[00:00:37] <dclarke> gotta run ..
[00:00:41] <dclarke> back eventually
[00:01:35] *** delewis has joined #opensolaris
[00:10:27] <dunc> PerterB, exactly the same technique didn't work on the box i want
[00:10:37] <dunc> so it's specific to that box all along
[00:10:42] <dunc> bollox
[00:11:50] <PerterB> is that to do with ssh keys or dodgy SATA controllers? :)
[00:13:25] <twincest> hm, really need an easier to use printer
[00:16:28] <sickness> dclarke: test ended :)
[00:16:32] <sickness> dclarke: system has not died :)))
[00:16:51] <dclarke> well done
[00:17:05] <dclarke> what sort of numbers did you get for the other passes
[00:17:35] <sickness> heh, really ugly :')
[00:17:42] <sickness> 238328 files  avg=0.004815 sec  total=1147.431109 sec  io_avg=0.405675 MB/sec
[00:17:45] <sickness> for the second test
[00:17:53] <sickness> and this for the third 238328 files  avg=0.002007 sec  total=478.242984 sec  io_avg=0.355966 MB/sec
[00:19:06] <dclarke> thats not too bad for that machine
[00:19:09] <sickness> anyway my highest zio* entry is 3982078 as of now, pretty small :)
[00:19:15] <dclarke> th second pass forces fragmentation
[00:19:20] <sickness> dclarke: 0.x is bad imho :/
[00:19:23] <dclarke> the third pass is just abusive
[00:19:26] <sickness> oh
[00:19:37] <sickness> so what's the record on those passes?
[00:19:57] <dclarke> flat out max into multiple channels of fibre and 44 spinfles
[00:20:04] <dclarke> spindles
[00:20:08] <dclarke> is 170 MB/sec
[00:20:16] <dclarke> then append is about 5MB/sec
[00:20:25] <sickness> oh, so few?
[00:20:26] <dclarke> final frag is 3MB/sec
[00:20:39] <dclarke> huh ?
[00:20:42] <sickness> so there's no way to achieve like 50MB/sec on the second test?
[00:20:48] <dclarke> no way
[00:20:52] <dclarke> not even to RAM
[00:20:52] <sickness> oh
[00:21:07] <sickness> well, so as you told, my numbers aren't so shitty after all :)
[00:21:13] <dclarke> this is a "forward looking" test that I am still working on
[00:21:32] <dclarke> the idea being to test large terabyte based SANs etc etc
[00:21:37] <sickness> anyway even a simple ls takes a bit more than the usual ls on that dir =)
[00:21:46] <sickness> and du -hs is taking a huge time too =)
[00:22:00] <dclarke> the next rev ( and time real soon now ) uses piles of dispatched POSIX worker threads
[00:24:13] <twincest> so with hg, i can commit changes to my local onnv, and then get diffs from that to send back?  (that's the point of a distributed scm?)
[00:24:57] <richlowe> Yes.
[00:25:08] <richlowe> assuming that really was a question, not a statement.
[00:25:21] <sickness> so this test didn't break zfs on snv_50 :)
[00:25:22] <sickness> nor did break memory usage :)
[00:25:32] <dclarke> its a simple test
[00:25:36] <richlowe> twincest: I tend to use Mq to actually deal with such things, but that's partly laziness on my part.
[00:25:37] <dclarke> no weird tricks
[00:25:43] <twincest> what's mq?
[00:26:01] <richlowe> an extension to hg that works kind of like quilt.
[00:28:25] <richlowe> the utility of it kind of depends on the scale of change, it makes general operation nicer (I think), but merging more of a pain in the ass.
[00:30:34] *** Tpenta has quit IRC
[00:34:16] *** Tpenta has joined #opensolaris
[00:35:01] <richlowe> I have hopes that workflow-ish niftiness could fall out of the hg-tools project.
[00:35:16] <richlowe> and that the evil hack I'm using right now dies a quick and righteous death ;)
[00:40:15] <jmcp> s/righteous/well-deserved/
[00:41:02] <twincest> is there any way to edit the SCA before printing it?  sadly it doesn't seem to be a PDF form
[00:41:43] <richlowe> jmcp: Hey, I think you have one of the more working copies.
[00:41:50] <richlowe> unless I screwed up then, too :(
[00:42:24] * jmcp chuckles
[00:43:42] *** fik has joined #opensolaris
[00:43:51] *** ChanServ sets mode: +o Tpenta
[00:43:56] *** tsoome has quit IRC
[00:45:35] <richlowe> jmcp: hey, nothing stopping anyone with an interest (you) and appropriate motivation (you) making it Not Suck.
[00:45:44] <jmcp> :)
[00:45:45] <jmcp> of course
[00:46:06] <jmcp> slight matter of some standards and chip documentation that I have to cram in the next 3 days though
[00:46:10] <jmcp> apart from that ... yeah
[00:46:15] <richlowe> Sounds fun.
[00:46:20] <richlowe> depending on the standards.
[00:46:36] *** Kernel86 has quit IRC
[00:47:03] <jmcp> richlowe: SAS2 - it's big. ..... *very* big
[00:47:08] <jmcp> and it's got UML in it :(
[00:47:09] <richlowe> twincest: I'm pretty sure you can't.
[00:47:47] <richlowe> if they take scanned copies, I guess you coud take a nice big screenshot, and edit that... but I don't know how you'd sign it. :)
[00:48:27] <twincest> i was planning on filling it in, printing it and scanning the signed copy :)
[00:48:35] <twincest> (my handwriting tends to be illegible)
[00:49:37] *** fik has quit IRC
[00:51:34] *** logic has quit IRC
[00:54:25] *** nbkk6fo__ has joined #opensolaris
[00:54:26] *** nbkk6fo_ has quit IRC
[00:56:26] *** koolniczka has quit IRC
[00:56:31] *** sickness has quit IRC
[00:59:17] *** fik has joined #opensolaris
[00:59:44] *** uebayasi has joined #opensolaris
[01:03:03] *** sommerfeld has joined #opensolaris
[01:03:04] *** ChanServ sets mode: +o sommerfeld
[01:12:45] *** sommerfeld has quit IRC
[01:13:22] *** sommerfeld has joined #opensolaris
[01:13:23] *** ChanServ sets mode: +o sommerfeld
[01:13:44] <sommerfeld> (oops, bad connection)
[01:14:05] *** stevel has joined #opensolaris
[01:14:05] *** ChanServ sets mode: +o stevel
[01:17:08] <richlowe> mornin' stevel.
[01:17:13] <PerterB> rats... my nge NIC which worked fine under 10u2 seems borked under b51 (it reports it's up dmesg but never receives any packets)
[01:17:15] <stevel> 'morning
[01:17:25] * stevel has been stymied by the power outlets being different
[01:17:40] <stevel> so i'm now SunRay'ing across the pacific ocean to MPK
[01:18:39] *** OnkelSchorsch has joined #opensolaris
[01:21:01] <Gman> hey stevel
[01:21:22] <stevel> hey glynn
[01:22:01] <Tpenta> ok gman I havent logged that bug yet, I'll get tere
[01:22:05] <Tpenta> hi steve
[01:22:13] <stevel> morning alan
[01:22:15] <Gman> Tpenta, no worries, it's not urgent
[01:24:05] <Gman> hrm, weekly summary time. [sigh]
[01:24:18] <stevel> ... he says enthusiastically
[01:24:59] <Gman> yeah, it's pretty tiring keeping it going
[01:27:01] <jmcp> stevel: didn't you get a plug adapter before you flew out?
[01:27:04] * Gman hardens up, and gets to work
[01:27:53] *** bunker has quit IRC
[01:27:54] <stevel> jmcp: nope.. i knew the hotel provided 2/3 prong US-style plugs, and the other Sun offices i'd been in had them as well - so i figured i was okay :-P
[01:27:59] <jmcp> darn
[01:28:12] <stevel> i'll pop out and buy one this afternoon or evening
[01:29:24] <jmcp> recovered from your flight yet?
[01:30:17] <stevel> yeah. stayed up the whole flight, landed in the evening, drank a bunch of beer and took a benadryl and slept like a dog
[01:30:24] <sommerfeld> stevel: where are you?
[01:30:32] <stevel> sommerfeld: seoul, korea for tech days
[01:31:15] <sommerfeld> ah.  korea seemed to be schizo about its preferred plug when I was there.   seemed to have all three of: US/japan style, continental euro style (two round pin) and UK lamp style.
[01:31:27] <sommerfeld> or, well, the hotel I was in was schitzo that way.
[01:31:40] <Tpenta> good afternoon bill
[01:31:47] <stevel> the country-wide standard seems to be the two round pins.
[01:32:06] <stevel> but the hotel i'm at provides the US style as well
[01:33:04] <stevel> ugh. trans-pacific sunray so slow
[01:33:21] <sommerfeld> stevel: opengrok on src.opensolaris.org seemed to be returning "Error: access denied (java.io.FilePermission /opengrok/data read)"
[01:33:51] <stevel> 'k. i'll look into it thanks
[01:34:50] <stevel> probably a stale NFS mount
[01:38:56] * Gman goes to deal with companion mail queue...sigh.
[01:39:13] * stevel wonders why the companion people don't deal with their own mail queue
[01:39:30] <Error_404> anyone use Spring (the j2ee framework) on occasion?
[01:40:08] <Gman> stevel, i have a good idea...
[01:40:50] <stevel> gman: as long as it doesn't involve me doing anything...
[01:40:53] <Gman> cab was up over 100 messages
[01:41:07] <Gman> stevel, nah, you shouldn't have to deal with this
[01:41:16] <stevel> sommerfeld: opengrok should be up now
[01:41:36] * Gman does enjoy having the site password on mailman though
[01:49:40] <Error_404> "spring" is such a common word it's hard to sift through all this crap trying to find a solution to my problem... all i want is to link to a bloody stylesheet
[01:53:02] <richlowe> stevel: I can take guesses, but not ones I'd state in public.
[01:54:26] *** nbkk6fo_ has joined #opensolaris
[01:54:26] *** nbkk6fo__ has quit IRC
[01:55:43] <twincest> i need more cpus :-(
[01:56:58] <stevel> richlowe: i talked to danek, he said the mod.c deltas thing was definitely a one-off, and that he made an exception for it
[01:57:05] *** dclarke has quit IRC
[01:57:17] <richlowe> stevel: I'm only interested, but did he say why?
[02:00:36] <sommerfeld> stevel: thanks
[02:00:52] *** dakrone has quit IRC
[02:01:58] *** dakrone has joined #opensolaris
[02:02:09] <richlowe> stevel: Has it been considered that merge turds in general break stuff?
[02:02:21] <richlowe> since they're generally killed, you probably won't have seen many thus far.
[02:02:25] <richlowe> and the ones you have seen, clearly...
[02:06:28] *** Gadzooks has joined #opensolaris
[02:10:40] *** Cybernd has quit IRC
[02:11:07] <Gman> richlowe, worth summarizing some of the bridge failures last week?
[02:11:28] * Gman undecided, but mostly on the 'no, i don't want to have to read that thread again' side
[02:13:39] *** Kernel86 has joined #OpenSolaris
[02:14:41] <richlowe> Gman: I only know of one failure.
[02:14:44] <richlowe> ... did I miss some? :)
[02:14:49] <richlowe> well, one last week, anyway.
[02:15:16] <Gman> yeah, those.
[02:15:32] <richlowe> I don't know.
[02:19:28] *** sala has joined #opensolaris
[02:19:51] *** Tpenta has quit IRC
[02:19:52] <stevel> richlowe: nope, he didn't say why
[02:20:16] <stevel> gman: i think you should have a section in the weekly summary titled
[02:20:22] <stevel> "What did stevel screwup this week?"
[02:20:32] <richlowe> I still maintain this isn't exactly your fault.
[02:20:35] <richlowe> as much as I'd like to claim otherwise.
[02:20:37] <Gman> heh, you won't get a quote this week
[02:20:50] <stevel> aw darn. it was nice to get the inaugural quote though ;-)
[02:21:05] <richlowe> Now, when you figure out how it actually happened, I may change my mind...
[02:21:10] <stevel> no, i don't think this is exactly mine either. this was a one-off case, so i think the bridge (as it is now, anyway) is fine.
[02:21:18] * jmcp catches up with the "reforming.*integration" thread ....
[02:21:20] * jmcp shudders
[02:21:23] <stevel> i'll just go make a one-off diffs exception for that putback and regenerate from that putback onwards
[02:21:24] <richlowe> stevel: now that, I disagree with.
[02:21:30] <richlowe> jmcp: reference?
[02:21:35] <richlowe> (sorry, I'm way behind on mail)
[02:21:46] <richlowe> stevel: I don't think that's correct.
[02:21:53] <jmcp> richlowe: Reforming the OpenSolaris project integration process (wasRe: how to make a new project?)
[02:21:55] <richlowe> though I may just as easily be wrong.
[02:22:00] <richlowe> damn secretive tonic folk.
[02:22:01] <jmcp> started by Josh Hurst
[02:22:04] <stevel> richlowe: how so?
[02:22:05] <richlowe> jmcp: the stupid rant on -discuss?
[02:22:07] <jmcp> yah
[02:22:10] <richlowe> jmcp: Oh, I binned that and it's author instantly.
[02:22:15] <richlowe> none of that crap for me ;)
[02:22:40] <richlowe> stevel: Well, you have the mod.c stuff, a merge turd(?) that broke the bridge.
[02:22:48] <richlowe> stevel: and you have the putback on the 3rd, a merge turd that broke the bridge.
[02:23:01] <richlowe> stevel: that's a pattern that suggests *something* needs to resist turds a little better...
[02:23:45] <richlowe> and obviously there's the changes to have the bridge check integrity per-putback, that you'd talked about last night.
[02:24:04] <stevel> oh yeah, the bridge integrity-check-post-putback thing is definitely necessary
[02:24:09] <stevel> the merge turd is harder
[02:24:18] <stevel> because really, that *is* just a delta
[02:24:19] <richlowe> It needs to be known what broke it, I think.
[02:24:24] <richlowe> Not quite 'just' a delta.
[02:24:31] <richlowe> the one in mod.c is a branch.
[02:24:38] <richlowe> I don't know about the one of the 3rd, obviously.
[02:25:04] <stevel> but even if it's a branch, the bridge should just work
[02:25:13] <richlowe> But that's somewhat my point, if any turd Danek misses breaks the bridge, *something* has to happen.
[02:25:24] <stevel> in the 2006-02-11 fma putback, what broke is the way the diffs didn't match up with reality
[02:25:43] <richlowe> and the 3rd was an undo that the bridge didn't take?
[02:25:44] <twincest> are there plans (even far inthe future) for the hg repo to be rw?
[02:25:47] <richlowe> or did it outright break things.
[02:25:50] <richlowe> twincest: Yes.
[02:26:21] <richlowe> ... there better be, anyway :)
[02:26:29] <stevel> i'm not 100% sure the undo was what broke the bridge
[02:26:43] <richlowe> stevel: well, nobody has said anything about what killed it on the 3rd.
[02:26:46] <richlowe> beyond that it happened.
[02:26:47] <stevel> because from what i can see, there was a mercurial exception that happened while it was renaming a file in the initial putback
[02:27:04] <richlowe> stevel: You could clean up from that though...
[02:27:15] <stevel> yeah, i don't have a proper explanation for what happened, so i haven't given one yet
[02:27:29] <stevel> i didn't think it was fair for me to blame mercurial since we are running a fairly old version (0.9)
[02:27:32] <richlowe> if hg returns failure, revert --no-backup and re-try, then give up?
[02:27:48] <stevel> richlowe: yeah, i put that check in there yeesterday :-P
[02:28:03] <stevel> that was def. my bad for not having done some sort of error check of $? from hg before
[02:28:12] <twincest> ooh, new b.o.o
[02:28:24] * richlowe has gotta see this.
[02:28:35] <twincest> (... coming this month)
[02:29:44] <richlowe> wonder what'll be new and crappy this time.
[02:29:57] <Gman> heh
[02:30:08] <richlowe> the notifications have left me *real* bitter. :)
[02:30:09] <Gman> richlowe, oh ye of little faith! ;)
[02:30:16] <twincest> i think it's amusing that we can see putbacks now which don't have a public bug :)
[02:30:36] <richlowe> twincest: that's because of stupidity too.
[02:30:45] <richlowe> slightly justifiable, for a change, though.
[02:31:59] <stevel> yes, there is a distinction between something being broken because of policy/process
[02:32:06] <stevel> versus being broken because of technical incompetence
[02:32:30] <richlowe> Yeah, there's "policy", "stupid policy", "implemented by idiot", and "bug"
[02:33:22] <richlowe> but I'm a cynic, so...
[02:33:35] <stevel> really. do tell. i could have sworn you were an optimist
[02:33:39] <stevel> ;)
[02:33:44] *** fik has quit IRC
[02:33:58] *** simford has joined #opensolaris
[02:36:53] <twincest> hmm on a related note, is there a better way to do this?  VERSION=hg-`hg heads -R /onnv/onnv-gate|head -1|awk '{print $2}'`
[02:37:52] <richlowe> Yes.
[02:38:01] <twincest> please elaborate :)
[02:38:11] <richlowe> I will when I remember how.
[02:38:12] <richlowe>  :)
[02:38:34] <twincest> i couldn't seem to find a command that just prints the checked out revision id
[02:38:40] <richlowe> hg tip --template '{rev}:{node|short}\n'
[02:38:51] <richlowe> remove the |short, to get the full nodeid.
[02:39:00] <twincest> aha.  thanks
[02:40:03] <richlowe> stevel: Hey, you can't claim the notification stuff was done by somebody actively thinking...
[02:40:39] <stevel> i can claim they were actively thinking
[02:40:50] <stevel> i will, however, make no claim as to the quality or process of their thoughts
[02:41:03] <Gman> i wish there was away to separate out the 'i can't get X working' mails on the list
[02:41:22] <twincest> x-doesnt-work-discuss@
[02:41:24] <Gman> most of the mails seem to be support related questions rather than development :/
[02:43:46] <Gman> i also hate the +1 mails
[02:44:39] * jmcp stops avoiding reading stuff
[02:46:00] <richlowe> Gman: how else do you expect the project/community stuff to work?
[02:46:13] <richlowe> justifying ones reasoning in the form of iambic pentameter?
[02:46:22] <stevel> i was going to argue haiku
[02:46:25] <twincest> i like that idea
[02:46:29] <stevel> if you can't form your project proposal in the format of a haiku
[02:46:39] <stevel> it has no reason to exist
[02:46:39] <delewis> Gman: I thought that was the purpose of opensolaris-help
[02:46:46] <delewis> (basic, end-user questions)
[02:46:56] <Gman> richlowe, it's a silly apache-ism.
[02:46:57] <delewis> that pertain getting Solaris to function correctly for use
[02:47:12] <Gman> delewis, yeah, though it's easier to find help where the experts are
[02:48:44] *** johnlev has joined #opensolaris
[02:48:59] <sommerfeld> stevel: still here?  could you send a message explaining how to recover from the repository problem which is comprehensible to people who aren't hg experts?
[02:49:20] <stevel> sommerfeld: sure, i can try
[02:49:25] *** dakrone has quit IRC
[02:49:26] <stevel> was the rollback && revert thing not enough?
[02:50:13] <johnlev> evenin'
[02:50:24] <sommerfeld> i suspect I didn't quite roll back the right thing.
[02:50:34] <sommerfeld> ... added 4 changesets with 20 changes to 156 files (+1 heads)
[02:50:35] <sommerfeld> (run 'hg heads' to see heads, 'hg merge' to merge)
[02:50:44] <stevel> doh
[02:51:02] <richlowe> load Mq.
[02:51:03] <richlowe> then
[02:51:06] <richlowe> hg strip 3100
[02:51:07] <richlowe> hg pull -u
[02:51:13] <richlowe> preserving your own changes, if necessary.
[02:51:18] <sommerfeld> "load Mq" ?
[02:51:26] <richlowe> add:
[02:51:27] <richlowe> [extensions]
[02:51:31] <richlowe> hgext.mq=
[02:51:32] <johnlev> strip is in mq?
[02:51:33] <richlowe> to ~/.hgrc
[02:51:35] <johnlev> weird.
[02:51:36] <richlowe> johnlev: sadly, yes.
[02:52:09] <richlowe> I really don't think it should be, it's generally useful, even without Mq.
[02:52:16] <richlowe> though Mq needs it to actually operate, obviously.
[02:52:29] <sommerfeld> If I didn't have an .hgrc, then those two lines should do it?
[02:52:51] <richlowe> I think so, yes.
[02:53:04] <richlowe> johnlev: and yes, I did mean the real gcore.
[02:53:25] <johnlev> richlowe: file a bug yet?
[02:53:38] <johnlev> it's really super broken
[02:53:45] <richlowe> Not yet, I was rather hoping to find the why.
[02:53:56] <richlowe> since it's always applicable this as well, obviously.
[02:54:34] <johnlev> probably not anything exciting. reading from certain spaces can and do make machines hang
[02:55:08] <richlowe> Yes, indeed.
[02:55:19] <richlowe> indicating to the code in question that it's intending to do that, however, would require work.
[02:55:25] <richlowe> and since it affects me too, I may as well look further into it.
[02:55:41] * johnlev sneaks off again
[02:55:42] *** starsoldier has joined #opensolaris
[02:57:33] *** yongsun_ has joined #opensolaris
[03:06:06] <Gadzooks> What fun to see GNOME 2.14/2.16 development on Solaris 8/9 desktops: http://www.blastwave.org/dclarke/GNOME/Screenshot_05.png
[03:12:49] *** johnlev has quit IRC
[03:13:22] * stevel wanders off in search of a power adapter
[03:14:18] *** schily has quit IRC
[03:15:03] *** Tpenta has joined #opensolaris
[03:15:36] *** ChanServ sets mode: +o Tpenta
[03:16:27] <delewis> hmm, is it just me or is GIMP in nv51 very slow starting up?
[03:16:34] <delewis> it's been 'initializing' for about 10 minutes.
[03:16:37] <delewis> and it isn't sleeping, either.
[03:17:27] <Tpenta> i've noticed it doesnt start on sparc for me in 51
[03:17:38] <Tpenta> it just rins the initialising graphic
[03:17:45] *** schily has joined #opensolaris
[03:18:30] <delewis> Tpenta: same here.
[03:18:40] <delewis> I'm on SPARC, as well.
[03:18:56] <Tpenta> it starts correctly on x86
[03:18:59] <delewis> interesting
[03:20:52] <Gman> OpenSolaris Weekly News #36 - http://mail.opensolaris.org/pipermail/opensolaris-discuss/2006-November/021477.html
[03:21:29] *** johnlev has joined #opensolaris
[03:23:38] <Gadzooks> What version of GIMP is it?
[03:24:16] <Tpenta> GNU Image Manipulation Program version 2.3.9
[03:24:37] <Tpenta> works find in x86, just not sparc
[03:26:21] *** aliquis has joined #opensolaris
[03:27:04] <aliquis> Crap, had to reboot it... First Acrobat Reader used all the cpu, killed it, then Opera used all the cpu, killed it, then Explorer used all the cpu.. Nothing wrong at all ;/
[03:28:13] <delewis> Explorer?
[03:28:21] <delewis> don't tell me you're using IE for Solaris :-)
[03:29:46] <darkcmd> there's an IE for solaris?
[03:29:47] <darkcmd> ?!
[03:29:56] <delewis> there was.
[03:29:59] <Gman> Tpenta, let me check that on my sparc
[03:30:02] <twincest> there was a long time ago.  it's been discontinued for some time
[03:30:03] <delewis> it's been discontinued for a numberof years.
[03:30:08] <delewis> IE4, IIRC.
[03:30:26] <twincest> wasn't there Outlook too, or did i imagine that?
[03:30:28] <delewis> it actually used Winsock, too, so you had TCP/IP emulation on top of TCP/IP emulation.
[03:30:45] <delewis> twincest: not that I recall, but I never actually was around when IE for Solaris was released.
[03:30:52] <delewis> there was also an HP-UX release, too.
[03:30:55] <Auralis> yes there was an outlook as well
[03:31:05] <Gman> Tpenta, i have 2.3.10 here, and it works fine
[03:31:06] <delewis> scary.
[03:31:22] <Auralis> its was again more then shit
[03:34:10] <aliquis> delewis: Wrong channel to begin with ;D
[03:35:49] <richlowe> Gman: it working for you guys on bits we won't see for at least two builds doesn't really help the bits we have now being broken :)
[03:36:02] <Gman> richlowe, yeah, though i don't think i've seen an open bug report
[03:36:21] <Gman> [not saying anything about our QA team...]
[03:36:47] <richlowe> Gman: it wouldn't matter, you'd close it as fixed is 2.16, same as the drag-and-drop segv ;)
[03:36:57] <richlowe> (what me? bitter?) :)
[03:37:18] <Gman> yeah, suffer in your jocks!
[03:37:40] <Gman> the integration into nevada of 2.12 probably wasn't very well done, i'll admit
[03:37:42] <aliquis> delewis: Also it was explorer as in the desktop manager part, not the browser, and it is because someone had linked to a theme but I suppose windows isn't made for "theme hacks" so of course it failed installing for like 8 times but I didn't wanted to reboot when a broken theme had messed up various system files. So in the end it seemed to work but the computer have behaved like shit since then. I where going to use
[03:37:42] <aliquis>  a system restore point but after a time it told me I had to little disk space to finish it, rebooted and removed a large file and uninstalled a couple of programs, but then I went back to system restore Windows had of course made place for the new restoration points from uninstalling programs by removing the old one so then I was fucked ;)
[03:38:09] <Gman> richlowe, we do need to manage how we track gnome latest while keeping other gnome in reasonably good nick
[03:38:16] <g4lt-U60> about the time I saw aieee/outleak for hp-sux, I was already a pine junkie
[03:38:17] *** rachel has joined #opensolaris
[03:38:37] <richlowe> Gman: neither option there sounds particularly good.
[03:39:01] <stevel> woot. got my power adapter
[03:39:24] <Gman> richlowe, it's the only way we're going to get desktop developers to solaris
[03:39:35] <richlowe> Gman: fixing-what-people-have seems good, but is time consuming and there aren't many of you.
[03:39:48] <richlowe> Gman: ignoring what you ship, because of what you don't yet ship is untennable, but what you're doing now.
[03:39:58] <richlowe> Gman: integrating as you go is also not so good.
[03:40:51] <Gman> the problem was the lack of qa cycles
[03:41:15] <Gman> and very obviously being able to integrate with outstanding bugs
[03:41:30] <Gman> [the actual polices in place are a poor fit for gnome]
[03:41:31] <richlowe> and to then mark them as fixed in something you don't plan to integrate for at least two more builds.
[03:41:41] <richlowe> (and at the time some were closed, more like 8 or 9 more builds, at least...)
[03:41:48] <Gman> yeah
[03:42:29] <Gman> the integration policy basically means you're not supposed to have p1s, p2s and p3s
[03:42:34] <johnlev> Gman: has anybody done a decent clock applet for gnome yet?
[03:42:38] <Gman> which is never the case
[03:42:39] <johnlev> namely one that can do different timezones
[03:42:43] <Gman> johnlev, heh, it's the same as ever
[03:42:51] <Gman> the current one can
[03:42:54] <Gman> we ported the patch across
[03:42:59] <stevel> johnlev: there's a world clock thing that has diff. timezones
[03:43:00] <Gman> [current probably being 2.16]
[03:43:39] <stevel> is there a way to swap ctrl/caps lock temporarily for the current session?
[03:43:53] <johnlev> stevel: coo yes.
[03:44:06] <stevel> all the korean keyboards here have the PC layout and the caps/ctrl thing is killing me
[03:44:11] <johnlev> now if only I could get a binary to install :)
[03:44:53] <Gman> sadly no thought has been put into the timezone feature in the clock applet
[03:44:54] <johnlev> believe it or not it's the one thing stopping me from switching from KDE.
[03:45:09] <richlowe> Gman: na, it's in the Nevada stuff, too.
[03:45:12] <Gman> [and our patches are nightmarish]
[03:45:17] <johnlev> gm152: or solaris generally :/
[03:45:43] <stevel> ahh..xmodmap, thank you
[03:45:51] <richlowe> setxkbmap would do it!
[03:45:54] * richlowe whistles
[03:46:13] <richlowe> as would the JDS folk fixing the keyboard settings stuff so it actually did something.
[03:46:23] <stevel> where does setxkbmap live?
[03:46:34] <stevel> aargh. damn this stupid one port policy to hell and back
[03:46:43] <richlowe> stevel: it doesn't, that's why I whistled.
[03:46:52] <stevel> ah
[03:47:00] * stevel is poor at detecting sarcastic whistles ;)
[03:47:01] <richlowe> stevel: I nagged alanc about it so much, he added me to the interest list on that CR, just so I could watch him ignore it more readily ;)
[03:47:03] <Gman> richlowe, yeah
[03:47:04] <Gman> see
[03:47:14] <Gman> we removed all the keyboard switching functionality now
[03:47:21] *** yusufg has joined #opensolaris
[03:47:28] <johnlev> panic[cpu0]/thread=fec20b00: assertion failed: ((iocmp)->b_wptr - (iocmp)->b_rptr) == sizeof (struct iocblk), file: ../../common/io/strsun.c, line: 218
[03:47:36] <johnlev> anyone else seeing this on recentish nevada?
[03:47:47] <richlowe> I haven't seen it.
[03:47:56] <richlowe> I'm running onnv-gate as of about the 2nd + patches.
[03:48:07] <johnlev> hrmph
[03:48:48] <richlowe> Now there's tref closed bins, I'm going to bring stuff up to more recently soon-ish though.
[03:48:54] <twincest> hmm, is it usual for vmware user cpu to show up as host system cpu, or is it spending a lot of time doing nothing?
[03:51:08] <richlowe> johnlev: I seem to be generally right about the gcore stuff, but that doesn't nesecarily mean I see a nice way to fix it.
[03:51:27] <richlowe> johnlev: if you want a CR, I'll file one, otherwise I'll hold it a little more so I can give it some more detail.
[03:51:46] <johnlev> well... I'm not going to do anything about it :)
[03:51:49] <johnlev> so more detail sounds good
[03:51:56] <Gman> timeless, nice reply to your usb code mail
[03:52:02] <Gman> 'i didn't read the code..'
[03:52:13] <richlowe> it didn't actually answer the question at all, either.
[03:52:20] <Gman> yeah ;)
[03:52:27] <richlowe> since "the code" in that case isn't usbether, but usba.
[03:53:21] <richlowe> johnlev: pity, I was hoping there was a neat way to fix this I just wasn't seeing. :)
[03:54:10] <johnlev> richlowe: tbh I haven't really thought about it...
[03:54:37] *** nbkk6fo__ has joined #opensolaris
[03:54:38] *** nbkk6fo_ has quit IRC
[03:54:38] <richlowe> Oh, not from you in specific, just in general. :)
[03:54:48] <johnlev> and certainly I can barely deal with the difficulties of typing after my flight :)
[03:54:56] <richlowe> if I derail myself, I tend to completely and utterly lose state on what I was doing before. :)
[04:05:55] *** silly_girl22 has joined #opensolaris
[04:08:32] <twincest> hm, lots of questions today:  how do i know if the current nightly closed bins is too old for the onnv-gate head?
[04:08:45] <richlowe> You don't.
[04:08:48] *** halton has joined #opensolaris
[04:08:50] <twincest> that's a shame
[04:08:55] <richlowe> In general, they won't be.
[04:09:09] <richlowe> in this case, there's the tref bits, that means they probably are, unless they're from the 3rd.
[04:09:26] <twincest> yes, i'm using 20061103
[04:09:28] <richlowe> 'last weeks' is fine, I've accidentally used older ones before, too.
[04:09:45] <richlowe> I track the closed-bins in hg, and keep things in sync that way too.
[04:10:02] <richlowe> unfortunately, that means you lose empty directories, which would make the packaging stuff unhappy, but since that doesn't work anyway yet, it's not such a loss.
[04:10:22] <richlowe> I was intending, at one point, not to have directory creation in usr/closed, to try and avoid the empty dirs in closed-bins, since I think it's silly.
[04:10:34] <richlowe> but for some reason, I never made the changes on the open side, and did so.
[04:12:15] <richlowe> twincest: if you want something vaguely automatable, /usr/ccs/bin/what closed/root_$(mach)/usr/lib/libike.so.1, and compare to the current date.
[04:12:23] <richlowe> (figuring libike is unlikely to open, and will tend to be there...)
[04:19:32] *** silly_girl22 has quit IRC
[04:27:00] *** eboutilier has joined #opensolaris
[04:32:50] <twincest> hmm: http://www.rafb.net/paste/results/MduO6n78.html
[04:33:04] <twincest> is that part of a normal build? :)
[04:47:18] *** Gadzooks has quit IRC
[04:49:18] *** sommerfeld has quit IRC
[05:02:05] *** jwtodd has joined #opensolaris
[05:12:22] *** johnlev has quit IRC
[05:16:59] *** OnkelSchorsch has quit IRC
[05:18:33] *** rafaeldt has joined #opensolaris
[05:18:38] *** OnkelSchorsch has joined #opensolaris
[05:21:46] *** icon has joined #opensolaris
[05:22:05] <icon> evening all
[05:22:36] * jamesd reordanises his icons
[05:22:47] <jamesd> er reorganises
[05:33:43] *** jwt has joined #opensolaris
[05:34:14] *** jwt has joined #opensolaris
[05:34:31] <icon> hrmm
[05:34:36] * icon wonders if Gman is around
[05:34:46] <Gman> hey hey ;)
[05:34:49] <icon> heya gman ;)
[05:35:07] <jwt> hey ... quick question. i've installed solb50 on an x86 amd system and all is good but i need to set TERM so that i can edit files sanely and such. thoughts?
[05:35:08] <jwt> thx.
[05:35:08] <icon> just updated to build 51 - quick question. im not seeing any gnome changes with respect to themes etc.
[05:35:13] <icon> do i need to nuke ~/.gnome?
[05:35:29] <icon> (updated from b46 to b41)
[05:35:34] <icon> err b51
[05:35:58] <icon> jwt, what shell are you using, and do you want this to be a permanent change?
[05:36:03] <jwt> bash
[05:36:11] <jwt> well, yes but
[05:36:30] <jwt> i'm going to build solb50 so that i get the amd64 build
[05:36:38] <jwt> and will then re-install on the same system.
[05:36:57] <icon> edit your ~/.profile and add a TERM=xxx; export TERM
[05:37:02] <jwt> at that time ... i'd like it to be permanent but i'm having trouble right now
[05:37:11] <jwt> literally xxx?
[05:37:16] <icon> whatever you want your term to be
[05:37:20] <Gman> icon, can you be more specific?
[05:37:26] <jwt> i wet it to sun100, etc. no joy.
[05:37:37] <Gman> open up the theme preferences dialog - is it showing a custom theme?
[05:37:46] <icon> nope, just nimbus
[05:37:48] <jwt> ahh, perhaps vt100 then. it has been awhile since i've been back on a sol box.
[05:38:07] <icon> icons look like they came from blueprint
[05:38:08] <jwt> gman ... this is james.todd (http://blogs.sun.com/gonzo) ... but moved on to a startup :)
[05:38:13] <jwt> hope all is well compadre.
[05:38:26] <Gman> oh, hey jwt, wondered where you got to
[05:38:27] * icon was thinking that there were theme changes around b50
[05:38:28] <Gman> hope that goes well
[05:38:33] <jwt> http://radarnetworks.com
[05:38:40] <jwt> mid year. having fun. learning tons
[05:38:43] <Gman> icon, hrm, yeah, though not sure whether they were backported to nevada
[05:38:48] <jwt> we're eval'ing solaris now :)
[05:38:48] <icon> ahh
[05:38:49] <Gman> however..
[05:38:58] <Gman> i do have a package for you to install i think
[05:39:05] <icon> Gman: groovy
[05:39:37] <Gman> http://www.gnome.org/~gman/SUNWgnome-themes-devel.tar.gz
[05:39:40] <Gman> http://www.gnome.org/~gman/SUNWgnome-themes.tar.gz
[05:39:53] <Gman> jwt, heh, funny that :)
[05:40:02] <jwt> damn ... it was TERM=vt100 ... been so long.
[05:40:17] <jwt> guess i just needed to verbalize that :)
[05:40:30] <jwt> thx. the startup essentials h/w benefits look sweet!
[05:40:35] * icon whistles
[05:40:40] <jwt> we hope to get a thumper or 2 as well
[05:40:43] <Gman> jwt, glad to be outside?
[05:40:44] <richlowe> the default $TERM on the console is somewhat odd, as I recall.
[05:40:45] <icon> i didnt realize SUNWgnome-themes was so large
[05:40:52] <Gman> icon, yeah, it's pretty huge
[05:40:59] <Gman> there's about 5 or 6 themes in it
[05:41:04] <icon> ahh
[05:41:05] <richlowe> but the bigger deal you may see is that columns/rows are set kinda oddly.
[05:41:25] <icon> Gman: anything special i need to do with these two other than the usual?
[05:41:26] <richlowe> also, any x86 Solaris since 10 should do the right thing regarding x64
[05:41:39] <Gman> icon, not really - the postinstall scripts should do it all
[05:41:44] <icon> excellent
[05:41:45] <Gman> i have a nice install script if you want it
[05:41:46] <jwt> dude ... i was *inside* for so long it is strange ... since '92. i am learning a ton. folks rock. very little b$. not dissing here just this is the env i like.
[05:41:46] <icon> thanks alot
[05:41:50] <icon> Gman: sure
[05:42:13] <richlowe> Gman: I thought your postinstall scripts didn't do much beyond make me wait 10 minutes to log in after an upgrade ;)
[05:42:24] <Gman> icon: http://www.gnome.org/~gman/inst-sunw-pkg.sh
[05:42:26] <jwt> and, there's no sure bets anywhere ehhhh ... and life is short so why not "go big or go home" .. me thinks.
[05:42:36] <Gman> icon, inst-sunw-pkg.sh SUNWgnome-themes*
[05:42:45] <icon> excellent
[05:42:46] <Gman> jwt, that's cool
[05:42:47] <icon> thanks a buncbh
[05:42:52] <icon> s/bunchbh/bunch
[05:42:55] <Gman> richlowe, heh, yeah
[05:43:06] <Gman> we have a patch in gnome-session, that checks if there's anything in the postrun queue
[05:43:11] <Gman> and doesn't let you start a session if there is
[05:43:14] <Gman> it's fricken evil.
[05:43:17] <richlowe> and shows me an abrasive little dialog. :)
[05:43:21] <jwt> anyways, sol/zfs/zones/etc is pretty much what we need plus getting some rocking h/w would be great. a few sun folks on staff as well.
[05:43:33] <Gman> richlowe, yeah, i think we need to kill it
[05:43:34] <richlowe> you may as well change the wording to "Go get a coffee" or something. :)
[05:43:51] <Gman> you saw the one we implemented for the ARC?
[05:43:54] <jwt> better let ya go. i'm grooving now back to building an amd64 sol from src. first timer for one who usually swims at the java layer :)
[05:43:59] <Gman> think we've managed to drop that sucker, and hope that no one notices
[05:44:02] * icon is suddenly reminded of ogrok
[05:44:03] <Gman> jwt, best of luck :)
[05:44:07] <jwt> likewise
[05:44:10] <richlowe> Gman: which?
[05:44:17] *** jwt has quit IRC
[05:44:21] <richlowe> Gman: hey, did you make the JDS session name right again yet? :)
[05:44:22] <icon> speaking of which, are there any ogrok guys on?
[05:44:23] <Gman> richlowe, it was one about interfaces - tagged onto the gnome-about dialog
[05:44:34] <Gman> richlowe, i fixed the bug locally
[05:44:41] <Gman> s/locally/in spec files/
[05:44:42] *** icon has quit IRC
[05:44:48] <Gman> however, seemingly the rel eng's use different scripts now :(
[05:44:59] <richlowe> Gman: 6459848.
[05:45:01] <richlowe> that one. :)
[05:45:10] <twincest> oh, apparently the build dislikes being on NFS without locking
[05:45:43] <Gman> richlowe, that's the one
[05:45:59] <Gman> richlowe, i should just check out the re-scripts and fix it
[05:46:06] <richlowe> Gman: hey, care to put me on the interest list for those two b.o.o bugs?
[05:46:11] <Gman> but laca promised me he was tidything them up
[05:46:14] <Gman> richlowe, sure
[05:46:17] * richlowe goes ID huntin'
[05:46:41] <richlowe> I guess I've filed about 12, total.
[05:46:42] <richlowe> maybe more.
[05:46:55] <Gman> which were the other bugs you wanted cc'ing on?
[05:46:58] *** icon has joined #opensolaris
[05:47:19] <icon> Gman: looks good
[05:47:31] <Gman> great ;)
[05:47:42] <icon> folders look tall, but ill get used to it ;)
[05:48:09] <Gman> heh, yeah
[05:48:14] <Gman> overall it's a good look though
[05:48:34] <icon> it is
[05:48:42] <icon> so is everything in svg now?
[05:49:07] <Gman> nope
[05:49:18] <Gman> svg would be too slow i think
[05:49:26] <icon> ahh
[05:49:27] <richlowe> "But JDS is slow anyway!"
[05:49:32] <icon> the resizes look great on these icons
[05:49:35] <icon> no distortion
[05:49:39] <richlowe> (I figured I'd get that out there early)
[05:49:58] * Gman kicks richlowe in the pants
[05:50:29] *** gnu2it2 has joined #opensolaris
[05:50:52] <icon> interesting
[05:51:00] <icon> gnome 'connect to server' doesnt support nfs?
[05:51:21] <icon> new spinner is much nicer on nautilus
[05:51:34] <richlowe> icon: if you have autofs running as the default, just get to stuff via /net :)
[05:51:45] <richlowe> icon: /net/<hostname>/<path to export>
[05:51:54] <icon> sure, but that defeats the purpose :D
[05:52:36] <richlowe> I suspect gnomes' connect-to-server stuff doing it would require gnome-vfs knowing about NFS
[05:52:59] <Gman> icon, yeah, nfs thing was pretty buggy
[05:53:01] <Gman> think it got removed
[05:53:50] <icon> ahh
[05:53:56] <icon> richlowe: ditto
[05:54:18] <icon> did b51 include initial tamarack support? or was it just HAL?
[05:54:37] *** halton has quit IRC
[05:55:37] <Gman> icon, tamarack
[05:55:43] <icon> nevermind its in, just threw in a quick dvd
[05:56:00] <icon> very nice... much cleaner than vold
[05:56:05] <Gman> yeah, it's sexy.
[05:56:13] * icon whistles
[05:56:16] <icon> joliet support too
[05:56:27] <richlowe> icon: yeah, tamarack is in 51, but the JDS-ish bits aren't.
[05:56:31] <icon> ahh
[05:56:40] <icon> very nice
[05:56:53] <icon> almost makes me sorry im going to have to throw 10u2 on here soon again ;)
[05:56:57] <Gman> soon, soon, soon! :)
[05:57:24] <icon> it will be interesting to see how nevada works under a parallels vm
[05:57:50] <icon> mac finally released the core 2 duo mbp's, so ill pick one up to dev with (finally)
[05:59:36] <jmcp> icon: boyd_ has been running nevada under parallels iirc
[05:59:45] <icon> has he had much luck?
[05:59:55] <jmcp> it's been working fine as far as I can remember
[06:00:02] <icon> para gets extremely finicky about the boot records on installation images
[06:00:09] <icon> ahh good
[06:00:28] *** polk__ has quit IRC
[06:00:47] <icon> okay folks im out for the night. sleep well\
[06:01:19] <jmcp> icon:  http://onesearch.sun.com/search/blog/index.jsp?qt=parallels+mac gets you a decent selection of blog entries about it
[06:01:51] <jmcp> http://blogs.sun.com/nilsn/entry/at_what_point_does_this1
[06:02:05] <jmcp> best one I can think of - mac running parallels running nevada running brandZ
[06:02:08] <twincest> hm, nightly -i lies, it rebuilds half of cmd every time
[06:03:24] *** stevel has quit IRC
[06:04:21] <richlowe> twincest: not exactly.
[06:04:23] <richlowe> but kinda.
[06:04:30] <icon> cool. thanks jmcp
[06:04:35] <richlowe> iirc, it rebuilds everything that's linked directly, rather than from objects.
[06:04:36] *** stevel has joined #opensolaris
[06:04:37] *** ChanServ sets mode: +o stevel
[06:04:43] <jmcp> icon: you're welcome.... it is kinda insane though :)
[06:04:48] <twincest> that sounds like a bug somewhere
[06:04:50] <icon> i have to agree :D
[06:05:01] <twincest> does it remove the binary after installing it or something?
[06:05:31] <richlowe> I really haven't dug into it at all, I just recall seeing what I assume you're seeing.
[06:05:57] <richlowe> digging too hard into the build involves Sun folk having to test stuff for you, and just generally sucks.
[06:06:03] <twincest> heh.
[06:06:04] * stevel coughs
[06:06:06] <richlowe> stevel'll agree with me.
[06:06:07] <richlowe> See.
[06:06:24] <Gman> sun employees have to spend at least 20% on opensolaris
[06:06:25] <twincest> well, i'll live, i only plan to do a full build once a year or so :)
[06:06:26] <Gman> so tough shit :)
[06:06:29] <richlowe> the lint stuff was totally worth doing though, and the package stuff too.
[06:06:43] <richlowe> though I wish I knew what happened wiht that, rainer and kupfer have been silent about it.
[06:06:45] * stevel adds a procmail recipe to fwd all mail from richlowe to gman
[06:06:54] <Gman> heh
[06:07:11] <Gman> if i was a kernel dude, i'd be *happy* to get richlowe's mail
[06:07:12] <richlowe> stevel: sounds good to me, I'll get the same "I don't know" responses ;)
[06:07:17] <stevel> haha
[06:07:28] <Gman> stevel, loser ;)
[06:07:35] * stevel could probably replace half his emails to richlowe with a procmail recipe that says "I don't know, let me ask kupfer"
[06:08:00] *** polk__ has joined #opensolaris
[06:10:05] *** halton has joined #opensolaris
[06:10:06] <twincest> stevel: bind it to a mutt macro ;-)
[06:10:22] *** halton has quit IRC
[06:11:10] <stevel> procmail would make it easier
[06:11:31] <Gman> richlowe knows my email
[06:11:49] <stevel> generate a random choice of 0 or 1. if 0, then send the email saying "I don't know, let me ask kupfer".  if 1, then send the email saying "I don't know, let me ask dduvall".
[06:11:55] <Gman> he has also figured out how to type /query Gman in an IRC client
[06:12:57] <richlowe> that tends to happen because of irssi's occaisional expansion a plain tab to /msg <something>
[06:13:04] <stevel> he's a smart guy
[06:13:08] *** nwf has quit IRC
[06:13:19] <Gman> :)
[06:13:28] *** eboutilier has quit IRC
[06:14:06] <g4lt-U60> he's a fart smeller?
[06:14:13] * Gman doesn't mind being a point of contact for opensolaris - was doing it for years with gnome
[06:14:53] * richlowe minds
[06:15:04] <LeftWing> heh
[06:15:05] <richlowe> if the infrastructure was sensible, most of this would be unnecessary.
[06:15:12] *** nwf has joined #opensolaris
[06:15:17] <Gman> well, yeah, obviously
[06:15:23] <richlowe> the fact a few employees are filling in because b.o.o and a bunch of other stuff is crap, is well, shitty.
[06:15:33] <Gman> it's good though because it makes me feel useful :)
[06:15:47] <stevel> richlowe: i *wholeheartedly* agree
[06:16:25] <Gman> hrm, almost sounds like a good proposal for a talk at the developer conference
[06:16:30] <richlowe> stevel: do your guys aliases take mail from mortals?
[06:16:38] <stevel> yeah
[06:16:39] <Gman> or at least, brainstorm session
[06:16:44] <richlowe> stevel: Interesting...
[06:17:22] <richlowe> though really, if you'd hurry up the tools stuff, and actually get more people talking to us, that'd work even better.
[06:17:28] * richlowe is no good at subtle hints
[06:18:12] <twincest> how's progres on opening manpages?
[06:18:19] <Gman> richlowe, heh
[06:18:32] <Gman> twincest, moving slowly
[06:18:41] <Gman> twincest, there's a lot of them, so they're doing it in stages i think
[06:18:56] <Gman> richlowe, if i knew how to help more, i would
[06:19:10] <Gman> not being in mpk doesn't help either
[06:19:58] <richlowe> Well, the stages are fairly obvious, as far as consolidation-by-consolidation.
[06:20:06] <richlowe> kill SFW because (iirc) they're already there.
[06:20:09] <richlowe> and you have ON, X, JDS.
[06:20:16] <richlowe> and everyone knows those last two don't count. ;)
[06:20:36] <Gman> heh
[06:20:53] <Gman> nah, i think they're doing the manpage release in stages
[06:21:10] <richlowe> actually, I was meaning that because if you look at who needs the pages most (distributions), most only actually distribute ON.
[06:21:26] <richlowe> other people who need manpages (arc cases, CRs), cheat and hope nobody notices.
[06:21:36] <richlowe> whereas actually redistributing them when you shouldn't is bad :)
[06:21:55] * Gman finds a mail
[06:22:01] <Gman> oh, right
[06:22:02] <Gman> yeah
[06:22:11] <Gman> more than happy to see ON out first
[06:22:20] <Gman> except for the fact that JDS are already there ;)
[06:22:37] <richlowe> I didn't know if you guys had stuff mixed between man and yourselves.
[06:22:40] <richlowe> (sun stuff v. non-sun stuff)
[06:23:18] <Gman> from program team -
[06:23:21] <Gman> about 30% complete, which is about a third of the whole project.
[06:23:21] <Gman> talking with the community about man pages and receiving feedback.  Will
[06:23:21] <Gman> focus on tools and gate work.
[06:23:42] <Gman> richlowe, i think we write most of it
[06:23:49] <Gman> though we're supposed to be getting someone in to help
[06:23:57] <richlowe> Gman: 30% is about a 3rd? no? really? :)
[06:24:08] <Gman> :)
[06:24:40] <Gman> so with those dates, looks like it will be 2007
[06:29:30] <Gman> http://www.opensolaris.org/os/announcements/
[06:29:35] <Gman> wow, that's pretty broken :)
[06:29:55] <Gman> [link 16]
[06:31:38] <Gman> hrm, i wonder if i submit a talk, whether i can get funding to travel
[06:33:50] <twincest> i want a zfs feature that automatically keeps a 'locale'-style database for a filesystem in realtime
[06:33:55] <twincest> s/locale/locate/
[06:34:29] <LeftWing> That doesn't really have to be an FS-dependent thing.
[06:35:04] <jmcp> twincest: in real time? do you have any idea how much overhead that'll cause?
[06:35:15] <LeftWing> heh
[06:36:24] <LeftWing> A better idea would be a transaction-log style output from the kernel (file renames, creates, deletes) that a userland daemon could get to when the system is less busy.
[06:36:46] <richlowe> file event notification to userland in general would be preferable.
[06:36:54] <LeftWing> Mmm.
[06:36:55] <jmcp> LeftWing: #define "less busy"
[06:37:37] <LeftWing> jmcp: #define "less busy" User(LeftWing).Logout.time() + 1      :P
[06:37:44] * jmcp snorts
[06:38:02] <LeftWing> Ooer, my Mac Mini just shipped.
[06:38:21] <g4lt-U60> richlowe, now THERE'S an idea, although sparc doesn't have nearly the overhead switching between kernelspace and userspace that X86 does
[06:39:05] <richlowe> jmcp: don't suppose you're feeling bored yet knowledgable?
[06:39:08] <LeftWing> Well, surely you could do it in a dtracey sort of way... Create a queue of notification events about relevant changes inside the kernel and just periodically collect them later from userspace...
[06:39:15] * jmcp looks slyly @ richlowe
[06:40:08] <richlowe> is that a yes, a no, or a maybe?
[06:40:16] <jmcp> guess :)
[06:40:18] <jmcp> depends on the question
[06:40:29] * LeftWing brb, postoffice.
[06:43:41] *** nwf has quit IRC
[06:46:01] * stevel is taking the bridge offline temporarily
[06:46:28] <stevel> (onnv-gate bridge that is)
[06:47:36] *** nbkk6fo__ is now known as rodrickbrown
[06:51:01] *** gm152 has quit IRC
[06:52:27] *** polk__ has quit IRC
[06:54:53] <twincest> yay, bfu worked
[06:55:00] <stevel> !@#(*$@)#$*@#$
[06:55:02] * stevel beats his head
[06:55:16] *** polk__ has joined #opensolaris
[06:55:30] <richlowe> that's an odd way to react to bfu working.
[06:55:42] * twincest was just thinking that
[06:56:03] <stevel> i figured out why my bridge pulled 1.1.1.1 instead of 1.1
[06:56:10] <stevel> because i f-ing told it to
[06:56:17] <richlowe> But that's *good*.
[06:56:24] <richlowe> the question is why it ignored the hell out of 1.2?
[06:56:26] <richlowe> isn't it?
[06:56:29] <twincest> (incidentally, vmware snapshots make bfu a lot less scary)
[06:56:38] <stevel> no, i explicitly told it 1.1.1.1 instead of 1.2
[06:56:38] <richlowe> twincest: as does liveupgrade.
[06:56:40] <stevel> i don't know why i did
[06:56:47] <twincest> i don't have enough disk space for that :)
[06:56:51] <richlowe> stevel: oh, you meant 1.2 I guess.
[06:56:55] <richlowe> ignore me, I'm dumb.
[06:57:07] <stevel> there were some malformed diffs in onnv-gate, so i built in a "blacklist" that let me map putbacks to something i could override
[06:57:21] <stevel> there are 4 putbacks that i had to override
[06:57:28] <stevel> one of which was cindi's putback on 2006-02-11
[06:57:36] <stevel> and i must have accidentally put in 1.1.1.1 instead of 1.2
[06:57:47] <richlowe> Out of interest, how were the other 3 bad?
[06:58:08] <richlowe> since checking them maybe good too. :)
[06:58:09] <stevel> i even said it explicitly:
[06:58:18] <stevel> # wrong revision number for mod.c & topo_xml.c - shoulda been 1.2
[06:58:18] <stevel> /ws/onnv-gate/public/log/2006/02/11.cindi => $history$/20060211.cindi
[06:58:22] <stevel> i wrote 1.2 there
[06:58:29] <stevel> but in the actual override, i typed 1.1.1.1
[06:58:31] <stevel> argh
[06:58:50] <stevel> hrm. 1 was due to "# missing revision information on ntwdt.c, danek said he was probably
[06:58:51] <stevel> # modifying his script around the same time and this putback got some
[06:58:51] <stevel> # one-off weird output to the diffs file"
[06:59:07] <stevel> another was due to "# missing revision information on mkdtemp.c"
[06:59:25] <stevel> and one was due to the whole bge retardedness
[06:59:42] <richlowe> Oh, great.
[07:01:49] <richlowe> so, are you rebuilding the full thing, or just back as far as the known brokenness?
[07:01:53] *** anthony79 has joined #opensolaris
[07:01:54] <richlowe> (do we know everything else is fine, btw?)
[07:07:03] *** Sieghard has joined #opensolaris
[07:07:26] <stevel> richlowe: i'm going to clone from the rev just before the 2006/02/11.cindi putback, and rebuild from there
[07:07:42] <stevel> i don't know that everything else is fine - i'm going to check once i clone
[07:08:54] *** Sieghard has quit IRC
[07:12:10] <stevel> it is a little bit reassuring to know that it was my manual mistake that caused it to pull the wrong rev
[07:13:00] <icon> does anyone know offhand how long the contents of /var/tmp typically last?
[07:13:05] <icon> (ie: is it purged at startup?)
[07:13:26] <jmcp> icon: the contents are purged when you purge them
[07:13:35] * icon nods
[07:13:40] <icon> just wanted to be sure
[07:13:41] <jmcp> it's /tmp that's made pure-n-clean on boot
[07:13:55] <icon>  /tmp also sucks up swap space, no?
[07:14:28] <jmcp> not as such
[07:14:46] <icon> well, /tmp shares disk with swap
[07:14:50] <icon> that might be more accurate
[07:17:25] <twincest> is there a description anywhere of what bugs should be P1, P2 etc?
[07:17:28] <icon> oh well, back to sleep.
[07:21:42] *** sickness has joined #opensolaris
[07:22:36] *** bougie has joined #opensolaris
[07:23:29] <richlowe> twincest: No.
[07:23:34] <twincest> oh well.
[07:23:56] <richlowe> 1 is highest, 5 is lowest.
[07:24:00] <richlowe> that's all you get, sadly.
[07:24:02] <Gman> bugster down at the moment - be warned
[07:24:15] <richlowe> yeah, don't even try when bugster is down.
[07:24:22] <richlowe> what happens with stuff then seems fairly random.
[07:24:31] <twincest> does that mean the bug i just submitted went disappeared into the void?
[07:24:33] <richlowe> either it bounces to Linda(?) and gets added anyway, or it doesn't.
[07:24:51] <richlowe> twincest: you'll find out if/when you get confirmation, pretty much.
[07:24:58] <twincest> fun
[07:25:19] <richlowe> the more you use b.o.o the more you'll like it :)
[07:25:24] <jmcp> twincest: p1 == system is down / repeatably panicing / data integrity issue
[07:25:28] <twincest> the requirement for a CR to integrate is much easier for people who can actually file bugs :)
[07:26:02] <richlowe> twincest: it's even more fun if you get to modify them.
[07:26:13] <richlowe> rather than sitting on them until you have the text as absolutely finished as you'd ever want, *then* file them.
[07:26:40] <twincest> mmh, well this is a two-line change, so i'm fairly sure what i want to say about it
[07:27:23] <stevel> oh wicked wicked rice wine. you've totally destroyed my afternoon of productivity
[07:27:39] <jmcp> stevel: afternoons are never supposed to be productive
[07:27:49] *** tsoome has joined #opensolaris
[07:27:53] <stevel> :)
[07:27:56] <twincest> (amusingly, the bug was caused by a missing check of return value... which someone had actually cast to (void) instead of fixing)
[07:29:06] <Gman> stevel, good to see you're enjoying your per diem :)
[07:29:41] <richlowe> stevel: and yeah, it's good to know it's just a typo.
[07:29:49] <richlowe> it'd be better to know it's the only thing wrong, too, though.
[07:30:00] <richlowe> twincest: eh, I'm just bitter about the whole situation :)
[07:30:13] <stevel> richlowe: indeed. which is why i'm still going to validate it against /ws/onnv-gate
[07:30:18] <twincest> i was around this time last year, i'm sure it won't take long ;-)
[07:30:40] <twincest> i am *finally* going to get my fix for 1214359 integrated though
[07:30:47] <richlowe> I'm wondering how much general unhappyness it'd take.
[07:32:18] <richlowe> stevel: Yeah, I know.
[07:32:40] <richlowe> twincest: I just find it funny they have all these goals for participtation, and absolutely nothing there to actually support it.
[07:32:50] <richlowe> it's funnier still that the so many haven't actually realized that yet...
[07:33:29] <twincest> well, i could understand that when it started
[07:33:42] <twincest> the fact that there's *still* no useful bug database is disappointing though
[07:33:58] <twincest> since it's fairly integral to developing anything
[07:34:01] <richlowe> there's more to it than that, but yeah.
[07:34:16] <twincest> probably, but i haven't done much yet, so it's all i've noticed :)
[07:34:17] <delewis> lawyers. ugh
[07:34:43] <richlowe> twincest: the rest isn't quite so bad.
[07:34:49] <richlowe> pretty much, you can write the code, and mostly test the code.
[07:34:54] <richlowe> and absolutely nothing else, right now.
[07:34:57] <delewis> we've at least got a public repo that's a mirror of the internal TW one.
[07:35:07] <twincest> am i expected to write text to be inserted into the manual page?
[07:35:21] <Gman> ideally your sponsor would pick up on that
[07:35:26] <twincest> okay
[07:35:27] <richlowe> twincest: Yes.
[07:35:36] <richlowe> Gman: they do, they say "and you'll need to provide manpage diffs"
[07:35:37] * delewis remembers when it was the case that we'd have to be patient to receive tarballs of ON releases.
[07:35:40] <twincest> .. diffs?
[07:35:43] <Gman> richlowe, harsh
[07:35:58] <richlowe> twincest: dump the formatted page to a file, cp the file, change it, diff it.
[07:35:59] <twincest> i meant text as opposed to diffs, since the manpages aren't very open yet :)
[07:36:02] <twincest> ah, okay
[07:36:13] <richlowe> Gman: actually I've done one set, someone else has done another and had me review it.
[07:36:24] <richlowe> Gman: it depends on whether they notice that the manpages aren't under a license that lets us change them or not. :)
[07:37:09] <richlowe> twincest: as I understand it, you provide content, the manpage folks do the formatting gunk.
[07:37:14] <Gman> richlowe, nod
[07:37:15] <twincest> nod
[07:38:02] <richlowe> delewis: different people, I believe.
[07:38:33] <richlowe> I'm *certainly* not bitching about anything stevel has done...
[07:38:48] <Tpenta> stevel has done wonders
[07:38:49] <stevel> (not at the moment, anyway) ;-)
[07:39:04] <richlowe> stevel: I haven't at all, I've just pointed out stuff that's wrong.
[07:39:08] <richlowe> at least, I don't think I have.
[07:39:15] <richlowe> if some of this *is* your fault, however... ;)
[07:39:16] <stevel> i'm kidding. ;)
[07:39:33] * stevel slices
[07:39:34] * stevel dices
[07:39:46] * stevel breaks the webpage :(
[07:39:49] <richlowe> Tpenta: yeah, it's all the other stuff that's stuck.
[07:39:52] <Tpenta> so you went to korea but you couldnt work a side trip down here :)
[07:40:14] <stevel> t'would be hard to justify :(
[07:40:19] <richlowe> Tpenta: there's going to come a time when someone is going to have to think about WebRTI, for instance.
[07:40:20] <Gman> richlowe, tonic team needs to expand to a few more people
[07:40:26] <richlowe> Tpenta: of course, by that time, it'll be way too late... but.
[07:40:39] <Tpenta> actually I think webRTI should not be too hard
[07:40:56] <richlowe> Tpenta: given it deals with NFS, I think that may not be correct.
[07:41:05] <richlowe> unless the rti checks in wx suggest something not actually the case.
[07:41:10] <Tpenta> i wasnt aware that it currently used nfs
[07:41:11] <richlowe> (they check them via NFS to a host I forget)
[07:41:21] <Gman> if i was a bit more of a developer, i'd ask my boss to work 50% on opensolaris.org stuff
[07:41:26] <Tpenta> i'd have to check the current nv wx code
[07:41:47] <Tpenta> gman: why would you want your boss working 50% on opensolaris? ;)
[07:41:55] <Gman> Tpenta, ;)
[07:42:17] <stevel> we could use some more hands
[07:42:28] <richlowe> it tries to use webrticli from webrti.sfbay, if that fails it looks on wizard.eng and onstc.eng
[07:42:32] <stevel> and in the case of b.o.o. we could use some more brains too
[07:42:37] <richlowe> I've been assuming the front end literally writes to those.
[07:42:40] <Gman> stevel, are you openly accepting people?
[07:42:59] <Tpenta> if the p-team meeting moves as suggested I might start joining them again
[07:43:01] <Gman> stevel, i might be able to convince karl to lend out some people
[07:43:04] <richlowe> Tpenta: the bug check that was added recently calls monaco straight via monaco.sfbay, too.
[07:43:13] <stevel> i don't think we're allowed to open reqs at the moment
[07:43:14] <richlowe> Tpenta: and can't even be sanely faked up from outside, really.
[07:43:22] <richlowe> though if you jump hoops with xslt and b.o.o you can get close.
[07:43:42] <Gman> stevel, i'm meaning people working with you while still being part of another group
[07:43:54] <richlowe> stevel: You can't fix most of what's wrong with b.o.o I thought?
[07:44:01] <richlowe> it'd been said that all requires BT2 changes.
[07:44:22] <stevel> richlowe: yeah, that's a different team.  not all of it requires BT2 changes.  the whole notification retardness for example
[07:44:42] * richlowe had been told that did
[07:44:44] <stevel> gman: ahhh... yeah that'd be nice - certainly on the ops side of things, it'd be nice to have more hands for maintaining the machines
[07:44:52] <stevel> well, it requires BT2 changes - but not *hard* BT2 changes
[07:44:57] <Gman> stevel, have to be local?
[07:45:09] <stevel> gman: i don't think so.  garypen and i are currently the two on it, and he's in the UK
[07:45:26] <Gman> stevel, ok, i'll ask karl, see if he can spare some people, if only on a part time basis
[07:45:39] <Gman> stevel, will tell him to get in touch with you & karyn?
[07:45:59] <stevel> well, let me check with bonnie and karyn first to make sure they're okay with outside hands
[07:46:09] <stevel> i don't see why they wouldn't be... but well.. .just to be safe
[07:46:16] <Gman> yeah, ok
[07:46:20] *** PosixC has quit IRC
[07:47:43] <richlowe> I'm still hoping the notifications aren't at all indicative of trust.
[07:47:47] <richlowe> if they are, you guys are screwed.
[07:48:23] <richlowe> though I guess they do serve a purpose, now I know when to ask someone to tell me what changed. :)
[07:49:38] *** bougie has quit IRC
[07:58:48] *** polk__ has quit IRC
[08:00:19] <Gman> http://www.youtube.com/watch?v=EQ02O67Nq_w
[08:00:20] <Gman> oh boy.
[08:00:23] <twincest> hmm, editing the roff and diffing doesn't work as well as you'd hope
[08:01:08] *** OnkelSchorsch has quit IRC
[08:01:54] <timeless> hi richlowe
[08:04:56] <timeless> so, is there any page somewhere that lists the CFLAGS/LDFLAGS for linking a solaris kernel module? P)
[08:05:21] <twincest> Nov  6 05:55:36 nvdev dls: [ID 943737 kern.info] NOTICE: Data-Link Services v%I% loaded
[08:05:22] <twincest> hm :)
[08:05:42] <timeless> v%I% eh?
[08:06:43] <Griff> Gman, you owe me 2 minutes of my life back
[08:06:43] <stevel> did you build that from the onnv-gate repository?
[08:06:54] <stevel> griff: you really want 2 minutes of Gman's life?
[08:06:55] <twincest> stevel: yes.  i can see the cause, but it's still amusing
[08:07:02] <stevel> twincest: yeah :) unexpanded SCCS keywords
[08:07:09] <Gman> Griff, heh
[08:07:16] <Gman> Griff, what's up dude?
[08:07:25] * Griff is happy
[08:07:43] <Griff> I'm waiting for OSS to come to me with some sun quotes
[08:07:58] <timeless> igm tgat;s a SCCS?
[08:08:07] <timeless> oh, that's a SCCS?
[08:08:10] <Griff> On Friday I had budget approval for an entry level SAN, which means I can use SUN servers again
[08:08:14] * timeless makes a mental note
[08:08:19] <Griff> Roll on the x4100s!
[08:08:24] <twincest> timeless: yes, sccs uses %X% keywords (where X is one letter)
[08:08:24] <Gman> oh, sweet :)
[08:08:39] <Griff> Open System Specialists sure are taking a while though
[08:08:40] <Gman> though i'm not sure how this means i owe you 2 minutes ;)
[08:08:48] <richlowe> timeless: Makefile.master has 'em.
[08:09:01] <twincest> here's a thought: will Hg keywords accidentally inserted into onnv-gate get expanded on my end?  or is it like svn (have to be enabled manually)?
[08:09:01] <Griff> oh, that damn youtube clip
[08:09:16] <Gman> ahhh
[08:09:19] <Gman> yes, it was duff.
[08:09:34] <richlowe> twincest: since hg doesn't do keywords natively, no.
[08:09:34] <timeless> richlowe: yeah i found that file
[08:09:43] <timeless> i couldn't figure out what i'm missing though :(
[08:09:46] <Gman> i thinking about making a video too
[08:09:56] <richlowe> twincest: if you were to manage to insert the right format (which I don't know, currently), maybe.
[08:10:16] <richlowe> timeless: there's a few bits undefining i386 in the amd64 case, and with gcc the -mno-red-zone thing.
[08:10:25] <Gman> [mostly about how everyone internally sucks for not getting more involved in opensolaris]
[08:10:47] <timeless> richlowe: i'm using forte or whatever
[08:10:54] <timeless> CFLAGS_amd64 = -xarch=amd64a -xmodel=kernel -Ui386 -U__i386
[08:10:56] <Griff> sounds good
[08:11:09] <Griff> I'll vote for you, though I doubt outsiders count
[08:11:16] <timeless> but my poor system panics because the local variables aren't relocated correctly or something
[08:11:25] <timeless> CFLAGS_PIC = -KPIC
[08:11:35] <timeless> CFLAGS= -D_KERNEL $(CFLAGS_$(ARCH)) $(CFLAGS_PIC) $(XCFLAGS) -g
[08:11:49] *** jwt has joined #opensolaris
[08:13:36] <jwt> hey ... i'm trying to build openoslaris from source on amd64 and the readme implies i should get sun studio 11. the link on http://opensolaris.org/os/community/tools/sun_studio_11_tools links to a for purchae product. any ideas on where i can find the free version used to build opensolaris.
[08:13:37] <jwt> thx.
[08:13:46] *** Griff is now known as Gr|ffious
[08:13:49] *** Gr|ffious is now known as Gr|ffous
[08:13:55] <richlowe> Studio 11 shouldn't be for purchase in general.
[08:13:58] <timeless> one of the links there is to  afre tarball
[08:14:07] <timeless> and if you follow some other path you get it for free
[08:14:12] <timeless> it just wants registration
[08:14:19] <jwt> k
[08:14:31] <twincest> the package version is free too if you register for SDN (free)
[08:14:54] <richlowe> and in fact, there's only the one link to SS11 on that page, and it's not for purchase.
[08:15:15] <richlowe> also, you want the SS11 with *exactly* the patches in the tarball image.
[08:15:24] <Gr|ffous> Now this might be a funny questing in a "sunish" channel, but does anyone know how the HP MSA1000's SANs are? The storageTek stuff is way too expensive for our budget
[08:15:25] <richlowe> certainly no less, don't know about more.
[08:15:58] <jwt> the url i gave? it goes to a reg page (that i built long ago interestingly enough) and once i get logged it indicates the product is for purchase and no other link is provided.
[08:16:00] <twincest> i hope they update it to fix the latest dbx  bug soon :)
[08:16:15] <richlowe> jwt: the URL you gave 404's
[08:16:18] *** polk__ has joined #opensolaris
[08:16:22] <jwt> my bad
[08:16:25] <richlowe> the correct all takes you to registration, then the tarball download.
[08:16:29] <richlowe> s/all/url/
[08:16:40] <richlowe> (well, link -> registration/login -> download)
[08:17:03] <twincest> i just followed the link, logged in and it shows me the download page
[08:17:05] <richlowe> and the very *moment* you hit the download page, the damn thing emails you.
[08:17:06] <richlowe> still.
[08:17:12] <twincest> (which isn't very surprising as i just downloaded it yesterday)
[08:17:27] <asyd> \_o<
[08:17:28] * richlowe acquires yet another confirmation email
[08:17:45] <stevel> richlowe: collect them all!
[08:18:00] <stevel> (sorry, retarded pokemon reference)
[08:19:27] <timeless> heh
[08:19:48] <Triskelios> you managed to make that too vague
[08:20:28] *** coolvibe has quit IRC
[08:20:44] *** laca has quit IRC
[08:20:47] * timeless got the reference :(
[08:20:54] <Gman> the quote was wrong
[08:20:58] <richlowe> indeed.
[08:21:41] * stevel apologies
[08:24:12] <stevel> apoligises even
[08:24:37] <stevel> apologises too
[08:24:38] <stevel> ugh
[08:25:06] <richlowe> So, the rice wine is good, huh? :)
[08:25:18] <Gr|ffous> hahaha
[08:25:39] <jwt> ahhhh, there's the binary this time. interesting.
[08:25:48] <jwt> onward now
[08:25:56] <Gr|ffous> I leave for a moment, come back and there's pokemon on screen. I thought Xchat2 had done something very wrong for a moment there
[08:25:59] <stevel> richlowe: oh totally :-)
[08:27:29] <jwt> v for vendetta ... kind'a sucky flick imho
[08:27:50] <jwt> watching it now as i try to get this build off the ground
[08:28:00] <LeftWing> V for Vendetta was great!
[08:28:06] <twincest> you know, if the solaris copyright statement (on boot) was changed to include the name of every contributor, it could start to rival the length of hp-ux's
[08:28:09] <jwt> really?
[08:28:16] <jwt> boring the crap out of me
[08:28:19] * Gr|ffous throws something heavy at jwt. Dude. It was great!
[08:28:32] <LeftWing> Yes, it's a brilliant redux of the core of Orwell's 1984 with a more modern spin.
[08:28:35] <jwt> well, it's not over yet and i'm still watching
[08:28:49] <LeftWing> It's all too appropriate in today's time, too. =P
[08:28:52] * LeftWing &
[08:28:55] <jwt> she's in prison now. feeding the rat, etc.
[08:29:16] <jwt> i can see the parallels.
[08:29:16] *** PosixC has joined #opensolaris
[08:29:31] *** PosixCompliant has joined #opensolaris
[08:29:48] *** PosixCompliant has left #opensolaris
[08:41:06] *** Gman is now known as GmanAFK
[08:42:20] *** rafaeldtinoco has joined #opensolaris
[08:42:21] *** rafaeldt has quit IRC
[08:44:57] *** karrotx has quit IRC
[08:54:55] *** robj has quit IRC
[08:55:31] <PosixC> yesterday I asked about a panic when booting installation DVD on niagra
[08:55:45] <PosixC> I was adviced to check md5sum
[08:56:06] <PosixC> and I did it thus : md5sum /dev/hdd (/dev/hdd is the cdrom)
[08:56:28] <PosixC> and the result was different than the md5sum of the iso
[08:56:39] <PosixC> However I am not sure at all
[08:56:40] <PosixC> because
[08:56:52] <PosixC> I made the same on linux and it was different
[08:57:21] <PosixC> also but the DVD there is good - passed medaicheck and I CAN install from it
[08:57:49] <Triskelios> /dev/hdd is not a solaris device...
[08:58:14] <Triskelios> or, I guess you mean you did the md5sum on a linux system?
[08:58:19] <PosixC> Triskelios, I checked the solairs DVD on linux
[08:58:26] <PosixC> right
[08:59:01] <PosixC> googling for it showed that due to DMA results can be different , any idea?
[08:59:45] <PosixC> for eaxmple: http://www.mail-archive.com/fedora-legacy-list at redhat dot com/msg01803.html
[08:59:48] <Triskelios> the results only differ if you had read errors...
[09:00:51] <PosixC> and I have :
[09:00:52] <PosixC>  hdparm /dev/hdd |grep dma
[09:00:53] <PosixC>  HDIO_GETGEO failed: Inappropriate ioctl for device
[09:00:53] <PosixC>  using_dma    =  1 (on)
[09:01:30] <PosixC> Triskelios, but I should see this read errors in the log/console, right ?
[09:01:40] <PosixC> in case there are such errors
[09:01:45] <Triskelios> PosixC: no, it could just silently corrupt data
[09:02:09] <PosixC> so maybe I will turn off dma on /dev/hdd ?
[09:02:30] <Triskelios> worth a shot. this isn't a linux help channel fyi...
[09:02:42] <PosixC> my problem is that I want to check if the problem is the DVD of the T2000 or the media on which the iso was burned
[09:03:04] <dwc-> "it could just silently corrupt data" ? that sounds like FUD to me ...
[09:03:09] <PosixC> so : what is a reliable way to check it? it seems to me a pure solaris question
[09:03:42] <Triskelios> dwc-: why? the dma transfer can succeed just fine with bad data
[09:04:18] <Triskelios> and that happens sometimes with flakey hardware or drivers
[09:04:48] *** Cyrille has joined #opensolaris
[09:04:49] <elektronkind> or dust
[09:05:23] <dwc-> that should produce a read error, or a write error when it was written
[09:05:50] <elektronkind> cd could have been fine when it was written :)
[09:05:53] <dwc-> if the media's bad, burn another one
[09:06:02] <dwc-> and do write verification
[09:06:23] <PosixC> " write verification" - you mean while burning ?
[09:06:26] <dwc-> yes
[09:06:55] <PosixC> is there no way to check media has the md5sum ? ?
[09:07:15] <dwc-> if it's the dvd of the t2000, try booting other media, or using the disk in another machine
[09:07:22] <dwc-> dd the data back off?
[09:07:31] <PosixC> in other OSs there is a "mediacheck" which in fact checks md5sum/sha1sum
[09:07:37] <dwc-> or readcd, or whatever is reliable for you
[09:12:57] <stevel> hrm. i've got a nightly running, a kernel build running, and an lucreate running
[09:13:04] <stevel> and i need to go back to the hotel
[09:13:09] <stevel> and it's raining out
[09:13:50] <stevel> should be fun walking back with my laptop
[09:17:30] <PosixC> dwc-, should I run dd if=/dev/hdd of=sparcSXCR50.iso
[09:17:46] <PosixC> or is it better to add some bs and count for better
[09:18:15] <PosixC> performance ? I don't know a lot about cdrom read i/o
[09:19:48] <richlowe> stevel: builds on the laptop?
[09:19:59] <stevel> richlowe: yeah
[09:20:07] <Tpenta> that sounds time consuming
[09:20:08] <stevel> nightly is running for the demo i'm giving
[09:20:21] <stevel> kernel build is running from some work i did on the laptop during my flight
[09:20:36] <richlowe> the demo nightly I can understand, I figure.
[09:20:47] <richlowe> I think I'd have waited on a build machine for the other though. :)
[09:20:59] <twincest> hey, i just built it inside vmware on a laptop
[09:21:00] *** |tsoome| has joined #opensolaris
[09:21:12] <richlowe> Well, *I* wouldn't have, but if I were you, I would. :)
[09:21:14] *** orange80 has joined #opensolaris
[09:21:44] <stevel> :) yeah, i probably could have waited on the kernel build
[09:21:47] <stevel> but oh well
[09:22:15] <richlowe> steal one of Tpenta's, he's more local than the others. :)
[09:22:39] <richlowe> and I recall the last thing he mentioned being rather nice.
[09:22:56] <Tpenta> the 4800 or the v40-z?
[09:23:35] * stevel saw an x4600 in the customer solutions centre here
[09:23:39] <stevel> wonder if i can get on it :-P
[09:25:05] *** logic has joined #opensolaris
[09:25:09] <Gr|ffous> ok, so I'm a novice here. I compile postfix and it's having linking problems. Desperate to not do the LD_LIBRARY_PATH trick, what am I mean to feed make with ?
[09:25:19] <Gr|ffous> *meant
[09:25:36] *** rafaeldtinoco has quit IRC
[09:25:49] <Gr|ffous> like this? make makefiles 'CCARGS=-DHAS_PGSQL -I/usr/local/include/pgsql' 'AUXLIBS=-L/usr/local/lib -lpq' 'LDFLAGS=-R/usr/local/lib/'
[09:25:56] <twincest> yes
[09:26:02] <Gr|ffous> the program didn't even compile with that :/
[09:26:09] <twincest> maybe you should show the errors
[09:26:15] <richlowe> Tpenta: 4800.
[09:26:16] *** rafaeldt has joined #opensolaris
[09:26:23] <Gr|ffous> if you're happy to help, sure
[09:26:50] <richlowe> Tpenta: sure, he probably needs an x86 build, but *I* wouldn't pass up the chance for a decently fast sparc build ;)
[09:27:36] <Gr|ffous> http://www.rafb.net/paste/results/H9fCFQ31.html
[09:27:37] <orange80> can anyone recommend a good used laptop that will run solaris well for sure and wifi too?
[09:29:49] *** |tsoome| has quit IRC
[09:30:06] *** |tsoome| has joined #opensolaris
[09:31:52] <stevel> hrm. i need a good Solaris performance/tuning slide-deck
[09:31:57] <stevel> anyone know of one off-hand?
[09:32:34] <richlowe> Hm, brendan may know of one?
[09:32:47] <Gr|ffous> any thoughts twincest?
[09:32:56] <Doc> brendan has one on his website
[09:33:07] <twincest> griff: that's not a linking problem
[09:33:11] <Doc> not sure about the legalities of using it tho... lemme find it
[09:33:29] <twincest> (also nothing to do with ld_library_path)
[09:33:36] <Gr|ffous> twincest, what's interesting, is that if I leave the -R stuff off, it works
[09:34:03] *** anthony79 has quit IRC
[09:34:04] <Gr|ffous> So maybe I'm clobbering some system default LDFLAGS?
[09:34:22] <twincest> LDFLAGS should have nothing to do with compilation.. that looks like a bug in the build system
[09:34:39] *** tsoome has quit IRC
[09:34:51] <stevel> thanks doc
[09:35:30] <Doc> stevel: http://www.context-switch.com/performance/dtrace_workshop01_slides.pdf  probably not exactly what you're after, but possiblt a good starting point
[09:35:32] * Gr|ffous tries again, knowing how computers are just so unpredictable..
[09:35:53] <stevel> thanks i'll take a look
[09:36:09] <twincest> is your postgres_ext.h in /usr/local/include/pgsql?
[09:36:27] <Gr|ffous> ok, so it's breaking the old way too now. I'll untar and start again
[09:40:39] <Gr|ffous> it's in /usr/include/pgsql
[09:41:35] <twincest> then why are you passing -I/usr/local/include/pgsql ?
[09:41:52] *** coolvibe has joined #opensolaris
[09:42:09] <Gr|ffous> hehe
[09:42:17] <Gr|ffous> because I'm blindly copying and pasting!
[09:42:19] <Gr|ffous> lol
[09:42:42] <twincest> nothing good can come of this
[09:42:49] <Gr|ffous> sorry
[09:43:06] <twincest> hmm, no Intel cpu frequency scaling in frkit :(
[09:43:56] <twincest> i don't suppose powernow source code is available?
[09:44:02] <stevel> twincest: yes there is
[09:44:08] <twincest> there is?  what module?
[09:44:08] <stevel> powernow supports speedstep
[09:44:13] <twincest> hm:  Probing powernow ... not applicable.
[09:44:17] <twincest> maybe vmware is interfering
[09:44:21] <stevel> at least it does on my machine
[09:48:03] *** jwt has quit IRC
[09:48:49] * richlowe isn't sure how vmware would attempt to emulate that.
[09:49:09] <twincest> well, the CPU is already doing it
[09:49:20] <twincest> i want solaris to notice so timekeeping is less confused :)
[09:49:30] <richlowe> Ah, you're seeing the stuff gisburn was complaining about.
[09:49:31] <richlowe> well, some of it.
[09:57:44] <stevel> the bridge is up to september 4th
[09:57:49] <stevel> only two months left to go :-P
[09:58:18] <twincest> hm, when changing a gettext string, what else do i need to do?
[09:58:19] <richlowe> and a really timeconsuming diff.
[09:58:48] <richlowe> 'time consuming' may as well be one word, right?
[09:59:20] <richlowe> twincest: Nothing that I know of, unless you speak a whole bunch of languages and want to fix the g11n bits, too ;)
[09:59:27] <twincest> heh.  ok
[10:06:21] *** sala has quit IRC
[10:08:59] <Triskelios> does anyone know how well nevada deals with laptop apic quirks?
[10:10:06] *** alanc has quit IRC
[10:10:07] <twincest> hm.  request-sponsor seems like a bad solution
[10:10:15] *** GmanAFK has quit IRC
[10:10:23] <richlowe> to what?
[10:11:02] *** alanc has joined #opensolaris
[10:11:05] <twincest> getting things putback
[10:11:10] <twincest> it's not very scalable
[10:11:12] *** GmanAFK has joined #opensolaris
[10:11:23] <richlowe> twincest: it hasn't had to scale particularly far, so far.
[10:11:31] <twincest> still
[10:11:42] <richlowe> well, depending on what you mean.
[10:11:57] <richlowe> the finding a sponsor bit isn't bad.  The rest depends on how much other stuff the sponsor has to do as well.
[10:12:09] <richlowe> since they have to do a whole bunch of the change-related stuff.
[10:12:21] <twincest> it seems like fixes should go back to the CR somehow, rather than through a mailing list
[10:12:32] <richlowe> twincest: that would work, except we can't change CRs.
[10:12:46] <twincest> i said it's a bad solution, not an unnecessary one :)
[10:12:50] <richlowe> and can neither see, nor put anything into either Evaluation nor Suggested Fix.
[10:12:57] <richlowe> I'm agreeing.
[10:13:06] <richlowe> except given the rate of change for b.o.o, the alternative would be several years away :)
[10:13:35] <stevel> are we back on b.o.o. again?
[10:13:38] <richlowe> if you're filing the CR, and you have the fix, stuff it into workaround.
[10:13:41] <richlowe> that's what everyone else does.
[10:13:42] <stevel> :)
[10:13:57] <stevel> richlowe: don't you sleep?
[10:14:06] <richlowe> twincest: last I mentioned something close to what you were suggesting, I was told it was part of the 'workflow' bits of the roadmap.
[10:14:12] <richlowe> and very little else.
[10:14:16] <twincest> oh good, a roadmap
[10:14:20] <twincest> i can stop worrying :)
[10:14:24] <richlowe> hah.
[10:14:45] <richlowe> it means very little beyond "we intend to, someday"
[10:15:18] <richlowe> stevel: isn't insomnia fun?
[10:15:49] <stevel> richlowe: ugh. i had a bout of insomnia once during my thesis work.  one of the worst experiences of my life
[10:18:48] <twincest> apparently secure-by-default turned off nfs/client
[10:18:52] <twincest> is that really necessary?
[10:19:05] <richlowe> it turns off everything networky except for ssh.
[10:19:09] <richlowe> either disables them, or makes them local-only.
[10:19:36] <richlowe> I thought the nfs services were meant to deal with themselves though, maybe that only applies to the server side of things?
[10:19:55] *** Tpenta has quit IRC
[10:20:21] <twincest> well, it wasn't mounting anything at startup, i assumed that was why
[10:21:19] *** B|nTaRa has quit IRC
[10:25:56] *** damienc has joined #opensolaris
[10:26:01] *** coolvibe has quit IRC
[10:27:23] <stevel> hrm.
[10:27:27] <stevel> i'm perplexed
[10:28:07] <stevel> why did my repo before i regenerated have 3104 revisions, but my new one only has 3070
[10:28:23] <stevel> somewhere along the way, 34 revisions disappeared
[10:28:27] <stevel> which is slightly worrisome
[10:31:11] <Gr|ffous> come now, you didn't really want ALL of them did you ;)
[10:31:24] <raph_ael> hello
[10:31:25] <stevel> kinda, yeah
[10:31:25] <stevel> :-P
[10:33:59] <stevel> ohhhhh
[10:34:00] <stevel> build tags
[10:34:01] <stevel> phew
[10:36:30] <richlowe> sure would be nice to figure out what to do about that stuff.
[10:36:47] <richlowe> since you regenerated anyway, I mean.
[10:38:31] <stevel> what do you mean?
[10:39:55] <richlowe> Nobody ever really said anything about whether that would be kept up going forward.
[10:40:09] <stevel> the build tags?
[10:40:17] <richlowe> Yes.
[10:40:46] <stevel> i assumed we were keeping them?
[10:41:22] <richlowe> That kinda depends on the 'we'.
[10:41:36] <richlowe> I've commented on it, as have you.  Danek, as best as I know, hasn't.
[10:41:47] <richlowe> neither has anyone else particularly authoritative.
[10:42:04] <richlowe> and I appear to have misplaced a FIXME
[10:42:07] <stevel> i don't see anything wrong with keeping them
[10:44:53] <Gr|ffous> If I'm trying to compile some software, and I don't see those arguements flying past amongst all the GCC lines as the software compiles, does that mean I've done something wrong in the Makefile?
[10:45:01] *** coolvibe has joined #opensolaris
[10:45:09] <Gr|ffous> sorry, it's the LDFLAGS that I'm not seeing
[10:45:38] <quasi> you should probably only see those when linking
[10:46:34] <Gr|ffous> I'm all but convinced that postfix's make system is broken *sigh*
[10:47:41] <quasi> really? I always thought it built pretty well, but admittedly it has been a while since I last needed to build postfix
[10:48:14] <Gr|ffous> I have spent 3 hours now, I really am feeling defeated
[10:48:34] <Gr|ffous> all just to try and get -R/usr/local/lib passed :/
[10:52:14] <Gr|ffous> It's the only progam I can remember installing that doesn't come with ./configure
[10:54:01] *** calumb has joined #opensolaris
[10:55:44] <Berny> morning
[10:56:04] <quasi> looking at a couple of makefiles, they do seem to have CFLAGS  = $(DEBUG) $(OPT) $(DEFS) and LIBS    =
[10:56:23] <quasi> if nothing else, you could always edit those ;)
[10:57:20] <Gr|ffous> I ended up editing the Makefile after it generated itself, that appears to have worked. No matter what I did, it wouldn't pickup my LDFLAGS. I tried exporting them, and specifing them as part of the make command.
[10:57:43] <Gr|ffous> $(OPT) seems to just map to CC=
[10:58:54] <Gr|ffous> my next problem is that /usr is RO, since I followed a guide that said I should configure /usr as a inherit-pkg-dir.
[10:59:02] <Gr|ffous> I fear the night isn't long enough for this
[11:02:46] <quasi> right, that one is pretty hard to get around
[11:02:59] <Gr|ffous> so it's not just me?
[11:05:18] <quasi> the /usr thing
[11:05:30] <Gr|ffous> oh, that
[11:06:00] <Gr|ffous> so what is the best practice there? I'm swimming in docs.sun atm trying to work that one out
[11:06:51] <quasi> sticking to sendmail ;)
[11:06:56] <stevel> lol
[11:07:15] <Gr|ffous> erm
[11:07:48] <Gr|ffous> I'm sure it's a common enough problem that people hit
[11:08:05] <Gr|ffous> I suppose that's where zfs snapshots & clones come into play
[11:08:29] <Gr|ffous> if /usr were it's own fs, then I could make it rw in the zone, without wasting too much space
[11:08:57] <quasi> or you could install postgres in the global zone
[11:09:02] <quasi> postfix
[11:09:27] * quasi really should have had that morning coffee a few hours ago
[11:10:43] <Gr|ffous> This is both my first postfix install-from-source, and my first stab at a zone. I'm sure if I shortcut it now, I'm just going to hit it again? /usr must get used quite a bit, or is the idea to install all new software into /opt perhaps?
[11:10:58] <Gr|ffous> in my case, both of those folders are rather large
[11:12:05] *** alobbs has joined #opensolaris
[11:12:16] <Berny> hmm, why does a zfs share pool do nothing? i.e. i cannot mount it from another box.
[11:12:44] <stevel> did you set it to be shareable?
[11:12:47] <Gr|ffous> nfs/server is running?
[11:12:48] <stevel> check the 'sharenfs' property
[11:12:49] *** GmanAFK is now known as Gman
[11:12:58] <stevel> zfs get sharenfs pool
[11:13:05] <Berny> NAME             PROPERTY       VALUE                      SOURCE
[11:13:06] <Berny> data/goettert    sharenfs       on                         local
[11:13:27] <quasi> Gr|ffous: the problem with dumping postfix in /opt is that it really wants to replace sendmail to be happy
[11:13:34] <Berny> and why does changing that property to something "real" take like minutes?
[11:13:44] <Berny> it's running for a few minutes now...
[11:14:25] <stevel> does showmount -e server show it's actually being exported?
[11:14:38] <Berny> nope
[11:15:33] <andersmo> /usr/{bin,lib}/sendmail is an application interface as much as anything when it comes to sending email on unix systems. That's why postfix wants to grab it. =)
[11:16:00] <stevel> try 'zfs unshare -a' and then 'zfs share -a'
[11:16:05] <Gr|ffous> at the risk of missing something big, is that really all that much of an issue for an MTA
[11:16:28] <Berny> stevel: as soon as that bloody property got set
[11:16:30] <Gr|ffous> this zone is for that purpose, I've never ever local mail on a unix sysmet
[11:16:33] <Gr|ffous> system
[11:16:44] <andersmo> Well, cron and friends send email and expect a sendmail to pipe it into?
[11:16:52] <stevel> berny: that should be pretty quick. how are you verifying that it's taking several minutes?
[11:16:55] <Berny> i can't even kill the zfs set sharenfs=blabla command
[11:17:04] <twincest> you can configure sendmail to deliver mail to postfix at localhost
[11:17:12] <Berny> stevel: because it's still running?!
[11:17:32] <stevel> what command did you type?
[11:17:38] <stevel> that should be almost instantaneous
[11:17:53] <Gr|ffous> I suppose the easy answer is just to remove the inherit-pkg-dir for that zone
[11:17:58] <Berny> zfs set sharenfs=rw= at 149 dot 203.91.0/24 data
[11:18:09] <andersmo> Yep, you can. At the cost of a bit of mess. Postfix' paths are configurable, everything can be configured in the config files. =)
[11:18:20] <stevel> do you have a lot of child filesystems off of data?
[11:18:24] <Berny> though a zfs set sharenfs=no takes about as long
[11:18:35] <Berny> so far 3 children
[11:18:39] <stevel> hrm
[11:18:42] <Berny> though a zfs set sharenfs=on takes about as long
[11:18:47] <Gr|ffous> andersmo, Mess? or extra disk usage?
[11:18:58] <Berny> ^C or ^Z don't do anything either
[11:19:04] <stevel> truss it
[11:19:11] <Berny> kill -9 doesn't bother it either
[11:19:12] <stevel> see what the hell it's doing
[11:19:19] <Berny> truss -p gives NO output
[11:21:58] <andersmo> Gr|ffous: just a matter of personal preference, really. I prefer wiping out sendmail and replacing it with postfix - but if removing sendmail is infeasible (as when you inherit the /usr file system into the zone), it makes sense to keep it around for local mail submission and let postfix handle the other stuff you want it to. =)
[11:22:56] <Berny> woohoo it finished
[11:23:27] <Gr|ffous> oh I'm after preference and advice, thanks :)
[11:23:52] <Berny> ok, off to the next problem
[11:24:20] <Gr|ffous> so you'd suggest keeping /usr inherited, but to put the various binaries else, and update the master.cf accordingly to make it all fit nicely into the zone?
[11:24:47] <Gr|ffous> this is a testzone, if I fudge it completely (and I expect to), I'll just start again.
[11:25:02] <Berny> now that it is shared... why do i get the individual filesystems mounted as owner root with 700 perms?
[11:25:38] <quasi> Gr|ffous: then you'll still be missing /usr/lib/sendmail and friends in the postfix version
[11:25:59] <Berny> pretty useless for a homelike dir :-\
[11:26:51] <Gr|ffous> quasi, Sorry, I'm not quite following what it is that you recommend then
[11:27:01] *** sylvain has joined #opensolaris
[11:27:30] <Berny> is it at all possible to share the pool and have the children show up as "normal" directories, or do i have to tweak the automount maps for that?
[11:27:31] <quasi> Gr|ffous: I'd probably drop in postfix in the global zone
[11:28:41] <quasi> or switch to not inheriting /usr
[11:28:48] <LeftWing> postfix in the global works quite well.
[11:28:55] <quasi> exactly
[11:29:28] <Gman> stevel, what version of planet are you running on grommit?
[11:29:30] <LeftWing> The sendmail wrapper is good enough for at least cron and other stuff.
[11:29:33] <Gr|ffous> and the difference between postfix and all future software that will hit the same issue; is that the sendmail replacement is important to have, while other third party software shouldn't clash with the standard software, and so can sit elsewhere?
[11:29:50] <stevel> gman: grommit.com/planet is old, i haven't updated it in a while.  planetidentity.org is running a newer one i think
[11:30:05] <LeftWing> Gr|ffous: Realistically every piece of software included in the OS is a different case.
[11:30:18] <Gr|ffous> I guess my challenge is that I plan on setting up a quite complex postfix setup, and this is my desktop that I'm playing with atm
[11:30:20] <stevel> i'm not sure how to check the version
[11:30:46] *** calumb has quit IRC
[11:30:51] <LeftWing> Gr|ffous: Well postfix was built by IBM to be a drop-in sendmail replacement in most respects.
[11:30:51] <stevel> but it was installed from a nightly that's got a timestamp of jan 15, 2006
[11:30:52] <Gman> well, given that planet has only had 1 official release :)
[11:31:03] <Gr|ffous> LeftWing? So are you saying that inheriting /usr in general, is a bad idea?
[11:31:14] <LeftWing> It depends what you want to do with the Zone.
[11:31:43] <Gman> stevel, ok, gives me a reference point
[11:31:44] <LeftWing> For very relocatable software, like databases for instance, you can inherit just about everything.
[11:32:00] <Gr|ffous> I guess I'm used to treating it like Xen, with That I build a standalone mail server, stand alone, and enjoy the benefits of virtualisation
[11:32:17] <stevel> gman: but we can use whatever version you want.  i don't mind updating to a newer release - i've just never had an incentive to
[11:32:18] <LeftWing> mmm
[11:32:28] <LeftWing> Gr|ffous: Well, you can always use full root zones, they just require more diskspace. =)
[11:32:32] <Gman> stevel, ok, will see what jdub says
[11:32:36] <stevel> k
[11:32:46] <LeftWing> Gr|ffous: JES likes to be in a full root zone, for instance.
[11:33:09] <Gr|ffous> LeftWing, how about if I leverage zfs snapshots somehow, that should save space?
[11:33:17] <Gr|ffous> oh ok, interesting
[11:33:38] <LeftWing> Gr|ffous: What Solaris version are you running?
[11:33:46] <Gr|ffous> snv_45
[11:33:59] <LeftWing> Does that have the ZFS cloning zone creation stuff in it yet?
[11:34:20] <Gr|ffous> I'm not sure, I just read about it on some blog somewhere
[11:34:31] <LeftWing> Worth taking a look at, it's a great idea. =)
[11:34:35] <Gr|ffous> my bigger frustration is likely to be that I can't LU solaris, and keep my zones
[11:34:48] <stevel> yeah, snv_45 has zfs clone
[11:35:19] <LeftWing> stevel: With the zoneadm additions I seem to remember being on the cards?
[11:35:46] <Gman> bash-3.00$ pkgtool --interactive --download build SFEbzr.spec
[11:35:47] <Gman> sweet.
[11:36:00] <stevel> zoneadm additions?
[11:36:29] <twincest> JDS is using bzr?
[11:36:30] <LeftWing> I'm probably imagining things... I seem to remember there being extra options to leverage ZFS cloning in zone creation.
[11:36:40] <Gr|ffous> hmm, well as this is my first try, I think I'll just give the zone it's own /usr
[11:36:44] *** Dar is now known as Dar_HOME
[11:36:58] <stevel> oh... i do recall that
[11:37:01] <Gr|ffous> I'm having enough trouble just compiling the software, and making pkg files is still WAY too hard
[11:37:08] <Gman> twincest, nope, was just going to suck down the latest planet
[11:37:09] <Gr|ffous> so it's going to be a bit of a mess the first time through anyway
[11:37:12] <stevel> i don't think it had putback yet though
[11:37:13] *** Dar_HOME is now known as Dar
[11:37:15] <Gman> but seemingly 2.0 is the one to get
[11:37:25] <LeftWing> stevel: Yeah, probably not.
[11:38:09] <Gr|ffous> speaking of zfs and all, has there been any progress on zfs root of late? I haven't heard anything for a while now
[11:38:54] <stevel> yeah .. people are working on it
[11:39:19] <Gr|ffous> I'm sure... any progress though? ;)
[11:39:52] <stevel> i don't know what the recent status is, sorry
[11:41:24] <coolvibe> Gr|ffous: the ZFS people have blogs, maybe you can find more info there
[11:41:27] <timeless> is there a safe way to test kernel drivers other than using vmware?
[11:41:39] <Gr|ffous> coolvibe, I'm reading Jeff's latest now :)
[11:41:42] <timeless> like, is it possible to have an app that emulates the behavior of modload?
[11:41:43] <coolvibe> timeless: qemu? ;)
[11:42:06] <timeless> coolvibe: does it work on opensolaris, and how much space would i need?
[11:42:11] <timeless> i have <1gb available
[11:42:28] <timeless> if it can use my existing opensolaris readonly, that'd be great
[11:42:50] <coolvibe> qemu should work, but I think it will be slow on opensolaris, since qemu doesn't have a driver for opensolaris, so it has to emulate a machine completely
[11:43:13] <LeftWing> I thought there was a kqemu-esque port going on for OS
[11:43:21] <timeless> all i need to do is safely figure out how to get this stupid usbethernet driver to load w/o panicing
[11:43:39] <coolvibe> LeftWing: I have't heard about it though, but you could be correct
[11:44:15] <LeftWing> I swear there was some noise about it on osol-discuss a whileago.
[11:53:20] <Gman> stevel, a slightly delicate question - is grommit running solaris? :)
[11:53:32] <stevel> gman: of course :)
[11:53:49] <Gman> stevel, can you check for the presence of bsddb.py please?
[11:54:33] <stevel> hrm
[11:54:35] <stevel> where would it be?
[11:54:43] <Gman> in /usr/lib/python/
[11:54:56] <stevel> nope, don't have it
[11:55:04] <stevel> where can i find it and install it?
[11:55:24] <Gman> seemingly our python packages aren't building it right now
[11:55:28] <Gman> and i need it for planet 2.0
[11:55:33] * Gman investigating
[11:57:12] *** logic has quit IRC
[12:11:19] * LeftWing is fond of ActivePython.
[12:16:09] <stevel> gman: i got planet-2.0 up
[12:16:18] <Gman> stevel, oh, so soon?
[12:16:20] <stevel> seems to run planet grommit fine
[12:16:29] <Gman> um, it does?
[12:16:38] <Gman> surprising - since it fails here :(
[12:16:42] <Gman> on lack of bsddb
[12:16:58] <Gman> stevel, want me to give you a public key? :)
[12:17:18] <stevel> oh, i installed bsddb
[12:17:32] <Gman> ahh, ok
[12:17:39] <Gman> supposedly it should be part of python2.4.4
[12:17:45] <Gman> so i'm trying to fix that bug here
[12:17:54] <stevel> ah. you're trying to fix the root of the problem ;)
[12:17:58] <Gman> but that's a good work around
[12:17:58] <stevel> i just worked around it
[12:18:03] <Gman> sweet
[12:18:13] <Gman> i've got all the template stuff from gnome
[12:18:17] <stevel> i'm trying to figure out this weird formatting problem that's always slightly bugged me - but never enough to do anything real about it
[12:18:22] <stevel> http://grommit.com/planet2/
[12:18:23] <Gman> so just a case of changing some configs
[12:18:31] <stevel> these random accented A's, and other weird characters
[12:18:43] <Gman> i can ask jeff, he's bound to know
[12:19:09] <stevel> yeah - how do you want to do it... do you want a shell on grommit and you ssh in and set it all up?  or do you want to play with it on your end - and then just send me the planet config and i'll have it run with my other planet?
[12:19:41] <Gman> either or really
[12:19:47] <Gman> i'll see if i can fix it locally
[12:19:52] <Gman> have a play and set up the files
[12:19:53] <Gman> then send you them
[12:19:55] <Gman> should be easy enough
[12:20:09] <stevel> sure thing
[12:20:12] <Gman> thanks though!
[12:21:51] <stevel> np
[12:21:56] * Gman absolutely pumped about this
[12:21:59] <Gman> it's gonna rock
[12:22:33] <Gman> <jdub> Gman: i bet that's being served up with the wrong locale
[12:22:37] *** rafaeldt has quit IRC
[12:22:40] <Gman> stevel, ^
[12:22:58] <Gman> <jdub> probably generated in utf-8 and served as iso-WHATEVER evi default
[12:23:12] <quasi> that's pretty common
[12:23:40] <stevel> hrm. wonder how i fix that
[12:23:54] <Gman> server settings?
[12:24:00] <stevel> probably
[12:24:42] <quasi> AddDefaultCharset if it is apache 2
[12:25:37] <stevel> nope, i've got AddDefaultCharset set to utf-8
[12:26:14] <stevel> oh wait no i don't
[12:26:15] <quasi> what is the content in? 8859?
[12:26:18] <stevel> i'm on crack
[12:26:22] <stevel> ahhhh
[12:26:23] <stevel> hey
[12:26:24] <stevel> that fixed it
[12:26:30] <stevel> well damn. that's been bugging me for like a year
[12:26:37] <Gman> sweet!
[12:26:40] <stevel> thanks quasi and gman :)
[12:26:45] <quasi> no problem
[12:27:17] <stevel> okay, i gotta run back to the hotel
[12:27:22] <stevel> dp should have gotten in
[12:27:43] <stevel> gman: email me whenever you have your planet config ready to go, and we can drop it in and try it out :)
[12:27:46] <stevel> 'night all
[12:28:07] *** stevel has quit IRC
[12:28:09] <Gman> night steleman
[12:28:12] <Gman> stevel even
[12:32:59] *** yongsun_ has quit IRC
[12:48:45] <twincest> does SUNWjmf come from JDS?
[12:49:14] <twincest> (trying to work out what i can remove without breaking ON build :)
[12:49:30] <Gman> no
[12:49:44] <twincest> hmm.
[12:49:46] <Gman> but i doubt anything in ON will need it
[12:50:02] <twincest> well, i didn't think ON would need the Java stuff i removed last time i did this
[12:50:09] <twincest> turns out ON includes more i thought
[13:00:12] *** bengtf has joined #opensolaris
[13:01:07] *** trs81 has quit IRC
[13:15:38] *** trs81 has joined #opensolaris
[13:21:31] <twincest> system      SUNWfsfw                        French freeware message
[13:21:36] <twincest> neat, an entire package for a single message
[13:26:06] <Gman> hrm, bdb is proving to be a pain in the arse to link in
[13:27:08] <Stric> bdb is a pain in the arse.
[13:28:56] <Gman> can't figure out how to link it into python2.4.4
[13:29:49] <Gman> can't quite figure out whether i have bdb in the first place ;)
[13:31:56] <quasi> SFWbdb perhaps?
[13:32:14] <Gman> isn't it installed on solaris already?
[13:32:42] <quasi> that's probably not berkely
[13:33:09] <Gman> there's stuff in /usr/lib/bdb
[13:33:20] <Gman> and there's /usr/lib/libdb.so.1
[13:34:53] <quasi> not on this box
[13:35:32] <twincest> hmm, i wonder if i unpacked the ON archives into an empty slice, what else would be needed to make it boot
[13:36:45] <twincest> are there any 'core' consolidations other than ON and NSW?
[13:36:46] <quasi> hmm, I do have SUNWbdb on the s10u3 box
[13:38:08] *** trs81 has quit IRC
[13:38:31] *** trs81 has joined #opensolaris
[13:39:32] <quasi> bugs.opensolaris.org seems rather dead
[13:41:04] <Triskelios> I'm trying to decide on a disk layout for a single-user laptop; tempted to go with just /var seperate
[13:43:43] <Triskelios> the downside is that if / is huge, neither /var nor swap get any advantage for being near the beginning of the disk
[13:44:06] *** uebayasi has quit IRC
[13:44:59] <quasi> you could always choose a diffrent ordering
[13:45:04] <quasi> of slices
[13:45:55] <coolvibe> hm, where does gdm in nevada look for the available session types?
[13:46:01] <Triskelios> yeah, but that gets annoying
[13:46:13] <coolvibe> not in /etc/X11/gdm, because nothing is there
[13:46:49] <Triskelios> probably /usr/dt/config
[13:47:04] <coolvibe> Triskelios: dtlogin and gdm use the same format?
[13:48:13] <Triskelios> I think so
[13:49:38] <Triskelios> hm, maybe it actually looks in /usr/share/xsessions , which have the desktop files which just point to the dtlogin session scripts
[13:49:45] <coolvibe> aha
[13:51:44] <coolvibe> I see .desktop files in /usr/share/xsessions... that must be it, thanks :)
[14:03:14] *** trs81 has quit IRC
[14:05:44] *** trs81 has joined #opensolaris
[14:33:08] *** trs81 has quit IRC
[14:33:29] *** trs81 has joined #opensolaris
[14:39:59] <coolvibe> grmbl, Blastwave KDE *still* does not have anti-aliasing
[14:47:34] *** laca has joined #opensolaris
[14:48:11] <twincest> solaris determines the root device from bootenv.rc, right?  nowhere else?
[14:49:18] <asyd> yup. But you can use grub to define the bootpath iirc
[14:49:46] *** _tsoome_ has joined #opensolaris
[14:54:17] <coolvibe> any blastwave people here?
[14:55:02] *** |tsoome| has quit IRC
[14:56:18] <asyd> more or less
[14:58:15] *** trs81 has quit IRC
[15:00:01] <coolvibe> asyd: why are the kdelibs not compiled with antialiasing fonts support?
[15:01:08] <coolvibe> I searched around on google, but found nothing conclusive about it
[15:01:32] <asyd> I can't tell you. But feel free to search (and ask) on the csw users mailing list
[15:01:51] <coolvibe> where is the archive? I don't see anything about that on the blastwave site
[15:02:22] <coolvibe> oh nvm
[15:02:24] <coolvibe> found it
[15:03:36] <coolvibe> hm, mailman... no way to search the damn archive :P
[15:05:05] *** laca is now known as lacaAFK
[15:05:23] <asyd> google ? ;p
[15:05:27] <asyd> (with site:...)
[15:05:29] <wilbury> nick otisPEE
[15:05:55] <coolvibe> asyd: blegh :)
[15:06:10] <coolvibe> I'll mail the list about it
[15:08:07] *** polk_ has joined #opensolaris
[15:11:30] *** polk__ has quit IRC
[15:11:40] <twincest> hmm, does grub always assume slice a is the root slice?
[15:13:52] <Error_404> *shrug*
[15:15:36] <coolvibe> twincest: not really, you can tell grub which slice is the root slice
[15:15:55] <twincest> how?
[15:16:15] <asyd> with the root command
[15:16:30] <twincest> but that only applies within a menu entry, doesn't it?
[15:16:43] <coolvibe> with groot=(device)
[15:16:55] <coolvibe> that sets the default root device
[15:17:02] <coolvibe> device == slice
[15:17:10] <asyd> twincest: what do you mean ?
[15:17:32] <twincest> asyd: the root command only applies after it loads grub.conf.  how does it know where to load grub.conf from?
[15:17:50] <asyd> ahh
[15:17:51] <coolvibe> twincest: when you install grub into the MBR, you can tell it where to look
[15:17:57] <asyd> the grub-install script do that
[15:18:07] <coolvibe> yes
[15:19:07] <twincest> ah.. that makes sense
[15:19:33] <coolvibe> or you specify when you install grub mannually
[15:19:44] <coolvibe> but grub-install is easier :)
[15:23:31] <twincest> aha, this makes sense now
[15:23:45] <twincest> i thought the rdsk argument was where to install it, but it's really the default root
[15:23:51] <coolvibe> yep
[15:24:17] *** karrotx has joined #opensolaris
[15:24:24] <coolvibe> namespace conflict between grub and solaris nomenclature :)
[15:28:05] <twincest> right, there we go.. managed to move it to a new disk
[15:28:49] *** karrotx has quit IRC
[15:32:01] *** rodrickbrown has quit IRC
[15:33:46] *** Gman is now known as GmanAFK
[15:46:19] *** lacaAFK is now known as laca
[15:48:55] *** trs81 has joined #opensolaris
[15:53:19] *** deather_ has joined #opensolaris
[15:54:33] *** trs81 has quit IRC
[15:54:48] *** trs81 has joined #opensolaris
[15:59:28] *** darkcmd has quit IRC
[16:02:50] *** sylvain has quit IRC
[16:08:01] *** RaD|Tz has joined #opensolaris
[16:08:11] *** aliquis has quit IRC
[16:09:52] *** deather has quit IRC
[16:11:28] *** glagasse has joined #opensolaris
[16:14:58] *** klocze1 has joined #opensolaris
[16:15:36] *** eboutilier has joined #opensolaris
[16:22:02] *** sommerfeld has joined #opensolaris
[16:22:02] *** ChanServ sets mode: +o sommerfeld
[16:31:43] *** nwf has joined #opensolaris
[16:47:17] *** rocks has joined #opensolaris
[16:47:42] <rocks> What's a good I/O benchmark tool to compare between two systems ?
[16:48:00] <elektronkind> iozone, filebench
[16:48:33] *** jafari has joined #opensolaris
[16:49:05] <rocks> elektronkind: I was looking at filebench, it seems to have binary for X86 solaris only, no sparc ...  is that true or I just didn't find the right site ?
[16:50:12] <rocks> ahh, x86 and spark in one installer
[16:56:10] *** zarathustra has joined #opensolaris
[16:56:12] <jafari> good morning all
[16:56:19] <jafari> how can i fix this error
[16:56:37] <jafari> top: no termcap entry for a `linux' terminal
[16:56:47] <jafari> it causes top to not run
[16:58:45] <elektronkind> try changing TERM to something else
[16:59:02] <elektronkind> TERM=xterm; export TERM
[16:59:03] <elektronkind> or
[16:59:11] <elektronkind> setenv TERM xterm
[16:59:29] <oxygene> elektronkind: it's possible to leave away the ";" in that statement
[16:59:44] <elektronkind> ah right... brain fart
[17:00:15] *** _tsoome_ has quit IRC
[17:00:55] *** sparc-kly_ has joined #opensolaris
[17:00:56] <jafari> ok thanks
[17:04:55] *** sparc-kly has quit IRC
[17:05:23] *** sparc-kly_ is now known as sparc-kly
[17:05:30] *** ChanServ sets mode: +o sparc-kly
[17:16:08] *** logic_ has joined #opensolaris
[17:19:56] *** esproul_ has quit IRC
[17:23:55] *** nwf has quit IRC
[17:27:32] *** jafari has quit IRC
[17:29:38] *** crib- has joined #opensolaris
[17:33:20] *** setui1 has joined #opensolaris
[17:34:51] <setui1> I have troubles with czech keyboard layout
[17:35:12] <setui1> We have some special characters which are writed
[17:35:19] <setui1> by multiple keys
[17:35:50] <setui1> but they aren't possible to be written as CAPITALS
[17:35:58] <setui1> even with turned on capslock
[17:36:24] <setui1> does anybody have similar problem?
[17:36:41] <setui1> I'm using JDS
[17:36:51] <setui1> and Xorg
[17:37:28] <setui1> Computer is HP nx6125 laptop
[17:41:56] *** crib has quit IRC
[17:43:43] <dwc-> not too many czech layouts here, I imagine, but there are some of the internationalization folks here
[17:52:09] <setui1> yeah
[17:52:19] <setui1> but they are almost on workstation
[17:52:22] <setui1> not on notebooks
[17:52:42] <setui1> so I think that problem will be in the keyboard
[17:52:49] <setui1> damn HP
[17:53:32] <quasi> 7bit ascii rulez! ;)
[17:54:14] <PerterB> why not save a bit and get rid of all those pesky lowercase letters too
[17:54:22] <setui1> so we must speak in words of the Big Sun
[17:54:29] <setui1> so unicode ..
[17:54:35] <setui1> no hah
[17:55:29] <setui1> maybe I have to create my own layout
[17:55:47] <setui1> but I have never do that
[17:55:55] <quasi> perterb: if anything, i'd support getting rid of upper case
[17:56:25] *** klocze1 has quit IRC
[17:56:36] <setui1> no
[17:56:39] <setui1> I cannot
[17:57:24] <setui1> I use my computer as notepad in school
[17:57:46] <setui1> but I must make a normal paper documents from it
[17:58:45] <setui1> How many people from here will be on Sun Tech Days in prague?
[18:05:07] *** johnlev has joined #opensolaris
[18:13:07] *** setui1 has left #opensolaris
[18:14:28] <sickness> evening all
[18:21:28] *** bougie has joined #opensolaris
[18:24:13] <bougie> hello
[18:26:01] *** logic_ has quit IRC
[18:26:01] *** ccitt has joined #opensolaris
[18:26:14] <ccitt> greets - disk mgmt question
[18:26:19] *** eboutilier has left #opensolaris
[18:27:12] <ccitt> have testbed Nexenta system, trying to add usb hdd and create zpool on it. zpool reports device busy, and format does not list it. anyone have any ideas?
[18:30:03] <elektronkind> !seen stevel
[18:30:04] <Drone> stevel (stevel!n=stevel at 192 dot 18.43.249) was last seen in #opensolaris on Mon 06 Nov 2006 11:29 GMT, saying ''night all'.
[18:34:06] *** coolvibe has quit IRC
[18:34:12] *** axxl has joined #opensolaris
[18:36:59] *** OnkelSchorsch has joined #opensolaris
[18:37:11] <djgregor> nice.... SXCR 51a works with my 30" apple cinema display with 0 configuration
[18:37:27] <djgregor> that was a lot better than the hell I had to go through to get it working on SuSE 10.0
[18:39:22] <OnkelSchorsch> yay. blastwave has flightgear binaries available.
[18:44:16] *** ndroux has joined #opensolaris
[18:44:58] *** axxl has quit IRC
[18:50:40] *** alanc is now known as alanc_away
[18:54:32] *** bunker has joined #opensolaris
[18:56:36] *** Cyrille has quit IRC
[18:58:18] *** damienc has quit IRC
[19:01:22] *** benr has joined #opensolaris
[19:01:22] *** ChanServ sets mode: +o benr
[19:08:09] *** orange80 has left #opensolaris
[19:09:32] <sickness> hi benr
[19:10:02] <benr> hey sickness
[19:17:14] <timeless> swift# zoneadm -z nexenta boot
[19:17:14] <timeless> WARNING: skipping interface 'lo' which may not be present/plumbed in the global zone.
[19:18:17] <timeless> what does that mean? :(
[19:22:29] *** dvorak has quit IRC
[19:22:38] *** dvorak has joined #opensolaris
[19:23:38] *** sahafeez has joined #opensolaris
[19:23:52] *** alanc-away is now known as alanc
[19:23:58] <timeless> hi alanc :(
[19:24:13] <alanc> *sigh* 37 new spams to xwin-discuss at os dot o since Friday
[19:24:14] <alanc> hello
[19:24:41] <timeless> any idea what that means?
[19:24:57] *** coolvibe has joined #opensolaris
[19:25:01] <alanc> the zoneadm/interface thing?  no clue
[19:25:38] *** calumb has joined #opensolaris
[19:25:44] <timeless> lo0:1: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
[19:25:51] <timeless> isn't that up and friendly enough?
[19:25:52] <alanc> I don't use zones much
[19:29:18] <oxygene> timeless: lo0 itself?
[19:30:04] <timeless> lo0:1: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
[19:30:13] <timeless> oops, sorry
[19:30:16] <timeless> that was from the zone
[19:30:26] <oxygene> "in the global zone"
[19:30:28] <timeless> lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
[19:30:28] <timeless>         inet 127.0.0.1 netmask ff000000
[19:30:31] <oxygene> hmm... okay
[19:30:34] <timeless> that's from the global zone
[19:31:11] <timeless> i wasn't intentionally answering your question, i just forgot that i had run ifconfig -a4 inside the zone :), i do know the difference
[19:31:18] <timeless> s/ans/mis-ans/
[19:32:08] *** alobbs has quit IRC
[19:32:24] <timeless> is there some log i can get?
[19:32:27] <timeless> this used to work
[19:34:14] *** johnlev has quit IRC
[19:35:04] *** slowhog has joined #opensolaris
[19:39:09] <timeless> actually, maybe that doesn't matter
[19:39:19] *** OnkelSchorsch has quit IRC
[19:39:58] *** calumb has quit IRC
[19:41:19] *** calumb has joined #opensolaris
[19:41:31] *** miffe has joined #opensolaris
[19:53:13] *** johnlev has joined #opensolaris
[20:02:26] <sickness> lupi_zones plugin <- what's this? so now it is possible to luupgrade zones? =)
[20:02:27] *** estibi has joined #opensolaris
[20:03:28] *** RaD|Tz has quit IRC
[20:05:07] *** RaD|Tz has joined #opensolaris
[20:10:46] *** axisys_ has joined #opensolaris
[20:17:21] *** ccitt has quit IRC
[20:23:23] *** alobbs has joined #opensolaris
[20:23:55] <timeless> i know it's slightly offtopic, but does anyone know of a link explaining how / why /usr/bin/test -h and /usr/bin/test -L differ?
[20:35:21] <PerterB> hysterical reasons, I think
[20:35:34] <timeless> is there a prefered flavor?
[20:35:35] <PerterB> BSD  precedent vs POSIX kind of thing
[20:37:53] <timeless> is that kind of stuff not listed in man pages? :(
[20:39:05] *** setui1 has joined #opensolaris
[20:45:00] *** RaD|Tz has quit IRC
[20:45:22] *** Dar is now known as Dar_HOME
[20:47:59] <estibi> timeless: my man page doesn't say anything so i guess it is compatibility reason
[20:48:34] <timeless> neither my snv_49 nor my nexenta explained it
[20:48:40] <timeless> but that doesn't explain which i should use :(
[20:51:57] <PerterB> FWIW freebsd test(1) has: -h file       True if file exists and is a symbolic link.  This operator is retained for compatibility with previous versions of this program.  Do not rely on its existence; use -L instead.
[20:52:40] <setui1> evolution
[20:52:53] <setui1> the same chroot
[20:53:07] <setui1> is there a diferent way to use it then in Linux
[20:56:23] <delewis> check SUSv3 when in doubt.
[20:56:30] <delewis> the specification is public.
[20:58:08] *** jamesb has joined #opensolaris
[20:58:52] <setui1> Does anybody have problems with importing Mercurial module?
[21:02:46] <richlowe> Nope.
[21:02:53] <richlowe> depending on what exactly you're referring to.
[21:03:27] <setui1> I just wanted to get opensolaris kernel (on-src)
[21:03:36] <setui1> so I had to use it
[21:03:52] <setui1> but when I use hg clone .....
[21:04:03] <setui1> ImportError: no
[21:04:11] <setui1> just python error
[21:04:25] <setui1> I looked to the bin directory
[21:04:31] <richlowe> twincest: ON has several java-ish components, but I'm not sure what java-ish things they make use of.
[21:04:34] <setui1> but hg wasn't symlink
[21:05:07] <setui1> and I couldn't find module named mercurial
[21:05:11] <setui1> just folder
[21:07:47] <gnu2it2> what is the md5sum for nv-51a?
[21:08:45] <delewis> gnu2it2: you shouldn't need one.
[21:08:51] <delewis> zip files have internal checksums
[21:09:31] <gnu2it2> hmm, ok so if it unzips without error then all is good ?
[21:09:37] <delewis> gnu2it2: yes
[21:09:58] <gnu2it2> sweet. I love the answers here
[21:09:58] *** paul_ has joined #opensolaris
[21:10:27] *** yarihm has joined #OpenSolaris
[21:10:36] *** rachel has quit IRC
[21:11:20] *** rachel has joined #opensolaris
[21:16:18] *** paul_ is now known as paul
[21:17:17] *** agliodbs has joined #opensolaris
[21:17:46] *** laca is now known as lacaAFK
[21:19:29] *** nwf has joined #opensolaris
[21:19:56] *** dapuzzl has joined #opensolaris
[21:21:41] <dapuzzl> hi
[21:22:55] *** coffman has joined #opensolaris
[21:23:37] *** bobbyz has joined #opensolaris
[21:25:37] *** dapuzzl has quit IRC
[21:30:29] *** logic_ has joined #opensolaris
[21:32:07] *** crib- is now known as crib
[21:32:20] *** jamesd_ has joined #opensolaris
[21:32:43] *** jamesd has quit IRC
[21:34:48] *** bengtf has quit IRC
[21:37:19] *** calumb has quit IRC
[21:40:37] *** |tsoome| has joined #opensolaris
[21:47:44] *** knightblader has quit IRC
[22:02:42] *** bondolo has quit IRC
[22:03:46] *** setui1 has left #opensolaris
[22:04:32] *** nwf has quit IRC
[22:20:56] *** kFuQ has quit IRC
[22:21:07] *** kFuQ has joined #opensolaris
[22:22:14] *** lacaAFK is now known as laca
[22:26:06] *** dunc has quit IRC
[22:29:38] *** benr has quit IRC
[22:31:01] *** LeftWing has quit IRC
[22:33:03] *** LeftWing has joined #OpenSolaris
[22:34:04] *** dunc has joined #opensolaris
[22:34:39] *** Teknix has quit IRC
[22:34:57] *** Teknix has joined #opensolaris
[22:36:08] *** GmanAFK is now known as Gman
[22:36:11] <Gman> hey hey hey
[22:36:28] <quasi> hey hey
[22:36:44] <richlowe> Gman: mornin'
[22:37:09] *** coffman has quit IRC
[22:37:37] <richlowe> Gman: You should be able to pretty easily generate a planet config from the OPML, btw.
[22:37:46] <Gman> richlowe, sweet
[22:37:47] *** axisys_ has quit IRC
[22:37:56] <Gman> stevel installed 2.0 on grommit
[22:38:03] <Gman> and i'm just working on the stylesheets
[22:38:06] <richlowe> I had some pretty horrid perl around somewhere, but I can't currently find it.
[22:38:07] <Gman> should be live real soon :)
[22:38:18] <richlowe> I kinda hoped something that did that would be in planet proper, but I didn't find anything at the time.
[22:38:25] <Gman> richlowe, the planet is going to be ruled with editorial power
[22:38:35] <Gman> so, not everyone gets on initially
[22:38:46] <richlowe> Gman: and I saw the planet bits, that's why I'm saying this. ;)
[22:38:52] * Gman wants to focus in on readability
[22:38:53] <Gman> nod
[22:39:11] <timeless> hrm
[22:39:20] <timeless> is there a best practices for setting up zfs volumes?
[22:39:21] <richlowe> Gman: start ith the blogs feed, trim the advertising and the java gunk, and you're set.
[22:39:28] <timeless> like "don't nest them very deeply"?
[22:39:29] <richlowe> Gman: oh, and those odd articles about hair cuts, too ;)
[22:39:49] <Gman> richlowe, :)
[22:40:13] <richlowe> though seeing that next to your image on planet gnome... ;)
[22:40:16] <timeless> i've managed to make a zfs hierarchy that's 11 deep
[22:40:17] <Gman> and the genesi crap
[22:40:24] <richlowe> Gman: that's what I meant by 'advertising.'
[22:40:25] <timeless> and i messed up on the order :(
[22:40:28] <Gman> richlowe, i'm totally going to do hackergotchies!
[22:40:30] <richlowe> timeless: I nest them as I need them nested.
[22:40:30] <Gman> richlowe, heh
[22:40:54] <timeless> richlowe: but you need them nested when? when you want to inherit permissions?
[22:40:57] <richlowe> timeless: (pretty much so I can be lazy and use -r and have it not do anything I don't want it to)
[22:41:10] <richlowe> timeless: I pretty much group things as I would directories, if that makes more sense.
[22:41:46] <timeless> basically i was trying to use it that way
[22:42:02] <timeless> but i kinda messed up, i have <4g free, and a misnamed and misnested 10g voume
[22:42:04] <Gman> richlowe, having issues trying to get a bdb linked python
[22:42:07] <timeless> volume even
[22:42:09] <Gman> it's a pain in the arse.
[22:42:16] <timeless> it needs to move one level down :(
[22:42:19] <sommerfeld> 11 deep !
[22:42:22] <richlowe> timeless: rename it.
[22:42:23] <sommerfeld> wow
[22:42:31] <richlowe> and yeah, I'm only 4 or 5 levels deep.
[22:43:05] <timeless> sommerfeld: well, root_pool/home/psuedousername/lxr-root/lxr-data/lxr_root_name/lxr_src_root/{directory-tree-here}
[22:43:20] <timeless> the lxr-root/lxr-data bit was also a goof
[22:43:53] <timeless> but zfs doesn't let me move around partitions like directories
[22:44:05] <timeless> so i've basically made a mess that's fairly hard to dig out of :)
[22:44:09] <richlowe> timeless: zfs rename
[22:44:20] <timeless> will that let me do a move?
[22:44:23] <richlowe> and actually, you kinda *can* mv them, it just doesn't do what you'd expect.
[22:44:28] <richlowe> and probably not at all what you'd want.
[22:44:35] * richlowe has been trying to decide if that's a bug for quite a while though.
[22:44:42] * timeless kinda presumed it was a relabel not a /bin/mv
[22:44:49] <sommerfeld> timeless: trying to share storage between related trees?
[22:45:12] <timeless> sommerfeld: trying to have managable volume sizes for the next time i get messed up
[22:45:17] <richlowe> timeless: it renames them, but since the dataset name is the full path, and the (normal) mountpoint is based on that name...
[22:45:36] <richlowe> zfs rename store/homes/richlowe store/richlowe should work, iirc.
[22:45:46] <timeless> ok, i'll try it :)
[22:46:16] <timeless> cannot rename 'root_pool/home/svn/lxr-root/lxr-data/world-all.0/world/stage/projects': out of space
[22:46:33] <timeless> :)
[22:46:46] <timeless> actually, even 'svn' is a misnomer'
[22:46:58] <timeless> since only some lxr roots are svn's :(
[22:47:55] <sommerfeld> so, boost quotas and try again
[22:48:01] <timeless> quotas?
[22:48:03] <timeless> what's that?
[22:48:18] <sommerfeld> you didn't set quotas here?
[22:48:19] <timeless> i litterally have <4g of free space and a 10g partition
[22:48:32] <timeless> s/partition/volume
[22:48:53] <timeless> (yes, i know what quotas are, i just have no need for them, i have physical disk space constraints)
[22:49:26] <richlowe> Hm, I wouldn't expect a rename to be consumptive.
[22:49:32] <richlowe> (maybe snapshots and such play into that?)
[22:49:36] <timeless> swift# zfs rename root_pool/home/svn/lxr-root/lxr-data root_pool/home/lxr-data
[22:49:36] <timeless> cannot rename 'root_pool/home/svn/lxr-root/lxr-data': out of space
[22:49:44] <timeless> doesn't look like it works :)
[22:49:57] <timeless> i'm guessing it's actually trying to play things safe and doing things the way one normally would
[22:50:22] <richlowe> works for me.
[22:50:22] <timeless> i don't have many snapshots
[22:50:40] <timeless> in fact, i have none anywhere near the 'svn' volume tree
[22:50:45] <richlowe> > zfs rename store/builds/notwork store/notwork
[22:50:50] <richlowe> > zfs rename store/notwork store/builds/notwork
[22:51:01] <timeless> but how much space is in your zpool?
[22:51:06] <timeless> and how big is your notwork?
[22:51:13] <richlowe> 6G available.
[22:51:17] <timeless> root_pool/home/svn                                                                        29.3G  3.86G   278M  /export/home/svn
[22:51:25] <richlowe> 489M
[22:51:35] <Error_404> wtf?
[22:51:38] *** linescanner has joined #opensolaris
[22:51:39] <timeless> can you mkfile a 4G file?
[22:51:52] <Error_404> i thought genesi killed their non-embedded products
[22:51:55] <richlowe> it's an ON tree that, well, doesn't actually function, but I'm keeping around anyway (don't ask).
[22:52:11] <timeless> i won't ask because i'm not curious enough
[22:52:30] *** lloy0076 has joined #opensolaris
[22:52:51] *** _dreams_ has quit IRC
[22:53:29] *** _dreams_ has joined #OpenSolaris
[22:53:30] <richlowe> either way, I don't have a pool with any dataset larger than the free space in the pool.
[22:54:22] <timeless> swift# zfs rename root_pool/home/svn/lxr-root/lxr-data/osso root_pool/home/cupid
[22:54:22] <timeless> swift# zfs list|grep cupid
[22:54:22] <timeless> root_pool/home/cupid                                                                      32.5K  3.86G  32.5K  /export/home/cupid
[22:54:22] <timeless> swift# zfs rename root_pool/home/cupid root_pool/home/svn/lxr-root/lxr-data/osso
[22:54:38] <timeless> works for me on tiny volumes
[22:54:46] * timeless suspects that volume is empty
[22:55:23] <timeless> anyway, i'm assuming that this is intentional
[22:56:00] *** sahafeez has quit IRC
[22:58:43] <timeless> /onnv/onnv-gate/usr/src/cmd/mdb/common/modules/zfs/ - mdb - modular debugger
[22:58:49] <timeless> that's really the modular debugger?
[22:59:30] * timeless finds zfs_main
[22:59:40] <richlowe> It's the zfs dmod.
[22:59:57] * timeless reads zfs_rename
[23:00:43] <sommerfeld> part of it.  /onnv/onnv-gate/usr/src/cmd/mdb is all of it
[23:00:56] <sommerfeld> (it == mdb, not the zfs mdb dmod)
[23:01:18] <timeless> yeah, i was just expecting a find for zfs.c to give me the zfs app, it didn't, so i was surprised, anyway, i'm past that :)
[23:02:01] <timeless> ioctl(zhp->zfs_hdl->libzfs_fd, ZFS_IOC_RENAME, &zc))
[23:02:13] <timeless> i guess that's where i jump into kernel land :(
[23:03:27] <sommerfeld> Yup
[23:03:51] <timeless> ok, at this point, i got lost
[23:03:56] <timeless> i can't find a consumer
[23:04:01] <richlowe> zfs_ioctl.c
[23:04:13] <richlowe> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/zfs_ioctl.c#1199
[23:04:16] <richlowe> then dmu_objset_rename
[23:05:24] * Gman starts listening to the joyent podcast
[23:05:29] <richlowe> (look at zfs_ioc_vec, if you want to go through how you got to there)
[23:05:43] <timeless> yeah i saw that as i was loading it
[23:05:47] * timeless is at dsl_dir_rename
[23:06:27] <timeless> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/dsl_dir.c#1087
[23:06:32] <timeless> /* is there enough space? */
[23:07:09] <timeless> i think i followed all the right hops
[23:07:42] *** _dreams_ has quit IRC
[23:07:47] <timeless> i suppose i could try to use dtrace to prove it
[23:08:19] *** bougie has quit IRC
[23:09:33] <timeless> actualy, i think i messed up
[23:09:40] <timeless> and i'm guessing it'll fail slightly after that point
[23:09:42] *** linescanner has quit IRC
[23:10:11] <timeless> dsl_dir_transfer_possible #1189
[23:12:16] <lloy0076> Is there an "easy" way to work out what Build Number I happen to have? I have an original ON consolidation which I've BFU'ed to opensol-20061030 (according to uname) but I'm sure how to correlate that to a build number (if that makes any sense to do at all).
[23:12:35] *** bondolo has joined #opensolaris
[23:13:56] <jmcp> lloy0076: check the flag days page http://www.opensolaris.org/os/community/on/flag-days/
[23:15:17] <timeless> anyway, i think zfs really doesn't want to let me do what i need to do :)
[23:15:19] <richlowe> it'd be easier to base it on the build schedule.
[23:15:46] <timeless> the reason for the nesting is mostly to avoid this problem in the future
[23:15:54] <timeless> the original deep nesting was just a botch
[23:16:13] <timeless> hrm
[23:16:14] *** jacotton has quit IRC
[23:16:39] <timeless> so is zfs forcing me to have enough space to move all nested volumes?
[23:16:45] <richlowe> I like how the top heads up mentions 'opensolaris contributors' explicitly, then links to tethys.eng.
[23:16:58] <timeless> heh
[23:24:17] <boyd_> Morning, all
[23:24:28] *** boyd_ is now known as boyd
[23:24:45] <boyd> Is the hg repo broken for anyone else?
[23:24:56] <richlowe> broken how?
[23:25:17] <boyd> Huh... well it *was* but worked on a re-try
[23:25:23] *** rocks has quit IRC
[23:25:26] <boyd> hg incoming | less
[23:25:27] <boyd> abort: no response from remote hg!
[23:25:27] <boyd> remote: Could not chdir to home directory /home/anon: No such file or directory
[23:25:28] <richlowe> boyd: that sounds like the automount snafu.
[23:25:30] <boyd> remote: abort: Not owner: None
[23:25:36] <richlowe> (or one of the automount snafus)
[23:25:40] <boyd> Looks like automount to me
[23:26:24] <boyd> richlowe: I didn't know they'd been having probs
[23:29:39] *** alanc_away has quit IRC
[23:31:16] *** laca is now known as lacaAFK
[23:38:36] *** stevel has joined #opensolaris
[23:38:36] *** ChanServ sets mode: +o stevel
[23:40:07] <richlowe> boyd: stevel would know more, probably.
[23:40:24] <boyd> No prob... I dodn't realise it was so transient..
[23:40:29] <boyd> s/dod/did
[23:44:53] *** solarisjon has joined #opensolaris
[23:45:25] *** gisburn has joined #opensolaris
[23:45:38] <gisburn> Greetings earthlings!
[23:45:50] <gisburn> I come in peace.
[23:45:59] <gisburn> (usually)
[23:46:04] <gisburn> (with exceptions)
[23:46:13] <gisburn> (for example stevel will suffer today)
[23:46:16] <gisburn> (somehow)
[23:46:19] <gisburn> stevel: ping!
[23:46:20] <gisburn> :-)
[23:46:36] <gisburn> stevel: bugs.grommit.org is down... ;-((
[23:46:48] * gisburn can't update his bugs...
[23:47:24] * jmcp chortles
[23:47:39] <jmcp> "we come in peace [shoot to kill].... Scotty, beam me up!"
[23:47:40] * gisburn purrs and looks in stevel's direction...
[23:47:41] <gisburn> *purr*
[23:47:49] <richlowe> jmcp: you suck for that.
[23:48:00] <richlowe> that's going to be stuck in my head for weeks :\
[23:48:03] <gisburn> richlowe: suck... what ?
[23:48:45] <Gman> startrekking across the universe!
[23:48:46] <gisburn> richlowe: (no, this joke was not adult-related)
[23:48:49] <jmcp> :-D
[23:48:57] <richlowe> gisburn: what jmcp typed were part of the lyrics from The Firm - Star Trekkin'
[23:49:02] <richlowe> one of the most annoying yet catchy songs ever.
[23:49:05] <jmcp> gisburn: as in, "teh suX0r"
[23:49:06] <Gman> yeah
[23:49:18] <jmcp> almost as bad as the crazy frog and the timelords songs
[23:49:30] <richlowe> and anything spitting image did.
[23:49:39] <Gman> or 'tragedy'
[23:49:56] * gisburn does the ritual "summon stevel" dance to get stevel's attention...
[23:50:08] <Gman> stevel is a busy man.
[23:50:12] * gisburn sacrifies pink flowers as part of the ritual
[23:51:19] <stevel> gisburn: should be back up now
[23:51:25] * gisburn grabs a taser and tortures Gman for fun until stevel comes back
[23:51:39] <gisburn> Gman: your're so... lucky... ;-(
[23:51:56] <gisburn> stevel: thanks! :-)
[23:57:00] <richlowe> stevel: did you not bring the bridge back up last night?
[23:57:32] <stevel> richlowe: nope, i wanted to finish the verification stuff first
[23:57:33] * jmcp heads off to a meeting
[23:57:38] <gisburn> Can anyone please remind me that I should visit the author of the gcc-?????????????-varargs stuff and throw him/her into a deep pit filled with komodo dragons ?
[23:57:45] <richlowe> stevel: ah, good idea.
[23:57:57] * richlowe only cared in as far as wondering if it broke again
[23:58:57] <timeless> richlowe: so, zfs is only moving metadata when i move a volume around, right?
[23:59:20] <richlowe> I would think so, at most, but I'm not certain.
[23:59:34] * richlowe is having lots of fun with private symbol table gunk, currently.
[23:59:47] <gisburn> Did Sun pay it's internet traffic bill this year ?





top