IRC [00:50:58] *** mikaeld has quit IRC [00:50:58] *** Bart_M has quit IRC [00:50:58] *** paul has quit IRC [00:50:58] *** razrX has quit IRC [00:50:59] *** unixconsole has quit IRC [00:50:59] *** gisburn has quit IRC [00:50:59] *** bondolo has quit IRC [00:51:00] *** Gman has quit IRC [00:51:01] *** PosixC has quit IRC [00:51:01] *** yusufg has quit IRC [00:51:01] *** schily has quit IRC [00:51:02] *** Vanuatoo has quit IRC [00:51:02] *** dwc- has quit IRC [00:51:02] *** soultan has quit IRC [00:51:02] *** Sporq has quit IRC [00:51:03] *** r3boot has quit IRC [00:51:04] *** nightswim has quit IRC [00:51:04] *** andersmo has quit IRC [00:51:04] *** kirma has quit IRC [00:51:04] *** junks has quit IRC [00:51:04] *** |joni| has quit IRC [00:51:05] *** alobbs has quit IRC [00:51:05] *** crib has quit IRC [00:51:06] *** zarathustra has quit IRC [00:51:07] *** Auralis has quit IRC [00:51:07] *** jcea has quit IRC [00:51:07] *** lasseoe has quit IRC [00:51:08] *** the-decider has quit IRC [00:51:08] *** timeless has quit IRC [00:51:09] *** spike723_ has quit IRC [00:51:10] *** svoboda has quit IRC [00:51:10] *** Dar_HOME has quit IRC [00:51:10] *** Peanut has quit IRC [00:51:10] *** phips has quit IRC [00:51:10] *** Stric has quit IRC [00:51:10] *** sniffy has quit IRC [00:51:10] *** wilbury has quit IRC [00:51:10] *** prg3 has quit IRC [00:51:11] *** lulf has quit IRC [00:51:11] *** raph_ael has quit IRC [00:51:11] *** Risky has quit IRC [00:51:11] *** jbalint has quit IRC [00:51:12] *** clee has quit IRC [00:51:13] *** g4lt-U60 has quit IRC [00:51:13] *** flynux has quit IRC [00:51:13] *** stevel has quit IRC [00:51:14] *** pixie_ has quit IRC [00:51:15] *** mustang has quit IRC [00:51:15] *** ProfMikey has quit IRC [00:51:15] *** twincest has quit IRC [00:51:17] *** trygvis_ has quit IRC [00:51:17] *** _soylent_ has quit IRC [00:51:17] *** Xaqueth has quit IRC [00:51:17] *** AbeFroman has quit IRC [00:51:17] *** cypromis has quit IRC [00:51:18] *** steleman has quit IRC [00:51:18] *** kb7sqi has quit IRC [00:51:18] *** quasi has quit IRC [00:51:20] *** oxygene has quit IRC [00:51:20] *** myrkraverk has quit IRC [00:51:20] *** xinkeT has quit IRC [00:51:21] *** Hunger- has quit IRC [00:51:22] *** switch has quit IRC [00:51:22] <alanc> something I said? 8-P [00:51:29] <djgregor> It feels a little more cozy in here for some reason [00:52:03] *** mikaeld has joined #opensolaris [00:52:03] *** Bart_M has joined #opensolaris [00:52:24] *** unixconsole has joined #opensolaris [00:52:24] *** gisburn has joined #opensolaris [00:52:24] *** bondolo has joined #opensolaris [00:52:24] *** paul has joined #opensolaris [00:52:24] *** Gman has joined #opensolaris [00:52:24] *** PosixC has joined #opensolaris [00:52:24] *** yusufg has joined #opensolaris [00:52:24] *** schily has joined #opensolaris [00:52:24] *** razrX has joined #opensolaris [00:52:24] *** Vanuatoo has joined #opensolaris [00:52:24] *** dwc- has joined #opensolaris [00:52:24] *** soultan has joined #opensolaris [00:52:24] *** kirma has joined #opensolaris [00:52:25] *** nightswim has joined #opensolaris [00:52:25] *** r3boot has joined #opensolaris [00:52:25] *** junks has joined #opensolaris [00:52:25] *** andersmo has joined #opensolaris [00:52:25] *** |joni| has joined #opensolaris [00:52:25] *** Sporq has joined #opensolaris [00:52:35] <gisburn> alien attack! [00:52:42] *** alobbs has joined #opensolaris [00:52:42] *** crib has joined #opensolaris [00:52:42] *** zarathustra has joined #opensolaris [00:52:42] *** Auralis has joined #opensolaris [00:52:42] *** jcea has joined #opensolaris [00:52:42] *** lasseoe has joined #opensolaris [00:52:42] *** prg3 has joined #opensolaris [00:52:43] *** clee has joined #opensolaris [00:52:43] *** sniffy has joined #opensolaris [00:52:43] *** Stric has joined #opensolaris [00:52:43] *** jbalint has joined #opensolaris [00:52:43] *** phips has joined #opensolaris [00:52:43] *** Peanut has joined #opensolaris [00:52:43] *** lulf has joined #opensolaris [00:52:43] *** g4lt-U60 has joined #opensolaris [00:52:43] *** Dar_HOME has joined #opensolaris [00:52:43] *** svoboda has joined #opensolaris [00:52:43] *** spike723_ has joined #opensolaris [00:52:43] *** wilbury has joined #opensolaris [00:52:43] *** Risky has joined #opensolaris [00:52:43] *** timeless has joined #opensolaris [00:52:43] *** raph_ael has joined #opensolaris [00:52:43] *** the-decider has joined #opensolaris [00:52:46] *** sparc-kly_ has joined #opensolaris [00:52:46] *** rodrickbrown has joined #opensolaris [00:52:46] *** LeftWing has joined #opensolaris [00:52:46] *** dvorak has joined #opensolaris [00:52:46] *** deather_ has joined #opensolaris [00:52:46] *** icon has joined #opensolaris [00:52:46] *** udos has joined #opensolaris [00:52:47] *** mazon has joined #opensolaris [00:52:47] *** TBCOOL has joined #opensolaris [00:52:47] *** Kernel86_ has joined #opensolaris [00:52:47] *** dduvall has joined #opensolaris [00:52:47] *** moazamraja has joined #opensolaris [00:52:47] *** Odin- has joined #opensolaris [00:52:48] *** Cyl has joined #opensolaris [00:52:48] *** hali has joined #opensolaris [00:52:48] *** Berny has joined #opensolaris [00:52:48] *** doownek has joined #opensolaris [00:52:48] *** noyb has joined #opensolaris [00:52:49] <unixconsole> nope.. it's probably not part of b49. I know I'm behind a little;( [00:53:07] *** |tsoome| has joined #opensolaris [00:53:07] *** rachel has joined #opensolaris [00:53:07] *** yarihm has joined #opensolaris [00:53:07] *** jamesb has joined #opensolaris [00:53:07] *** miffe has joined #opensolaris [00:53:07] *** slowhog has joined #opensolaris [00:53:08] *** sickness has joined #opensolaris [00:53:08] *** Kernel86 has joined #opensolaris [00:53:08] *** boyd has joined #opensolaris [00:53:08] *** Saltsa has joined #opensolaris [00:53:08] *** richlowe has joined #opensolaris [00:53:08] *** DataStream has joined #opensolaris [00:53:08] *** claudiush has joined #opensolaris [00:53:08] *** elektronkind has joined #opensolaris [00:53:08] *** libkeiser has joined #opensolaris [00:53:08] *** heffnerd has joined #opensolaris [00:53:08] *** Trident has joined #opensolaris [00:53:08] *** Drone has joined #opensolaris [00:53:13] *** stevel has joined #opensolaris [00:53:13] *** pixie_ has joined #opensolaris [00:53:14] *** mustang has joined #opensolaris [00:53:14] *** ProfMikey has joined #opensolaris [00:53:14] *** twincest has joined #opensolaris [00:53:15] *** flynux has joined #opensolaris [00:53:15] *** cypromis has joined #opensolaris [00:53:15] *** steleman has joined #opensolaris [00:53:15] *** kb7sqi has joined #opensolaris [00:53:15] *** xinkeT has joined #opensolaris [00:53:15] *** myrkraverk has joined #opensolaris [00:53:15] *** trygvis_ has joined #opensolaris [00:53:15] *** _soylent_ has joined #opensolaris [00:53:15] *** quasi has joined #opensolaris [00:53:15] *** Xaqueth has joined #opensolaris [00:53:15] *** Hunger- has joined #opensolaris [00:53:15] *** switch has joined #opensolaris [00:53:15] *** AbeFroman has joined #opensolaris [00:53:15] *** oxygene has joined #opensolaris [00:53:15] *** irc.freenode.net sets mode: +o stevel [00:53:25] *** richlowe has quit IRC [00:53:28] *** richloo has joined #opensolaris [00:53:44] *** richloo is now known as richlowe [00:55:03] *** rafaeldt has joined #opensolaris [00:56:04] <alanc> talk about a conversation killer [00:56:37] *** rafaeldt has left #opensolaris [00:56:48] <g4lt-U60> yeah, netsplits do that [00:57:03] <richlowe> stevel: better ways to deal with local workspaces, or opensolaris.org workspaces? [00:57:51] <stevel> richlowe: either. both. [00:58:13] *** nexrafa has joined #opensolaris [00:59:35] *** jwt has joined #opensolaris [00:59:54] <richlowe> stevel: local with Mq == qpop, strip, pull, qpush, (insert Mqmerge gunk if needed). local without Mq/opensolaris.org == if you don't care about history diff against current onnv-gate, get correct one, apply patch. If you do care about history, outgoing -p or export, save a copy of the log from your base, walk the new gate forward applying your diffs and merging as they were before. [01:00:25] <richlowe> I can't think of anything better or less manual. [01:00:32] <jwt> opensolb50 build on amd64 failure ... any quick suggestions? ... [01:00:43] <jwt> pwd .../usr/src/uts; dmake all [01:01:22] <jwt> Current working directory /export/testws/usr/src/uts/i86pc/genassym [01:01:22] <jwt> -o obj64/genassym ../../i86pc/ml/genassym.c [01:01:22] <jwt> + /opt/SUNWspro/bin/cc -D_KERNEL -D_SYSCALL32 -D_SYSCALL32_IMPL -D_ELF64 -D_DDI_STRICT -Dsun -D__sun -D__SVR4 -Di86pc -D_MACHDEP -DC2_AUDIT -I../../i86pc -I../../intel -Y I,../../common -o obj64/genassym ../../i86pc/ml/genassym.c [01:01:31] <stevel> crap, gotta run and give a preso [01:01:35] <jwt> ld: fatal: file genassym.o: unknown file type [01:01:36] <Gman> good luck :) [01:01:37] <stevel> richlowe: can you send that to the list? (if you don't mind)? [01:01:37] <jwt> ld: fatal: File processing errors. No output written to obj64/genassym [01:01:39] <jwt> *** Error code 1 [01:01:42] <jwt> dmake: Fatal error: Command failed for target `obj64/genassym' [01:01:43] <jwt> Current working directory /export/testws/usr/src/uts/i86pc/genassym [01:01:44] <jwt> -o obj64/genassym ../../i86pc/ml/genassym.c [01:01:46] <jwt> + /opt/SUNWspro/bin/cc -D_KERNEL -D_SYSCALL32 -D_SYSCALL32_IMPL -D_ELF64 -D_DDI_STRICT -Dsun -D__sun -D__SVR4 -Di86pc -D_MACHDEP -DC2_AUDIT -I../../i86pc -I../../intel -Y I,../../common -o obj64/genassym ../../i86pc/ml/genassym.c [01:01:50] <jwt> ld: fatal: file genassym.o: unknown file type [01:01:53] <jwt> ld: fatal: File processing errors. No output written to obj64/genassym [01:01:56] <jwt> *** Error code 1 [01:02:05] <jwt> dmake: Fatal error: Command failed for target `obj64/genassym' [01:02:06] <jwt> dmake: Warning: Initial write of statefile failed. Trying again on /tmp/.make.state.10519.nea4Iu [01:02:06] <jwt> Current working directory /export/testws/usr/src/uts/i86pc/genassym [01:02:06] <jwt> -o obj64/genassym ../../i86pc/ml/genassym.c [01:02:07] <jwt> + /opt/SUNWspro/bin/cc -D_KERNEL -D_SYSCALL32 -D_SYSCALL32_IMPL -D_ELF64 -D_DDI_STRICT -Dsun -D__sun -D__SVR4 -Di86pc -D_MACHDEP -DC2_AUDIT -I../../i86pc -I../../intel -Y I,../../common -o obj64/genassym ../../i86pc/ml/genassym.c [01:02:10] <jwt> ld: fatal: file genassym.o: unknown file type [01:02:13] <jwt> ld: fatal: File processing errors. No output written to obj64/genassym [01:02:14] <jwt> *** Error code 1 [01:02:17] <jwt> dmake: Fatal error: Command failed for target `obj64/genassym' [01:02:18] <jwt> thx [01:02:31] <richlowe> stevel: it was here in the hopes of someone having better ideas. [01:02:36] <richlowe> but sure. [01:08:32] *** andres__ has joined #opensolaris [01:09:49] <richlowe> jwt: the workspace you're attempting to build appears to be at least partially read-only. [01:10:31] *** uebayasi has joined #opensolaris [01:16:28] <jwt> ahhhh [01:16:41] <jwt> interesting. running as root though in /export/testws per the readme. [01:16:46] <jwt> *ponder* [01:16:48] *** sahafeez has joined #opensolaris [01:20:54] *** Tpenta has joined #opensolaris [01:21:10] *** ChanServ sets mode: +o Tpenta [01:22:39] <alanc> amusing observation earlier today: CDE has more Java apps than JDS does now [01:22:48] <jwt> hey tpenta. jwt==james.todd!=anymore blogs.sun.com/gonzo [01:23:07] *** lloy0076 has joined #opensolaris [01:23:31] <lloy0076> Giving my kernel -kd puts me into the kernel debugger...which I don't know how to use :( [01:23:40] <Tpenta> hi james [01:23:47] <Tpenta> hows tricks gonzo? [01:23:49] <jwt> how's the brewing down under? [01:24:00] <lloy0076> aha [01:24:06] <Tpenta> i need to put another one on [01:24:07] <lloy0076> My Ubuntu recognises my solaris partitions. [01:24:16] <jwt> pretty good. trying to get my first amd64 opensol build badge (of honor) :) [01:24:43] <Gman> alanc, hopefully there'll be more java-gnome in the future [01:24:44] <jwt> running b50 x86 32bit on a amd64 box. [01:25:01] * Triskelios wonders if he should wait 3 hrs for b51 or just install b48... [01:25:25] <alanc> hopefully the java-gnome apps will be less maligned than the gloriousness that is sdtaudiocontrol [01:25:34] <Gman> :) [01:25:37] *** sommerfeld has quit IRC [01:26:16] <LeftWing> sdtaudiocontrol is fantastic. [01:26:25] * Triskelios is also wondering why the normal 2mb/s down from sdlc has been reduced to a 100kb/s trickle [01:26:28] <alanc> just as long as you don't make bug buddy a java app - "Hey, your program crashed, please wait 30 seconds for us to start a JVM to tell you it died!" [01:26:29] <LeftWing> Pinnacle of achievement in comptuer science. [01:27:41] *** lloy0076 has left #opensolaris [01:30:57] <Error_404> oh, good... linux just spewed swap data all over my solaris partition [01:31:37] <LeftWing> Ouch. :| [01:31:56] <Error_404> meh, it was a fresh install, it's only mildly annoying [01:33:04] <Error_404> that'll teach me i guess [01:33:23] *** rodrickbrown has quit IRC [01:33:44] <Error_404> oh, it didn't write swap all over the partition... it was just the wrong fdisk ID [01:34:17] <richlowe> jwt: you don't need to build as root, either. [01:34:30] <richlowe> but dmake being unable to create it's state files in the normal places tends to mean you can't write to the workspace. [01:36:23] *** coolvibe_ has joined #opensolaris [01:36:24] <richlowe> Gman: just ARC mono, and be done with it :) [01:36:32] <Gman> no. [01:36:35] *** coolvibe has quit IRC [01:37:12] *** karrotx has joined #opensolaris [01:37:47] <Stric> delewis: you found any solution to that SUNW_1.22.2 not found thing? [01:40:10] *** yarihm has quit IRC [01:41:41] <jwt> ahhhh [01:41:42] <jwt> really [01:41:44] <jwt> hmmmm [01:41:49] <jwt> i'll retry then [01:42:21] *** rydis has joined #opensolaris [01:42:26] <jwt> hard to believe root can't write somewhere though [01:44:03] *** estibi has joined #opensolaris [01:46:36] <Stric> Error_404: linux doesn't use something as swap unless you've run mkswap on it first to poke a "swap signature" there [02:01:00] <Error_404> what, uhh, is the solution if ./install-jds fails? [02:01:19] <Error_404> fex, if it catches sigHUP accidently [02:01:25] <gisburn> Error_404: commit suicide. [02:01:26] <Error_404> by that point, the script is useless [02:02:04] *** LeftWing has quit IRC [02:03:53] <Error_404> mind you, come to think of it it's a server, i don't really much care [02:04:05] <Error_404> so long as dbus gets in properly [02:06:16] *** yongsun_ has joined #opensolaris [02:06:34] *** LeftWing has joined #OpenSolaris [02:07:09] *** yongsun_ has quit IRC [02:12:19] *** maitrella has joined #opensolaris [02:16:50] *** tamlin has joined #opensolaris [02:17:17] *** alanc is now known as alanc-away [02:18:43] <tamlin> While I haven't used it myself yet, anyone got expeerience with zfs and large files (100GB in this case)? [02:19:24] <gisburn> which character can be used to add comments to a patch *.diff file ? [02:19:42] *** jwt has quit IRC [02:20:14] <Triskelios> gisburn: why not just add the comments before or after the diff output? [02:23:13] <gisburn> Triskelios: I'd like to do that, howver there is a special character which introduces comment lines. [02:24:04] *** alanc-away is now known as alanc [02:24:28] * gisburn looks at alanc for help [02:24:36] <alanc> not if you put it before the diffs - everything before the first thing that looks like a diff is a comment [02:24:56] *** nwf has joined #opensolaris [02:25:05] * alanc is leaving for real this time [02:25:08] *** alanc is now known as alanc-away [02:27:01] <gisburn> no. [02:27:19] <gisburn> a simple +, - or . causes an error for gpatch [02:27:22] <LeftWing> No? [02:27:37] <LeftWing> heh [02:27:38] <richlowe> LeftWing: I was wondering the same thing. [02:28:09] <LeftWing> You could try throwing hashes on the beginning of each line. I believe that's what Mercurial does in its bundles (which are essentially diffs) [02:28:10] <tamlin> What if you put a leading hash, or semicolon, on those lines? [02:28:29] *** slowhog has quit IRC [02:29:45] <tamlin> If in dubt though, I'd suggest looking at the source code for gpatch, to really verify that 1) it's got such a character you are looking for, and 2) to see what it is. [02:30:27] <richlowe> bundles aren't diffs, exported patches are. :) [02:30:34] <richlowe> But yeah, that's what hg does, it's what BitKeeper did too. [02:30:48] <richlowe> Mq just does exactly what alanc said, and stores metadata bare above the first /^diff/ line. [02:31:37] <gisburn> so # at the beginning of each single comment line, right ? [02:31:58] *** LeftWing has quit IRC [02:32:07] *** LeftWing has joined #OpenSolaris [02:33:14] <richlowe> it depends on context, but largely, Yes. [02:33:24] <richlowe> why it's me reading the gpatch source rather than you is beyond me, however. [02:33:24] *** lacaAFK is now known as laca [02:34:00] <tamlin> richlowe: I suspect you're just too nice. :-) [02:36:46] *** Error_404 has quit IRC [02:37:29] <sahafeez> anyone want to make an offer on an ultra 80 (4x450) 4gb ram, 2 elite3d-m6&m3 before it goes on ebay? [02:37:48] <Stric> shipping costs? :) [02:38:00] <sahafeez> hahah [02:38:03] <sahafeez> no [02:38:10] <Stric> :( [02:38:37] <sahafeez> shipping + $300 [02:38:44] <tamlin> sahafeez: On what continent is it? [02:38:51] <sahafeez> usa [02:38:56] <richlowe> heh, shipping is probably not going to be pleasant, either. [02:39:01] <tamlin> Count me out! :-) [02:39:12] <sahafeez> it is not bad in the us. ups ground is like $50 [02:39:25] * Stric is in northern .se where polar bears roam the streets (err. maybe not.. but almost) [02:39:31] <tamlin> I'm in europe. [02:39:50] <richlowe> Hm, though they're apparently only 65lb [02:39:55] * richlowe was expecting them to be heavier. [02:39:58] <tamlin> Stric: I'm in middle/southern .se. [02:41:10] <tamlin> sahafeez: Just out of curiosity, how much power doe that thing require/use? [02:41:23] <sahafeez> normal plug [02:41:28] <sahafeez> shipping to the uk is $124 [02:42:24] <richlowe> > Input Power: 670W [02:42:27] <richlowe> says the syshbk [02:42:32] <sahafeez> tamlin, the powersupply is labeled 870w [02:42:39] <sahafeez> 875w sorry [02:42:52] <sahafeez> richlowe, the books wrong. [02:42:52] *** LeftWing has quit IRC [02:42:53] <tamlin> OK, that's reasonable. [02:43:13] <tamlin> (for such an old machine, I mean) [02:43:18] <sahafeez> i love the box it is great and fast. [02:43:18] *** LeftWing has joined #OpenSolaris [02:43:35] <sahafeez> but i am trying to have less stuff so i am going to just dual boot an x86 box [02:44:02] <tamlin> sahafeez: Come clean - you want to shave the power bill. ;-> [02:44:37] <sahafeez> naw. power in CA is $$$ however land is more so..no room in the house [02:45:39] <sahafeez> i would rather have the u80 becasue of the sun keyboard [02:46:36] <richlowe> sahafeez: oddly, I pasted the wrong bits. [02:46:57] <richlowe> sahafeez: it claims 380W max, on a 670W psu [02:47:07] <tamlin> 4x? [02:47:18] <sahafeez> the label engraved into the metal of the powersupply say 875w [02:47:53] <Stric> same on blade1000.. but it's nowhere that hot [02:48:27] <delewis> heh, Blade 1000s run fairly cool. [02:48:30] <sahafeez> it pushes some hot air but if you have good air flow in the room it is fine [02:48:31] <delewis> mine runs at about 26C [02:48:41] <delewis> it is pushing a lot of heat out, though [02:48:46] <delewis> decent ventilation and cooling, obviously. [02:48:53] <richlowe> delewis: which sensor are you looking at? [02:49:00] <delewis> richlowe: CPU [02:49:07] <tamlin> 26c with ~20c inlet? [02:49:17] <delewis> DIE temp, anyway. [02:49:19] *** LivingCorpse has joined #opensolaris [02:49:41] <delewis> tamlin: roughly [02:50:07] <delewis> air should be 20C, but it's probably closer to 22 or 23C with all the hardware I have in here. [02:50:54] <delewis> $ /usr/sbin/prtpicl -v | grep :Temperature [02:50:55] <delewis> :Temperature 78 [02:50:55] <delewis> :Temperature 71 [02:51:00] <delewis> I assume those are in fahrenheit. [02:51:03] <richlowe> what does prtpicl -vc temperature-sensor say about it? [02:51:05] <delewis> :-) [02:51:08] <richlowe> hah, you beat me to it. [02:51:21] <richlowe> delewis: see how cpu-ambient matches up more with reality though? [02:51:24] *** nate_a has joined #opensolaris [02:51:27] <richlowe> (or hoped reality) [02:51:58] *** LeftWing has quit IRC [02:52:01] <Stric> ambient 27, 25.. die 70, 79 [02:52:24] *** LeftWing has joined #OpenSolaris [02:52:25] <delewis> wait [02:52:35] <delewis> are these temperatures in Fahrenheit or Celsius? [02:52:44] <richlowe> I'd assume one of each. [02:52:47] <tamlin> 78 I believe is F. [02:52:49] <richlowe> 27F seems pretty unlikely... [02:52:52] <delewis> I would certainly hope so [02:52:59] <delewis> I was just making sure. :-) [02:53:00] <Stric> I doubt it's below freezing in here [02:53:01] <richlowe> otherwise the cooling is *way* better than I thought :) [02:53:13] <delewis> richlowe: still, I'd say the cooling in the SB1000 is fairly decent. [02:53:20] <delewis> my peecee runs *far*, *far* hotter. [02:53:29] <delewis> closer to 50C with a single 1.4GHz k7 [02:53:45] <tamlin> Whooo! RIMM? ;-> [02:53:56] <delewis> tamlin: no, SDRAM which is even more sad. [02:53:59] <Stric> delewis: btw, did you find any solution to that SUNW 1.22.2 thingie with useradd? [02:54:00] <delewis> pc133 at that. [02:54:09] <delewis> Stric: yes, remove the -24 kernel patch [02:54:17] <delewis> (and serveral other patches) [02:54:20] <delewis> there's like three total [02:54:24] <Stric> meh [02:54:29] <delewis> if you have the latest 10_Recommended, it should revert to -17 [02:54:41] <tamlin> delewis: PC133 on a 1.4GHz CPU?! [02:54:52] <delewis> tamlin: sure [02:55:00] <delewis> but that system is actually in the closet now :-) [02:55:03] <tamlin> sad... [02:55:07] <delewis> I'm converting most of my network over to SPARC :-) [02:55:24] <Stric> delewis: oh well.. I don't need useradd that much.. thanks though.. [02:55:47] <delewis> Stric: yes, I definitely would not go installing the latest patches arbitrarily. [02:55:58] <delewis> I think they're getting ready to roll out S10u3 and things are a bit, well.. buggy. [02:56:07] <delewis> I also had a problem with zones after I installed all of the latest patches [02:56:09] <tamlin> It bugs me that the shipping is so bloody much - else I could have seriously considered the box. [02:56:17] <delewis> apparently various important files weren't be copied to the zone on install [02:56:18] <richlowe> Stric: if you search b.o.o for that version string you'll see two CR's, one fixed, one won't fix. Relating to TSol stuff and that version. [02:56:21] <delewis> (or symlink'd) [02:56:47] <delewis> richlowe: yes, but what's sad is that these patches (that cause this problem) are being released to Solaris 10 customers. [02:57:08] <tamlin> Sun knowing about the issues too? [02:57:23] <delewis> tamlin: apparently if these issues originally appeared in the S10u3 express program. [02:57:28] <delewis> it sort of pisses me off. [02:57:56] <delewis> this has taught me to stay with the patch clusters [02:58:10] <Stric> I installed the latest solaris 10 patches and kinda hoped they would fit.. [02:59:13] <tamlin> Have you tried to contact Sun in any way, or some oth the OpenSolaris forums, to explain the error (and severity of it)? I can't for my life believe they knowingly release an update that breaks stuff. [02:59:34] <delewis> tamlin: I don't have a contract and things with 10 only tend to get escalated if you're an actual customer. [02:59:49] <delewis> this might be an exception, though, as it does effect *everyone* that uses Solaris 10 [02:59:59] <richlowe> may as well try. [03:00:00] <delewis> and not just fringe users that are doing something strange. [03:00:14] <tamlin> Screw "contract". This is something they should listen to from anyone now! [03:00:21] <Stric> unfortunately for me, this is a non-profit student organization.. at work we're still at sol9.. [03:00:38] <Stric> .. but fortunately for me, if stuff breaks it doesn't matter too much :) [03:01:01] <delewis> well, fortunately, I'm in the same position [03:01:21] * richlowe suspects "only gets escalated if you're a customer" maybe overly pessimistic [03:01:26] <delewis> just a hobbyist that unknowingly installed all of the latest patches on a fresh Solaris 10 install on my Ultra 2. [03:01:38] * tamlin knows it's overly pessimistic [03:01:46] <delewis> though, if I were running a system in production, I'd be screaming for bloody murder. :-) [03:01:59] <delewis> and even then, I'd probably be using the patch clusters anyway. [03:02:15] <tamlin> If you ran an Ultra 2 in production, _I'd_ be screaming. ;-) [03:02:33] <delewis> hey, you'd be surprised how a decent Ultra 2 configuration will perform nowadays. [03:02:49] <delewis> my Ultra 2 with 2x300MHz procs and 1.5GB of memory flies. [03:03:00] <rydis> They make decent file servers. Just add an FC card or two. ;) [03:03:07] <delewis> rydis: that's the plan :-) [03:03:14] <delewis> (whenever the A5200 and Silkworm arrive) [03:03:32] <tamlin> delewis: I stand down. That indeed seems it could be more than useful. [03:03:36] <jamesd_> tamlin, try a 2x 300mhz u2 with 2 GB of ram and dual 36GB drives.. perferably with fibre or 711 disk box. [03:03:39] <Stric> we use the server equiv, 10x400MHz in a e4500.. 7G ram.. got it for free of course :) [03:03:53] <delewis> well, actually, it'll replace my aging peecee (that has a Cyrix PII-clone with 128MB of memory) that servers as my router, smtpd, pop3d, bind, httpd, etc. [03:04:11] <delewis> Stric: hey, I'm getting one of those, too, any day now. [03:04:18] <delewis> though, mine is *only* an 8-way :-) [03:04:28] <tamlin> Stric: "Got it for free"? Yeah, I see how that could be a good deal. :-) [03:04:38] <delewis> tamlin: E4500s are fairly affordable, nowadays [03:04:47] <delewis> I paid $550 for mine (with an additional $150 freight fee) [03:04:54] <delewis> 8-way with 8GB of memory [03:05:00] <rydis> Our 4500 seems to have a busted clock card, unfortunately, so we can't actually use it. :-/ [03:05:05] <delewis> rydis: ouch [03:05:08] <Stric> 80MB cpu cache :P [03:05:10] <delewis> that's a pretty cheap FRU, though. [03:05:27] <delewis> clock boards typically go for $50 or less on Ebay [03:05:29] <tamlin> Stric: 80MB?! [03:05:40] <delewis> Stric: 4MB of cache per processor [03:05:41] <delewis> not 8MB :-) [03:05:42] <Stric> 10x8.. [03:05:45] <Stric> delewis: 8. [03:05:54] <delewis> oh, I must be thinking of the workstation-class UltraSPARC-II variants. [03:05:58] <Stric> 0 0 0 400 8.0 US-II 10.0 [03:05:58] <delewis> silly me. [03:06:03] <rydis> delewis: Really? Interesting. We just sold our old raised floor, so that'd pay for that nicely. :) [03:06:04] <delewis> ah, you've got the 8MB variant [03:06:11] <delewis> not all UltraSPARC-IIs were 8MB [03:06:18] <delewis> those are the saphire black ones if I'm not mistaken [03:06:30] * LeftWing has 8MB cache in his US-II 480Mhz chips. [03:06:44] <delewis> I only have 2MB in my Ultra 2 [03:06:55] <delewis> 8MB in my SB1000, though :-) [03:07:12] <LeftWing> The IIi's that I have are all 2MB now. The difference between 256/512k and 2MB is amazing. [03:07:23] <delewis> LeftWing: that much of a difference, eh? [03:07:26] <jamesd_> i need more ram in my blade 1500, zfs just doesn't like 512MB [03:07:36] <tamlin> "only" 8MB... I remeber when 32 *bytes* cache was state-of-the-art. :-) [03:07:41] <LeftWing> delewis: Yeah. cp over NFS no longer kills the machine. [03:07:46] <delewis> I don't think the larger cache size in the UltraSPARC-III really helps a lot [03:07:55] <delewis> that is, the UltraSPARC-III systems that have CU extensions disabled [03:08:12] <delewis> which means the various prefetch instructions are disabled, though, that might not have any effect on cache effiency. [03:08:58] <delewis> tamlin: 256KB/512KB is still the standard in the peecee world :-) [03:09:13] <tamlin> jamesd_: I've been toying with an idea for a plain filesrver - store cache compressed (using one of the few *really* fast compression algo's). Would take some testing though, to verify it's worth the CPU cost.. [03:09:15] <LeftWing> Heh. That's not really true. [03:09:22] <delewis> I even have 8MB of cache in my pSeries [03:09:23] <LeftWing> My Pentium M (from a year or two ago) has 2MB. [03:09:28] <delewis> which has 2xPOWER3s [03:09:53] <delewis> LeftWing: true, but I've rarely seen peecees that are equipped with those processors. [03:10:04] <LeftWing> What do Core CPUs get? [03:10:09] <LeftWing> Surely at least 1MB. [03:10:18] <delewis> my university re-furnished all of the labs with some P4 variant [03:10:19] <delewis> 512KB of cache [03:10:31] <delewis> those aren't dual core or anything, though [03:10:34] <delewis> standard P4 [03:10:34] <LeftWing> P4 is a waste of time. [03:10:42] <delewis> it's a giant DSP, and a terrible one at that ;-) [03:10:51] <tamlin> I suspect the Intel Core's get 512K L2, and their old 16/16 L1. [03:11:07] <tamlin> (I'd love to be proven wrong, though) [03:11:22] <LeftWing> On an unrelated note, buying from Apple has been a positive experience in the last two days. [03:11:30] <delewis> L1 caches on SPARC tend to be 16K, right? [03:11:33] <LeftWing> Ordered a Mac Mini on Sunday, and it showed up this morning. [03:11:42] <delewis> D$, etc. [03:11:50] <tamlin> LeftWing: P4 isn't a waste of time only, it's a waste of space and power too. :-) [03:12:14] *** LeftWing has quit IRC [03:12:47] *** LeftWing has joined #OpenSolaris [03:12:56] <tamlin> Funny that it took years for Intel to see that P3 was a better architecture, and reverting to it with P4m. [03:13:13] <delewis> Intel tends to make mistakes like that. [03:13:19] <delewis> the Itanium and Itanium2 ring a bell. [03:13:30] <tamlin> Thay have indeed been know to make such mistakes. :-) [03:13:32] *** LeftWing has quit IRC [03:14:01] <rydis> The "better" Core 2 Duos seems to have 4 megs of L2 cache. [03:14:06] <gisburn> delewis: I disagree. The original Itanium arch was quite nice... but the x86 emulation and other junk watered the design. [03:14:09] <tamlin> Man, I can't even count how many CPU sockets they have plowed through in just 5 years... [03:14:18] <gisburn> delewis: but now they are getting rid of this junk again. [03:14:20] <delewis> gisburn: of course [03:14:22] *** fik has joined #opensolaris [03:14:45] <delewis> gisburn: a lot of it is still there in AMD64 [03:14:55] <delewis> they've at least gotten rid of the hardware segmentation management cruft. [03:15:00] <gisburn> delewis: which is bad for SPARC since they're getting smaller, faster and far less power-hungry IA64 chips in the near future. [03:15:03] <delewis> but all of the annoying compatiblity memory modes are still there. [03:15:26] <gisburn> delewis: not really... much of the complex stuff is done in the firmware now. [03:15:35] *** stevel has quit IRC [03:15:38] *** LeftWing has joined #OpenSolaris [03:15:42] <tamlin> (or microcode) [03:15:47] <gisburn> no [03:15:53] <gisburn> firmware for sure [03:16:08] <gisburn> suse staf is cursing that since their old kernels are running... SLOWLY. [03:16:09] <delewis> what does firmware have to do with segment management? [03:16:10] * LeftWing screams. [03:16:34] <tamlin> All is emulated in f/w? I didn't know that, but I like it. [03:16:37] <gisburn> delewis: some of the memory access instructions are now trapped and emulated. [03:16:55] <gisburn> tamlin: no, just the obsolete junk [03:16:55] <delewis> gisburn: in 64-bit all of that should be gone [03:17:04] <delewis> 32-bit compat. mode yes [03:17:12] *** stevel has joined #opensolaris [03:17:12] *** ChanServ sets mode: +o stevel [03:17:13] *** fik has quit IRC [03:17:20] <tamlin> gisburn: Yes, obviously that's what I referred to, too. [03:17:55] *** jlc has joined #opensolaris [03:17:57] *** schily_ has joined #opensolaris [03:18:29] <gisburn> schily_: congratulations. [03:18:41] <gisburn> schily_: you got your "sync" builtin. [03:18:56] <tamlin> Anyway, comparing a peecee to a properly designed system when it comes to *real* loads, isn't that like comparing a stable of formula one cars with a shipping company? [03:19:27] <delewis> if that. [03:19:35] <delewis> more like a bunch of 'riced-up' imports. [03:20:00] <delewis> at least the formula-one race cars were built for a purpose that they serve properly. [03:20:05] <tamlin> Still, the analogy is pointing in the right direction. [03:20:08] *** stevel has quit IRC [03:20:15] *** yongsun_ has joined #opensolaris [03:21:21] *** stevel has joined #opensolaris [03:21:21] *** ChanServ sets mode: +o stevel [03:21:24] <delewis> in any case, my Ultra 2 handles a load far better than my 2xPIII Xeon 866MHz /w 1GB of memory. [03:21:42] <delewis> I'll be curious to see what an E4500 can do. [03:22:01] <richlowe> delewis: double your power bill :) [03:22:02] <delewis> it's not really a fair comparison to bring my SB1000 to it, as I use that interactively. [03:22:08] <tamlin> Really? "A" load? :-) [03:22:09] <delewis> richlowe: hope not. ;-) [03:22:16] <delewis> tamlin: :-P [03:22:49] <delewis> that and the A5200 may be my un-doing. [03:22:58] <delewis> 22 10,000 RPM disks must suck up some serious power. [03:23:16] <delewis> I'll know Wednesday, though :-) [03:23:19] <tamlin> Like two vacuum cleaners? [03:24:07] <tamlin> (now, sorry, much be way smaller power consumption) [03:24:36] *** johnlev has quit IRC [03:24:50] <tamlin> Plain 7200 uses... 20W? 10k = 30W? 22* 30W = *half* a vacuum cleaner. [03:25:03] <tamlin> Still, running it around the clock... [03:25:50] <tamlin> Errors in my estimates? [03:26:09] <delewis> let me see if Seagate mentions it in the specs [03:26:41] <tamlin> (I just shot from the hip, so I may be wrong) [03:26:53] <delewis> ugh. can't remember the Seagate models Sun uses in the A5200. [03:26:56] <delewis> sunsolve here I come. [03:27:48] <tamlin> I must say, A5200 looks like an Amiga model number for some peripheral thingie. :-) [03:28:23] <tamlin> (I suspect I've been involved in too much, too long) [03:28:57] <delewis> hmm [03:29:01] <delewis> I don't see a power rating in the specs [03:29:49] <tamlin> Not even at Seagates site? [03:30:33] *** halton has joined #opensolaris [03:31:45] <delewis> it looks like ~17W to me [03:31:50] <delewis> http://www.seagate.com/support/disc/specs/fc/st318203fc.html [03:32:07] <delewis> and that's startup power reqs. [03:32:16] <tamlin> So I was actually overestimating it? [03:32:32] <LivingCorpse> delewis is the smartest person here [03:32:46] <delewis> LivingCorpse: yeah, right. :-) [03:33:37] * gisburn kills LivingCorpse [03:33:50] * delewis wonders when #opensolaris became so brutal [03:33:55] * gisburn prepares the dead remains of LivingCorpse for dinner [03:34:02] <delewis> probably when the ksh93 PSARC case began :-) [03:34:05] <LivingCorpse> i thought server people were all nice [03:34:10] <tamlin> Still, I think the "normal" for 7k2 drives is ~20W, so I'd rather count high for 10k. [03:34:12] <gisburn> you wish [03:34:25] *** schily has quit IRC [03:34:43] <delewis> tamlin: according to the specs, with +12V and 1.4 amps that would be ~17W [03:34:51] <delewis> and again that's spin-up time [03:34:58] *** kleppari has quit IRC [03:35:04] <tamlin> Server people... "nice"? /me thinks someone is unaware of BOFH. ;-> [03:35:21] <delewis> server people tend to get angry when hardware dies [03:35:26] <delewis> or when users do stupid things [03:35:53] <delewis> like forkbomb the system [03:35:59] <tamlin> delewis: I, putting on my "user" hat, tends to get pissed off when h/w dies. :-) [03:36:28] <tamlin> (PC's, the curse of the century) [03:36:41] <delewis> curse of the millenium is more like it. [03:36:54] <LivingCorpse> windows ME is my favorite OS [03:36:56] *** kleppari has joined #opensolaris [03:36:59] <tamlin> LOL! [03:37:15] <delewis> LivingCorpse: I'm afraid to know what you hate most then. [03:37:16] <Triskelios> what's your favourite drug? [03:37:43] <tamlin> I love it too, as it suggest Microsoft can "Go Windows Yourself"! :-) [03:38:36] <tamlin> (I don't know what the verb Windows means, but I must assume it's malicious in this context) [03:40:55] <gisburn> dduvall: ping! [03:42:57] <gisburn> when I checked-out the sources via SVN and unpacked the closed-bins stuff... how do I run nightly (e.g. options and NIGHTLY_OPTIONS) to match the gatekeepers checks ? [03:43:56] <richlowe> You can't, precisely. [03:44:01] <gisburn> umpf [03:44:04] <richlowe> what you need to do is look at $SRC/tools/env/gatekeeper.sh [03:44:04] <tamlin> Speking of checking out; how large is a full checkout now (# of entries and total file size)? [03:44:12] <richlowe> but add the stuff that prevents bringovers and package builds. [03:44:20] <gisburn> tamlin: ask me in 10mins [03:44:24] <richlowe> gisburn: really, adding the ABI checks and findunref to your current env file maybe enough. [03:44:36] <richlowe> let me double check though. [03:45:10] <richlowe> gisburn: take gatekeeper.sh, drop -p -u and -U [03:45:19] <richlowe> be aware that does both debug and non-debug, lint, check, the lot however. [03:45:35] <richlowe> I'm not *certain* the ABI checks function, but I believe they do. [03:45:37] <tamlin> That's a LOT! [03:45:42] <richlowe> if they don't, tell me, and i'll see if I can fix them. [03:45:46] <Tpenta> if you are going to do non-debug, you WILL need to make a change to usr/src/Makefile [03:45:53] <richlowe> Tpenta: not true. [03:46:00] <richlowe> Tpenta: if he wants non-debug closed bins, he will. [03:46:07] <richlowe> Tpenta: but just for checking it, that doesn't actually matter. [03:46:10] <Tpenta> its true if you want to build both togther [03:46:17] <gisburn> umpf [03:46:18] <richlowe> No it isn't, as long as you don't care about the closed-bins mix. [03:46:20] <gisburn> I only want to do debug. [03:46:23] <richlowe> Trust me, I've been doing it for *months* [03:46:36] * Tpenta was assuming you'd want a consistant build [03:46:49] <richlowe> Tpenta: I thought he just wanted the gate checks, and to make sure both builds worked. [03:46:58] <richlowe> if he wants consistency, then yes, he should use your stuff. [03:46:59] <Tpenta> ahh [03:47:42] <gisburn> Hear me: Todays FIGHT TO THE DEATH: In the left corder: RICH-THE-LOWE (=LION) and in the right corner T-PENTA, the T-PENTAGRAM, equipped with satanic fighting powers! [03:48:10] <tamlin> As judge, GCC C compiler? ;-P [03:48:15] <richlowe> gisburn: it's hardly an argument. :) [03:48:34] * gisburn smacks tamlin [03:48:35] <Gman> stevel, woo, have it working finally - now just need to dick around with some css [03:50:08] <richlowe> gisburn: but yeah, one of my goals is to get all the checks and such working outside, so if any don't, let me know :) [03:50:18] <richlowe> (except package stuff, which rainer is doing) [03:50:25] <richlowe> and warlock, which I can't have. :) [03:50:32] * richlowe mutters [03:51:28] <tamlin> warlock? Proprietary? [03:52:59] <richlowe> not available outside Sun even, that I've found. [03:53:09] <richlowe> lock_lint does mostly the same stuff, but not the same way. [03:53:17] *** Error_404 has joined #opensolaris [03:53:25] <Error_404> okay, maybe i do care if gnome dies [03:53:30] <stevel> gman: awesome. [03:53:31] <Error_404> apparantly irssi depends on glib [03:53:40] <stevel> shoot it over when you're done and we'll try it out [03:55:37] *** tamlin is now known as tamlin-afk [03:59:09] *** nwf has quit IRC [03:59:29] *** nwf has joined #opensolaris [04:00:28] <gisburn> tamlin-afk: svn checkout of OS/Net: $ du -ks . [04:00:29] <gisburn> 848296 . [04:01:55] <richlowe> depending on the version of svn, that will be oversized. [04:01:59] <richlowe> (closed to double, actually) [04:02:08] <richlowe> since svn stashed unmodified copies in the .svn/ directories [04:06:10] <richlowe> Yeah, my onnv-clone is 402M [04:11:02] *** stevel has quit IRC [04:12:49] *** tamlin-afk is now known as tamlin [04:14:38] *** tamlin has quit IRC [04:17:18] *** rodrickbrown has joined #opensolaris [04:17:39] <jlc> 1 more day and the political adds are done..... [04:18:10] <Triskelios> ha. I don't watch TV [04:18:11] <Error_404> i'm canadian [04:18:30] <jlc> lol [04:18:52] <jlc> I watch hero's and monday night footbal [04:18:55] <jlc> l [04:28:24] *** jlc has quit IRC [04:37:13] *** dclarke has joined #opensolaris [04:37:22] <dclarke> evening [04:38:19] <Error_404> ahoy [04:38:30] <dclarke> hey ... [04:38:40] <dclarke> I am officially a miserable little shite [04:38:49] <Error_404> what'd you do? [04:38:50] <dclarke> I have your box of stuff here .. still [04:39:00] <Error_404> oh... meh, that's not a huge deal [04:39:17] <dclarke> maybe by Christmas ? [04:39:22] <Error_404> lol [04:39:33] <dclarke> I have been around the clock trying to get this new GNOME all tested [04:39:36] <dclarke> http://www.blastwave.org/dclarke/GNOME/Screenshot_15.png [04:39:37] <Error_404> perhas i'll sign up for genesi's OSW developer program [04:39:44] <Error_404> those guys have only been waiting a couple years [04:40:19] <dclarke> well I can definately get you an EFIKA unit [04:40:24] <dclarke> they are cool little toys [04:40:49] <Error_404> clear something up... [04:41:01] <dclarke> uh huh ? [04:41:04] *** yongsun_ has quit IRC [04:41:05] <Error_404> EFIKA != 8641D [04:41:11] <Error_404> real, or malarky? [04:41:22] <dclarke> its best to consult the docs [04:41:33] <dclarke> lets look at the www.blastwave.org/sponsors page [04:41:38] <Doc> all of us? [04:41:44] <Error_404> Doc: yes [04:41:51] <Error_404> Doc: flood it out, see what happens [04:42:02] <dclarke> Freescale MPC5200B PowerPC [04:42:13] <Error_404> ahh, so it's the embedded chip, as i thought [04:42:32] <dclarke> http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=MPC5200 [04:42:54] <dclarke> http://www.freescale.com/files/graphic/block_diagram/MPC5200BLOCKDIAGRAM.gif [04:43:13] <Error_404> and what of the tetrapower 970? what's the likelyhood we'll get our hands on a couple of those? [04:43:26] <dclarke> quad 970 ? [04:43:30] <dclarke> geez .. [04:43:35] <Error_404> because that's basically one-off from IBM's kit [04:43:43] <dclarke> thats a bigger ticket item I think [04:43:48] <Error_404> is it quad 970? i thought it was just dual [04:43:59] <dclarke> dual - dual-core I thought [04:44:02] <dclarke> not sure [04:44:06] <Error_404> so, unlikely then [04:44:09] <Error_404> :) [04:44:33] <dclarke> lets just focus on a T-shirt and ODW for Christmas [04:45:02] <Error_404> reason I ask is becasue as it stands nothing that we can get (with the exception of your pSeries that IBM didn't send you, it just showed up one day) is PPC64 [04:45:44] <Error_404> believe me, I'm more than grateful for anything thrown my way [04:46:57] <dclarke> the EFIKA is 32 bit PPC [04:47:06] <dclarke> but the future is 64-bit [04:47:16] <dclarke> our current Polaris kernel is 32 bit [04:47:23] <dclarke> its a hell of a situation [04:47:32] <dclarke> and I am busy as hell here working on GNOME [04:48:10] <Error_404> new version of gnome going in to blastwave? [04:48:17] <dclarke> I think I'll see if I can figure out how to rip some tracks off a CD and then convert them to mp3 here [04:48:23] <dclarke> new GNOME ? yes [04:48:29] <dclarke> GNOME 2.16.1 [04:48:32] <dclarke> top of the line [04:48:35] <dclarke> no feature spared [04:48:40] <dclarke> all the bells and whistles [04:49:18] <dclarke> goto http://art.gnome.org/screenshots/gnome216 [04:49:29] <dclarke> see that screenshot at the top there ? [04:49:40] <dclarke> thats the new GNOME from Blastwave [04:49:58] <dclarke> runs on Solaris 8 and 9 and 10 and 11 and x86 and AMD64 and Sparc [04:50:11] <dclarke> sort of a grand unified GNOME for the Solaris world [04:50:21] <dclarke> I'm still testing the hell out of it [04:50:30] *** jbk has joined #opensolaris [04:52:37] <richlowe> Gman: nice post. [04:52:47] <Gman> : [04:52:47] <dclarke> where ? [04:52:48] <Gman> ) [04:53:35] * dclarke looks for post .. [04:54:40] <LivingCorpse> dclarke: Will you be my best friend FOREVER? [04:54:54] <dclarke> forever and ever [04:55:12] <LivingCorpse> Indeed [04:55:30] * dclarke is reminded of a line from a David Bowie movie [04:55:39] <Stric> Anyone have a clue how to make a sunray "reboot" and try to connect to its primary server again? if I kill the session with utsession -k -t <blah>, it will just keep the ip&serverinfo and reconnect to the same server again instead of its primary [04:56:04] <Stric> the purpose is to clean up after all of them jumped over to their backup server.. [04:56:11] <jmcp> Stric: control alt (moon key) (moon key) iirc [04:56:16] <Stric> jmcp: .. from remote [04:56:38] <jmcp> oh [04:56:39] <Stric> our terminals are spread out all over campus, I don't want that much excercise :) [04:56:43] <jmcp> soft! [04:56:54] <jmcp> do you have users attached to the terminals? [04:57:00] <Stric> (and the sunrays are -way- better than the old indys running linux and X -query) [04:57:03] * jmcp pictures Agent Smith from The Matrix [04:57:08] <Stric> let's assume I don't. [04:57:13] <LeftWing> You can poke them with utfwsomethingorother [04:57:33] * dclarke runs man utfwsomethingorother [04:57:39] <LeftWing> Damn straight. [04:57:44] <jmcp> Stric: well as far as I'm aware your options are (a) wander around and whack the sunrays, or (b) ..... find an intern to do it for you [04:57:45] <Error_404> jmcp: i was picturing cube slaves chained to the sunrays (which are in turn chained to the table with those steel cables that I doubt work very well) [04:57:52] <Stric> utfwsync restarts the auth server I think.. so they will probably abandon that server and all jump to some other :) [04:58:04] <jmcp> Error_404: so was I :) [04:58:05] <Error_404> in tattered rags [04:58:22] <Stric> unfortunately, this is for a non-profit org.. so I'm the slave :P [04:58:28] <dclarke> with red swingline staplers [04:59:30] <dclarke> Gman == Glynn ? [04:59:33] <Gman> yeah [04:59:42] * dclarke *** waves *** [04:59:50] <Triskelios> dclarke: you've been here... how long? and you don't know that? =P [05:00:03] <dclarke> I have been here since the beginning [05:00:12] <dclarke> since the very beginning of the Matrix [05:00:15] <dclarke> looking for .. [05:00:16] <dclarke> the one [05:00:27] <LeftWing> Stric: There's a utfw* command you can use to prod individual MAC addresses. [05:00:53] <Stric> utfwsync threw them all out, so every ray jumped over to the other server.. [05:00:59] <LeftWing> heh [05:01:13] <dclarke> so er .. Gman .. what are your thoughts on this ? :: http://art.gnome.org/screenshots/gnome216 << -- see top Screenshots [05:01:52] *** tamlin has joined #opensolaris [05:01:55] *** OnkelSchorsch has joined #opensolaris [05:02:10] <Gman> dclarke, looks good [05:02:25] <Gman> i would imagine the volume control applet is broken because you need a patch that hasn't gone upstream [05:02:27] <LeftWing> GNOME keeps getting better to look at, I must say. [05:02:34] <Error_404> there, I've upgraded JDS on my server for some reason [05:02:35] <dclarke> it runs ( walks briskly ) on Solaris 8 Sun Ultra 2 with 200 MHz procs [05:02:52] <jbk> speaking of gnome, might i ask a question? [05:02:57] <dclarke> yeah .. I didn't fire up xine yet [05:02:57] <Gman> cool [05:03:03] <Gman> it'll miss hal/tamarack love [05:03:10] <Gman> which will cripple quite a few apps [05:03:19] <dclarke> there is no escaping that [05:03:22] <tamlin> jbk: Asking if you may ask a question is redundant. [05:03:37] <dclarke> the future rolls forwards but I need to support that massive Solaris 8 base out there [05:03:37] <jbk> well i'm new here, don't want to be rude :) [05:03:39] <richlowe> Gman: we're all missing hal/tamarack interaction with JDS. [05:03:46] * dclarke stubborn buggers the lot of them [05:03:54] <richlowe> well, I guess I could install the 2.16 vermillion bits. [05:03:55] <Error_404> jbk: you don't really need to ask to ask a question [05:03:56] <Error_404> just ask it [05:04:01] <tamlin> dclarke: Many boxes, eh? [05:04:09] <dclarke> tamlin : tons [05:04:33] <dclarke> tamlin : let me get my stats [05:04:53] <OnkelSchorsch> hey. does CSWsbcl for x86 work for any of you folks? I'm getting an error from sbcl, that it isn't installed properly... :( [05:05:17] <jbk> but basically, i've noticed that between the version of gnome in solaris 8 and what was shipped with solaris 10 (as well as at least build 33 of nevada), gnome terminal seems broken--i see it on multiple irc clients where when it's updating the status bar (with a different background color), it ends up changing the entire background for the window to the status bar backgrond color.. any ideas how to debug that? [05:06:10] <Gman> the gnome in solaris 8 and the gnome in solaris 10 is quite different [05:06:13] <richlowe> do you happen to be running it in screen? [05:06:18] <Gman> and a different terminal emulator - zvt vs vte [05:06:22] * richlowe saw that for a while, then it went away with the vermillion bits. [05:06:26] <tamlin> jbk: If so, that seems like a higher-level error, The X-server itself not collating rect's? [05:06:31] <dclarke> and the GNOME from Blastwave is different again [05:06:41] <Gman> it is? [05:06:42] <jbk> hmm..ususally it is screened [05:07:00] <richlowe> Gman: from 9 and 10, yeah, not vte v. zvt, I assume. [05:07:01] <jbk> but i think (though i might just be remembering wrong) i tried it outside of screen as well [05:07:04] <dclarke> tamlin : on November 1st I had 24,229 software downloads [05:07:16] <dclarke> 18,021 were Sparc [05:07:25] <Gman> richlowe, yeah [05:07:53] <tamlin> dclarke: And that 18k figure effectively states "8"? [05:07:53] <dclarke> and 4408 of those were to Solaris 8 users [05:08:06] <dclarke> no .. Solaris 8 == 4408 [05:08:13] <dclarke> Solaris 9 == 5453 [05:08:19] <dclarke> Solaris 10 == 5085 [05:08:25] <tamlin> OK. Enough to NPT leave nehind, from my POV. [05:08:25] <Error_404> yum, vegetarian chili from a can [05:08:26] <dclarke> its a nice distribution [05:08:35] <tamlin> s/NTP/NOT/ [05:08:42] <dclarke> Solaris 10 is growing slowly in the Sparc world [05:08:50] <dclarke> its growing like mad in the x86 world [05:09:00] <tamlin> Obviously! [05:09:14] <rodrickbrown> dclarke we are not allowed to use csw at work :( [05:09:17] <dclarke> tamlin : there is a large revenue base still dragging along Solaris 8 [05:09:18] <jbk> hmm.. i don't suppose there are any plans to backport any of that to solaris 10? [05:09:35] <dclarke> jbk : backport ? [05:09:50] <Error_404> wait... solaris runs on sparc now? [05:09:53] <dclarke> jbk : Solaris 10 runs circles around Sol 8 [05:09:59] <dwc-> either that or the solaris10 people keep reinstalling it when they mess up! [05:10:00] <jbk> vermillion [05:10:03] <jmcp> Error_404: nah, only when you're not looking [05:10:07] <jbk> i've got 10 on my workstation at work [05:10:20] <jbk> i don't know if i could get away with putting nevada on it [05:10:29] <dclarke> Well .. I also see Solaris 11 ( Nevada ) users there at 1719 packages [05:10:30] <rodrickbrown> running SOlaris 10 here in the flesh : [05:10:31] <rodrickbrown> :-) [05:10:31] <jbk> they're kinda touchy about that [05:10:38] *** bobbyz has joined #opensolaris [05:10:39] *** alobbs has quit IRC [05:10:51] <dclarke> rodrickbrown: http://www.blastwave.org/dclarke/GNOME/Screenshot_15.png [05:11:04] <dclarke> rodrickbrown: that will be released shortly .. its GNOME 2.16.1 [05:11:54] <dclarke> wow .. Solaris 10 just dominates the x86 stats I have [05:11:59] <tamlin> dclarke: I think I'm still hosed (private ISP) at a few 8 boxes. Thay are public, and have ~46k usrs IIRC. I just now came to think I'd like to migrate to 10 here - and get sun to sponsor to display "a working pilot case". Sounds insane? [05:12:00] <rodrickbrown> looks cute [05:12:04] <dclarke> by a factor of .. 3 to 1 [05:12:11] <Error_404> I've had NV on my U2.... it was... not so good [05:12:39] <dclarke> Error_404: I am runnign snv_46 here on an Ultra 2 .. runs fine [05:12:51] <richlowe> the newer JDS bits are kinda bad, everything else is just fine. [05:12:55] <Error_404> hmm... [05:13:00] <dclarke> tamlin : I recently installed a cluster in a telco with 13,000+ users [05:13:01] <Error_404> i blame my U2 then i s'pose [05:13:06] <richlowe> and JDS isn't that bad, if you don't see the drag-n-crash bug. [05:13:15] <dclarke> tamlin : they went straight to Solaris 10 from Solaris 8 [05:13:24] <dclarke> tamlin : it went .. very very well [05:13:29] <tamlin> dclarke: On existing Sun h/w I presume? [05:13:37] <dclarke> tamlin : new hardware [05:14:03] <tamlin> dclarke: Still, it wasn't some x86 thing? [05:14:10] <dclarke> tamlin : the idea was to migrate away from a pile of older Sun UltraSparc gear to UltraSparc III and IV gear [05:14:32] <dclarke> I pitched out the Opteron but they were not convinced [05:14:33] <jbk> the one issue i saw as we've been doing our upgrades was an app that was referencing the result of readdir() after calling closedir() (i.e. app bug) that happened to work in 8, but crashed in 10 [05:15:20] <dclarke> tamlin : Solaris 10 on Opteron is a damn good thing [05:15:54] * OnkelSchorsch seconds that [05:16:02] <OnkelSchorsch> and touches his ultra 20 m2 :) [05:16:31] <jmcp> OnkelSchorsch: not unless you're in the privacy of your own home, I hope .... [05:16:32] <dclarke> geez .. I don't have any slick hardware here in front of me [05:16:37] * jmcp drools @ his u20-m2 [05:16:42] <tamlin> dclarke: Then you can't speak about plain "PC" architecture I assume? We're talking about Sun arch? [05:16:51] <dclarke> I am currently messing with VMWare on RHEL4 [05:17:12] <dclarke> tamlin : I have Sol10 running on plan jane PC gear big time [05:17:18] <jmcp> dclarke: I noticed that blastwave's xsane isn't built with libusb support. This is a problem :( [05:17:30] <dclarke> jmcp : its being fixed [05:17:35] <Error_404> i want to pick up one of those tyan barebones servers.. the 1U units w/ a board & nothing else [05:17:40] <dclarke> jmcp : file another bug report [05:17:44] <jmcp> dclarke: cool, thankyou [05:17:45] <jmcp> ok [05:17:52] <dclarke> jmcp : that will really get us moving [05:18:03] * dclarke not that I sleep now anyways :-) [05:18:04] <Error_404> next time i've got $1000 kicking around i think [05:18:23] <dclarke> tamlin : want to see plain jane ? [05:18:33] <tamlin> dclarke: You are serious when you say a plain "PC" handles all the I/O, and interrupts, and still works? (my experience with a plain "PC" arch is that interrupts add up, and all goes to hell) [05:18:57] <dclarke> tamlin : I have a five year old HP Kayak .. it runs fine [05:19:08] <dclarke> its one of my trusted production boxes [05:19:21] <dclarke> and it runs Sol10Update 2 on 400MHz Pentium II procs [05:19:25] <dclarke> never fails [05:19:30] <Stric> Ok, I tried the 'route add rayip 127.0.0.1' and wait a while.. lo and behold, it restarted :P [05:19:50] <dclarke> I also have Sol10 on really wild kit [05:19:57] <tamlin> With 46k users split on 2 boxes? I can't belive the I/O will work. [05:19:59] <dclarke> like .. as in I have no idea who built it [05:20:06] <dclarke> whoa ! [05:20:10] <dclarke> what ? [05:20:13] <dclarke> no no [05:20:23] <dclarke> you can not stick 46,000+ users on that [05:20:26] <dclarke> geez [05:20:33] <tamlin> Also, all homedirs are on NFS. [05:20:36] * dclarke throws a brick at tamlin [05:20:45] <tamlin> No way in hell a peecee can handle that. [05:20:50] <dclarke> exacly [05:20:55] <jmcp> dclarke: do I need to create yet another darned account to log a blastwave bug? [05:20:56] <dclarke> no way .. no day [05:20:59] <tamlin> *phew* [05:21:01] <gisburn> bye [05:21:13] <Error_404> jmcp: yes, it pumps up the stats [05:21:16] <jmcp> ok [05:21:17] *** jbk has quit IRC [05:21:20] <dclarke> tamlin : the new bigger Sun X4600 type gear is what you need [05:21:28] <Error_404> "29 billion users in the continental US alone!" [05:21:32] <dclarke> jmcp : what ? [05:21:49] <dclarke> jmcp : did you lose a login id ? [05:21:52] *** gisburn has left #opensolaris [05:22:17] <jmcp> nope ... just haven't bothered to create one on blastwave.org [05:22:18] <dclarke> wtf ? [05:22:22] <dclarke> okay [05:22:32] <dclarke> and gisburn was joking .. I can't kick people [05:22:45] <Error_404> suuuuuure [05:22:55] <tamlin> dclarke: It's currently running on 2 Sun machines (old ones, Solaris 8) with NFS-mounted homedirs. I'm looking for a way to them to migrate (if only to get the benefit of zfs) to 10. [05:22:58] <dclarke> tamlin : think 8 way big Sun Opteron gear [05:23:36] <dclarke> tamlin : I used a pile of Sun fibre storage distributed across a pile of Sun V490's for 13,000+ users [05:23:37] <tamlin> dclarke: A reason to not keep the current working h/w? [05:23:41] <Error_404> and then think "send it to Error_404" [05:23:54] <dclarke> tamlin : what is it? E10K ? [05:23:54] *** rodrickbrown has quit IRC [05:23:56] *** rodrickbrown has joined #opensolaris [05:24:19] * dclarke please not E4500's [05:24:28] <dclarke> tamlin : E4500's ? [05:24:39] *** jbk has joined #opensolaris [05:24:48] <jbk> grr [05:25:04] <tamlin> dclarke: To be honest, I don't know what it is today. I know what it started out as, and that was basically desktop but rack mounted. :-) After that, it... evolved, and I no longer know what the machines are. [05:25:24] <tamlin> Gimme a second... [05:25:25] <dclarke> tamlin : you have no idea ? geez .. [05:25:41] <dclarke> well .. it ain't workstations I can tell you that [05:25:54] <tamlin> lol [05:25:55] <LivingCorpse> servers are better than workstations [05:26:05] <tamlin> Gimme a second to logon and see what I can deduce. [05:26:19] <dclarke> just run uname -a [05:26:32] <dclarke> and maybe prtconf -v | grep Memory [05:26:44] <tamlin> "sun4u sparc SUNW,UltraSPARC-IIi-cEngine" [05:26:47] <jmcp> dclarke: http://www.blastwave.org/mantis/view_bug_page.php?f_id=0001920 [05:26:54] * Stric has a rack mounted (sideways) blade1000 as zfs-nfs server for cdimage.debian.org [05:27:08] <dclarke> $ uname -a [05:27:09] <dclarke> SunOS enterprise 5.9 Generic_118558-05 sun4u sparc SUNW,Ultra-Enterprise [05:27:09] <dclarke> $ uptime [05:27:09] <dclarke> 5:26am up 577 day(s), 16:07, 4 users, load average: 0.21, 0.32, 0.45 [05:27:19] <dclarke> jmcp : let me go look [05:27:37] <tamlin> LOL! 1GB (1024 Mbytes) from "prtconf -v | grep Memory" [05:28:00] <dclarke> tamlin : that is not supporting 46K users [05:28:08] <dclarke> tamlin : not even 46 users [05:28:16] <dclarke> tamlin : maybe 4 [05:28:27] <dclarke> tamlin : whats wc -l /etc/passwd say ? [05:28:32] <tamlin> uname -a reports "5.8" even. [05:28:53] <dclarke> so .. Solaris 8 [05:29:40] <tamlin> dclarke: You do mean wc ls? [05:29:58] <dclarke> I mean wc -l /etc/passwd [05:30:05] <dclarke> thats a lower case letter L [05:30:08] <tamlin> Just wc -l passwd = 26. [05:30:19] <dclarke> so then .. no users to speak of [05:30:31] <Stric> and 'getent passwd|wc -l' ? [05:31:13] <LivingCorpse> servers rule the world [05:31:30] <Stric> 100x diff for me.. 22 vs 2837 on one machine.. [05:31:41] <tamlin> Still running here... [05:31:50] <dclarke> whats uptime say ? [05:31:58] <tamlin> 18598 [05:32:07] <dclarke> huh ? [05:32:08] <tamlin> 363 days [05:32:15] <Stric> dclarke: ever heard of NIS or LDAP? [05:32:20] <dclarke> hey .. party in 2 days !! [05:32:31] <dclarke> Stric : no [05:32:35] <dclarke> Stric : never [05:32:36] <dwc-> sun even has this cool NIS+ thing [05:32:45] <dclarke> Stric : I reject change and forward motion [05:32:46] <tamlin> 363 days uptime, 18598 from "getent passwd|wc -l". [05:33:34] <dclarke> I think after I passed the exams I swore to not look at NIS+ again [05:33:41] <jbk> hahaha [05:34:08] <tamlin> As an ISP, I think you cannot. :-) [05:34:47] <dclarke> I have this nice cozy little environment that is growing slowly [05:35:13] <dclarke> $ getent passwd | wc -l [05:35:13] <dclarke> 132 [05:35:26] <dclarke> everyone knows everyone .. its cozy [05:35:56] <dclarke> jmcp : got that bug report .. it will be the last straw ... thanks !! [05:36:31] <jmcp> ta [05:36:54] <dclarke> geez ... speaking of Solaris 8 .. I'm stuck here watching paint dry [05:37:00] <dclarke> patches going in [05:37:24] <tamlin> I just checked: "ls -l .. |wc" in just *my* user dir says over 20k. [05:37:30] <rodrickbrown> dclarke, its almost faster [05:37:37] <rodrickbrown> to rebuild a box from scratch via Jet [05:37:45] <rodrickbrown> than apply the lastest 10Rec [05:38:08] <dclarke> ya know .. I have never used jet [05:38:09] <rodrickbrown> Sun really has to do something about this slow ass patching [05:38:13] <dclarke> thought about it [05:38:21] <dclarke> rodrickbrown: they did [05:38:32] <dclarke> rodrickbrown: they bought out aduva [05:38:40] <Stric> now it's only painfully slow :P [05:38:41] *** Gman is now known as GmanAFK [05:38:46] <tamlin> "aduva"? [05:39:04] <tamlin> Why not simply apply a bindiff? [05:39:07] <dclarke> aduva? some company that has brilliant patch technology [05:39:31] <jbk> but is it fast? :) [05:39:34] <dclarke> rodrickbrown: besides .. this is Solaris 8 I am patching .. [05:39:43] <tamlin> (for the known previous binaries, I mean, and only use bindiffing for 2-8 of'em) [05:39:50] <dclarke> jdk : it wil still be slow [05:40:07] <dclarke> jbk : sorry [05:40:29] <dclarke> jbk : but .. you can play tetris while it patches [05:40:48] <dclarke> speaking of which .. what games do I have here ... lemme see [05:40:58] <jbk> heh.. our night staff will be happy about that :) [05:41:12] <tamlin> To put the writing on the wall, it actually seems Solaris 8 is a stable as Windows 2000 is in the MS world, and many, many are reluctant to move away "just because". [05:41:34] *** hile_ has joined #opensolaris [05:41:46] <LivingCorpse> windows 2000 advanced server? [05:41:52] *** hile_ has quit IRC [05:42:06] <jbk> well perhaps with zfs root + cloning it might allow for some alternate patching techniques for 10 at least (once it's there) [05:42:19] <tamlin> LivingCorpse: Never tried it. [05:43:00] <jamesd_> who knows they may even take advantage of the sqlite server solaris 10 shipps with to deal with smf ... to speed things up [05:43:05] <dclarke> http://blastwave.informatik.uni-erlangen.de/csw/users/michael/images/games.png [05:43:33] <dclarke> hrmmm .. there are a few old games here ... [05:43:40] <tamlin> LivingCorpse: I have just experience with "Windows 2000" unlimited. No names, no extras, no nothing. If it in marketing material has some extension, that's their problem. [05:44:05] <jbk> jamesd_: wasn't that tried prior to solaris 10 but pulled prior to GA? [05:44:13] <jbk> i seem to vaguely recall something about it [05:44:16] <jamesd_> not that i heard [05:44:24] *** klepparii has joined #opensolaris [05:44:27] <jbk> could just be on crack :) [05:45:05] <tamlin> No, I recall having seen it too. If it was ever implemented is another issue. [05:45:51] <richlowe> jbk: I recall that too, though I'm not sure about it being sqlite backed. [05:45:57] * dclarke goes looking for this : http://blastwave.informatik.uni-erlangen.de/csw/users/michael/images/comi.png [05:46:39] <tamlin> Ah, MI3! [05:47:35] <dclarke> MI3 ? [05:47:38] <dclarke> huh ... [05:47:56] <tamlin> Monkey Island 3, your screenshot. [05:48:01] <dclarke> oh .. [05:48:09] <dclarke> is *that* what that is .. [05:48:19] <dclarke> well .. I was trying to install it here [05:48:24] <tamlin> You haven't played it?! [05:48:31] <dclarke> nope [05:48:40] <tamlin> Shame! [05:48:41] <dclarke> we just have the packages to run some games [05:48:59] <dclarke> I am so busy with stuff like .. well .. GNOME testing [05:49:18] <jamesd_> you guys didn't even mention the best game on unix/linux... freeciv [05:49:36] <tamlin> Yeah, GNOME testing must be way more important than actually experiencing MI3... [05:49:39] <dclarke> freeciv .. is that open source ? [05:49:57] <tamlin> FreeCiv is GPL. [05:50:07] <jamesd_> dclarke, yes.. you better check that, since its included at blastwave ;-) [05:50:17] <dclarke> never played it [05:50:30] <Stric> freecraft is free too.. but then blizzard came and said "DIE! DIE! DIE!" [05:50:41] <tamlin> Note: FreeCiv is as most such conversions, onlu multi-player (AFAIK). [05:50:54] <dclarke> http://www.blastwave.org/packages.php/freeciv ah yes .. now I recall it [05:50:55] <jamesd_> dclarke, http://www.blastwave.org/packages.php/freeciv [05:51:19] <tamlin> Stric: Tha name after that became something *quite* different IIRC? [05:51:37] <Stric> Stratagus [05:51:44] *** kleppari has quit IRC [05:51:47] <tamlin> Right. [05:52:02] <jamesd_> hmm ularn isn't in blastwave.... [05:52:04] <dclarke> jamesd : I really should play a few of these games [05:52:17] <dclarke> ularn ? [05:52:21] <jamesd_> dclarke, not really you are too damm busy at it is. [05:52:36] <sahafeez> great. i updated the bios on an amd64 board and now it it bricked [05:52:42] <dclarke> speaking of which .. I have to step away for a bit [05:52:45] <Triskelios> CSWscummvm, btw --> [23:45:45] dclarke goes looking for this : http://blastwave.informatik.uni-erlangen.de/csw/users/michael/images/comi.png [05:52:58] <dclarke> sahafeez: that sucks ... :-( [05:53:08] <tamlin> dclarke: If you're heading the Monkey Island way, I really suggest DOSBox and MI1&2 too. Hell, while at it, why not Space Quest. [05:53:16] <jamesd_> dclarke, http://www.users.bigpond.com/jolds/ularn.htm [05:53:20] <sahafeez> completely [05:53:27] <jamesd_> ULarn is a dungeon type adventure game similar in concept to HACK, ROGUE or MORIA, but with a different feel and winning criteria. [05:53:44] <tamlin> sahafeez: Please say you did uopdate from "DOS"? [05:53:45] <Stric> sahafeez: we have an amd64 machine with bios "BETA RELEASE: System Bios Revision = V00.09.08" .. it was part of the "look at this new shiny thing called opteron" program.. we don't dare to upgrade the bios :) [05:53:57] * dclarke remembers the original Adventure .. played on a terminal [05:54:08] <Triskelios> tamlin: I hope you didn't mean dosbox in association with the MI games [05:54:17] <jamesd_> dclarke, yeah me too.. its my son's favorite unix game [05:54:20] <sahafeez> yes [05:54:21] <tamlin> dclarke: Me too! Running off a PDP-11 I think. [05:54:32] <Stric> it's quite identical to a sun v20z.. but without the logos and with stoneage bios :) [05:54:32] <dclarke> XYZZY [05:55:00] <tamlin> Triskelios: Yes, I did mean DOSBOX for MI1 and 2. [05:55:13] <Triskelios> tamlin: ew, why not scummvm? [05:55:51] <tamlin> Triskelios: They are in scummvm now too? Oh, I didn't know that! Jot my "suggestion" down to ignorance then. [05:56:04] <Triskelios> tamlin: they are the first thing supported in scummvm [05:56:07] <Triskelios> since like 2001 [05:56:16] * dclarke **poof** - you are standing at a small wooden shed. A small stream flows down a hill towards open forest [05:56:58] <tamlin> dclarke: xyzzy was the *original* adventure. I don't know, but I might have been playing that off a TOPS-10 machine. :-) [05:57:02] <Stric> detonate shed [05:57:11] *** laca has quit IRC [05:57:33] <dclarke> tamlin : welcome to the old man club .. Jamesd has milage also [05:58:00] * dclarke these new whipper snappers .. whadda they know [05:58:31] <dclarke> we had CPM and 8 inch floppies [05:58:36] <dclarke> and we were happy ! [05:58:40] <dclarke> happy I tell you ! [05:58:57] <tamlin> Me too, until that floppy got an error... [05:59:01] <dclarke> but you try to tell *that* to kids today .. [05:59:12] <dclarke> ha ha :-) [05:59:13] <sahafeez> well guess i need a new amd64 mb [05:59:14] <Stric> Amstrad 3" floppies were kinda nice too. (no, not 3.5") [05:59:27] <tamlin> Try telling the kids of today to "backup"... [05:59:30] <dclarke> sahafeez : really ? [05:59:55] <dclarke> sahafeez : no way to yank the PROM or ? [06:00:05] <tamlin> When I say "kids", I mean some sysadm's too... [06:00:14] <sahafeez> brickage. bricked. shed it mortal coil. it snuffed it. [06:00:19] <sahafeez> its an pc. [06:00:20] <dclarke> tamlin : you don't need a backup when you have lots of disk .. right ? :-\ [06:00:42] <dclarke> sahafeez : nice space heater you have there :-( [06:01:03] *** rafaeldt has joined #opensolaris [06:01:05] <dclarke> sahafeez : that sucks ... [06:01:15] <sahafeez> its my fault. it may have committed suicide since i put xp on it. [06:01:28] <jbk> i always found the HP K boxes made excellent space heaters :) [06:01:41] <dclarke> oh .. XP on an Opteron ... thats just .. wrong [06:01:45] <tamlin> dclarke: Of course not. They migrate data magically. Isn't that in the manual (serious sarcasm, as it seems you too had a boss voice). [06:01:59] <Triskelios> our sparcserver 1000 is named spaceheater.acm.jhu.edu [06:02:15] <dclarke> Triskelios: is it running ? [06:02:26] *** rafaeldt has quit IRC [06:02:33] <dclarke> Triskelios: thats sun4d I think ? [06:02:36] <sahafeez> i wonder if i can get a new socket 754 board.. [06:02:41] <tamlin> dclarke: *a boss voicing his opinion [06:02:49] <Triskelios> dclarke: yup [06:02:53] <Triskelios> (to both) [06:03:10] <Triskelios> someone stopped sshd though [06:03:19] <tamlin> sahafeez: a NEW? I dubt it. [06:03:21] <dclarke> tamlin : I had an argument with an IT Director in which I had to tell him that if you can not put it onto a tape and walk out the door with it .. then it ain't a backup [06:04:00] <dclarke> socket 754 ... I wonder what I have here [06:04:08] <tamlin> dclarke: Backups are stored in safes, in a bank. [06:04:35] *** GmanAFK has quit IRC [06:04:36] *** OnkelSchorsch_ has joined #opensolaris [06:04:46] <dclarke> with a bare metal restore test .. once in a while [06:05:03] <tamlin> Best of luck... [06:05:22] *** alanc has joined #opensolaris [06:05:22] <dclarke> sahafeez: what is "AMD Opteron(tm) Processor 246" ? [06:05:31] <dclarke> is that socket 754 or ? [06:05:34] <Stric> without socket information [06:05:37] <dclarke> geez .. I dunno [06:05:40] <Stric> err. no [06:05:42] <tamlin> Anyone familiar with NT4 (or 3.51)'s backup program? [06:05:47] <sahafeez> no idea. new to the 64bit amd [06:05:48] <Stric> 246 must be non-754 [06:05:49] <dclarke> tamlin : yes [06:05:50] <sahafeez> yes [06:05:52] <Stric> 754 is single cpu only [06:05:55] <Error_404> is there a such thing as a 754 opteron? [06:06:04] *** alanc is now known as alanc_away [06:06:16] <Stric> hm. now when I think about it.. no. :) [06:06:17] <dclarke> tamlin : why ask about NT4 ? [06:06:37] <Stric> opterons are 939, 940 etc.. [06:06:46] <Triskelios> maybe he has a funny anecdote to tell about the backup program [06:07:02] <dclarke> huh .. well .. it will have to wait [06:07:14] * dclarke steps away [06:08:02] <tamlin> dclarke: Imagine the "joy" when I needed a restore, and the person having done all the backups always was careful to do "Eject" from the file menu, and never gave it a though why it took so long? Check that menu. Eject is a menu item from Erase... [06:08:14] <bobbyz> What's the recommended way to keep system patches up-to-date with the sun x86 solaris 10 06/06 release? [06:08:27] <bobbyz> I'm using it on just one machine [06:09:24] <tamlin> dclarke: Need I spell it out? The IDIOT had erased ALL backup tapes before they were sent to the safe. [06:10:03] *** nexrafa has quit IRC [06:10:52] <sahafeez> so anyone want to buy an Ultra 80? [06:10:53] <tamlin> bobbyz: There's no easy answer to that. How many machines? [06:11:00] <bobbyz> tamlin: 1 machine [06:11:15] <LivingCorpse> why would somebody use outdated NT4? [06:11:20] *** solarisjon has quit IRC [06:11:32] <sahafeez> new motherboard @ frys is $50 [06:11:32] <bobbyz> 1 machine and preferably using the CLI. Is it smpatch I need to reeed up on or something else? [06:11:33] <LivingCorpse> is it even useful anymore? [06:11:43] *** karrotx has quit IRC [06:11:54] <tamlin> sahafeez: Again, had it been closer, it's be mine by now. :) Seems ebay may be your better choice. [06:12:20] <sahafeez> well thought i would ask a few times here before i list it as the people change.. [06:12:21] <tamlin> LivingCorpse: This was "back in the day". [06:12:32] <sahafeez> i will ask tomorrow then its going on ebay [06:12:42] *** OnkelSchorsch has quit IRC [06:13:28] <LivingCorpse> is solaris better than netware? [06:13:52] <sahafeez> they are 2 completely different things. [06:13:59] <tamlin> LivingCorpse: That is indeed an interesting question, that has no answer. [06:14:02] <sahafeez> you are asking is car like a boat [06:14:46] <tamlin> Though I'd venture to say: YES! [06:14:56] <tamlin> :-) [06:15:07] <Error_404> sahafeez: if you're smuggling mexicans, a '57 chevy is very much like a boat [06:15:11] <Error_404> apparantly [06:15:14] <Error_404> err s/mexicans/cubans [06:15:21] <g4lt-U60> sahafeez, how much you looking for for it? [06:15:36] <tamlin> $300 IIRC [06:15:56] <sahafeez> well it has 4x450. 4gb of ram. 18gb scsi, dvd and 2xelite3d(m6&m3) [06:15:58] <tamlin> 4-way [06:16:22] *** andres__ has quit IRC [06:16:35] <sahafeez> and right now has a working install of gentoo linux with xorg working in dual head ;) [06:16:40] <tamlin> What kind of RAM is it in it? [06:16:52] <sahafeez> 4gb ecc as far as i know [06:17:12] <g4lt-U60> U80 has that weird config like the U60, right? [06:17:27] <sahafeez> what do you mean? [06:17:30] <tamlin> I mean, PCx, DDRx or something lese (i.e. RAM speed) [06:17:50] <sahafeez> ah, its a sparc tamlin. that means nothing [06:17:53] <g4lt-U60> tamlin, let me guess, you don't own a sparc? [06:18:12] <g4lt-U60> sahafeez, does it take memory from U30/60? [06:18:23] <sahafeez> yes [06:18:26] <sahafeez> but it is full. [06:18:32] <sahafeez> 4gb was the max i think [06:18:36] <tamlin> Correct. I don't own a single Sparc box. I do have an O2. :-) [06:18:39] <g4lt-U60> yeah, sounds right [06:19:06] <g4lt-U60> tamlin, almost all sparc memory until very recently is the same type, 128-pin IIRC [06:19:18] <sahafeez> 4 GB maximum (with 256-MB DIMMs, in pairs) 576 bits wide; 512 bits data, 64 bits ECC 1.78-GB/sec throughput there you go tamlin [06:19:30] <g4lt-U60> the SB100 is the first machine to take DIMMs [06:19:49] <g4lt-U60> sahafeez, actually, that type is called DSIMM [06:20:02] <sahafeez> da-sim? [06:20:13] <g4lt-U60> dual SIMM [06:20:14] <sahafeez> in my best hick voice [06:20:23] <jbk> it's what they call them in chicago :) [06:21:31] <sahafeez> so this women goes into the hospital for an emegancy c-setion on her twins. he husband is out of town but his brother from the south (deep south alabama area is with her). [06:21:55] <sahafeez> when she wakes up the doctor tells her that the brother named the twins - one boy one girl [06:22:20] <sahafeez> she is quite upset about that - ask what are the names. the doctor says, well the girl is denise. [06:22:33] <sahafeez> she thinks that good. maybe i am upset for nothing. [06:22:34] <g4lt-U60> 3...2...1 [06:23:06] <tamlin> What's the maxed out read-speed of these modules (just out of interest - latency is usually more interesting to me)? [06:23:16] <sahafeez> so whats the boys name. the doctor looks and say "da-neffu" [06:24:02] *** nbkk6fo_ has joined #opensolaris [06:24:03] *** rodrickbrown has quit IRC [06:24:10] <sahafeez> tamlin - http://sunsolve.sun.com/handbook_pub/Systems/U80/spec.html [06:24:25] <sahafeez> come on. that was funny [06:25:32] <tamlin> OK, so it is 1.78GBps. [06:25:43] *** lisppaste3 has quit IRC [06:26:29] <sahafeez> no bad for 1999.. [06:26:50] <tamlin> Basically what my old O2 had, and INDEED not bad! [06:27:21] <tamlin> Only in 200x that PC's even got over 1GBps I think, no? [06:28:01] <tamlin> That PC's today easily pull 4GB/s is still wicked, I think. [06:28:10] <nbkk6fo_> System Configuration: Sun Microsystems sun4u Netra t 1120/1125 (2 X UltraSPARC [06:28:12] <nbkk6fo_> System clock frequency: 110 MHz [06:28:12] <nbkk6fo_> Memory size: 2048 Megabytes [06:28:15] *** nbkk6fo_ is now known as rodrickbrown [06:29:00] *** lisppaste3 has joined #opensolaris [06:29:54] <tamlin> What IS impressive is 110MHz, and it could still use this RAM bandwidth. [06:30:34] <g4lt-U60> tamlin, uhm, look at the SS10, you could get that in 50MHz and still use the same memory [06:31:02] <g4lt-U60> (yes, they did make a microsparc MBUS module, why do you ask?) [06:35:17] <tamlin> As you might have noticed, I didn't ask - I reflected on history (meaniong it seems inventions all happened in the past, and today it's just... improvements). [06:36:50] <LivingCorpse> i have a Cray 1 [06:38:51] <tamlin> LivingCorpse: You know, having met the people I have over IRC, I see no reason to not believe you, however unlikely. [06:39:10] *** johnlev has joined #opensolaris [06:39:40] <richlowe> johnlev: mornin'(?) [06:40:15] <johnlev> richlowe: evening, I'm in the valley this week :) [06:40:42] <jamesd_> tamlin, don't forget its not just a 32 or 64 bit buss, it was probably over 500bits like my u2 is. [06:42:41] *** slowhog has joined #opensolaris [06:43:00] *** yongsun_ has joined #opensolaris [06:45:41] <richlowe> johnlev: after thinking more, I think I'd go for something savecore-like, rather than deal with what would effectively be another target. [06:46:10] <johnlev> richlowe: well [06:46:41] <johnlev> richlowe: I think I can get away with it almost. from mdb's POV it's just a little bit of extra tricky in ->activate for mdb_kvm [06:46:48] <johnlev> s/tricky/trickery/ [06:47:07] <johnlev> richlowe: and it seems a bit wrong to add administrative overhead just to make the source be less hassle [06:47:26] <richlowe> Hm, if you can get away with it within the kvm target, I guess that works too. [06:48:58] *** tamlin has left #opensolaris [06:50:55] <johnlev> richlowe: well hopefully it won't work out more tricky than I expect it to be :) [06:51:20] <richlowe> heh, I'm adding more FIXME's than I'm removing, my opinion may not be trustworthy :) [06:51:59] *** dclarke has left #opensolaris [06:52:04] <johnlev> that sounds like progress to me :) [06:54:52] <Error_404> I never understood the difference between netra & ultra [06:55:09] <richlowe> The netra stuff is NEBS certified, I think. [06:55:17] <jbk> i believe the netras can run off dc [06:55:19] <richlowe> or whatever the term is. [06:55:20] <jbk> and nebs [06:55:36] <jbk> if you want to stick it inside a cabinet of a phone switch :) [06:56:51] <Error_404> the government is that picky about phonelines? [06:56:58] <jbk> more telcos [06:57:18] <jbk> but there's (In the US at least) *lots* of regulations regarding phone service [06:58:05] <jbk> especially for local phone service [06:58:16] <Error_404> yeah, i guess 911 is important [06:58:40] <Error_404> beyond that... i mean, calling gramma's not *that* important [06:58:57] <jbk> not just that.. there are requirements as to how long service can be out, and just a whole bunch of stuff [06:59:11] <jbk> and in general telcos tend to be (sometimes overly imo) cautious [07:00:30] * jbk still has nightmares of trying to connect an ultra-2 to a siemens EWSD [07:01:16] <jbk> as i recall, the ewsd only spoke x.25 externally via rs-232 [07:01:17] <jbk> however [07:01:29] <jbk> we had to get special adapters to sheild the ground from teh switch [07:01:52] <jbk> then run that through an rs-232->rs-443(?) converter because the built in rs-232 ports wouldn't do hdlc or something like that [07:01:53] *** sahafeez has quit IRC [07:02:51] <jbk> (usually just mentioning x.25 is enough to send shivers down people's spine :P) [07:03:02] *** Tpenta has quit IRC [07:03:42] <Error_404> i don't know what that is, so *shrug* [07:04:05] <jbk> that's probably not a bad thing :) [07:04:24] <g4lt-U60> x.25 r0x0rz ;P [07:04:48] * jbk looks at g4lt-U60 funny [07:05:11] * g4lt-U60 looks funny for jbk [07:05:31] <dwc-> ack. x.25 [07:05:54] <jbk> it was a brand new phone switch too (this was in 99) [07:06:03] <jbk> complete with 9-track 6250dpi tape drive :) [07:09:37] *** |tsoome| has quit IRC [07:23:53] *** rodrickbrown has quit IRC [07:24:03] *** rodrickbrown has joined #opensolaris [07:28:31] *** bougie has joined #opensolaris [07:28:43] *** lloy0076 has joined #opensolaris [07:34:23] <djgregor> hmm.... this seems to reliably panic SXCR 51: route add <some IP> localhost; ping <some IP> [07:34:53] *** |tsoome| has joined #opensolaris [07:45:58] <lloy0076> Does one have to install the BFU archives in order? [07:46:19] <lloy0076> Or can I, say, install the BFU archives for 50 straight over a 48 install (from DVD)... [07:46:24] * lloy0076 hopes he's making sense [07:47:41] <djgregor> anyone know details on bug 6348581 (it's not available from bugs.opensolaris.org) [07:48:30] <johnlev> lloy0076: not really [07:49:22] * lloy0076 hmmm [07:49:23] <richlowe> not unless there's a flag day inbetween, no. [07:49:35] <richlowe> http://www.opensolaris.org/os/community/on/flag-days/all/ [07:49:41] <richlowe> in general, check, and do what they tell you :) [07:49:44] <lloy0076> Ok, I think I'll do an incremental update. [07:49:56] * lloy0076 knew there was a reason to keep the old bfu archives :P [07:50:04] <richlowe> the old archives won't help you. [07:50:09] <richlowe> but you may need a newer SX:CR. [07:50:12] <richlowe> (I don't think you do, in this case) [07:50:18] <johnlev> djgregor: fix for that was integrated into build 44 [07:50:19] <richlowe> djgregor: out of interest, what is it? :) [07:50:42] <lloy0076> richlowe: I had a system that I'd gotten to build 50 but for some reason ON nightly 20061103 caused a kernel panic. [07:50:57] <lloy0076> richlowe: My intention is to BFU the system back to 50 using the same steps I did to get it there :) [07:51:05] <richlowe> djgregor: Oh, I see it in the putback logs. [07:51:16] <richlowe> djgregor: if it's hidden, that's most likely intentional, and you'll have to figure it out from the diffs. [07:51:29] <richlowe> (I may no claims as to whether it's *sensible*, just that it's probably on purpose) [07:52:18] <johnlev> it is. [07:52:41] <richlowe> it looks to be loopback specific. [07:52:47] <richlowe> if I'm reading the comments above the changes correctly. [07:52:50] *** bougie has quit IRC [07:53:15] <richlowe> johnlev: well, the only other option was it being filed somewhere strange, I'm guessing it's keyworded as security though. [07:53:38] <richlowe> which is also why I stated the difference between sensible and intentional :) [07:53:58] <djgregor> I found what I needed in the google cache for that bug. :-) [07:54:08] <djgregor> It doesn't seem to be what I'm running into [07:54:27] <djgregor> it looks like ire_add is what eventually does the dereference [07:54:48] *** estibi has quit IRC [07:55:41] *** lloy0076 has quit IRC [07:56:14] <djgregor> richlowe/johnlev: thanks for the info, btw [07:56:22] <twincest> apparently 300M + 512M RAM is not enough to build ON [07:56:35] <richlowe> beyond the sorting options and such, google is better at finding stuff on bugs.opensolaris.org than bugs.opensolaris.org, by the way. [07:56:43] <richlowe> handy hint, for anyone who cares. [07:57:33] <richlowe> twincest: it should be enough to build, if you're gentle with DMAKE_MAX_JOBS. [07:57:37] <richlowe> twincest: lint will not run, however. [07:57:52] <richlowe> (it'll probably run out of VM, then segv) [07:58:35] <richlowe> the kernel cross checks drop between 500 and 800M in /tmp, and use about as much real memory again. :) [08:10:18] <twincest> hm, my contributor agreement number is surprisingly low. [08:11:15] *** slowhog has left #opensolaris [08:11:54] <Error_404> twincest: what's your number? [08:12:45] <richlowe> twincest: above or below 20? [08:13:26] <Error_404> are there free tshirts involved for those <20 or something? [08:13:45] *** |tsoome| has quit IRC [08:14:03] <richlowe> Nah, I just don't expect it to be >= 30. [08:15:36] <Error_404> mine's 51 [08:15:52] <richlowe> really? [08:16:03] <Error_404> yes [08:16:14] *** sahafeez has joined #opensolaris [08:16:15] <twincest> rich: 50 [08:16:27] <richlowe> The highest I've seen on request sponsor has been 18, I think. [08:16:39] <richlowe> Well, that cheers me up :) [08:16:54] <sahafeez> so trying to boot 50 on a dual p3. get to "ERROR: Cannot find windowing software, Exiting to shell" [08:16:55] <Error_404> twincest: heh, neat... i got my agreement all signed up right after yours i s'pose [08:17:19] <richlowe> it's only taken twincest the better part of 18 months. :) [08:19:46] <Error_404> what cheers you up, btw? that solaris has 51 non-sun contributors? [08:19:54] <richlowe> Yeah. [08:19:57] <twincest> hmm, and i never got a confirmation of that bug [08:20:04] <richlowe> like I said, I was expecting the number to be much lower. [08:20:05] <twincest> i suppose it did go into the void [08:20:14] <richlowe> twincest: It maybe worth asking someone to check. [08:20:33] <richlowe> last time I did this I ended up filing a duplicate, because it didn't actually make it anywhere for quite a while. [08:20:41] <richlowe> (though I did ask someone to check, and they said "No", and it still got there twice.) [08:21:19] <richlowe> johnlev: if BT2 is down or read-only, do bugs-my-mail submissions bounce to their originator/specified user, or vanish? [08:22:03] <johnlev> richlowe: well they don't bounce back to me :) [08:22:08] <johnlev> no idea... [08:22:47] <richlowe> Well, I don't know if anyone actually uses bugs-by-mail. [08:22:53] <richlowe> beyond the b.o.o stuff. [08:30:32] *** dlg has joined #opensolaris [08:31:02] <dlg> is it just me or are some of the opensoalris sites not working? [08:32:07] <twincest> which ones? [08:32:14] <richlowe> src isn't. [08:32:19] <dlg> i cant get to the src browser [08:32:27] <dlg> and mercurial thingy isnt happy [08:32:43] <richlowe> mercurial thingy tends to be intermittent. [08:32:45] <twincest> src is down, was mentioned on -discuss, and i guess steve is still fixing onnv-gate [08:32:46] <richlowe> try that one again. [08:32:55] <twincest> at least his last message said 'i'll fix this later' [08:32:56] <richlowe> but steve hasn't pushed the fixed onnv-gate yet, but the broken one is accessible. [08:33:05] <dlg> hrm [08:33:08] <dlg> i might go home then [08:33:28] <richlowe> dlg: there's an intermitant fault with the way the mercurial stuff works, where an automount will time out, when that happens the failure mode seems fairly random. [08:33:39] <richlowe> generally either a permission denied from ssh, or a binary not found (i forget it's name) [08:33:53] <twincest> sounds like something should be mounted less automatically :) [08:33:55] <richlowe> but there'll be a fairly lengthy pause before it. [08:34:23] <richlowe> in that case, it'll generally work on the next attempt. [08:34:39] <dlg> hrm [08:34:40] <dlg> ok [08:34:41] *** Dar_HOME is now known as Dar [08:35:15] <dlg> there isnt a tarball of the source up anywhere anymore is there? [08:35:52] <richlowe> only for the builds, not the snapshots. [08:36:08] <richlowe> Hm, I wonder if stevel has realized he'll have to regenerate all the build bundles, too. [08:36:11] <dlg> the code i want to look at wont have changed [08:38:30] *** maitrella has quit IRC [08:42:18] <dlg> does anyone here have a copy of the tree? [08:42:36] <dlg> would you mind emailing me a few files? [08:43:07] *** coolvibe_ has quit IRC [08:44:11] <richlowe> http://polaris.blastwave.org/browser/on/trunk [08:44:13] <richlowe> try looking there? [08:44:23] <richlowe> (I'm not sure when Cyril last updated, I don't think it was that long ago, though) [08:44:26] <sahafeez> hey, how would i copy one disk to another - lets say 2 hard drivers. dd? [08:44:44] <richlowe> 2 weeks ago, I guess. [08:45:14] <twincest> http://dlc.sun.com/osol/on/downloads/20061023/on-src-20061023.tar.bz2 [08:45:29] <dlg> heh, genesi [08:45:50] <richlowe> wish I knew how I broke vread. [08:45:51] <dlg> twincest: thanks [08:46:01] <richlowe> made it do what (I think) it should, and it now fails most confusingly. [08:47:59] <sahafeez> how long do you guys thing it will take to dd if=/dev/sdb of=/dev/sda for an 18gb disk? [08:48:12] <dlg> sahafeez: a while [08:48:26] <sahafeez> as in 10 hours? 1 hour? [08:50:52] *** |tsoome| has joined #opensolaris [08:51:34] *** lloy0076 has joined #opensolaris [08:51:39] *** Cyrille has joined #opensolaris [09:04:21] <twincest> where's the full ON cstyle guide? [09:04:30] *** sylvain has joined #opensolaris [09:06:19] <richlowe> opensolaris.org/os/community/on [09:06:27] <richlowe> if the link is still dead, give me a while, and I'll hunt a direct link. [09:06:34] *** sahafeez has quit IRC [09:06:51] <lloy0076> Well, I appear to be back in the land of the "working" systems again. [09:06:54] <richlowe> cstyle -cPp will get most of it. [09:07:03] <twincest> yep, 404 [09:07:12] <richlowe> really? [09:07:17] <richlowe> I just loaded it. [09:07:20] <twincest> We're sorry. The resource "/os/community/on/cstyle/" could not be found. [09:07:24] <richlowe> developing-in-on 404's, probably. [09:07:37] *** sahafeez has joined #opensolaris [09:07:38] <richlowe> twincest: ah, no, that'd be cstyle.el [09:07:45] <richlowe> twincest: "Cstyle Guide" [09:07:46] <twincest> all i really wanted was the line length max :) [09:07:49] <richlowe> 80 [09:07:57] <richlowe> cstyle(1) will complain about that bit at least. [09:08:00] <twincest> tx [09:08:10] <richlowe> the only things I've noticed cstyle -cPp not getting are matching bracedness on if/else [09:08:25] <twincest> apparently cat.c has several other cstyle errors.. [09:09:01] <richlowe> twincest: -c isn't used by default. [09:09:06] <richlowe> I use it, because I always forget that bit :) [09:09:41] *** halton has quit IRC [09:10:53] <twincest> hmmz, no source code for tail? [09:11:26] <twincest> that's silly. [09:12:04] <richlowe> it's one of a few fairly silly ones. [09:12:51] <richlowe> od is the other that springs to mind. [09:13:15] <twincest> maybe i should file '/usr/bin/tail should be replaced with open BSD tail' [09:13:22] <twincest> (open as in open source, not openbsd) [09:13:37] <richlowe> Roland had said that was on his list of things to do. [09:13:55] <Triskelios> twincest: as opposed to closed BSD? ;P [09:14:11] <twincest> Trisk: as oppose to replacing it for some other reason [09:14:16] *** dunc_ has joined #opensolaris [09:14:19] <sickness> openbsd rU|3z :))) [09:15:07] <richlowe> twincest: the issue with most of them is making sure it does exactly what it did before, without being able to see the source for 'before', and without currently having any of the ON tests. [09:15:18] <lloy0076> sickness: OpenBSD may rule but Zope is a right royal PITA to install [09:15:21] *** razrX has quit IRC [09:15:46] <twincest> so.. do i risk filing another bug, or ask an employee to do it for me? :) [09:15:58] <twincest> or just submit the patch and hope the sponsor does it.. [09:16:00] <richlowe> ask an employee if BT2 is up, then file it? :) [09:16:52] <richlowe> may as well ask if it's already there, and you just haven't been forwarded the notification, too. [09:17:38] * twincest awaits Gman [09:18:06] <richlowe> the bug interface of choice. [09:20:08] <twincest> is /usr/xpg6 conforming to single unix specification issue 6? [09:20:55] *** [1]Cyl has joined #opensolaris [09:22:21] *** johnlev has quit IRC [09:23:21] *** razrX has joined #opensolaris [09:30:56] <lloy0076> Confound it! [09:31:19] <lloy0076> Why, oh why, does Solaris' default tar have to silently just truncate files whose paths are more than 100 characters long... [09:31:27] <twincest> tar -E [09:31:48] <twincest> the problem is the old tar file format only supports filenames that long [09:32:04] <lloy0076> man tar doesn't have a -E option. [09:32:16] <lloy0076> ...but tar -E shows it. [09:32:23] <twincest> it does here.. " E Write a tarfile with extended headers." [09:33:23] <lloy0076> Does it work on the read-back side too? Because I'm not familiar with Zope (which insists on having stupidly long paths in its install), I get obfuscated errors such as "Cannot find *-config.zcml" :( [09:33:36] <richlowe> use gtar instead? [09:33:45] <lloy0076> richlowe: Such is what I resorted to :) [09:33:46] <twincest> as long as the destination tar supports it, extended header swill be handled automatically [09:34:27] <twincest> (in particular, older versions of GNU tar and GNU pax do *not* support it - they have their own non-standard extension) [09:37:39] *** Gman has joined #opensolaris [09:39:33] * Gr|ffous waves at Gman [09:39:33] *** Cyl has quit IRC [09:39:34] *** [1]Cyl is now known as Cyl [09:39:40] <Gman> hey Gr|ffous [09:40:01] <Gr|ffous> I have to admit, I wasn't expecting to read about haircuts in my RSS feed today ;) [09:41:25] <Gman> you're not the only one who said that ;) [09:41:38] <Gr|ffous> must have been a great cut! [09:41:43] <Gman> yeah, seriously [09:41:53] <Gman> i'll blog about something more normal next time [09:42:05] <Gman> best cut was actually going into a place in krabi, thailand [09:42:11] <Gman> no english [09:42:21] <Gman> fun, even got a shave out of it too :) [09:42:58] <Gr|ffous> oh nice [09:43:07] <Gr|ffous> are you doing the MOvember thing this month? [09:44:00] <Gman> no, but a mate is [09:44:16] * Gman hopes he grows a nice handlebar [09:45:28] *** lloy0076 has left #opensolaris [09:45:35] <Gr|ffous> It's funny at work, most men are doing it. I had a meeting with our director, on casual Friday, completely unshaven, to ask for more money for an IT project [09:45:52] *** lloy0076 has joined #opensolaris [09:45:53] <Gr|ffous> the thing was, he looked just as bad, and he gave it to me, so it was a funny day all round! [09:46:02] <Doc> try and talka few of the chicks into it! [09:46:30] *** MaGre has joined #opensolaris [09:46:37] <Gr|ffous> erm, and why exactly would we want that? [09:48:42] <twincest> -e file True if file exists. (Not available [09:48:42] <twincest> in sh.) [09:48:46] <twincest> damn, sh is stupid [09:48:57] <twincest> Gman> hey! can i nag you about bugs? :) [09:49:02] *** PosixC1 has joined #opensolaris [09:49:15] <Doc> twincest: so just use /usr/bin/test [09:49:19] <Gman> twincest, so long as it's not drag and drop in nautilus [09:49:24] <twincest> Doc: yeah, but still, it's stupid [09:49:24] *** PosixC1 has quit IRC [09:49:53] <twincest> Gman: no :) i filed a bug yesterday along the lines of '/usr/bin/passwd getresponse does not handle control-d' but i didn't get a confirmation - was wondering if it's just waiting in the queue, or it got lost somewhere.. [09:50:12] <Gman> twincest, let me have a look for you [09:52:12] *** razrX_ has joined #opensolaris [09:52:57] <Gman> twincest, bugster seems slow right now, some patience required [09:53:53] *** stevel has joined #opensolaris [09:53:54] *** ChanServ sets mode: +o stevel [09:54:05] <richlowe> stevel: just the man I wanted to annoy. [09:54:17] <twincest> gman: no rush [09:54:20] <richlowe> stevel: you're going to need to regenerate all the build bundles when you do the gate, if you hadn't realized that already. [09:54:30] *** razrX has quit IRC [09:54:30] <stevel> ughhhhhh [09:54:48] <Gman> hey steve [09:54:56] <Gman> twincest, it's back up - looking [09:55:13] <Gman> twincest, it's not in the triage queue [09:55:23] <twincest> :-( [09:55:46] <richlowe> stevel: Yeah, that was my reaction when I realized, too. [09:55:55] *** razrX_ is now known as razrX [09:55:57] <richlowe> twincest: file it again, if it dups ask someone to close it :) [10:01:13] *** nwf has quit IRC [10:01:29] *** jamesd__ has joined #opensolaris [10:02:28] *** jamesd_ has quit IRC [10:02:57] *** jamesd__ has quit IRC [10:04:03] *** nwf has joined #opensolaris [10:07:21] <richlowe> dlg: source browser is back. [10:07:24] *** BoeB has joined #opensolaris [10:07:41] *** BoeB has left #opensolaris [10:08:33] <stevel> i just restarted it [10:08:34] <dlg> richlowe: thanks [10:08:35] <stevel> sorry [10:08:38] <twincest> rich: and what if it gets lost again? :) [10:08:54] <dlg> its ok, im just going to keep rubbishing java [10:08:58] <stevel> i took off for korea without telling anyone the username/password to the glassfish domain [10:09:18] <Doc> Short Description : BugTraq Production and Reporting databases are not in sync.Email is also affected. [10:09:39] <Doc> twincest: that's probably why you havent got an email yet [10:09:42] <twincest> ah. [10:10:03] *** Gman has quit IRC [10:10:05] <twincest> i'll wait a bit longer then [10:10:06] *** alanc_away has quit IRC [10:10:09] <Doc> so hold off on raising it again for a few days at least [10:10:16] <Doc> they are actively working on the problem at the moment [10:10:22] <twincest> i have other bugs to fix in the meantime anyway [10:10:51] *** alanc_away has joined #opensolaris [10:10:55] *** Gman has joined #opensolaris [10:12:59] *** mega has joined #opensolaris [10:20:47] <twincest> heh, trust gisburn to start a two page discussion on request-sponsor :) [10:21:02] <richlowe> it's not his fault people chose to respond on there. [10:21:11] <richlowe> it's not like stevel didn't start discussion on onnv-notify. ;) [10:21:18] <richlowe> (and someone or other kept moderating it through, too) [10:21:20] <twincest> i know, but it just had to be one of his mails that started it ;-) [10:27:09] <raph_ael> hello [10:30:11] <Berny> morning [10:31:14] *** schily__ has joined #opensolaris [10:31:36] <Gman> twincest, ok, back online and checking your bug...again ;) [10:32:25] <Doc> gman: bugster is borked, so i wouldnt spend too much time on it [10:32:52] <Gman> oh, ok? [10:33:03] <Gman> was sort of working previously [10:33:13] <Doc> production and reporting databases are out of sync, and as such emails are not being sent [10:33:20] * Gman sees mail [10:33:21] <Gman> thanks Doc [10:34:49] *** coolvibe has joined #opensolaris [10:36:26] <Gr|ffous> I've created a zone, in ZFS namespace, but it didn't create its own zfs fs. I'm on snv_45, was this perhaps added after this version? [10:37:09] *** lloy0076 has left #opensolaris [10:40:44] *** schily_ has quit IRC [10:43:28] *** halton has joined #opensolaris [10:54:39] *** stevel has quit IRC [10:59:08] *** damienc has joined #opensolaris [11:03:35] *** lplatypus has joined #opensolaris [11:04:04] <twincest> hm, i think my first project will be to make all of cmd/ report system errors properly [11:04:11] <twincest> the number of commands that don't is surprising [11:08:14] *** yongsun_ has quit IRC [11:11:58] <Gr|ffous> I'm still stuck on this, is there some special trick to make zoneadm create a zfs file system for you? [11:12:11] <jmcp> Gr|ffous: it just did it for me on snv_48 [11:12:28] <Gr|ffous> does it matter if the zfs fs is already nested? [11:12:32] <jmcp> Gr|ffous: dunno [11:12:47] <jmcp> Gr|ffous: you need to set the dataset property for the zone. man zonecfg [11:13:21] <Gr|ffous> I have angel/zones created, and mounted in /export/zones. I've set the zone path to /export/zones/fullzone. What else do I need to do? [11:14:01] <Gr|ffous> I may be misunderstanding what a dataset is then, isn't that for exporting a zfs FS from the global zone INTO the non-global [11:14:17] <twincest> isn't that what you want to do? [11:15:02] <Gr|ffous> The way I've been reading this, when I create my new zone (fullzone), it will create a zfs FS in the global zone, and the fs will contain my newly installed zone [11:15:10] <Gr|ffous> ie, that zone sits in it's own FS [11:16:08] <twincest> oh, this has to be the best error message: [11:16:14] <twincest> 1/root@nvdev:~/onnv-gate/usr/src/cmd>bdiff a [11:16:15] <twincest> bdiff: arg count [11:16:42] <Gr|ffous> http://docs.sun.com/app/docs/doc/819-2450/6n4o5mde8?a=view <== "How to install a Configured Zone"/2. That says "Install the zone, automatically creating a ZFS file system if the zonepath is on ZFS." [11:16:55] <richlowe> twincest: aren't useful and informative error messages great? [11:17:38] <Gr|ffous> so I have angel/zones created (zfs...), and I'm expecting it to create angel/zones/fullzone [11:17:40] * richlowe still doesn't quite get why bdiff exists in general. [11:17:51] <twincest> i guess because diff is old and sucky [11:18:16] <richlowe> you'd think teaching diff not to suck would be better than having bdiff slice and dice, then fork diff [11:18:40] <twincest> it wouldn't surprise me if bdiff was written for 16-bit systems [11:18:45] <twincest> it looks fairly old [11:19:41] <richlowe> ah, arg mismatch. [11:22:43] <twincest> oh, <fatal.h> is just fun as well [11:22:46] <twincest> who wrote this crap :) [11:23:29] <dlg> you could just steal a diff from somewhere else [11:24:04] *** nbkk6fo_ has joined #opensolaris [11:24:06] *** rodrickbrown has quit IRC [11:24:09] <twincest> "just" [11:24:22] <dlg> why not? [11:24:27] <richlowe> diff isn't so bad. [11:24:35] <richlowe> bdiff is pointless, and could probably be rolled into diff. [11:24:42] <richlowe> patch is intolerably awfuly :) [11:24:45] <richlowe> 'awful' [11:24:48] <richlowe> (and closed) [11:25:18] <twincest> i might think about replacing diff or bdiff when i have more than 0 putbacks of experience :) [11:25:54] <dlg> http://www.openbsd.org/cgi-bin/cvsweb/src/usr.bin/diff/ [11:25:56] <dlg> omg free [11:26:20] <twincest> i'm quite aware there are existing free diffs, you'll notice i suggested doing the same with tail earlier today [11:26:28] <richlowe> diff(1) is free. [11:26:33] <richlowe> replace patch! [11:26:34] <twincest> yes, but sucky [11:26:41] <richlowe> twincest: not compared to patch. [11:29:04] <dlg> can you guys fix up ksh a bit while you're at it [11:29:27] <twincest> there's already a ksh93 integration project [11:31:19] *** nwf has quit IRC [11:31:41] * jmcp thwaps dlg [11:31:50] <richlowe> jmcp: eh? [11:32:02] <jmcp> richlowe: asking for fixes to ksh [11:32:06] <jmcp> dlg: silly!! [11:32:15] * richlowe is still puzzled [11:32:17] <dlg> heh [11:32:38] <twincest> i have to say i don't like this 4 space continuation rule in cstyle [11:32:41] <jmcp> richlowe: .... gisburn will get you [11:32:44] <richlowe> I've come to the conclusion this Xorg segv is triggered by me having useful yet unsaved state. [11:32:53] <twincest> for readability i much prefer matching the continuation with the relevant opening brace [11:32:55] <richlowe> it's the only coherent pattern I've found. [11:32:56] <jmcp> twincest: dare i say it, but too bad. that's ON's style [11:33:02] <twincest> jmcp: yeah, i know [11:33:03] <Gr|ffous> has anyone else had the issue that fc-cache nearly cripples their system when a new zone is setup? [11:33:06] <twincest> i just don't like it :) [11:33:19] <twincest> (fortunately the rest of cstyle is pretty much my preferred code style) [11:33:21] <richlowe> it'll die when I'm in the middle of something and have all the state on screen but not yet noted down, no other time. [11:33:36] <jmcp> richlowe: sounds like it's watching you [11:34:30] <lplatypus> though <fatal.h> is well obfuscated, I can deduce that solaris didn' [11:34:35] <richlowe> twincest: you may think that. [11:34:42] <lplatypus> ..didn't support multithreading in 1988 [11:34:45] <richlowe> cstyle also triggers my paranoia ;) [11:35:42] <twincest> do i need a separate CR to bring an existing file into cstyle conformance or can i do it as part of an existing CR for that file? [11:36:03] <richlowe> twincest: existing. [11:36:07] <twincest> phew [11:36:45] <jmcp> twincest: if you have an existing CR for it then make the file cstyle-happy before you RTI. otherwise, file a new CR [11:37:32] <richlowe> jmcp: since we don't RTI, it's pretty safe to do it while there. [11:37:40] <jmcp> details, details .... [11:37:47] <twincest> (yet) [11:38:39] <richlowe> I'm not sure if that's optimism or cynicism. [11:38:55] <twincest> kind of both. i hope it'll happen within 10 years :) [11:39:08] <richlowe> I want the facilities to be there, but not to actually have to deal with them :) [11:40:38] <twincest> i wonder if anyone's written a vim macro to make these 4-space continuations automatically [11:40:59] <dlg> its there, but its a bit long [11:41:11] <twincest> where? [11:41:12] <dlg> Atabtabtabspacepsacespacespace [11:41:25] <twincest> that would be funny, but it's wrong :) [11:41:35] <dlg> maybe less tabs [11:41:41] <dlg> or more [11:41:42] * quasi is getting slightly tired of opengrok now that I get results for each file from /loficc/ /zfs-crypto/ /onnv/ [11:41:52] <richlowe> quasi: path:onnv [11:42:21] <richlowe> click-through stuff, yes, it's annoying. [11:42:36] <quasi> richlowe: cool, thanks (would be grat to have as a dropdown) [11:43:14] *** ppires has joined #opensolaris [11:43:32] <ppires> hi :-) [11:47:22] <jmcp> good evening [11:49:09] *** mustang has quit IRC [11:51:29] <Gman> hrm, i guess stevel has gone for the evening [11:51:51] <twincest> hmm, bdiff also uses intimate knowledge of struct FILE [11:52:10] <twincest> i guess feof hadn't been invented :) [11:52:24] *** deather_ is now known as deather [11:52:27] *** ppires has quit IRC [11:53:16] <oxygene> twincest: sounds like some careful cleanup is necessary [11:53:38] <richlowe> flamethrower. [11:53:45] <twincest> well, i suspect rich is right and removal of bdiff entirely is necessary [11:53:50] <twincest> but this is a nice easy project for now :) [11:54:25] *** sahafeez has quit IRC [12:02:21] *** Xh4 has joined #opensolaris [12:03:23] *** raph_ael has quit IRC [12:04:37] <lplatypus> wouldn't it be hard to remove bdiff from solaris, given the commitment to backwards compatibility? [12:04:59] <twincest> it could just be a wrapper for diff, it doesn't do anything special that diff couldn't [12:05:06] <twincest> (just needs to ignore its extra arguments) [12:05:15] <richlowe> it could just be diff built to take those arguments. [12:05:29] <lplatypus> yeah that sounds better [12:05:32] <richlowe> it is just a wrapper for diff, it just cuts a file to chunks and feeds it to diff. [12:05:41] <twincest> i mean a wrapper that just does exec(diff) [12:05:47] <twincest> rather than all the magic [12:06:53] <lplatypus> diff is no longer limited to 3500 lines right? [12:08:32] <twincest> hmm, maybe not, it works okay for a 6M input file. but if someone fixed diff, why is bdiff still there at all? [12:09:00] *** kimc has joined #opensolaris [12:09:05] <lplatypus> exactly [12:09:22] <twincest> i mean, why wasn't it removed when diff was fixed [12:09:25] *** raph_ael has joined #opensolaris [12:09:32] <lplatypus> hmm nexenta seems to be missing man pages for solaris commands [12:10:43] <twincest> hm, i wonder if it's because diff could still use a lot of memory [12:10:49] <twincest> i suspect these days that isn't much of a nissue [12:11:18] <richlowe> lplatypus: it's quite happy with 64,000 lines. [12:11:44] *** mega has quit IRC [12:12:22] <twincest> how does one go about removing bdiff's functionality then? i suppose it needs an ARC case? [12:15:45] *** sparc-kly_ is now known as sparc-kly [12:15:52] *** ChanServ sets mode: +o sparc-kly [12:20:47] *** PosixC has quit IRC [12:21:01] *** PosixC has joined #opensolaris [12:21:02] *** PosixCompliant has joined #opensolaris [12:26:15] *** PosixC has left #opensolaris [12:31:39] *** bunker_work has joined #opensolaris [12:35:34] *** MaGre has quit IRC [12:35:38] *** Gman is now known as GmanAFK [12:36:34] *** kimc has left #opensolaris [12:37:31] *** MaGre has joined #opensolaris [12:44:53] *** gbon121 has joined #opensolaris [12:48:21] <twincest> what does *cmd* mean in a CR? (like "*ls*: ls causes machine to explode") [12:51:21] <jmcp> twincest: means that the submitter wanted to draw attention to it [12:51:28] <twincest> ah [12:51:51] <jmcp> or *dis* is borked for 32bit [12:52:29] <jmcp> basically making you think "ah, it's a command that they're referring to" [12:54:45] <twincest> hmm, is ON allowed to use C99 features? (__VA_ARGS__ in particular) [12:54:52] <jmcp> not sure [12:54:58] <timeless> doubtful [12:55:01] <jmcp> I think there are some cases where it's ok, but not generally [12:55:12] <twincest> apparently it's compiled with -xc99=%none, but i wonder if that's changable on a per-directory basis [12:55:18] <timeless> given that if you try using them, the compilers will complain about lots of things that are currently ignored [12:58:09] <richlowe> most doesn't, some does. [13:01:59] <richlowe> I recall there being a bug reference in one of the Makefile's related to why it's not used in uts [13:02:02] <richlowe> but I can't seem to find it. [13:02:25] <timeless> http://opensolaris.org/jive/message.jspa?messageID=34654 [13:04:28] *** Xh4 has quit IRC [13:08:57] *** vogelfaull has joined #opensolaris [13:09:02] <vogelfaull> why [13:09:33] <timeless> http://www.opensolaris.org/os/community/tools/gcc/bug_fixing_notes/;jsessionid=C3DB15C0DDD4406AA1C7C6C3A5D186BF [13:09:36] <vogelfaull> why [13:09:38] <timeless> was the page i was thinking of [13:09:47] <vogelfaull> why [13:12:35] *** bengtf has joined #opensolaris [13:18:31] <twincest> >cstyle -cPp acctcms.c|wc -l [13:18:32] <twincest> 148 [13:18:36] <vogelfaull> why [13:19:01] <richlowe> drop the -c, and things improve. :) [13:19:09] <richlowe> like I said, I just use it as a reminder for myself, and type it out of habit :) [13:19:14] <vogelfaull> why [13:19:15] <Cyrille> <pulp fiction>say why, say why one more time, I dare you, I double dare you</pulp fiction> [13:19:27] <vogelfaull> why [13:19:44] <twincest> rich: -c is still meant to be adhered to though, isn't it? [13:19:50] <vogelfaull> why [13:19:51] <richlowe> they don't improve by much though. [13:20:03] <richlowe> twincest: I believe so, but the normal checks don't include it. [13:20:12] <twincest> well, may as well fix it while i'm here [13:20:15] <richlowe> so it's fairly routinely noisy. [13:20:22] <vogelfaull> why [13:20:27] <richlowe> that'll make the diff nice and hard to review... :) [13:20:43] <vogelfaull> why [13:22:01] <twincest> all it does is change a printf to include errno [13:22:06] <vogelfaull> why [13:22:21] <dlg> solaris doesnt have err.h? [13:22:30] <vogelfaull> why [13:23:43] <vogelfaull> why [13:23:47] <vogelfaull> why [13:23:50] <vogelfaull> why [13:23:54] <vogelfaull> why [13:25:01] *** jmr has joined #opensolaris [13:28:08] *** jmr has left #opensolaris [13:28:57] *** uebayasi has quit IRC [13:29:42] *** jamesd_ has joined #opensolaris [13:30:07] *** jamesd_ is now known as jamesd [13:30:33] *** ChanServ sets mode: +o jamesd [13:31:53] <vogelfaull> jameed sucks [13:32:40] *** vogelfaull has quit IRC [13:33:25] <Cyrille> that was a useful contribution... [13:35:09] <twincest> dmake lint in cmd/acct makes me sad [13:39:02] *** lin has joined #opensolaris [13:41:10] <jamesd> damm... i didn't get a chance to thank him for his support in person [13:43:41] *** calumb has joined #opensolaris [13:44:45] * dlg yawn [13:46:25] <twincest> would this code be acceptable in ON? #define FOOSZ sizeof(((struct T *)0)->a_member) -- i'm fairly sure it's well-defined according to the C standard [13:47:04] <dlg> hrm [13:47:24] <dlg> if you want that size, wouldnt you have an instance of the struct floating around at the time [13:47:34] <dlg> so you can just got sizeof(t->a_member)? [13:47:37] <dlg> s/got/go/ [13:47:39] <twincest> apparently not, the existing code creates a dummy struct to do it [13:48:04] <dlg> weird [13:48:19] <dlg> is the member a buffer or a type? [13:48:32] <twincest> it's a char array [13:48:45] <twincest> (utmpx fields, to be exact) [13:48:46] <dlg> is its size specified by a macro? [13:48:58] <dlg> their size should be well known... [13:49:14] <twincest> the manpage doesn't seem to offer a size [13:49:23] *** lin is now known as linma [13:50:42] *** sylvain has quit IRC [13:50:44] <twincest> neither does SUS, afaik [13:51:00] <dlg> damnit, thats what i lost when i reformatted [13:51:08] *** linma has left #opensolaris [13:51:15] *** sylvain has joined #opensolaris [13:51:28] *** mega has joined #opensolaris [13:51:31] <twincest> SUS? you can download current versions, at least [13:51:44] <dlg> yeah, i had a mirror of it on my laptop [13:51:50] <dlg> so i could easily look it up and appear smart [13:52:27] <twincest> perhaps this code should be using strlen at the point of use, instead of max sizes [13:52:52] <dlg> no, i dont think those fields are proper strings [13:53:12] <twincest> fixed size space padded or something? [13:53:33] <dlg> its very platform dependant [13:53:42] <dlg> but generally theyre a fixed width, zero padded [13:53:57] <dlg> but you dont need a nul to terminate it if the string in the field takes the full length [13:54:17] *** linma has joined #opensolaris [13:54:23] <dlg> theyre the reason strncpy and strncat behave like they do [13:58:19] <dlg> i still find it very strange you would need the size of one of those buffers without an instance of them being around [13:58:33] <twincest> it uses the size for the size of other arrays [13:58:40] <twincest> struct foo { char s[LSZ]; }; [13:59:10] <dlg> how ugly [13:59:55] <twincest> indeed [14:02:20] *** sil3nt|warri0r has joined #opensolaris [14:04:20] <twincest> hmm: FILE *fopen(), *holptr; [14:04:26] <twincest> this code is *fun* [14:04:31] <timeless> heh [14:04:48] <timeless> i don't think i've seen that syntax recently [14:04:56] <timeless> you have access to more amusing code than me [14:05:01] <timeless> i didn't think that was possible [14:05:14] <twincest> you too can view this code at src.opensolaris.org .. but i'd recommend against it ;-) [14:05:39] <timeless> i have access to gnome, mozilla, among others at mxr-test.landfill.bugzilla.org [14:05:52] <timeless> but the scariest stuff, you can't view, it's on http://swift/ :) [14:06:12] <timeless> and the world is a better place for not being able to see it [14:08:09] <twincest> ooh, this is fun, it defines its own 'struct hours' that includes the initial part of struct tm, then uses struct tm* and struct hours* interchangably [14:08:31] <twincest> this compiles fine because it's using K&R function declarations [14:15:15] <timeless> heh [14:15:42] *** laca has joined #opensolaris [14:17:38] *** PosixCompliant has quit IRC [14:17:45] *** PosixC has joined #opensolaris [14:19:54] <PosixC> shold b52 be released today ? I saw that b51 was 25-Oct-2006 1 [14:28:18] * jmcp sleeps [14:28:51] *** LeftWing_ has joined #OpenSolaris [14:29:07] *** Snake007uk has joined #opensolaris [14:36:50] *** lplatypus has quit IRC [14:38:21] *** lplatypus has joined #opensolaris [14:40:54] *** klocze1 has joined #opensolaris [14:44:58] *** trede has joined #opensolaris [14:46:03] *** LeftWing has quit IRC [15:06:32] <twincest> so, is this a bug or just very bad code? http://www.rafb.net/paste/results/4P7biJ77.html [15:06:46] <trede> sxcr is same concept as sx , only abit newer and more untested ? [15:08:10] *** edp has joined #opensolaris [15:08:14] *** PosixC has quit IRC [15:08:41] <timeless> that's scary [15:09:03] <timeless> shadowing variables is generally a bad idea [15:09:16] <twincest> ah, never mind, that was my fault [15:13:55] *** hile_ has joined #opensolaris [15:19:45] *** MaGre has quit IRC [15:23:42] *** maitrella has joined #opensolaris [15:26:50] *** gbon121 has quit IRC [15:30:06] <twincest> "/usr/xpg4/bin/awk -f nonsuch1 nonsuch2" <- this command produces curious output [15:33:49] *** eboutilier has joined #opensolaris [15:34:14] *** sil3nt|warri0r has quit IRC [15:35:48] <Stric> a question regarding zfs.. if I have a raidz2 of 7 disks and a recordsize=128k .. is that per disk or in total? [15:38:16] <Stric> trying to figure out why we're getting around 7x10MB/s of IO from ~5MB/s of nfs traffic (reading from zfs, sending out) [15:38:31] <OnkelSchorsch_> hm. I think that you can set it per filesystem. not per pool or per disk. [15:39:00] <Stric> OnkelSchorsch_: yeah, I can set it per filesystem.. but is the recordsize a span over the entire stripe or is a stripe 7x128k ? [15:39:37] <OnkelSchorsch_> sorry. can't help you there :) [15:39:59] * timeless wonders why you aren't allowed to chgrp symlinks [15:40:56] *** karrotx has joined #opensolaris [15:42:12] *** karrotx has quit IRC [15:42:58] <PerterB> Stric: there's a known issue relating to that but I can't find the details just now... something along the lines of the zfs readahead cache causing excess IO in that kind of situation [15:44:54] *** |tsoome| has quit IRC [15:47:45] *** |tsoome| has joined #opensolaris [15:47:54] *** nbkk6fo_ has quit IRC [15:48:31] <libkeiser> Stric: I think 128k has to be total stripe size. iirc, the major performance advancement in raidz and raidz2 is that the io scheduler has the flexibility to pick any subset of spindles large enough to meet the raid level's redundancy requirement (3 for raidz, 4 for raidz2) based upon current i/o loading [15:51:54] <PerterB> ah, this was the bug I was thinking of: http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6437054 [15:53:33] *** deather_ has joined #opensolaris [15:57:43] *** mrdeviant has joined #opensolaris [15:58:52] <twincest> usr/src/cmd/man/src/util/instant.src/sqlnet.log < what is this? [16:06:00] *** halton has left #opensolaris [16:06:52] <trede> can solaris boot from a zfs raid ? [16:07:22] <PerterB> no [16:08:29] <Stric> PerterB: running nv48 now which has lower vdev prefetch at least.. but it's still way off :/ [16:09:25] <jbk> you know zfs really spoils you for day to day stuff :) [16:09:46] *** deather has quit IRC [16:10:06] <PerterB> Stric: ah ok, was worth a shot :) [16:10:09] <trede> thanks for help peterb [16:10:38] <Stric> according to http://blogs.sun.com/roch/entry/the_dynamics_of_zfs, there's a few more readaheads as well.. [16:10:38] *** glagasse has quit IRC [16:11:45] <jbk> hmm.. our web proxy seems overloaded today :( [16:14:16] *** RaD|Tz has joined #opensolaris [16:15:07] *** sommerfeld has joined #opensolaris [16:15:08] *** ChanServ sets mode: +o sommerfeld [16:15:45] *** PosixC has joined #opensolaris [16:16:04] *** PosixCompliant has joined #opensolaris [16:16:49] *** trede has quit IRC [16:16:54] *** trede has joined #opensolaris [16:18:06] <Stric> PerterB: but it could be that "minimum 64k" reads.. [16:20:16] *** deather_ is now known as deather [16:26:32] *** mdj has joined #opensolaris [16:34:13] *** bengtf has quit IRC [16:34:59] *** mrcranky has joined #opensolaris [16:35:22] <asyd> I wonder how openssl is rfc4158 complitant [16:37:35] *** PosixCompliant has left #opensolaris [16:42:21] *** trede has quit IRC [16:42:33] <Stric> http://src.opensolaris.org/source/search?q=zfs_prefetch_disable&defs=&refs=&path=&hist= might be something for me.. since all io will be random io and caching is only sane on the nfs clients, not the zfs-nfs backend.. [16:42:44] <Stric> (1G ram, 1.5TB working set.. might not fit in ram :) [16:46:00] *** linma has quit IRC [16:48:19] *** |tsoome| has quit IRC [16:51:38] *** PosixC has quit IRC [16:51:58] *** klepparii is now known as klepari [16:52:03] *** klepari is now known as kleppari [16:54:57] *** Michael has joined #opensolaris [16:56:02] <Michael> hello, everyone! [16:56:21] *** PosixC has joined #opensolaris [16:56:22] *** PosixCompliant has joined #opensolaris [16:56:25] *** boxix has joined #opensolaris [16:56:31] *** PosixCompliant has left #opensolaris [16:57:06] <boxix> hello, where can i get the inode size of a ufs partition ? [16:57:57] <Stric> mkfs -m filesystem [16:57:58] <PosixC> I am trying to copy a file which is on solaris to linux; that folder in which the file is is nfs mounted fom linux ; running "share" shows root=myMachine / rw "" [16:58:11] <twincest> or df -t filesystem [16:58:20] <PosixC> yet I get an error : cannot chown target file [16:58:24] <boxix> thnaks Stric [16:58:29] <PosixC> operation not permitted [16:58:47] <PosixC> As far as I saw, adding root=myMachine to the share [16:58:54] <PosixC> should solve such a problem [16:58:59] <PosixC> is not it so ? [17:00:27] <PosixC> and the mount from linux is " type nfs (rw,soft,addr=..." [17:01:07] *** vmhobbes has joined #OpenSolaris [17:01:44] <bobbyz> Is it typical for root's home directory to be '/' instead of '/root' with solaris? I'm coming from using linux, and that seems odd [17:01:46] *** bengtf has joined #opensolaris [17:01:58] <dwc-> yes [17:02:28] *** Dar is now known as Dar_HOME [17:03:23] <bobbyz> weird. Do people never 'su -' then? I just noticed all of the 'dot' file sitting in / on my solaris system because of my use of 'su -' [17:03:36] *** bunker_work has quit IRC [17:04:18] <kleppari> bobbyz, you can move /root if you want to.. [17:04:30] <kleppari> err, root's home dir, that is [17:05:10] <bobbyz> kleppari: yeah, I think I'll do that. I just find it odd that it defaults to '/'. If you didn't move it, it seems like users might be able to read some files they shouldn't have access to :) [17:05:22] <dwc-> I don't really run much as root that makes .files .... lsof, ssh, and rsync are about it [17:06:16] <dwc-> oh, I guess .sh_history [17:06:53] <kleppari> bobbyz, [17:07:02] <kleppari> bobbyz, no, they wouldn't [17:07:16] <kleppari> they might see that they're there, but that's all [17:07:19] *** alanc_away is now known as alanc [17:07:26] <bobbyz> yeah, I noticed most of the perms are 600 [17:08:11] <alanc> root's home directory is / because that's where the first Unixes put it years ago, and when they tried to move it to /root for Solaris 10 too many of the test suites broke, so they chose a cluttered / over not being able to test the system as well [17:08:29] <alanc> they're supposed to be working on fixing the test suites so they can change it in the future [17:09:12] <bobbyz> alanc: Thanks, that explains it. I think I'll probably go ahead and move it to /root on my own then because that's what I'm used to :) [17:09:59] *** alanc is now known as alanc_away [17:17:01] *** RaD|Tz has quit IRC [17:23:03] *** laca is now known as lacaAFK [17:36:10] <hile_> arguably, there shoudln't BE anything but maybe .profile-EIS in root's home directory anyway [17:38:26] *** mrcranky_ has joined #opensolaris [17:43:35] *** PosixC has quit IRC [17:45:54] *** mrcranky has quit IRC [17:47:27] *** johnlev has joined #opensolaris [17:55:11] <sickness> i'm back [17:57:49] *** Hunger- has quit IRC [17:59:26] *** Hunger- has joined #opensolaris [17:59:46] *** |tsoome| has joined #opensolaris [18:00:10] *** |tsoome| has quit IRC [18:00:20] <quasi> damn [18:01:03] <twincest> johnlev: i don't suppose you would be the right person to ask about a strange file in the onnv-gate tree? [18:01:14] <johnlev> probably not. what file? [18:01:41] <twincest> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/man/src/util/instant.src/sqlnet.log [18:01:44] <Error_404> install_backdoor.c ? [18:01:54] *** Michael has quit IRC [18:01:58] <Error_404> that's normal... we don't talk about it [18:02:04] *** |tsoome| has joined #opensolaris [18:02:54] *** Michael has joined #opensolaris [18:03:16] *** Michael has quit IRC [18:03:33] <twincest> it looks like someone was using an sqlnet program and committed their logfile by accident :) [18:03:44] <johnlev> twincest: indeed it does... file a bug please [18:03:48] <jwtodd> hey .... trying to build opensolaris on an amd64 machine. per README.opensolaris when i try to build the kernel (nearly last step) i get an error: [18:03:49] <jwtodd> % cd usr/src/uts [18:03:49] <jwtodd> % dmake all [18:03:49] <jwtodd> dmake: warning: Can't find `/Makefile.noget': no such file or directory [18:03:49] <jwtodd> dmake: Fatal error in reader: .../Makefile.master, line 1007: Read of include file `/Makefile.noget' failed [18:03:50] <twincest> okay [18:03:51] <jwtodd> thoughs? thx. [18:04:02] <twincest> jwtodd: the readme doesn't tell you do 'dmake all' does it? [18:04:04] <jwtodd> first time solaris builder [18:04:12] <twincest> i thought it used nightly [18:04:23] <johnlev> twincest: that's an old old file... [18:04:27] <twincest> johnlev: hm, any idea what category that would be in? [18:04:38] *** dunc_ has quit IRC [18:04:40] <johnlev> twincest: on b.o.o? no :/ [18:04:56] <johnlev> but it doesn't matter much [18:04:59] <twincest> consolidation:os-net-misc looks reasonable [18:05:05] <jwtodd> up above at the toplevel i do "nightly ./opensolaris.sh &" and let it run [18:05:27] <jwtodd> and i just want a kernal image (iso) to write to cd [18:05:28] *** Snake007uk has quit IRC [18:05:29] <twincest> jwtodd: nightly builds the kernel, you don't need to do it yourself [18:05:40] <jwtodd> YES! [18:05:44] <jwtodd> ok ... good ... good. [18:05:47] <twincest> jwtodd: if you want to run dmake for whatever reason though, you can. but you have to run "bldenv -d myfile.sh" first [18:05:56] <jwtodd> now, where the heck is it? [18:05:56] <twincest> (it sets up the environment dmake for ON needs) [18:06:02] <jwtodd> cool. thx. [18:06:13] <twincest> jwtodd: in proto/root_$(mach) [18:06:41] <jwtodd> toplevel? in /export/testws? [18:06:46] <twincest> yes [18:07:06] <jwtodd> searching for proto [18:07:12] <jwtodd> THERE SHE BE :) [18:07:17] <jwtodd> phewwww [18:07:29] <twincest> nightly runs dmake install which "installs" the system into proto/ [18:07:39] <twincest> sort of like DESTDIR=/proto make install for a normal makefile [18:07:57] <jwtodd> k. one more if i may ... what do i need to write to cd to boot disk of this build? [18:08:26] <jwtodd> i better mention what i'm trying to do. [18:08:28] *** boxix has quit IRC [18:08:30] <twincest> i have no idea about that. i don't even know if you can boot ON on its own.. i usually install it over SX [18:08:40] <jwtodd> ok ... [18:09:23] <jwtodd> now, if on the same box i want to run this kernel ... do i simply cp -r proto/root_i386/* to / [18:09:29] <jwtodd> and restart? [18:09:36] <twincest> no! [18:09:52] <twincest> http://opensolaris.org/os/community/on/devref_toc/devref_5/ [18:09:58] <jwtodd> full brakes on .............................. screechhhhhhhh [18:09:58] <twincest> scroll down to BFU [18:10:18] <jwtodd> 5.1.2 [18:10:21] <twincest> (read the warnings carefully - once you bfu, you can never do a normal upgrade) [18:10:33] <twincest> oh, i meant 5.3 [18:10:37] <twincest> down further :) [18:11:19] <jwtodd> well now ... intersting. [18:11:26] <jwtodd> i'll read this. thx. [18:11:48] <jwtodd> now, upleveling a bit .. are prebuilt amd64 bins available somesomewhere? [18:11:59] <jwtodd> or am i going the right route? [18:12:13] <twincest> http://dlc.sun.com/osol/on/downloads/current/ [18:12:16] <twincest> 'ON BFU archives' [18:12:34] <twincest> or can use SX or one of the other binary distributions, which include ON along with the other stuff [18:12:58] <twincest> whether you're going the right route depends on what you're trying to do [18:13:11] <jwtodd> i didn't see an amd64 bin when looking [18:13:17] <twincest> amd64 is part of x86 [18:13:35] *** Cyrille has quit IRC [18:13:41] <jwtodd> interesting. so i have it built in already then? [18:13:48] <twincest> yes [18:13:49] <jwtodd> perhaps i'm over thinking this. [18:13:56] <jwtodd> well, that's good news. [18:14:04] <twincest> there is no amd64 version of solaris. x86 has both 32-bit and 64-bit, the same as the SPARC version [18:14:09] <jwtodd> can i confirm i'm running a 64b os? [18:14:27] <twincest> 'isainfo' should return 'amd64 386' on a 64-bit system [18:14:38] <jwtodd> YES!!!! [18:14:44] <jwtodd> i have arrived. the promised land. [18:15:00] <jwtodd> k twincest. nothing to see here. our work is done. [18:15:05] <jwtodd> much appreicated. [18:15:39] <jwtodd> i went the SX route iirc, thinking i'd have to build a amd64 build. looks like i am done and can now swim back up the stack to app land. [18:15:42] <jwtodd> thx again!!! [18:15:51] <jwtodd> happy to have solaris in our house :) [18:15:54] <twincest> np [18:16:32] <hile_> hiya twincest [18:16:35] <jwtodd> good day all. thx again! [18:16:38] <twincest> evening [18:18:08] * hile_ goes to run some errands [18:19:25] <twincest> hm, there needs to be a more direct way to enter bugs so i don't have to wait for the triage queue before requesting a sponsor :) [18:19:41] *** ndroux has quit IRC [18:21:09] *** bougie has joined #opensolaris [18:21:10] *** ndroux has joined #opensolaris [18:21:12] <johnlev> twincest: no kidding... [18:21:48] <twincest> (i also feel sorry for the person who has to process the 10 rfes i submitted today) [18:21:50] <bougie> hello world :) [18:24:38] *** pizdec has joined #opensolaris [18:26:38] <Stric> hrm. about what time would a sx featuring current (after b51) zfs code be released? mostly interested in the zfs_prefetch_disable option cause it's killing our performance :( [18:27:11] <twincest> i think the current code is b53, but i don't know why i think so, that i could be making it up [18:27:17] <twincest> which would mean about a month [18:27:40] <Stric> topic doesn't claim any 53's :) [18:27:47] <twincest> it's not 53 yet, i mean it will be [18:27:53] *** mega has quit IRC [18:27:58] <Stric> 52 was dumped? [18:28:28] <twincest> no, but a build lags behind the current ON code [18:28:37] <twincest> (but as i say, i could be wrong about the exact build) [18:29:14] *** alanc_away is now known as alanc [18:29:17] <alanc> http://www.opensolaris.org/os/community/on/onnv_schedule.txt [18:29:18] <twincest> http://www.opensolaris.org/os/community/on/onnv_schedule.txt agrees with me though [18:29:25] <Stric> :) [18:30:24] <alanc> the dock for delivering packages to the build 52 ISO's closed yesterday, so the ISO's should have been built last night and be going through sanity testing today before release inside Sun Wed or Thursday and to the community about a week later [18:31:14] <Stric> it's the 3rd from bottom at http://dlc.sun.com/osol/on/downloads/current/on-changelog-20061103.html I'm very interested in.. [18:31:24] <alanc> JDS has built and published their build 53 packages - ON code freeze will be next week [18:31:54] <alanc> desktop consolidations code freeze the week before ON, which is the week before the WOS/ISO dock closes [18:32:08] <twincest> Stric: live on the edge, BFU it ;-) [18:32:19] *** hile_ has quit IRC [18:32:39] <Stric> twincest: if it blows up, cdimage.debian.org is no more :P [18:33:03] <twincest> that's okay, the debian users can BFU it too [18:33:22] <Stric> that's kind of the usual package method anyhow.. [18:37:58] *** RaD|Tz has joined #opensolaris [18:39:26] <twincest> stric: is ftp.acc.umu.se meant to display welcome.msg when one connects? [18:39:58] *** alanc is now known as alanc_away [18:40:07] <Stric> it does through http [18:40:34] <twincest> oh, a web 2.0 ftp server ;-) [18:40:50] *** eboutilier has quit IRC [18:40:53] <Stric> ftp sucks for file transfers anyway.. [18:41:08] <Stric> we get mails every now and then about we being evil and doing portscans on their computers :P [18:41:13] <Stric> (it's called "active ftp") [18:41:52] <g4lt-U60> on the contrary, it shows who the idiots are [18:41:53] <quasi> Stric: like we used to get at $ISP for hacking people on port 53 ;) [18:42:55] <Stric> quasi: oh you evil bastard :) [18:43:22] <twincest> is there a mailing list problem? casper replied to my message on request-sponsor but i haven't seen the follow, even though it's available via jive [18:43:48] <quasi> Stric: we did debate just following peoples requests and stop responding to them from our dns'es [18:43:56] <Stric> quasi: :) [18:44:15] *** logic_ has quit IRC [18:44:23] <Stric> quasi: we tell the people who complain about the same.. "it's because you contacted us.. stop doing that if you want to.." [18:44:38] *** damienc has quit IRC [18:45:04] <quasi> Stric: we dropped the idea because we'd end up spending too much time adding ips to the filter ;) [18:45:15] <Stric> :P [18:46:30] *** LivingCorpse has quit IRC [18:47:06] *** yippi has joined #opensolaris [18:59:45] *** hile_ has joined #opensolaris [19:00:56] *** logic_ has joined #opensolaris [19:11:30] *** slowhog has joined #opensolaris [19:16:58] *** yippi has quit IRC [19:17:43] *** axisys has joined #opensolaris [19:19:36] *** yippi has joined #opensolaris [19:22:21] *** yippi has quit IRC [19:22:46] *** solarisjon has joined #opensolaris [19:30:55] *** lacaAFK is now known as laca [19:33:38] *** trygvis_ is now known as trygvis [19:38:52] <Error_404> *yawn* [19:39:28] *** Azureus has joined #opensolaris [19:39:35] *** alanc-away is now known as alanc [19:40:21] *** sylvain has left #opensolaris [19:42:18] <Azureus> does anybody knows if b52 of osol should be out today? [19:43:10] <Error_404> man would that be annoying [19:43:18] <Error_404> i just moved to b51 last night [19:43:31] <delewis> Azureus: if you really need a certain putback really bad, just grab a working copy of the repo. [19:43:55] <Error_404> mind you, I moved from b43 or some such so I don't really follow builds very much [19:44:11] <Azureus> ok.ok. just wanted to know. [19:44:18] <sickness> I update at every dvd iso update :P [19:44:26] *** slowhog has left #opensolaris [19:52:21] *** Drone has quit IRC [19:53:09] <alanc> Azureus: not even people inside Sun will get build 52 ISO's today [19:53:43] *** calumb has quit IRC [19:54:22] *** Drone has joined #opensolaris [19:54:51] <Azureus> alanc, thnks for the info. [19:55:27] <alanc> quoting my answer from earlier today: <alanc> the dock for delivering packages to the build 52 ISO's closed yesterday, so the ISO's should have been built last night and be going through sanity testing today before release inside Sun Wed or Thursday and to the community about a week later [19:56:30] *** sommerfeld_work has joined #opensolaris [19:57:00] <Azureus> alanc, I was not aware of this answer (I assume in this IRC); thnks again [19:57:09] <alanc> it was before you logged in [19:57:22] *** sommerfeld_work has quit IRC [19:57:34] <alanc> (you weren't the only one to ask today) [19:58:47] <Azureus> alanc, maybe it should be consiedered putting it in the banner because I assume there are many who are interested [19:58:58] <Azureus> because for example [19:59:23] <Azureus> I hesitated between using the nightly from 3 of november or to wait to b52 [20:00:22] <alanc> oh, do you mean build 52 code drops or bfus? that's a different story - I was talking SX:CR ISO images [20:00:51] <twincest> the current available code is already later than what will be in b52 [20:01:01] <Azureus> alanc, I am talking about opensolris code , not SXCR [20:01:17] <alanc> sorry, confused - ignore me then, ask stevel [20:01:28] <Azureus> alanc, ok, thnks [20:03:54] <alanc> it's already one of those days - I should just go back to bed [20:05:31] <twincest> is he even making build-specific code drops anymore? [20:12:24] *** yippi has joined #opensolaris [20:14:52] *** axisys has quit IRC [20:15:23] <Error_404> oh fer f* sakes... [20:15:48] <Error_404> I've had a domain in my name for all of what, a week now? I'm already getting telemarketing calls from hosting vendors [20:26:06] <sommerfeld> actually, new registrants are more likely to be looking for hosting. kinda like how you get spam after buying a house for heating oil, etc., [20:29:26] <jbk> grr.. why do I have a feeling i'm hitting yet another ASL bug w/ veritas.. [20:35:50] <jbk> and we have a winner [20:35:51] <jbk> libvxemc.so:claim_device() : SKIPPED [20:42:37] <|tsoome|> ASL's are written by HW vendors [20:42:47] *** RaD|Tz has quit IRC [20:43:57] <jbk> well i know what the bug is, but no one can give me a straight answer as to how it's triggered [20:44:34] *** Teltariat has joined #opensolaris [20:44:47] <jbk> the bitmasks they use against the inquiry data are wrong, and in some instances misidentify regular luns as 'special' luns (i.e gatekeeper or vcmdb devbices) [20:51:15] * jbk anxiously awaits rootable zfs :) [20:51:39] <timeless> is that coming in 52? i suppose you don't like the existing hacks? [20:52:05] <twincest> why is the only manual page for va_start for the kernel DDI version? [20:52:16] <jbk> well [20:52:23] <jbk> 1. i need sparc support [20:52:37] <jbk> 2. it needs to be integrated into a regular solaris 10 update before i can use it [20:53:00] <hile_> man -s 3c va_start ? [20:53:09] <twincest> No entry for va_start in section(s) 3c of the manual. [20:53:15] <quasi> jbk: don't hold your breath - I saw someone saying perhaps U5 ... and we're not even at U3 yet [20:53:20] <jbk> twincest: try man stdarg [20:53:35] <twincest> ah, thanks [20:53:40] <jbk> i had heard (and hoped) U4... but.. [20:54:02] <timeless> swift% plocate va_start [20:54:03] <timeless> /usr/share/man/man9f/va_start.9f [20:54:08] <timeless> that's not helpful? [20:54:15] <twincest> timeless: that is the DDI documentation [20:54:16] <jbk> that's kernel, not userland i believe [20:54:19] <timeless> ah [20:54:22] <jbk> the userland docs are under stdarg [20:54:33] * timeless needs to learn all the numbers and letters :( [20:54:50] <timeless> yeah, stdarg(3EXT) [20:54:53] <quasi> jbk: you'd need grub on sparc first [20:55:12] <jbk> though dunno why at least man -s 3ext va_start doesn't bring it up... [20:55:30] <jbk> ick :( [20:55:49] * timeless looks for the magic to get windex's [20:56:00] <alanc> catman -w [20:56:01] <|tsoome|> catman -w [20:56:05] <|tsoome|> :) [20:56:29] <timeless> swift% ssh root@localhost /usr/bin/catman -w [20:56:31] * timeless waits [20:57:50] <quasi> timeless: it isn't the fastest thing in the world [20:58:28] <jbk> not too bad here.. [20:58:41] <timeless> quasi: it didn't work very well [20:58:44] <jbk> > time catman -w [20:58:44] <jbk> real 16.5 [20:58:44] <jbk> user 2.1 [20:58:44] <jbk> sys 3.4 [20:58:45] <timeless> it didn't index a bunch of places [20:58:55] <quasi> ah, right - takes a lot less with -w [20:58:59] <timeless> root's man path is a lot smaller than my user's [20:59:12] *** Teltariat has quit IRC [20:59:19] <twincest> also, is it just me, or is b.o.o being a bit sluggish today? [20:59:22] *** yippi has quit IRC [20:59:43] *** yippi has joined #opensolaris [21:01:15] <PerterB> I think the stdarg(3EXT) thing came about to distinguish ANSI vargs from K&R (both have a va_start) [21:01:31] *** sahafeez has joined #opensolaris [21:01:32] <PerterB> but that distinction is pretty obsolete these days [21:01:44] <twincest> you'd expect va_start to be linked to either K&R or ISO documentation though [21:02:26] *** nwf has joined #opensolaris [21:04:16] <quasi> I find it in Kernel Functions for Drivers va_arg(9F) [21:09:09] *** nwf has quit IRC [21:09:12] *** Tpenta has joined #opensolaris [21:09:32] *** ShadowHntr has joined #opensolaris [21:09:41] *** ChanServ sets mode: +o Tpenta [21:12:07] *** Symmetria has joined #opensolaris [21:12:10] <Symmetria> hey all [21:12:21] <Symmetria> hrm [21:12:50] <Symmetria> am I correct in my thinking that this line will run the script every 6 hours: [21:12:51] <Symmetria> 0 0,6,12,18 * * * /mirror/mirror_scripts/mozilla.sync.sh [21:12:57] <Symmetria> in terms of cron [21:13:11] <elektronkind> yah [21:13:24] <Symmetria> hrm its spawning more often than that, so Im confused [21:19:34] <PerterB> have a look in /var/cron/log to be sure, and check other users' crontabs [21:20:27] <twincest> i wonder what happened with that rctl bug that broke cron [21:20:34] <twincest> there was a brief discussion about it, then it disappeared [21:25:11] *** jwt has joined #opensolaris [21:25:31] <jwt> hey hey ... running a shiny new sxb50 and happy. [21:26:01] <jwt> suggestions on how best to pimp out this build in the same spirit as i do w/ ubuntu and apt? [21:26:49] <twincest> you want your build to have sex with people for money? [21:26:58] <jwt> but of course :) [21:27:01] <jwt> noooooooooooooo [21:27:12] <jwt> i want to add in sudo, etc. [21:27:23] <jwt> is blastwave the place? [21:27:40] <twincest> it's okay. i prefer to build things myself [21:27:41] <jwt> guess i should say i want to bling out my sol build :) [21:27:49] <jwt> pimp'n == bling in my dictionary though [21:27:51] *** bobbyz has quit IRC [21:27:52] <delewis> jwt: you really don't need sudo in Solaris. [21:28:00] <delewis> Solaris has a least-privilege model. [21:28:01] <jwt> ahhhhh .... do tell [21:28:35] <jwt> k. i can live w/ that. it is nice though to sudo for only one command when needed. no? [21:28:36] <elektronkind> an random privilege model would be more fun [21:28:46] <delewis> http://www.samag.com/documents/s=7667/sam0213c/0213c.htm [21:29:14] *** maitrella has quit IRC [21:29:19] <delewis> jwt: no reason installing third-party freeware on your system when there's a Sun-supported solution that's far more comprehensive. [21:29:26] <hile_> delewis: true. [21:29:52] <jwt> i'm down w/ that. [21:30:17] <quasi> delewis: rather old link [21:30:32] <jbk> also, if you put your stuff under smf, you can do some really neat stuff with rbac [21:30:43] <jbk> instead of sudo [21:30:47] <jwt> i'll just add what i need when i need one by one then. that will work. just got used to ubuntu (syn)apt(ic) goodness. [21:30:59] <delewis> ugh. [21:31:02] <delewis> docs.sun.com is broken [21:31:13] <delewis> otherwise I'd give you a link to the document that contains RBAC documentation. [21:31:25] <jwt> yeah. we *really* want to try out all the sol10 goodness. [21:31:47] <jwt> fwiw, i used to work at sun for quite some time ('92-06 mod 1 dot.com year). [21:31:50] <jwt> i'm a fan! [21:32:02] <jwt> and believe mo *nix is a good thing for all. [21:32:16] <delewis> jwt: the RBAC model originally came from Trusted Solaris, which is the Solaris version with all the high-end security features. [21:32:17] <jwt> and now we have sol in the house! time to unleash the beast [21:32:23] <jwt> NICE!!!! [21:32:41] <jwt> wow, i heard that stuff was rolling down. excellent! [21:32:59] <delewis> it's been rolling down since Solaris 8 and 9 [21:33:05] <twincest> does current ON have the new TX stuff? [21:33:06] <jwt> not that it matters much ... this is/was me ... http://blogs.sun.com/gonzo [21:33:20] <delewis> twincest: it should.. [21:33:30] <delewis> the latest SX releases have TX [21:33:52] <jwt> k. so in general, get used to smf/rbac as a starting point? when coming from a linux model? [21:33:54] <delewis> you can look at the flag day page on OpenSolaris.org to see whenever it was integrated [21:34:18] <delewis> jwt: Linux has something like RBAC but's overly complex and doesn't work very well, period -- SELinux [21:34:25] <delewis> most vendors tend to stay away from it. [21:34:45] <delewis> actually, SELinux is closer towards TX than just RBAC [21:34:49] <twincest> hm... i don't have the impression that selinux is like rbac [21:34:55] <delewis> but regardless, RBAC is the way to go in the real world. [21:34:56] <twincest> linux's rbac would probably be sudo :) [21:35:01] <hali> i don't think anyone outside the nsa has actaully ever use selinux in prod [21:35:02] <delewis> twincest: yeah :-) [21:35:22] <delewis> hali: Red Hat is at least shipping an SELinux enabled kernel nowadays, but it's not turned-on by default. [21:35:23] <twincest> linux has had TX-type stuff long before SELinux though - RSBAC was doing it years ago [21:35:26] <jwt> just looking for a relatively smooth ops path for all *nix boxen yet play w/ sol above and beyond where we can. i don't intended to unify the stacks. more train the fingers. [21:35:32] <twincest> for some reason the selinux people reimplemented it from scratch [21:35:48] <delewis> twincest: and Trusted Solaris was around way before then ;-) [21:35:59] <dwc-> delewis: did they turn that off by default now? The RHEL and CentOS boxes have been coming with it on .... [21:36:29] <delewis> dwc-: well, it may be turned on, but there's no policies by default [21:36:50] <dwc-> hm, on mine there were [21:37:02] <delewis> dwc-: what RHEL release are you using? [21:37:04] <dwc-> like... when I replaced the mysqld from their rpm with a new version, mysqld stopped working [21:37:07] <twincest> hmm, i think there's a bug in libtol(3LIB) - it formats with a vertical line through the middle of half the text [21:37:10] <delewis> ugh. [21:37:26] <dwc-> because removing the redhat rpm'd version removed the selinux rules [21:37:26] <delewis> you have to love privilege models that break backwards compatibility. [21:37:50] <delewis> or require the application to be aware of it [21:37:59] <delewis> dwc-: ouch. [21:38:05] <dwc-> rhel4, upd 1~3 or so [21:38:24] <elektronkind> SEpoop [21:38:24] <dwc-> it only bit us a couple times ... then we just turned selinux off by default :p [21:38:32] <delewis> like most sane people do [21:38:57] *** pizdec has quit IRC [21:39:54] *** pizdec has joined #opensolaris [21:39:59] <jwt> thx for the rbac link ... i'm convinced. good day all. [21:40:33] *** bobbyz has joined #opensolaris [21:41:07] *** bengtf has quit IRC [21:41:50] *** bengtf has joined #opensolaris [21:44:48] *** pizdec has quit IRC [21:48:13] *** jwt has quit IRC [21:48:32] <hile_> ugh seLinux [21:48:57] <hile_> i don't even know anyone who uses that on machines that are NOT UNCLAS machines [21:54:29] *** ShadowHntr has quit IRC [21:55:58] *** yippi has quit IRC [21:58:30] *** GmanAFK is now known as Gman [21:58:33] <Gman> hey hey hey! [21:59:13] *** Symmetria has quit IRC [22:01:55] *** djgregor has left #opensolaris [22:02:00] <twincest> hey gman [22:02:28] * twincest files a 7-line bug reports for a 1 character change [22:02:49] <Gman> twincest, did you get that bug id notification mail yet? [22:02:53] <twincest> nope [22:02:59] *** mrcranky_ has quit IRC [22:03:01] <twincest> nor any of the others i reported today :( but i'm patient [22:03:03] <Gman> ok bugster is up again, so i'll check [22:05:31] <Gman> twincest, can you msg me the summaries of them? [22:05:48] <Gman> or just provide your email address so i can cross check? [22:05:49] <twincest> uhm. i don't remember the exact summaries. but i can suggest some keywords :) [22:05:54] <twincest> river at attenuate dot org [22:05:57] <Gman> ok, cool [22:06:10] <Gman> 6491015 [22:06:38] <Gman> 6490985 [22:07:02] <Gman> 6490777 [22:07:16] *** hell` has joined #opensolaris [22:07:26] <Gman> 6490754 [22:07:45] <Gman> 6490721 [22:08:07] <Gman> 6490715 [22:08:31] <Gman> that's the current ones - have added you to the interest list [22:08:37] <Gman> now need to hunt out your other one [22:08:51] <twincest> great, thanks :) [22:10:38] <Gman> 6490935 is the passwd one - have added you on that too [22:10:43] <Gman> i think that's everything? [22:12:37] * twincest waits for diff to run.. [22:12:38] <sahafeez> Gman, drag and drop - do not care anymore. selling the ultra ;) [22:12:41] <twincest> hg = not super speedy [22:12:51] *** mrcranky_ has joined #opensolaris [22:12:52] <Gman> sahafeez, hahaha [22:13:39] <sommerfeld> twincest: it's immensely faster than teamware [22:13:55] <sommerfeld> twincest: more like "ON is huge" [22:13:57] <sahafeez> speaking of which anyone want to buy an ultra 80 fully loaded? [22:14:06] <twincest> heh, true [22:14:57] <twincest> gman: hmm, i had changes for 10 commands: wc, touch, passwd, awk, mv, grep, eqn, diff3, cut, chmod [22:15:00] <twincest> which did i miss? :) [22:15:48] <LeftWing_> sahafeez: AU$25? [22:16:06] <Gman> 6490925 [cp/mv/..] [22:16:14] <Gman> twincest, seemingly some were already triaged [22:16:19] <sahafeez> what is that? like $2 [22:16:28] * LeftWing_ slaps sahafeez. [22:16:29] <twincest> ah, okay, i thought your list was for both [22:16:32] <LeftWing_> You take that back! =) [22:16:43] <sahafeez> haha [22:16:47] <Gman> 6490919 - strange file... [22:16:56] <sahafeez> $300 us + shipping [22:17:02] <twincest> (i can't see any of these on b.o.o, you see) [22:17:11] <Gman> 6490855 awk [22:17:15] * LeftWing_ runs away to an exam. [22:17:19] <sahafeez> 4x450, 4gb ram, 2 x elite 3d (m6&m3) [22:17:41] <Gman> 6490848 = awk/nawk error msgs [22:17:42] <dunc> would anyone be up for helping me try and diagnose a strange SATA problem? I'm not sure if it's a driver problem or what, but I'm stumped. I have some results from testing various things, but not as much as I would like because all my spare tat seems to have dies now [22:17:59] <dunc> s/dies/died/ [22:18:03] <Gman> 6490789 = grep [22:18:26] <Gman> 6490780 = cut [22:18:28] <sahafeez> i am going to try to get x86 on my dual p3 tonight. be ready to help as it blew up last time ;) [22:18:39] <Gman> ok, got the others [22:19:05] <dunc> it's a card listed as supported on bigadmin, but it hangs whenever i connect both disks. one at a time is fine, but both together and it doesn't even make it through a normal boot [22:19:30] <dunc> seems to work totally fine on amd64 however, but i can't for the life of me find another working 32bit box here to test with [22:19:49] *** dwc- has quit IRC [22:19:52] *** bl4ckth0rn has joined #opensolaris [22:19:53] <sahafeez> bad cable [22:20:09] <dunc> nah either disk on either channel is fine by itself [22:20:14] <dunc> i tried all 6 permutations [22:20:33] <dunc> well 7 if u include no disks :) [22:20:36] *** Tpenta has quit IRC [22:21:53] <dunc> if i boot -v I can see it find the 1st disk, and then just hang there (the box doesn't lock up though, it just never goes any further) [22:22:01] <dunc> i think it's when devfsadm runs [22:22:12] <dunc> hence i was wondering if it is a driver problem [22:24:31] <dunc> i was quite peeved at having bought a supported card and it not work, but now that I know it works in a 64bit box, I'm wondering if it's a 32bit specific thing, because the place where I saw it announced that this card works, the guy is talking about on 64bit [22:24:35] *** Symmetria has joined #opensolaris [22:26:35] <twincest> hmm, wonder how long until b.o.o updates now :) [22:27:30] <Gman> twincest, yeah, the bug database does seem to have had a beating over the last couple of days [22:28:07] <alanc> the "reporting" bug db crashed and is still resyncing to the live db - our internal web tools for viewing and the email notifications are out of sync too [22:28:39] <alanc> (the bug db is kind of like the ON gates - a master for changing, and a read-only clone for everyone else to read from) [22:29:27] <timeless> it's kinda funny [22:29:36] <timeless> i'm used to systems where the readonly clones are *fast* [22:30:06] <timeless> since you have fewer people w/ write bits, you can afford to have fewer machines dealing w/ the writers [22:30:34] <timeless> does the bug db sync constantly or 'nightly'? [22:30:41] <alanc> it usually is, but the resync is apparently slow when people are constantly writing to the master [22:30:49] *** Azureus has quit IRC [22:31:05] <alanc> the internal clone syncs every 5 minutes or so [22:31:15] <timeless> locking is so much fun :) [22:31:21] <alanc> I think bugs.os.o syncs to that at a much slower rate though [22:34:01] *** Xaqueth has quit IRC [22:34:12] *** mrdeviant has quit IRC [22:34:16] <twincest> sounds a lotlike mysql replication ;-) [22:37:28] <kleppari> doesn't sun plan to reveal comments and such on bugs any time soon now? [22:37:38] <twincest> who knows [22:37:47] <Gman> we plan to - just not sure when it's going to happen [22:38:15] <alanc> someone mentioned a bugs.os.o refresh coming soon - not sure how much it will have though [22:38:22] *** bl4ckth0rn has left #opensolaris [22:38:27] <Gman> yeah, seemingly next month [22:38:36] <Gman> no idea what it contains either [22:38:50] <twincest> it would be really helpful even if more info was available for news bugs (+ the ability to comment on them) [22:39:07] <kleppari> I've often found myself without a project, fixing bux is a nice way to occupy time.. [22:39:24] <kleppari> and bugs.os.o isn't good enough :/ [22:39:41] <Gman> kleppari, anytime you want more info, let me know and i can provide it in the meantime [22:39:47] <Gman> [obviously not the ideal situation...] [22:39:56] <kleppari> Gman, thanks, i'll keep that in mind :) [22:42:12] *** gisburn has joined #opensolaris [22:42:22] <gisburn> !seen kupfer [22:42:25] <Drone> kupfer (kupfer!i=kupfer@nat/sun/x-f72355ceab7909f5) was last seen in #opensolaris on Tue 31 Oct 2006 00:41 GMT, saying 'gisburn, sorry, "beats me" means "I have no idea"'. [22:42:37] <gisburn> umpf [22:42:43] <gisburn> f*ck [22:43:04] <twincest> okay, the retry time on sun's outgoing smtp servers is insane [22:43:27] <twincest> i just got a mail sent about like 6 hours ago [22:44:12] * gisburn grabs twincest and strangles him for fun to vent his anger [22:44:23] <gisburn> grrrr [22:44:42] <hile_> that's some feat, Roland :) [22:44:43] <gisburn> Gman: is kupfer around ? [22:45:16] <Gman> email him [22:45:57] <gisburn> Gman: is he still sick@home%with:flu ? [22:46:09] <Gman> i have no clue, i'm not his mum. [22:46:14] <dunc> lol [22:46:20] <gisburn> heh [22:46:59] <gisburn> Gman: didn't you made some cookies for him recently ? [22:47:51] <alanc> Gman's about as far away from kupfer as you are [22:48:31] <gisburn> alanc: I know.... I am just trying to tease hima little bit... with zero effect... ;-( [22:49:24] [22:51:02] <gisburn> ;-( [22:54:30] *** axisys has joined #opensolaris [22:55:14] <axisys> i attached a 3510 to the T200.. but dont see it in format [22:55:27] <axisys> booted w/o -r tho [22:55:32] <axisys> i guess that is the problem [22:56:29] <Stric> hrm.. btw.. anyone know if there's some way to tell zfs not to use any/so much data cache? I'd prefer more metadata cache instead.. [22:58:53] <quasi> axisys: devfsadm -v [23:00:22] <gisburn> Stric: does zfs support directio ? [23:00:29] <axisys> quasi: oh yeah i could have done that.. duh! [23:00:33] <axisys> quasi: thnx [23:00:48] <Stric> gisburn: hm. dunno.. [23:01:22] <quasi> axisys: remember you're not on windows anymore ;) [23:01:29] <axisys> quasi: yep :-) [23:01:41] <axisys> quasi: check this out i still dont see the 3510 [23:02:01] <axisys> quasi: i have a fiber card and a connection to the 3510 [23:02:03] <axisys> hmm [23:02:20] <quasi> luxadm probe [23:02:31] <g4lt-U60> waht's the canonical reading toool for [w|u]tmpx? [23:03:14] <axisys> No FC devices found [23:03:16] <axisys> hmm [23:03:31] *** yarihm has joined #OpenSolaris [23:03:43] <axisys> shit i have not run the fc cable yet [23:03:48] <axisys> dont laugh [23:04:00] <quasi> too late ;) [23:04:58] <gisburn> axisys: maybe we should thank someone that you're not responsible for the fly-by-wire system in the A380... =:-) [23:05:08] <trygvis> hehehe [23:05:23] *** bougie has quit IRC [23:05:34] <gisburn> tower: A380 took off for the maiden flight [23:05:53] <quasi> gisburn: fly by wireless - it's the way of the future ;) [23:05:54] <gisburn> plane: uhm... someone forgot to run the control cable yet [23:05:59] <dunc> g4lt-U60, last i think [23:06:20] <gisburn> quasi: yeah, that wold be good - crash-by-handy [23:07:08] <quasi> gisburn: at this rate, who knows if they will ever finish any A380s before their customers cancel the orders [23:07:22] <Stric> gisburn: I think it will be difficult to persuade nfsd to use directio though :) [23:07:46] <gisburn> Stric: nfsd does not use directio unless the nfs client requests it. [23:08:06] <gisburn> Stric: but I was thinking about the directio mount option [23:08:20] <twincest> forcedirectio does not work with zfs [23:08:27] <twincest> (does it?) [23:09:05] <Stric> neither zfs(1M) nor zpool(1M) mentions any direct io.. [23:09:40] <gisburn> ZFS, the latest and best development in filesystem technology [23:09:44] <gisburn> lacks quotas [23:09:47] <gisburn> lacks directio [23:09:54] <gisburn> lacks proper resource usage [23:09:56] <Stric> prefetches too much :) [23:10:07] <gisburn> lacks support for storing metadata on seperate disks [23:10:13] <twincest> it's not correct to say it lacks quotas, they're just different [23:10:16] <gisburn> ... should I continue ? [23:10:28] <AbeFroman> but it contains 200% more awesomeness [23:10:30] <dlg> gisburn: if it sucks for you then dont use it [23:10:35] <richlowe> metadata separation is pretty specific to SAM-FS [23:10:38] <richlowe> if you want SAM-FS, use SAM-FS [23:10:45] <jbk> hey at least it doesn't corrupt itself when you fill it up 100% [23:10:46] <richlowe> (if I'm confusing samfs and qfs, I'll accept that...) [23:10:51] <jbk> *cough*vxfs*cough* [23:11:10] * Stric prefers the admin interface to zfs/zpool over newfs/metadb/metainit.. kinda hard not to :P [23:11:13] <gisburn> twincest: if multiple users share one disk (which is common, very common) then per-user quotas are almost the only option [23:11:21] <richlowe> *filesystem* [23:11:22] <richlowe> not *disk* [23:11:39] <gisburn> richlowe: even filesystem [23:12:06] <richlowe> I was merely pointing out that 'disk' means absolutely nothing as far as this goes. [23:12:14] <twincest> um, i have multiple users sharing one disk and ZFS quotas would work fine [23:12:41] <sommerfeld> and, really, per-user quotas break down horribly when there are multi-user projects. [23:12:46] <richlowe> The biggest thing I've heard around this is mail. [23:13:10] <richlowe> In that it's not necessarily trivial to pump every mbox to a different FS. [23:13:24] <Stric> symlinks? [23:13:25] * Stric runs [23:13:33] <sommerfeld> well, /var/mail [23:13:36] <twincest> there are people who still don't do delivery-to-homedir? :) [23:13:40] <dlg> richlowe: which type of mail? [23:13:58] <Stric> I'm guessing regular /var/mail/$UserMbox [23:14:03] <dlg> ultimately a users mail goes into their own directory [23:14:15] *** kmaraas has joined #opensolaris [23:14:21] <sommerfeld> fixed by either delivery-to-homedir or by alternate mail services (imap, etc.,) [23:14:22] <dlg> and by own, i mean a separate directory somewhere owned by them [23:14:45] <dlg> i agree with sommerfeld [23:14:50] <richlowe> Yeah, but that's still the only quota-ish complaint I can somewhat understand. [23:15:08] *** d3vi1 has joined #opensolaris [23:15:21] <dlg> im more upset about quotas including the space taken by snapshots [23:16:10] <d3vi1> hi guys. Can anyone help me with a bug? [23:16:15] <d3vi1> I forgot about-it [23:16:20] <twincest> no. there is to be no helping with bugs. [23:16:24] <d3vi1> and I need to contact the owner [23:16:34] <d3vi1> it's already in the bugs.opensolaris.org db [23:16:57] <d3vi1> it's from the 0606 beta [23:17:01] <axisys> gisburn: i need one of those fly by wireless [23:17:02] <gisburn> sommerfeld: fact is that per-user+per-group quotas are mandatory features in many environments and the lack of this feature makes many people shake their head and drop the ZFS idea. [23:17:04] <sommerfeld> dlg: yeah. i have a half-composed rant about physical vs logical quotas and the like. [23:17:22] <dlg> sommerfeld: im half listening [23:17:23] <sommerfeld> i.e., quotas before vs after compression [23:17:24] <dlg> ;) [23:17:53] <sommerfeld> who pays for redundancy, who profits from compression, etc., [23:18:00] <dlg> i can see arguments both ways [23:18:24] <d3vi1> and was moved to the bugtracker by someone at sun... but with incorrect information... IIRC, only the owner (on the sun side) can modify that bug [23:18:36] <sommerfeld> d3vi1: do you have a bugid? [23:18:41] <d3vi1> sure [23:18:59] <d3vi1> 6406183 [23:19:11] <dlg> sommerfeld: the problem is that you know too much [23:19:14] <dlg> and so these things matter [23:19:30] <dlg> however, the only thing a user is really worried about is how much data they can store [23:20:22] <trygvis> is putting something in /etc/defaultrouter supposed to work in a zone? [23:20:36] <dlg> ie, they dont care if their data is being compressed or not, they just want a real number to work against [23:20:38] <delewis> trygvis: no [23:20:45] <delewis> zones have their own defaultrouters [23:21:04] <trygvis> I meant putting something in that file in the zone [23:21:24] <delewis> trygvis: in that case, yes [23:24:13] *** laca is now known as lacaAFK [23:25:23] <gisburn> dlg: I wish the zfs team would take such problems more seriously. You can't dictate the use of the filesystem in the way the current zfs does it. It limits the usage far to drastic... ;-( [23:26:08] <dlg> gisburn: like i said before, if it doesnt suit you then use something else [23:26:14] <dlg> there are choices [23:26:19] <gisburn> yes [23:26:20] <gisburn> HP [23:26:33] <jbk> haha [23:26:34] <delewis> HP? [23:26:37] * delewis chokes [23:26:38] <twincest> which filesystem does HP have? [23:26:42] <delewis> I think you at least mean IBM :-) [23:26:49] <twincest> advfs? [23:26:53] *** deedaw has joined #opensolaris [23:26:57] <delewis> twincest: VxFS and UFS [23:27:01] <gisburn> dlg: at least they now got a contract which Sun wolud have won easily if Sun germany was a little bit smarter. [23:27:02] <jbk> well i believe they oem vxfs [23:27:09] <delewis> jbk: the boot slice is still UFS [23:27:11] <jbk> then charge you through the nose for various features [23:27:12] <twincest> delewis: i coulda sworn those run on Solaris [23:27:20] <delewis> twincest: they do :-) [23:27:24] <delewis> but HP-UX bundles VxVM [23:27:25] <dlg> gisburn: that's their loss. why do you seem so bitter? [23:27:43] <richlowe> Gman: hey, thanks. [23:28:05] <Gman> richlowe, thanks for what? :) [23:28:10] <gisburn> dlg: ---> /msg [23:28:13] <richlowe> Gman: interest list. [23:28:22] <jbk> but hp-ux is dieing -- the only people buying itaniums are those forced to because they can't/won't move off hp-ux, and hp's not really doing anything with hp-ux [23:28:28] <Gman> richlowe, ahh, mails are starting to filter through then? :) [23:28:39] <delewis> jbk: it'll last a bit longer. [23:28:48] <jbk> i'm sure it will [23:28:50] <delewis> it's used quite a bit in the healtcare sector or various emergency management solutions. [23:28:54] <jbk> we still have solaris 2.5.1 systems :( [23:28:57] <richlowe> Gman: yeah, you even hit a couple I didn't directly ask for :) [23:28:59] <delewis> and until those vendors port it, it'll last :-) [23:29:06] <Gman> richlowe, oops :) [23:29:08] <delewis> I think it's safe to say every hospital at least has *one* HP-UX system. [23:29:08] <jbk> but it's not really advancing or growing [23:29:19] <delewis> we had just bought two before I left my old job [23:29:24] <delewis> and that was in 2005 [23:29:37] <kleppari> delewis, wrong, IBM machine at dad's hospital :P [23:29:49] <delewis> kleppari: I said one [23:29:50] <delewis> not all [23:30:06] <delewis> we were pretty much exclusively an IBM shop unless a vendor pushed something on us [23:30:12] <kleppari> it has no hp-uxes :P [23:30:22] <delewis> which is usually GE [23:30:46] <jbk> yeah, GE has a thing for IBM [23:30:54] <delewis> no, GE uses Solaris for imaging [23:30:59] <delewis> er, their imaging solutions [23:31:12] <delewis> either optical storage/IBM storage and Sun hardware running Solaris 8 [23:31:14] <jbk> hmm.. our ex cio thought they could do no wrong [23:31:23] <delewis> (yes, 8 with Sybase) [23:32:01] <delewis> and Veritas cluster [23:33:22] <delewis> basically, an outdated software stack that's usually mis-configured. [23:33:42] <delewis> I remember one time when maintenance was working in the datacenter to install some power strips, one of them pulled the plug on the Sun rack that contained the GE stuff. [23:33:46] <delewis> well part of it, anyway [23:34:01] <delewis> the cluster didn't failover [23:34:32] <delewis> and when the failed system powered back up, some part of the VxVM configuration was screwed. [23:35:02] <jbk> was it frozen? [23:35:03] <delewis> what should've been at most a couple of minute's worth of downtime to failover turned into 30 minutes [23:35:14] <jbk> (the cluster)? [23:35:36] <jbk> or was it just one of those app turf wars where one group wouldn't concede control of the app to the cluster? [23:35:53] <delewis> jbk: if I recall correctly, the system that failed had to be powered back on again and fixed (whatever was broken with VxVM), because the resources wouldn't transfer to the secondary node [23:36:08] <delewis> jbk: no, GE had brought the systems in and exclusively administered them [23:36:15] <trygvis> I wonder .. how am I supposed to set the default gw? [23:36:16] <sommerfeld> or was the rack power miswired so that losing one circuit killed all of some resource? [23:36:17] <delewis> no one else had authority to administer them [23:36:25] <jbk> ahh, that was my next guess :) [23:36:32] <jbk> i've run into all three scenarios [23:36:43] <delewis> sommerfeld: if that's the case, it was still GE's reponsibility because that whole section was their's, basically. [23:37:23] <delewis> this meant anyone wanting to see a patient's X-ray image, CAT scan, etc. for the duration of the downtime could not [23:37:34] <delewis> we were getting calls to the datacenter within the first minute's worth of downtime :-) [23:37:53] <delewis> 5 minutes later, the IT manager gave us a visit :-) [23:38:09] <jbk> i'm glad it's not just my company that has those issues at least :) [23:38:11] <gisburn> dduvall: ping! [23:38:32] <dduvall> gisburn: pong [23:38:52] <gisburn> dduvall: http://bugs.grommit.com/show_bug.cgi?id=36 [23:39:10] <gisburn> dduvall: I think check_rtime is somehow on drugs... ;-( [23:39:30] <alanc> but are they good drugs? [23:40:21] <sommerfeld> delewis: yes, clearly. [23:40:26] <alanc> 6488392 PSARC 2006/609: Xserver provider for DTrace - Integrated in Build: snv_53 [23:41:23] <axisys> so i can just connect 3510's SFP FC0 and FC4 to T2000's two FC card and it will give me two controllers? [23:41:26] <gisburn> dduvall: do you have any idea what I can do in this case ? [23:41:33] <axisys> i dont need to use loop [23:42:04] <dduvall> We get these from time to time. I haven't been able to figure out what's going on, and since they always go away after a build (even if they come back again), I mostly just ignore them. [23:42:10] <dduvall> That's bad, but it's what I got. [23:42:21] <dduvall> If they're coming from your code, make sure that the warnings are in fact bogus. [23:42:42] <dduvall> Not all of check_rtime's output is false positives. [23:43:12] <gisburn> dduvall: I've checked with "truss" which files are used. [23:43:36] <gisburn> dduvall: check_rtime uses /usr/bin/libcmd.so.1 instead of ${ROOT}/usr/bin/libcmd.so.1 [23:43:45] <dduvall> Oh, I see. [23:43:46] <dduvall> Hm. [23:44:09] <dduvall> You might be right that it's a bug in check_rtime, that it picks up the wrong library. [23:44:18] <dduvall> If so, you'll have to get it fixed. [23:44:39] <gisburn> dduvall: the problem goes away when libcmd from ksh93 gets installed in the underlying system [23:46:22] <gisburn> dduvall: but I guess this doesn't count as "fix", right ? [23:47:00] <dduvall> No. I'd ping Rod Evans about it -- he may be able to fix it (or suggest a fix) quickly. [23:48:45] <gisburn> dduvall: email ? [23:48:54] <dduvall> sorry: rod.evans at sun dot com [23:49:21] <alanc> I thought gisburn knew the pattern for sun e-mail addresses by now [23:49:46] <jbk> here something i've been meaning to ask for a while -- i have a small utility i've written that listens for ethernet multicast packets that are snap encapsulated.. right now, i'm doing a DL_BIND_REQ w/ a sap of 0, set the multicast addr, then push pfmod and filter to get the right packets.. anyone know if there's a better way? [23:50:22] <alanc> though I do remember having to retrain him to stop mailing me at @eng.sun.com when Sun IT forced us out of the eng domain [23:51:12] <dunc> on the HCL page, it says that my controller worked with "Solaris 10 ATA/ATAPI-6 drivers", during device configuration, it says it's found a disk which is ATA/ATAPI-7 supported. Could this be a problem? [23:51:17] <gisburn> alanc: I'd like to have a conformation before creating a bugzilla account.