January 26, 2007  
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31

[00:01:24] <Gman> stevel, jds has a hack to have a different notification mail?
[00:05:56] <Gman> jmcp, check your mail
[00:06:50] *** yarihm has quit IRC
[00:09:19] <richlowe> Gman: different how?
[00:09:42] <Gman> richlowe, format mostly - i'll forward
[00:10:24] <richlowe> I can see them in the archives.
[00:10:35] <richlowe> all I see different are the svnisms, and the irritating attachment of svn status
[00:11:01] <Gman> the subject of the mail is different
[00:11:35] *** Arnald has quit IRC
[00:11:36] <Gman> one uses the first line of the commit log, the other is the repository that's been changed
[00:16:45] *** flowtech has quit IRC
[00:17:41] * jmcp checks
[00:17:44] <jmcp> Gman: :-D
[00:18:04] <jmcp> thankyou!
[00:20:56] <Gman> jmcp, i'm sure you would have picked up on it sooner than later :)
[00:21:35] <jmcp> early notice is always appreciatd though
[00:21:43] *** hile_ has quit IRC
[00:23:22] *** Starless_ has joined #opensolaris
[00:23:24] *** delewis has quit IRC
[00:26:31] *** flowtech has joined #opensolaris
[00:27:53] <alanc> today's internal psarc mail is starting to make psarc-ext look calm
[00:28:10] <sommerfeld> yes.
[00:28:25] <sommerfeld> particularly the "please vote tomorrow" thing that's the current target of the clue bats
[00:31:27] *** Starless_ has quit IRC
[00:32:18] *** Starless_ has joined #opensolaris
[00:32:45] <Starless_> ...nobody else saw that, did they?
[00:33:37] <sommerfeld> all I saw was that you left and rejoined.
[00:34:29] <Starless_> Cool, I was trying something too clever with my client, multiple windows, exec and catting my log and I was afraid I'd flooded the channel.
[00:34:44] <Starless_> :}
[00:35:05] <sommerfeld> <-- Starless_ has quit ("BitchX: born to raise hell")
[00:35:22] <Starless_> *relieved* Thanks
[00:37:26] *** tsoome1 has quit IRC
[00:41:37] *** jsubl2 has joined #opensolaris
[00:43:55] *** hell` has quit IRC
[00:52:41] *** lloy0076 has joined #opensolaris
[00:53:08] *** darrenm has quit IRC
[00:53:15] <lloy0076> The GRUB boot menu doesn't appear on a Dell Latitude 510 with the disk that has ON55 on it, yet it does appear with my ON54 disk...
[00:53:26] <lloy0076> This is where the new "Developer" option got added.
[00:54:13] <lloy0076> This isn't particularly important because I'm just putting Solaris on the laptop to see "what works" however I suppose that would be a bug of some sort.
[00:56:32] <salmandr> i've gotten at least 2 panics in module flowacct over the last couple days
[00:56:38] <salmandr> this is on snv_54
[00:56:43] <salmandr> where's the best place to report those, if any?
[00:57:04] <comay> bugs.opensolaris.org
[00:57:29] <comay> please file a bug, include a full stack trace and what you did to reproduce the panic.
[00:57:58] <salmandr> how do i get a stack trace? i've got core files in /var/crash/hostname
[00:59:31] <lloy0076> pstack core
[01:00:55] <alanc> does pstack work on kernel cores or just app cores?
[01:01:32] <salmandr> it doesn't seem to work on either unix.0 or vmcore.0
[01:01:33] <salmandr> pstack: cannot examine /var/crash/sfs01/vmcore.0: no such process or core file
[01:01:38] *** uebayasi has joined #opensolaris
[01:01:51] <salmandr> excuse my newbiness, i've never had solaris panic before :)
[01:02:13] *** klocze2 has quit IRC
[01:02:52] <alanc> I've been using it for 12 years now - doesn't happen often to me either
[01:04:15] * lloy0076 ponder
[01:04:31] <salmandr> i was checking out exacct + flow accounting a few days ago, on this test box
[01:04:35] <lloy0076> IIRC it might.
[01:04:45] <salmandr> i got busy with other stuff and didn't turn off flow accounting
[01:04:54] <salmandr> then just today i noticed the box had freshly rebooted
[01:05:19] <lloy0076> salmandr: To attach mdb to a core...
[01:05:29] <lloy0076> salmandr: mdb -k unix.* vmcore.*
[01:05:36] <lloy0076> salmandr: I'm reading from Solaris Internals
[01:05:46] <lloy0076> salmandr: ::showrev will show revision
[01:05:53] <lloy0076> salmandr: ::status will show status
[01:05:57] <comay> salmandr: cd /var/crash ; echo ::stack | mdb 0
[01:06:33] <lloy0076> salmandr: And ::panicinfo
[01:07:21] <salmandr> thanks, should i attach all 4?
[01:07:28] <salmandr> showrev, status, stack, & panicinfo?
[01:07:29] <lloy0076> I think only two.
[01:07:57] <lloy0076> salmandr: As I said, I'm reading from a book. I'm not sure what to look for but you might see something that catches your eye.
[01:08:32] <lloy0076> salmandr: Sun's guide to MDB is at: http://docs.sun.com/app/docs/doc/816-5041/6mb7ae3j0?a=view
[01:08:43] <salmandr> k
[01:09:09] <salmandr> well the panic was a null pointer deref in flowacct
[01:09:57] <salmandr> neither ::stack nor ::panicinfo mean anything to me unfortunately, hopefully someone can make something of them
[01:10:04] <sommerfeld> you want mdb -k <number> (0, 1, 2, 3, ...)
[01:12:49] <richlowe> you shouldn't need the -k
[01:13:14] <sommerfeld> it will figure out unix.$N and vmcore.$N
[01:14:31] *** sickness has quit IRC
[01:16:33] *** sickness has joined #opensolaris
[01:17:13] <salmandr> luckily either way gives me the same output, heh
[01:18:46] *** gdamore has joined #opensolaris
[01:19:01] <comay> salmandr: what was the the first reference to flowacct in the stack?
[01:19:02] <gdamore> hi!
[01:19:07] <uebayasi> hi :)
[01:19:32] <gdamore> uebayasi-san: another NetBSD fugitive? :-)
[01:19:39] <salmandr> comay: flowacct_invoke_action+0x35(2, e2f26f00)
[01:20:26] <CIA-22> bm42561: 6500027 after partial unconfigure of XSB (failure due to lack of memory) cannot reconfigure XSB
[01:21:07] <uebayasi> not from NetBSD, but i'm leaving this "NetBSD" company i'm working for ;)
[01:22:02] <gdamore> heh. :-)
[01:22:03] <CosmicDJ> whats wrong with netbsd?
[01:22:07] <alanc> gdamore: haven't seen you around much lately
[01:22:19] <gdamore> i was on vacation last week -- ski trip to tahoe
[01:22:24] <uebayasi> CosmicDJ: NOTHING!!!
[01:22:27] <alanc> ah
[01:22:45] <alanc> I've heard vacations are good things
[01:22:48] <gdamore> unfortunately, i've been in loonix hell lately.
[01:22:55] <uebayasi> HEH
[01:23:29] <gdamore> my current gripe of the week: loonix framebuffer support _sucks_ for doing anything besides X. :-)
[01:23:57] <gdamore> i might actually get stuck doing major surgery on intelfb as a result.  i'd rather be doing solaris stuff.
[01:26:50] *** rodrick-brown has joined #opensolaris
[01:27:21] *** hspaans has joined #opensolaris
[01:27:34] <hspaans> g'day
[01:29:55] <hspaans> wasn't SUNWvts for validation?
[01:30:18] <sommerfeld> hardware diagnostics
[01:30:56] *** swa_wtf has quit IRC
[01:30:59] <hspaans> also for production machines?
[01:31:01] <gdamore> not really diagnostics.
[01:31:11] <gdamore> but for customer validation of delivered hardware
[01:31:52] * gdamore has written SunVTS test suites, and thinks it is pretty lame.
[01:32:02] <jbk> heh
[01:32:10] <jbk> it's still used quite a bit
[01:32:11] <hspaans> as in get hardware do validation and disable it
[01:32:26] <jbk> yeah, you're not going to run it while your production apps are running
[01:32:35] * alanc has had long fights with SunVTS engineers about the uselessness of their keyboard tests
[01:32:39] <gdamore> no, you wouldn't want to do that, at all.
[01:32:40] <hspaans> guess what jbk ;-)
[01:32:44] <jbk> more likely if you got new hardware or replacement hardware
[01:32:56] <gdamore> alanc: the synch. serial tests are "known" to fail frequently. :-)
[01:33:08] <alanc> these aren't even serial tests
[01:33:21] <gdamore> i wrote the vts test suites for the dca and vca crypto boards.
[01:33:28] <alanc> they open /dev/kbd and send the ioctls to set the LEDs and then read back if they were set
[01:33:37] <gdamore> heh.  pretty damned lame.
[01:33:40] <jbk> in fact, once some hsPCI+ assemblies arrive, i'll be running it for a few days
[01:33:42] <wamty> Is it legal to use dns other than provided by ISP?
[01:33:53] <alanc> they got totally pissed when we did virtual keyboard and their tests passedd with no keyboard attached
[01:33:58] <gdamore> wamty: not really the best place to ask, but i don't see why not.
[01:33:59] * hspaans thinks a 15k with suncluster and veritas and emc isn't the place to have sunvts running
[01:34:15] <alanc> wouldn't listen to the argument that they never tested the keyboard, only the software driver's ability to set a variable and read it later
[01:34:16] <jbk> not while live :)
[01:34:22] <hspaans> wamty: resolver? no not reallt
[01:34:38] <hspaans> jbk: business critical ;-)
[01:34:39] <gdamore> alanc: haha.  well, the vts tests only casually test hardware.
[01:34:53] <jbk> that seems.. very bad
[01:35:03] <gdamore> (which doesn't mean that they don't eat up all available cpu in the process. :-)
[01:35:04] <jbk> at least if you're doing the cpu/memory tests
[01:35:11] <hspaans> jbk, no that seems as free lunch for me tomorrow
[01:35:31] <alanc> their tests against things they can actually get usable data back from, like disk and memory, seem much more useful
[01:35:48] <alanc> but a keyboard isn't really something you can have an automated software test of
[01:36:04] <gdamore> yes, but even those tests are limited.  the factory has a much more comprehensive and brutal test suite that is not Solaris-based, IIRC.
[01:36:23] <CIA-22> Tom Erickson: 6513863 java_api tests assume that PATH includes current directory (remove unneeded file)
[01:36:24] * hspaans is never going to joke about standby when he leaves a sun building
[01:36:47] <gdamore> alanc: well, you can test whether the keyboard is connected, and reporting basic configuration info (country code, etc.) even if you can't test keyswitches without human interaction
[01:37:06] <gdamore> its sort of like the video or audio tests -- without a human there to confirm, its hard to say much useful.
[01:37:41] <gdamore> i.e. "not known to be bad != known good"
[01:38:00] <alanc> yeah - at least we got HCTS to stop using texteroids as their test of functioning video hardware
[01:38:17] <alanc> though when we switched to Xorg and had no DPS, they didn't have much choice 8-)
[01:38:28] <gdamore> heh.
[01:38:47] <gdamore> we _had_ to do crypto testing as part of FIPS 140-2 certification
[01:39:06] <alanc> now they run x11perf quickly to at least make sure most basic operations can be run a few thousand times without crashing
[01:39:41] *** swa_wtf has joined #opensolaris
[01:39:52] <gdamore> our customers are asking for an OpenGL test suite.  the best we have right now is glgears
[01:40:55] <alanc> just hope they don't decide glgears is also your performance benchmark
[01:41:04] <gdamore> hmmm... i'm disappointed that the laptop community never nominated any "members".
[01:41:40] <alanc> ooh, we have an opening for a summer intern in our group
[01:41:54] <gdamore> our customers have a pretty basic 3D need -- the creator runs well for them.  unfortunately there is no graphics hardware that we can put in a mobile sparc machine that has _any_ 3d support
[01:42:19] <rodrick-brown> why not use x86?
[01:42:30] <gdamore> military customers with legacy sparc apps
[01:43:08] <gdamore> said customers do not have source, and are _stuck_ with Sun's OpenGL (i.e. can't use XiG server, either)
[01:43:09] <alanc> there's already hundreds of companies selling x86 laptops - the sparc laptop market is much less crowded
[01:43:38] <rodrick-brown> can solaris do processer affinity on network i/o?
[01:43:41] <gdamore> we might be doing a non-SPARC based machine in the future, but it will be targetted as an extreme high end mobile server, not a typical laptop
[01:43:48] <alanc> pretty much just Tadpole (gdamore's company) and Naturetech last I saw
[01:44:22] <alanc> Sun doesn't even resell them anymore
[01:45:00] <gdamore> rodrick-brown: i'd be surprised if it didn't, with interrupt binding and the ability to use NUMA-like memory affinity, it should be possible to set it up.  but i've never looked much into it
[01:45:26] <rodrick-brown> YEah didnt even know it was possible on linux, till today
[01:45:33] *** jafari has joined #opensolaris
[01:45:36] <rodrick-brown> our Market data team is so pro linux :)
[01:46:36] <sommerfeld> the fireengine stack in s10 introduced a form of processor affinity on connections
[01:46:53] <uebayasi> mmm military market
[01:49:45] <sommerfeld> there's also the "soft rings" support which allows the processor fanout to happen in drivers rather than in ip
[01:49:54] <e^ipi> alanc: you're the X guy...
[01:50:12] <e^ipi> alanc: do you know anything about beryl/compiz/xgl ?
[01:50:13] <alanc> yeah
[01:50:23] <alanc> a little
[01:50:33] <e^ipi> because it'd be nice if i could tell my linux-using collegues to shove it up their ass
[01:50:44] <e^ipi> what's the likelyhood it'll work in solaris?
[01:51:24] *** lloy0076 has quit IRC
[01:51:29] <e^ipi> personally i think it's flashy nonsense,  but I've had people seriously use it as a reply to the question "what can linux do that solaris can't"
[01:51:36] <alanc> nrubsig had Xgl working on Solaris 2 years ago, but Xgl isn't all that interesting - it's just an Xnest-like X server that sits on top of your normal X server and uses OpenGL to draw everything
[01:51:48] <alanc> beryl/compiz should work on Solaris with a little work
[01:51:56] <alanc> work we should be doing soon
[01:52:32] <e^ipi> IIRC it depends on xorg 7, is why i asked you
[01:52:38] <alanc> the only reason it's not running today is lack of time - all the support should be there - I think blastwave could make packages now
[01:52:50] <alanc> oh yeah, not everyone has that on their solaris desktops yet
[01:52:53] <sommerfeld> for someone who isn't following this, what's the one-sentance summary of what these packages are?
[01:52:59] <gdamore> that could be said to be true of a great deal of software. :-)
[01:53:04] <alanc> sommerfeld: Eye Candy
[01:53:22] <alanc> though the accessibility team is looking at beryl to replace gnome-magnifier
[01:53:26] <sommerfeld> alanc: i figured *that* much out.
[01:53:53] <alanc> they are compositing managers - they take your apps and apply various transformations/effects to them
[01:53:58] <gdamore> isn't "Eye Candy" the description of most of the lines of code coming out of the Linux arena for the last few years?
[01:53:59] * gdamore ducks
[01:54:13] <rodrick-brown> gdamore, E *cough*
[01:54:14] <alanc> for instance, you can set beryl so that when you close a window, it burns off the screen
[01:54:21] <alanc> with flames and such
[01:54:37] <rodrick-brown> E is funny lets release a bloated wm that looks pretty, rewrite it from scratch to only do it again
[01:54:39] <alanc> or you can get nice smoothed magnification at OpenGL speeds
[01:54:58] <e^ipi> http://youtube.com/watch?v=ZD7QraljRfM <-- that's beryl
[01:55:01] <rodrick-brown> I dont get why people think UNIX need's X
[01:55:07] <rodrick-brown> why not just kill X like apple did
[01:55:16] <gdamore> the funny thing was not long ago I heard someone describe E as a "light-weight wm"
[01:55:17] <rodrick-brown> and adopt something new and more robust?
[01:55:17] <alanc> I'm working right now on Xorg 7 for nv_58
[01:55:18] <steleman> it doesn't. vt100 does great.
[01:55:41] <alanc> rodrick-brown: because unlike Apple, no one else is going to assign a team of hundreds to replace X
[01:55:42] <e^ipi> rodrick-brown: many people have tried to replace it, none of them took off
[01:55:46] <gdamore> heh.  I've been working on a graphics app that runs direct on the framebuffer.  it is cool.
[01:55:55] <rodrick-brown> apple's UI is so sleek and performance is just as snappy as windows
[01:56:20] <gdamore> X's problem isn't performance.  Its bloat.
[01:56:23] * Auralis thinks osx gui is neither snappy nor elegant
[01:56:27] <alanc> rodrick-brown: I bet apple's UI cost them more than all the Linux & Unix companies spend combined on X
[01:56:30] *** dlynes_laptop has joined #opensolaris
[01:56:46] <e^ipi> it's really the only thing apple does right
[01:56:51] <e^ipi> mach? c'mon....
[01:56:58] <e^ipi> what a horrid kernel choice
[01:56:59] <Stric> osx on an xserve is so fun.. the gfx card isn't that fast, so the UI is just painfully slow :P
[01:57:06] <gdamore> Mozilla/Firefox is the biggest bloat hog.  Without it, my machine runs much zippier. :-)
[01:57:12] <uebayasi> someone suggested somewhere to use OpenGL API as in-kernel framebuffer API
[01:57:41] <e^ipi> Stric: plus the kernel spends most of it's time in IPC & VM ops... OSX "server" is kinna a joke, performance wise
[01:57:46] <gdamore> ueb: yah, I've heard that.  NetBSD was talking about something like that at one point.
[01:58:03] <alanc> uebayasi: yep, they suggested that, and Xgl was early work towards it, but not enough people wanted to work on it for it to reach critical mass, so it's mostly abandoned now
[01:58:08] <rodrick-brown> I think X is a disgrace no offense, I seen Windows Vista for the first time last week
[01:58:15] <rodrick-brown> and its really not bad
[01:58:31] <alanc> and there's not much Vista does you can't do in X
[01:58:52] <e^ipi> beryl was pretty much a response to vista
[01:59:15] <e^ipi> MS: "flashy UI effects! hooray!" so the beryl team one-upped them
[01:59:17] <rodrick-brown> alanc I dont know vista is really snappy with all the eye candy you get
[01:59:36] <gdamore> alanc: only if you add OpenGL to X.
[01:59:40] <alanc> have you tried a recent Fedora or Ubuntu with compiz/beryl?
[01:59:41] <rodrick-brown> e^ipi, beryl is ugly it might be good code but they need better designers :)
[01:59:51] <e^ipi> alanc: it's  really quite snappy
[01:59:52] <rodrick-brown> alanc naw not yet
[01:59:58] <alanc> it's not that bad
[02:00:32] <e^ipi> though it had the weird effect of permanently breaking java gui apps
[02:00:45] <alanc> for those who haven't seen any screen shots of beryl/compiz: http://www.beryl-project.org/features.php
[02:00:50] <e^ipi> i tried beryl, and mysteriously netbeans stopped working ever again
[02:00:55] <alanc> (beryl is a fork of compiz - same idea, different people)
[02:01:08] <e^ipi> it would load... but not draw anything inside it's parent window
[02:01:09] <gdamore> so, what's the recommended way to get code these days?  Hg, or SVN?  I've had a hard time following all the Hg snafus.
[02:01:48] <e^ipi> alanc: so anyways, point being it should theoretically work in solaris?
[02:02:15] <Gman> gdamore, depends which code you're trying to get obviously :)
[02:02:15] <alanc> should, if you build Xorg 7 on your own today or wait a few months for us to deliver all the bits integrated
[02:02:41] <gdamore> OS/Net, of course.  Nothing else is terribly interesting (to me :-)
[02:02:43] <Gman> gdamore, if it's ON, then the hg gate is back up and working fine
[02:03:10] <e^ipi> oh, yeah... speaking of ON, i gotta remember to talk to Tpenta
[02:03:17] <e^ipi> but first, i run out for coke
[02:05:43] * gdamore is disappointed no community recognized him as a "Contributor".  Not that he knows what that means. :-)
[02:06:32] *** jamesd has quit IRC
[02:06:41] <alanc> would you like the X community to list you as "Contributor to campaign to shame SPARC graphics into doing the right thing" ?   8-)
[02:06:54] <gdamore> hahaha. :-)
[02:06:57] <Auralis> lol
[02:07:02] <Gman> gdamore, you've submitted patches, right?
[02:07:03] <gdamore> too bad i don't think it worked.
[02:07:07] <gdamore> Gman: yes.
[02:07:15] <Gman> then you should be on that list, definitely
[02:08:04] <gdamore> heh.  I've contributed code to every major release of Solaris since Solaris 8, inclusive.
[02:08:07] <Gman> just depends on the person in the particular community you're most associated to who compiled the list
[02:08:18] <sommerfeld> solaris hasn't had any major releases since solaris 2.0
[02:08:20] <gdamore> nobody from the laptop community mentioned me. :-)
[02:08:21] <Gman> gdamore, i'd be inclined to mail sch
[02:08:36] * Gman thinks that list isn't the best way to do it
[02:08:37] <gdamore> nobody from the laptop community mentioned "anyone", it appears
[02:09:32] <gdamore> sommerfeld: well every "minor" release, then.  sorry, i forgot the "major" vs. "minor" distinction.
[02:10:20] <gdamore> there will never ever be another major release of Solaris, because whatever the next non-SVR4.2 OS is, I'm sure marketing won't allow them to call it Solaris. :-)
[02:10:51] * gdamore was very amused when SunOS 4.1.3 was retroactively labeled Solaris 1.x.
[02:11:32] <gdamore> anyway, time to feed the kids while Hg clone runs.  ttfn.
[02:11:44] <alanc> sch has been poking community leaders - not all have responded
[02:12:03] *** sommerfeld has quit IRC
[02:12:49] *** jafari has quit IRC
[02:15:16] *** piefra has quit IRC
[02:16:56] *** slowhog has quit IRC
[02:17:00] *** jamesd has joined #opensolaris
[02:17:00] *** ChanServ sets mode: +o jamesd
[02:17:50] <CIA-22> yw161884: 6221685 Wrong arg used in ddi_dma_addr_bind_handle() after ddi_dma_mem_alloc()in amr driver; 6267612 amr issues bogus 'package timeout' message; 6309908 amr sleeps for resources when it shouldn't
[02:18:18] *** slowhog has joined #opensolaris
[02:18:43] *** slowhog has quit IRC
[02:19:16] *** slowhog has joined #opensolaris
[02:22:30] <Gman> alanc, amusing mail from mali :)
[02:23:00] <alanc> heh
[02:23:11] <alanc> wonder if that should have gone to desktop-discuss
[02:23:54] *** simford has joined #opensolaris
[02:24:01] <alanc> it's following on several other recent email threads lately about whether gnome-screensaver is worth the effort, especially if it requires forking from the community like we already had to with xscreensaver
[02:24:26] <Gman> i'm *really* surprised we'd have to fork
[02:24:37] * Gman thinks it sounds more like a communication breakdown
[02:25:34] <MikeE> alanc is that because of some of the sunray extensions for xscreensaver?
[02:25:46] <MikeE> (and they want to do the same with gnome-screensaver?)
[02:25:55] <alanc> apparently previous discussion with the community ended with them saying that splitting gnome-screensaver into a GUI side and a setuid-root-authentication engine like we did with xscreensaver was not acceptable to them, and our PAM was broken if it required root access
[02:26:16] <MikeE> hmmm interesting perspective...
[02:26:19] <alanc> MikeE: nope - there's no sunray specific code in xscreensaver that I'm aware of
[02:26:23] *** yippi has joined #opensolaris
[02:26:57] <MikeE> hmmm like like mikee was missinformed... they have lots of pam-funk that they possibly use to glue the pieces together then
[02:27:21] *** dunc has quit IRC
[02:27:28] <alanc> yeah - xscreensaver calls PAM, and Sun Ray puts pam modules on the stack to do their voodoo
[02:27:58] <MikeE> isn't there a new in-program/library "sudo like" function that is new with S10?
[02:28:03] <alanc> Sun Ray usually just pushes for more standard PAM code with less knowledge of the environment, not more
[02:28:26] <MikeE> isn't something like that therefore capable of providing "is that the right-password" type of tests without requireing suid/root components?
[02:28:28] *** piefra has joined #opensolaris
[02:28:37] <alanc> MikeE: doesn't sound like anything I've heard of
[02:29:12] <MikeE> i remember reading that on some slide somewhere.... probvably very S10 only, so that could put a damper on its use/popularity
[02:29:12] <alanc> closest in Solaris to anything sudo-like I know of is RBAC, which has been in since Solaris 8, but still doesn't help here
[02:29:26] <Gman> alanc, wasn't that original split only due to the fact that jwz didn't like gtk and thought it was grossly insecure?
[02:30:07] <alanc> the process split is because gtk refuses to run setuid, because the gtk maintainers believe it's insecure
[02:30:31] <dwc-> I've never really been sure why my screensaver needs to be gtk-ified
[02:30:36] <alanc> jwz not liking gtk is simply why the changes to use it went upstream
[02:30:39] *** stevel has quit IRC
[02:30:40] <Gman> right, cause by jwz's insistence he wouldn't accept any other patch
[02:30:55] <alanc> dwc-: gtk is the only accessible toolkit in Solaris
[02:31:30] <alanc> so it needs gtk so that blind users or users who need screen magnifiers can actually unlock the screen
[02:31:44] <Gman> i don't think it was clear at the time whether or not that was valid
[02:32:01] <Gman> but given everyone wanted the code in xscreensaver the split was done to satisfy jwz
[02:32:09] <Gman> [or at least, that's what i thought the chain of events were]
[02:32:22] <Gman> gnome-screensaver is different because the maintainer isn't a dick )
[02:32:31] <alanc> jwz has softened his stance a bit over the years: http://www.jwz.org/xscreensaver/toolkits.html is his current position
[02:32:35] <comay> salmandr: are you still around?
[02:32:36] <dwc-> mm, sounds like a sorta reasonable reason
[02:32:49] <salmandr> comay: yeah
[02:33:02] *** ___schily has joined #opensolaris
[02:33:25] <alanc> the process split is the only way to have a setuid-root application with a gtk GUI - http://www.gtk.org/setuid.html
[02:33:32] <comay> i believe the issue you found is a known bug: 6509271
[02:33:48] <comay> it's due to be fixed in the next few builds
[02:34:22] <salmandr> comay: i saw that when i was filing it, but i couldn't tell if it was a dupe or not
[02:34:27] <alanc> (and because later versions of gtk just call exit() when they detect they're in setuid apps)
[02:34:32] <salmandr> i mentioned it and put it in the "related bugs" field
[02:34:44] <salmandr> good to hear it'll be fixed soon :)
[02:35:02] *** gustav3d has joined #opensolaris
[02:35:34] * brendang_ is off to SVOSUG
[02:35:35] <comay> it looks very similar - i've recategorized it for evaluation
[02:35:46] *** wamty has quit IRC
[02:35:47] <brendang_> assumnig that I can find Santa Clara, and don't get lost in traffic
[02:36:07] <Gman> follow the 101
[02:36:08] <Gman> easy :)
[02:36:15] <brendang_> I printed the map :)
[02:36:22] <salmandr> comay: great, thanks
[02:36:49] <Gman> brendang_, i think you can see it from the highway :)
[02:37:05] *** yongsun_ has joined #opensolaris
[02:37:08] <comay> what time is svosug tonight? 7:30?
[02:37:47] <alanc> Gman: not from the 101, but from the montague expressway
[02:37:59] <Gman> alanc, ahh
[02:39:16] *** lloy0076 has joined #opensolaris
[02:40:04] <alanc> the tower you can see from 101 with the Sun logo is not the building you want to go to
[02:40:18] <alanc> the SCA campus is all shorter buildings (2-3 stories)
[02:40:30] <MikeE> alanc , took me a while to find it... http://docs.sun.com/app/docs/doc/816-5166/6mbb1kq1c?a=view
[02:41:02] <alanc> oh that
[02:41:21] <MikeE> not sure if its being used much out there... i would think not...
[02:41:28] <alanc> hmm...
[02:41:31] <MikeE> not yet at leat..
[02:41:40] <lloy0076> Gawd, Microsofts' fdisk is dumb.
[02:42:05] *** lloy0076 has quit IRC
[02:45:19] *** Theoden-Nexenta has left #opensolaris
[02:46:57] *** jsubl2 has quit IRC
[02:50:10] *** __schily_ has quit IRC
[03:00:07] *** edwardocallaghan has joined #opensolaris
[03:00:12] <edwardocallaghan> Hi all
[03:05:22] <jbk> hi
[03:05:49] *** comay has quit IRC
[03:06:54] <CIA-22> Christopher Horne: 6452574 we should improve management of the /dev/.devlink_db file
[03:08:00] *** Plazma has joined #opensolaris
[03:12:38] <edwardocallaghan> OK good night
[03:12:41] <edwardocallaghan> :p
[03:12:44] *** edwardocallaghan has left #opensolaris
[03:15:22] *** Plazma has quit IRC
[03:25:49] <CIA-22> rz201010: 4421990 allow changing keyboard bell frequency
[03:26:57] <alanc> heh - waited years for the kernel to provide an interface for that, and soon we'll have two within weeks of each other
[03:28:24] *** gm152_ has joined #opensolaris
[03:30:15] <jbk> so it'll be twice as good, right?
[03:30:34] *** CosmicDJ has quit IRC
[03:30:39] <alanc> yep
[03:31:02] <alanc> one will let you set the default, so that even the console beeps outside X go at a different volume/frequency
[03:32:07] <alanc> the other will let X sound a specified frequency for a specified time instead of the current "ioctl(bell_on); sleep(duration); ioctl(bell_off);" it does now that makes your Xserver seem to lock up when you accidentally cat a binary to your xterm and it plays 100 beeps in a row
[03:33:24] *** ibb has quit IRC
[03:38:59] *** gm152 has quit IRC
[03:42:56] *** slowhog has left #opensolaris
[03:44:38] *** jsubl2 has joined #opensolaris
[03:45:52] *** Theoden-Nexenta has joined #opensolaris
[03:47:58] *** dunc has joined #opensolaris
[03:48:23] *** dunc has joined #opensolaris
[03:48:40] <UnixTitan> I <3 sqlite3.
[03:48:51] <UnixTitan> just wrote a complete image databaser in around 2 and a half hours.
[03:50:21] <e^ipi> see how high it can scale
[03:50:34] <e^ipi> connect 100,000 users to it
[03:54:04] <e^ipi> :)
[03:54:27] <e^ipi> my phonebook is written in qt4 + sqlite
[03:54:33] <e^ipi> it's quite nice to work with
[03:56:02] <rodrick-brown> sqlite is cool
[03:57:39] <e^ipi> not ultra powerful, but that's not what it's for
[03:58:16] <e^ipi> though i've been tempted to plug it in to glassfish before
[03:58:19] <e^ipi> because that would be funny
[04:00:04] *** UnixTitan has quit IRC
[04:00:49] <gustav3d> funny in what way ?
[04:01:56] *** UnixTitan has joined #opensolaris
[04:02:01] <e^ipi> big memory hungry j2ee server, being driven by a tiny, fragile little sql "server"
[04:02:26] <gustav3d> ok =)
[04:02:33] <Gman> hrm, the osvug thing must be on soon, right?
[04:04:40] <Gman> seemingly not
[04:04:49] <gustav3d> e^ipi: what lighter alternatives do you recommend ?.
[04:05:18] <e^ipi> to... what? glassfish?
[04:05:22] <e^ipi> anything other than glassfish
[04:05:23] <gustav3d> yes.
[04:05:47] <e^ipi> tomcat, geronimo, jboss...
[04:05:50] <e^ipi> oracle
[04:05:51] <e^ipi> heh
[04:06:01] *** sahafeez has quit IRC
[04:06:20] <jbk> somehow 'lighter' and 'oracle' seem wrong said in the same breath
[04:06:50] <gustav3d> somehow tomcat is not the same kind of product.
[04:07:38] <CIA-22> rw148561: 6508652 pcmcia has memory leak after TTCP test
[04:07:43] <jbk> of course i love what oracle's expert told me to do today
[04:07:56] <e^ipi> ?
[04:08:44] <jbk> i'm supposed to configure a switch to essentially forward ethertype 0800 packets (i.e. IP) while not forwarding ethertype cafe (vcs llt)
[04:09:21] *** Ludo_ has quit IRC
[04:09:31] <jbk> so that way I can use ipmp
[04:09:53] <jbk> instead of the vcs agent designed to do the same thing for vcs hb links
[04:10:27] <e^ipi> i have no idea what you just said.
[04:11:32] <jbk> for veritas cluster server, you have two separate networks for sending heartbeats and broadcasting cluster configuration changes
[04:12:15] <gustav3d> seperate as in Virtual or physical ?
[04:12:27] <jbk> if using switches, ideally, each network goes to a separate switch, so the failure of one switch does not disrupt the other network
[04:12:32] <gustav3d> ah ok
[04:12:36] *** rodrick-brown has quit IRC
[04:12:43] *** rodrick-brown has joined #opensolaris
[04:12:48] <jbk> but the minimum requirement is separate broadcast domains (i.e. vlans)
[04:13:13] <gustav3d> switches uses VRRP or so to replace the failed one ?
[04:13:24] *** logic has joined #opensolaris
[04:13:35] *** Odin-LAP has quit IRC
[04:13:36] <jbk> no, you just have two redundant networks
[04:13:39] <gustav3d> ah ok
[04:13:50] *** Ludo_ has joined #opensolaris
[04:13:58] <jbk> but so, oracle rac piggybacks on these for it's cache coherency traffic
[04:14:35] <jbk> however, in their infinite wisdom, oracle decided that they needed to write their own cluster software (crs), but would allow it to integrate with other cluster software (such as vcs or sun cluster)
[04:14:55] <gustav3d> out of topic abit: does oracleDB provide cluster that has not a single point of failure ?.
[04:15:33] <jbk> well do you want it continously available, or where it can just startup on another machine if one dies?
[04:16:02] <jbk> (it can do either, depending on how much money you want to spend)
[04:16:17] <gustav3d> well that answers it. thanks
[04:16:38] <gustav3d> i will never work on a project with oracle budgets anyhow.  just interested in general.
[04:16:43] <jbk> but so, oracle requires an ip address to use on one of these piggybacked connections between every node of the cluster
[04:17:20] <jbk> so veritas provides an agent that will move the ip addresses on all the nodes to a working link if the one it's on happens to fail somewhere in the cluster
[04:17:23] <jbk> this works fine
[04:17:27] <jbk> but the moment you do it
[04:17:45] <jbk> oracle decides it needs to reboot all but one of the nodes in the cluster
[04:17:54] <gustav3d> sigh
[04:18:05] <jbk> it's annoying
[04:18:28] <jbk> as crs (oracle's cluster) is poorly documented, and not even oracle seems to understand it (not too surprising from talking to our dbas)
[04:18:43] <gustav3d> thats abit scary =)
[04:19:12] <jbk> but so their solution was to somehow cause the IP traffic to be bridged between the vlans, while leaving the vcs hb traffic (which is a different ethertype) isolated
[04:20:41] <gustav3d> they certenly not make it easy. better hope the complex solutions they have work.
[04:22:14] <jbk> well it's stupid, so i'm hoping i can get a better answer
[04:22:36] <jbk> as neither veritas nor oracle can point to anything in their config that says you must configure the interconnect in such a convoluted manner
[04:22:37] <Doc> CRS is hardly what i'd call a cluster
[04:22:45] <Doc> it's what oracle call a cluster, but i wouldnt trust them to know
[04:22:46] <jbk> well they claim it is
[04:22:57] <boyd> Doc: :)
[04:23:00] <jbk> i didn't say i agreed :)
[04:23:34] <jbk> i think they'd be better to stick to databases
[04:24:08] <jbk> they also store the config in a dbf file (when all you have is a hammer..)
[04:24:30] <jbk> i wish we could rip it out
[04:25:43] *** jsubl2 has quit IRC
[04:25:53] *** alanc is now known as alanc-away
[04:26:02] *** logic_ has quit IRC
[04:26:13] <gustav3d> myself is to start learning about mysql cluster's , or is there any other decent alternatives to consider ?
[04:26:30] *** Odin-MAC has joined #opensolaris
[04:27:25] * boyd wishes yet again for more functional parity between OpenSSH and SunSSH
[04:27:59] <dlg> boyd: just use openssh
[04:28:13] <jbk> or, you can always now submit your own improvements
[04:28:26] <boyd> But then I lose LDAP and KRB features IIRC
[04:29:07] <boyd> jbk: True, but it would be nice to not have to port every damn feature from a related program
[04:29:38] <dlg> boyd: via pam?
[04:30:02] <boyd> I can't recall what the reason is.. It may have been PAM
[04:30:12] <Gman> silicon valley users group meeting starting now
[04:30:23] <Gman> [though can't actually here anything]
[04:30:27] <boyd> You there Gman
[04:30:28] <boyd> ?
[04:30:40] <Gman> boyd, i'm dialling in
[04:30:41] <boyd> Oh, audio? It's always crappy. I've given up
[04:30:59] <Gman> i'll hang up if i can't hear
[04:31:03] <jbk> i missed the local one here due to work :(
[04:31:18] <Gman> http://mail.opensolaris.org/pipermail/opensolaris-discuss/2007-January/023597.html
[04:31:20] <jbk> though chances are, it was mostly coworkers and sun employees
[04:31:25] <Gman> ^^ for anyone wanting to dial in
[04:31:43] <Gman> jbk, doubt it, if it's an opensolaris meet, you're more than welcome to attend
[04:32:10] <jbk> well, i'm in a technology backwater, and work for the single largest employer (and largest sun customer) in the area
[04:32:55] <jbk> though in some ways its nice, we all know everyone in the local sun office
[04:33:04] <Gman> boyd, you're not missing much yet, have only heard one dude say 'testing' so far :)
[04:33:10] <Gman> jbk, nod
[04:33:13] <boyd> Heh
[04:33:20] <boyd> At least you heard it :)
[04:33:29] <dlg> we auth at work via krb using openssh
[04:33:29] <dlg> on sol 9 and 10
[04:33:29] <dlg> krb and nis
[04:33:29] <dlg> if i wasnt lazy id try ldap
[04:33:56] <boyd> openssh sshd too?
[04:34:14] <CIA-22> John M Martin: 6367939 First JDS login after reboot scrambles the display on two card configurations
[04:34:26] <boyd> I may be recalling incorrectly, or it may have changed since I last looked
[04:35:31] *** Odin-MAC is now known as Odin-LAP
[04:35:45] <jbk> i just wish the host access control in ldap was better
[04:36:54] <Gman> anyone happen to be in SCA so they can dial into the conf call? :)
[04:37:23] <jbk> nope
[04:37:39] <jbk> <- flyover state :)
[04:39:48] * boyd reads osol-discuss. Cheers at the tantalizing hint that we may see the back of Jive. Worries that the replacement will be as bad. Wonders how we got to the stage where the popular interfact to forums is a web app that has less functionality than the clients we used 10 years ago
[04:40:25] <Gman> boyd, looks like derek has successfully fixed it
[04:41:04] <boyd> No, he's fixed the most recent showstopper. He hasn't fixed all the other issues with Jive. Like thread breakage 4 e.g
[04:41:23] <Gman> same thing
[04:41:40] <boyd> Huh?
[04:41:48] <jamesd> is everyone online for the   usergroup meeting concall?
[04:41:57] <Gman> jamesd, i am, not hearing jack though
[04:42:08] <jamesd> me either
[04:42:08] *** laca has quit IRC
[04:42:11] <jbk> i hear typing
[04:42:14] <jamesd> http://blogs.sun.com/aland/entry/svosug_this_thursday_why_joyent  ... for anyone that needs it
[04:42:16] <Gman> boyd, suspect there won't be too much effort to fix non-showstopper stuff
[04:42:29] <jamesd> my phone is on mute so i'm good
[04:42:35] <Gman> me too
[04:42:44] <boyd> Hence my desire to ditch the whole POS
[04:42:47] * Gman half thinking of asking who's online and whether sca is there
[04:43:12] <Gman> jamesd, sent a mail to osol-discuss about it in any case
[04:43:30] <jamesd> yeap just got done reading it now
[04:44:32] <Gman> ooh, activity :)
[04:44:37] <jbk> \o/
[04:46:00] <Gman> boyd, the sound seems good
[04:46:07] <Gman> though benr hasn't started talking yet
[04:46:14] <jbk> we use cisco's iptv.. just wish there was a solaris client
[04:46:36] * jbk almost never uses his dell laptop (has to run windows)
[04:47:40] <boyd> Gman: Thanks for the info
[04:50:06] <Gman> benr's uploading the slides for tonights talk
[04:51:54] <Gman> hahahaha
[04:52:02] <Gman> brendan gregg you nut.
[04:52:07] <jamesd> roflao
[04:52:19] <boyd> okok I'm calling
[04:52:20] <jamesd> "Hi GREG!!!"
[04:52:31] <Gman> 'I'm brendan gregg and i'm an alcoholic'
[04:52:44] <Gman> [the group is doing introductions]
[04:52:48] <boyd> It's a holiday here... I must have better things to do on a beautiful friday afternoon
[04:53:05] * boyd wonders if he will just be able to read the gman transcript :)
[04:53:56] <Gman> all sorts of dudes present - from sun seems to be josh berkus, steve lau, michelle olson, alan dubuff, alan burlison, gary pennington, ..
[04:54:11] <Gman> george wilson from zfs
[04:54:16] <Gman> [if this isn't useful, i can stop]
[04:54:31] <boyd> I thought stevel was from Singapore
[04:54:34] <Gman> mark mabeal [?]
[04:54:45] <Gman> boyd, no, he lives in sf with 2 beagles
[04:54:50] <Gman> [and obviously his wife]
[04:54:57] <boyd> .. and a ruined piano
[04:55:01] <Gman> :)
[04:55:02] <jbk> haha
[04:56:05] <Chile`> there's no way to reduce java's ram usage, is there? I still don't quite get why firefox can manage to run in 30 megs, yet sun studio takes 370mb with no loaded projects/files, and another instance of java that apparently starts by default takes up another 80mb of ram.
[04:56:29] <jbk> you can control the heap size
[04:56:56] <Chile`> will that actually help?
[04:57:23] <jbk> well most everything lives in the heap afaik
[04:57:41] <Auralis> the main reason for the mem hunger of java is that it starts for every thing a new vm
[04:58:15] <boyd> I never understood why it's made that way
[04:58:24] <Chile`> Auralis: still though, why does NetBeans take up 3x the ram of VS2005 to display an empty window?
[04:58:31] <Chile`> I just don't get whats actually *in* that ram
[04:58:48] <dlg> magic beans
[04:58:49] <jbk> did they ever give out a url for the slides?
[04:59:12] <jbk> i missed it if they did (but them i'm doing a few other things too)
[04:59:21] <Gman> hrm, benr didn't upload the slides
[05:00:30] <boyd> We need a live IRCer at the meeting
[05:00:37] <Gman> yeah
[05:00:48] <Gman> joyent, widely considered ruby on rails experts
[05:00:57] <Gman> everyone is currently a director at the company
[05:01:04] <boyd> That'
[05:01:13] <boyd> That's so they don't need to pay dental :)
[05:01:14] <Gman> involved in both nevada and opensolaris [community]
[05:01:29] <Gman> incubator for innovation - try to get smart people working with smart people
[05:01:44] <Gman> try to be open with secrets
[05:02:09] <Gman> providing 5 products - shared business hosting, joint connector [web 2.0], accelerator hosting [containers] and something else i can't remember
[05:02:15] *** yongsun__ has joined #opensolaris
[05:02:18] <jbk> haha
[05:02:37] <Gman> shared business hosting on dell machines [boos], ufs
[05:03:14] <e^ipi> aww, octave won't  build with studio
[05:03:15] <e^ipi> :(
[05:04:02] <Gman> lots of traditional rules of an isp are being broken because they want customers to run fun and cool stuff
[05:04:21] <Gman> joyent connector, web 2.0 shared calendaring, email and contacts
[05:04:22] <boyd> e^ipi: Sparc or x86?
[05:04:53] * boyd notes that e^ipi has a shorter nick every time he looks
[05:05:01] <Gman> implemented in ruby on rails, currently deployed on 5 containers [zones] on 4 systems 2 t1000's, 2 x4100's
[05:05:06] <e^ipi> x86
[05:05:17] <Gman> nothing is done in the global zone
[05:05:18] <dlg> Gman: hows the shared calendaring done?
[05:05:19] <e^ipi> it's not going to get shorter than this
[05:05:25] <jbk> soon it'll be -1
[05:05:29] <e^ipi> heh
[05:05:30] <Gman> dlg, dunno :)
[05:05:33] <boyd> You mean i
[05:05:42] <jbk> err 1
[05:05:45] <e^ipi> that's root(-1)
[05:06:09] *** Chile` has quit IRC
[05:06:10] <boyd> Yeah, but I figure the distinction:pixel ratio is better
[05:06:21] <e^ipi> and then i'll just be .
[05:06:24] <Gman> why not nfsv4 - no big community adoption right now, only one guy blogging away, no obvious advantages, nfsv3 is well understood
[05:06:44] <Gman> maybe i should move this onto another channel ;)
[05:06:45] <boyd> Interesting... I thought it benchmarked better
[05:06:51] <jbk> i thought so too
[05:06:58] <e^ipi> let's try cranking down these optimizations
[05:07:07] <jbk> of course i can understand the troubleshooting
[05:07:08] <Gman> #opensolaris-ossvug
[05:07:32] <jbk> i stumbled upon /var/run/nfs4_domain (or whatever it's called)
[05:08:27] <jbk> but, you'd think just the inclusion of compound operations would help a lot
[05:08:53] <boyd> and delegations
[05:09:00] *** yongsun_ has quit IRC
[05:11:08] <jbk> i wonder if we should ask them to repeat the audience questions
[05:11:36] <boyd> Do you have a way to ask that?
[05:11:50] <jbk> other than interrupting, no
[05:12:14] <boyd> Live semi-transcripting in #opensolaris-ossvug from Gman
[05:12:58] *** nbkk6fo_ has joined #opensolaris
[05:12:59] *** rodrick-brown has quit IRC
[05:16:58] <nbkk6fo_> http://www.youtube.com/watch?v=0qEotHQgUsg&eurl=
[05:29:17] *** alfism has quit IRC
[05:39:20] *** Odin-LAP has quit IRC
[05:43:51] <e^ipi> huh...
[05:44:03] <e^ipi> octave uses fortran, C, *AND* C++, all in the same program
[05:46:20] <jbk> ouch
[05:46:45] <steleman> it's a test harness for studio
[05:47:29] <e^ipi> test... harness...
[05:49:22] <e^ipi> on the plus side, it looks like i finally managed to get it built
[05:49:25] <e^ipi> minus readline
[05:49:40] <steleman> you dont have readline ?
[05:50:34] <e^ipi> i have readline
[05:50:45] <e^ipi> but octave's build scripts can't find it for some reason
[05:50:53] <steleman> oh
[05:52:12] <e^ipi> even though it somehow managed to find itself lib_perf
[05:58:29] *** deather__ has joined #opensolaris
[06:13:37] *** polk__ has quit IRC
[06:14:29] *** gm152_ has quit IRC
[06:14:45] <e^ipi> aww, it won't build w/ amd64
[06:15:54] <steleman> wut's it saying
[06:16:46] <e^ipi> wrong elf class
[06:16:54] <steleman> ohhh
[06:17:01] *** deather_ has quit IRC
[06:17:07] <e^ipi> at the moment i found a 32-bit compiled readline, so i'm working on that
[06:17:20] <steleman> you have some -L/<something>/lib -l<something> instead of -L/<something>/lib/amd64 -l<something>
[06:17:27] <e^ipi> i guess
[06:19:06] *** CapRiCoRN^80 has joined #opensolaris
[06:24:14] *** freakazoid0223 has quit IRC
[06:27:01] *** freakazoid0223 has joined #opensolaris
[06:32:39] *** tg has joined #opensolaris
[06:32:44] <tg> helo
[06:33:17] *** tg has quit IRC
[06:33:21] *** tg has joined #opensolaris
[06:34:24] *** polk__ has joined #opensolaris
[06:34:46] *** ddunham has joined #opensolaris
[06:50:52] *** dunc has quit IRC
[06:51:50] *** dunc has joined #opensolaris
[06:52:29] *** gustav3d has quit IRC
[07:00:32] *** scx has joined #opensolaris
[07:07:36] *** Bart_M has quit IRC
[07:08:38] *** Theoden-Nexenta has left #opensolaris
[07:09:06] *** karrotx has joined #opensolaris
[07:10:21] *** scx is now known as noone
[07:10:45] *** noone is now known as scx
[07:11:51] *** linma has joined #opensolaris
[07:12:00] *** scx has left #opensolaris
[07:12:53] *** nbkk6fo_ has quit IRC
[07:12:54] *** nbkk6fo__ has joined #opensolaris
[07:23:05] *** DebolazX has quit IRC
[07:27:47] *** Risky has quit IRC
[07:28:27] *** Risky has joined #opensolaris
[07:35:29] *** linma has quit IRC
[07:35:52] *** linma has joined #opensolaris
[07:41:09] *** rachel_ has joined #opensolaris
[07:44:40] *** Gman is now known as GmanAFK
[07:52:09] *** jesse-jads has quit IRC
[07:54:24] <CIA-22> tz204579: 4358464 praudit text tokens don't output standard control characters.; 6510139 *praudit* dumps core on stdin fuzz testing; 6512519 praudit displays arbitrary data token strings with four spaces between characters
[08:08:32] *** jteo has quit IRC
[08:08:35] *** jteo_ has joined #opensolaris
[08:11:48] *** CapRiCoRN^80 has quit IRC
[08:20:07] *** lloy0076 has joined #opensolaris
[08:20:28] <lloy0076> Stupid burn folder functionality in Gnome Nautilus caused my stupid OS to reboot.
[08:20:29] * lloy0076 sigh
[08:22:25] <g4lt-mordant> se, let schillyware near your computer and it'll slag it in no time ;P
[08:23:02] <e^ipi> speaking of joerg, has schillix been updated since it was created?
[08:24:59] <lloy0076> @#@$#@ I hate @!#$#@!# Windows
[08:25:27] <lloy0076> Reboot it. That'll fix it.
[08:25:28] <e^ipi> so don't use it
[08:25:45] <lloy0076> I don't.
[08:26:32] <g4lt-mordant> lloy0076, ues solaris ;P
[08:26:37] <lloy0076> g4lt-mordant: I do.
[08:26:48] <lloy0076> I'm setting up a Windows box for a friend.
[08:26:52] <lloy0076> And it's just stupid.
[08:27:01] <g4lt-mordant> then use sunpci ;P
[08:27:32] <lloy0076> I'd use a LART club on said friend, except said friend knows lots of unsavoury characters who could LART me even harder :P
[08:35:34] *** mazon is now known as Mazon
[08:37:43] *** tsoome has joined #opensolaris
[08:41:06] *** Risky has quit IRC
[08:43:40] <lloy0076> #@!$$#@$#@
[08:43:43] <lloy0076> I hate Dell.
[08:43:54] <lloy0076> I don't *want* a f@##@$$#@ Dell 520
[08:43:58] <tsoome> welcome to club;)
[08:43:59] <lloy0076> I want to find the info on a Dell 510
[08:44:17] <lloy0076> But do you think the wankers' website will lead me to info about it...
[08:44:36] <dlg> lloy0076: support.ap.dell.com
[08:44:47] <dlg> look for the manuals
[08:45:29] <lloy0076> Oh, I downloaded the only likely NON wireless drivers for network from there.
[08:45:46] <lloy0076> Needless to say LoseDows 2000 still thinks there's no network card installed.
[08:46:17] <lloy0076> (I should note that SXCR 55 doesn't see the network either :()
[08:46:56] <lloy0076> Why don't they have a: "This is the CD we package with the stupid damned notebook. Download it from this link.":
[08:47:07] <g4lt-mordant> that'd be convenient
[08:47:08] <lloy0076> Rather than splitting it up into six trillion bits?
[08:47:49] <lloy0076> Of course, Windows isn't telling me *what* it thinks the network card is, apart from the fact that it thinks there isn't any network devices there.
[08:47:56] <lloy0076> I have a Netowrk adapter but no network controller.
[08:50:45] <Griff> hi folks. I'm trying to build software under solaris 8 for 8-10 (blastwave). It works under 8, but under 10 ldd is giving me this: libz.so.1 (SUNW_1.1) =>  (version not found). What does this mean exactly?
[08:51:15] <lloy0076> I could be uncharitable and point out that it means that ldd cannot find libsz.so.1 :P
[08:51:21] <lloy0076> I think you're asking WHY can't it find it?
[08:52:03] <Griff> well I know the can't find it errors... this is slightly different. Is it finding it, but it's too old? The whole solaris abi compatability is meant to cover this kind of thing isn't it
[08:52:28] <Griff> and given that I'm going from 8 to 10, and specifically not the other way around.... I'm know quite understanding what is going on
[08:52:37] <lloy0076> I don't think blastwave and Sun with a capitabl have the same philosophy.
[08:53:25] <Griff> well I know that all packages are meant to work from solaris 8-10
[08:53:32] *** Griff is now known as Gr|ffous
[08:53:43] <lloy0076> Gr|ffous: Meant to.
[08:53:50] <Gr|ffous> :)
[08:53:57] <lloy0076> Gr|ffous: Currently blastwave will randomly crap out my Gnome desktop.
[08:54:07] <Gr|ffous> well is there some way that I can troubleshoot this better?
[08:54:20] <Gr|ffous> are you using 2.16.2?
[08:54:34] <lloy0076> Gr|ffous: You could try to find a "newer" version of what it's complaining about...
[08:54:43] <lloy0076> Gr|ffous: And then report what you find.
[08:55:15] *** broadcast has quit IRC
[08:55:15] *** Tpenta has quit IRC
[08:55:27] *** ofu has quit IRC
[08:55:29] <Gr|ffous> how do I check the version of the library?
[08:55:34] <lloy0076> [I've always found: But this old version doesn't work with some random new version bugs easier to debug if the reporter has found what version it *does* work with]
[08:56:09] *** Tpenta has joined #opensolaris
[08:56:16] <lloy0076> Gr|ffous: Not sure.
[08:56:35] <Gr|ffous> bash-3.2$ grep libz.so.1 /var/sadm/install/contents /opt/csw/lib/libz.so.1=libz.so.1.2.3 s none CSWzlib
[08:56:48] *** ddunham has left #opensolaris
[08:56:49] <lloy0076> Gr|ffous: Btw, if I sound grumpy it's because I'm fighting with a Win2K laptop without the original install disks and do you think I can get the network drivers to work? At least you've got an error.
[08:56:50] * lloy0076 sigh
[08:57:00] <Gr|ffous> I also have /usr/lib/abi/abi_libz.so.1 f none 0755 root bin 53604 55062 947116767 SUNWzlib
[08:57:25] <Gr|ffous> heh, that's fine... if you don't like bw, help me, and I'll continue the crusade to make it better ;)
[08:58:02] <lloy0076> Ok, Fedora thinks this thing does have a Broadcom BCM4401
[08:58:05] <lloy0076> I've got the drivers.
[08:58:20] <lloy0076> But STUPID WIN2K installs the network drivers but not the network components.
[08:58:24] <Tpenta> muryama-san's drivers for that one
[08:58:29] <lloy0076> *ponder*
[08:58:34] <Tpenta> for solaris anyway
[08:58:36] <lloy0076> Ah...maybe it needs the Win2K disk.
[08:58:53] <e^ipi> oh, Tpenta
[08:58:55] <e^ipi> there you are
[08:59:11] <lloy0076> Makes sense. Windows didn't detect the NIC so it didn't install the network stack.
[08:59:14] * lloy0076 flop
[08:59:28] <e^ipi> so i grepped through the libc source looking for defined but not implemented i18n stuff...
[08:59:38] <e^ipi> there's only a handful of functions... like, 150 or so
[08:59:43] <lloy0076> i18n is overrated. Everyone speaks English.
[09:00:16] <e^ipi> lloy0076: my point is that i don't understand why sun wouldn't just re-implement those functions, seeing as how they're required for sun-libc & not opensource
[09:00:37] * lloy0076 sigh
[09:00:52] <lloy0076> Apparently, even though I have the right NIC drivers installed, there's no network controllers available.
[09:06:41] *** cmihai has joined #OpenSolaris
[09:06:46] <lloy0076> Ah, of course.
[09:06:55] <lloy0076> That other RANDOM device is the wireless network device.
[09:06:58] <lloy0076> !@##$
[09:07:03] <lloy0076> Why didn't I think of that?
[09:07:03] <cmihai> I figured out why my Solaris killed ssh but responded to ping after LU
[09:07:14] <cmihai> SunOS cmsol 5.11 snv_55b i86pc i386 i86pc :)
[09:07:32] <g4lt-mordant> yep, that would be where the problem is
[09:07:41] <cmihai> Guess what, when I came to work this morning it sported a neat screen saying: "Language"
[09:07:53] <e^ipi> that'll teach you for swearing
[09:07:54] <cmihai> Asked me to select English, F2, then everything worked. Weird.
[09:08:10] *** qdk has quit IRC
[09:10:40] <cmihai> So, is that language thing a new "feature"
[09:11:05] *** Starless_ has quit IRC
[09:13:26] <g4lt-mordant> cmihai, not really, it happens every time you do a fresh install of solaris in interactive mode
[09:13:58] <cmihai> Well, I don't recall that doing live upgrades before. Seen it with containers, seen it with fresh installs... Oh well.
[09:14:14] <cmihai> g4lt-mordant: but is it normal for ssh to be closed until I answer that?
[09:14:35] <g4lt-mordant> no, THAT's new.  interactive installs usually start sshd
[09:14:36] <lloy0076> omgiwkmn
[09:14:51] * lloy0076 kicks windows
[09:16:00] *** razrX has joined #opensolaris
[09:16:55] *** LordKing has joined #opensolaris
[09:17:12] *** Cyrille has joined #opensolaris
[09:17:21] <razrX> mornin'
[09:17:28] <cmihai> morning
[09:17:55] <razrX> hey cmihai, did the lu go okay?
[09:18:42] *** bengtf has quit IRC
[09:18:51] <cmihai> razrX: it locked me out ;)
[09:19:06] <cmihai> razrX: funny thing was, when it rebooted (all went fine) it asked for the Language :)
[09:19:11] <cmihai> Only started ssh _after_ that.
[09:20:39] <razrX> do you know which steps to take to get that sysidtool to fire up again?
[09:21:08] <razrX> i saw that language thing too but had some keybaord problems and ended up with a wrong language selection
[09:21:29] <razrX> now when i log onto console my keyboard is kinda fucked up
[09:21:41] <razrX> i noticed there are 2 smf services
[09:21:52] <tsoome> x86?
[09:21:56] <razrX> sysidtool:net and :system
[09:22:00] <razrX> tsoome: yup
[09:22:17] <tsoome> sys-unconfig ;)
[09:22:24] <razrX> no, don't wanna do that
[09:22:28] <tsoome> or for keyboard just use the eeprom command
[09:22:47] <razrX> aparently, something from sysidtool checks for a /etc/.UNCONFIGURED file
[09:23:12] <razrX> if it exists, you need to fill in everything again (seems the same as sys-unconfigging a box)
[09:23:41] <razrX> but my config should be known so i just like to know how to fire up that tool again to see what it exactly wants from me
[09:24:18] <razrX> tsoome: can't remember what my keyboard setting was
[09:24:24] <razrX> i know i can set it thru eeprom
[09:24:34] <razrX> currently it's set to Latin-American
[09:24:39] <razrX> that's not correct iirc
[09:25:08] <razrX> i could always boot to my previous BE (snv_54) and see what was set there can i not ?
[09:25:21] <razrX> i don't want to guess
[09:25:32] <razrX> just want a plain us-ascii keyboard settings
[09:27:01] <tsoome> you dont need to boot, just mount the / and look into boot/solaris/bootenv.rc
[09:27:21] <cmihai> Hm... I wonder what's new in svn_55b
[09:27:28] <cmihai> iso seems to be a LOT bigger
[09:27:31] <razrX> lol
[09:27:39] <razrX> my conclusion exactly
[09:27:44] <razrX> almost 1gig bigger
[09:28:00] <cmihai> Well, a devel guide pops up when I start firefox
[09:28:01] <razrX> netbeans and studio 11 ?
[09:28:04] <cmihai> Says something about netbeans and studio
[09:28:08] <cmihai> But it's not there
[09:28:23] <cmihai> razrX: yup, but they didn't get installed (via LU anyway)
[09:28:53] <razrX> if you actually used the 55b installer instead of luupgrading then you'd have a developer setting in grub i think which enables studio and netbeans
[09:28:56] <razrX> i think
[09:29:08] <razrX> since we followed the same path, we don't have it
[09:29:25] <cmihai> A "developer setting that enables them".. that sounds stupid
[09:29:28] <cmihai> Oh well.
[09:29:52] <razrX> tsoome: keyboard-layout 'Latin-American' and kbd-type is set to 'US-English'
[09:29:58] <cmihai>  789M  on DeveloperTools on the DVD
[09:30:05] <razrX> i only need to change the layout but to which setting?
[09:30:08] *** ofu has joined #opensolaris
[09:30:11] <ofu> good morning
[09:30:16] <razrX> mornin
[09:30:21] <cmihai> I guess it's nice that they're on the disk.
[09:30:39] * razrX agrees
[09:30:41] *** Vratha has joined #opensolaris
[09:31:30] <tsoome> well, for a white man with decent desktop, there are some apps in menu to set up keyboard layout etc, but I have no idea if they are working in gnome....
[09:31:31] <cmihai> Not so nice for people on 128kbps hehe
[09:31:34] *** jesse-jads has joined #opensolaris
[09:39:59] *** lplatypus has joined #opensolaris
[09:41:56] *** qdk has joined #opensolaris
[09:43:44] *** bondolo has quit IRC
[09:46:58] *** bengtf has joined #opensolaris
[09:56:05] *** pogma has quit IRC
[09:56:22] *** pogma has joined #opensolaris
[09:56:46] <Gr|ffous> I wonder what the going record is for the most attempts to compile a program?
[09:56:50] <Gr|ffous> I'm up there...
[09:59:27] <dlg> most ive seen is 900ish to get a working kernel on a box
[10:00:06] <Gr|ffous> linux I assume?
[10:00:13] <dlg> no
[10:00:25] <dlg> linux isnt work that much effort
[10:00:36] <dlg> some guy porting to a new arch
[10:00:40] <dlg> openbsd
[10:00:41] <Gr|ffous> polaris maybe?
[10:00:44] <Gr|ffous> oh..
[10:00:51] <Gr|ffous> man that's an epic effort
[10:01:45] <Gr|ffous> ok, so I don't win then ;)
[10:01:51] <dlg> no...
[10:05:11] *** uebayasi has quit IRC
[10:05:47] <dlg> http://miod.online.fr/solbourne/
[10:05:49] <dlg> 600ish sorry
[10:06:48] <Gr|ffous> thanks dlg :)
[10:10:04] *** hspaans has quit IRC
[10:10:06] *** alanc_away has quit IRC
[10:10:06] *** yippi has quit IRC
[10:10:07] *** alanc-away has quit IRC
[10:10:37] *** alanc_away has joined #opensolaris
[10:10:44] *** alanc-away has joined #opensolaris
[10:11:42] *** robelkin has joined #opensolaris
[10:12:01] *** GmanAFK has quit IRC
[10:13:56] <raph_ael> hello
[10:15:24] *** bunker_work has joined #opensolaris
[10:22:02] *** pogma has quit IRC
[10:24:01] *** Starless_ has joined #opensolaris
[10:26:40] *** robelkin has quit IRC
[10:29:15] *** broadcast has joined #opensolaris
[10:29:35] <jteo_> hello *
[10:29:43] *** robelkin has joined #opensolaris
[10:30:26] <e^ipi> ahoy
[10:32:57] <lasseoe> hm
[10:33:07] <lasseoe> seems the System Handbook on Sunsolve is broken
[10:33:08] <lasseoe> Forbidden
[10:33:09] <lasseoe> You don't have permission to access /handbook_private/ on this server.
[10:33:34] <lasseoe> oh now it works, weird
[10:36:01] <lloy0076> lasseoe: It's a CONSPIRACY
[10:36:20] <jteo_> i prefer the term setup.
[10:36:49] *** Risky_ has joined #opensolaris
[10:38:22] *** robelkin has quit IRC
[10:38:37] <Cyrille> that's because sunsolve personally hates him.
[10:38:47] *** simford has quit IRC
[10:44:30] * lloy0076 hmmm
[10:44:42] <lloy0076> I hope they haven't sent someone to his place to silence him in person
[10:54:51] *** Starless_ has left #opensolaris
[11:04:45] *** yongsun__ has quit IRC
[11:06:34] *** gallium has joined #opensolaris
[11:08:17] *** LordKing has quit IRC
[11:16:25] <Berny_> morning
[11:16:53] <Berny_> does anyone know what the average power consumption of an 3510 with 12 disks is?
[11:17:31] * dlg laugh
[11:17:38] <dlg> sunsolve should tell you
[11:17:45] <Berny_> doesn't
[11:17:46] *** lloy0076 has quit IRC
[11:18:01] <Berny_> all it says it's got two 420W power supplies
[11:18:29] <quasi> so <420W
[11:18:38] <Berny_> a little more precise? :-)
[11:18:47] <Berny_> 0 is also less than 420 :-P
[11:19:37] <quasi> probably depends on how busy your disks are
[11:20:02] <Berny_> happily spinning most of the time...
[11:21:19] <Berny_> well if i assume max consumption to be about 400, avg should be somewhere around 300?
[11:22:07] <dlg> whens sun going to build a hsm around zfs?
[11:22:21] * dlg want a hsm that can power unused disks off
[11:22:33] <Berny_> hmm
[11:22:50] <Berny_> set the unused disks to unavail ;-)
[11:23:28] <dojtoll> Done anyone know if the E3000 was shiped with FA-CL as internal disks?
[11:26:13] <quasi> Berny_: that sounds like a reasonable assumption
[11:26:19] *** benignbala has joined #opensolaris
[11:26:38] <benignbala> hello, can anyone tell how to dtrace c programs?
[11:27:49] <Berny_> .oO(ok now the next thing what's the conversion factor from VA to watt again?)
[11:29:07] *** Kmays has joined #opensolaris
[11:30:04] <cmihai> Berny_: it's not exactly the same thing, but it's equal to watts
[11:30:25] *** Risky__ has joined #opensolaris
[11:30:42] <cmihai> http://www.powervar.com/Eng/ABCs/CalcVAWATTS.asp
[11:31:49] <andersmo> benignbala: http://docs.sun.com/app/docs/doc/817-6223/6mlkidlms#hic
[11:31:51] <Berny_> cheers
[11:32:08] <benignbala> andersmo: thanks :)
[11:32:21] <benignbala> andersmo: i got it.
[11:34:05] *** CosmicDJ has joined #OpenSolaris
[11:34:26] *** Tpenta has quit IRC
[11:36:13] <Berny_> damn that's gonna cost some real money... :-\
[11:36:39] *** Risky_ has quit IRC
[11:37:12] <cap_> Berny_, there is no conversion factor between VA and W
[11:37:31] <cmihai> Yeah, too bad we can't pay with make-belief moneyt
[11:37:34] *** CosmicDJ has quit IRC
[11:37:50] * Gr|ffous rubs his aching eyes
[11:39:19] <Berny_> cap_: so if i have a total power consumption of 8500W what size of ups (in VA) would i choose?
[11:40:20] <cmihai> I'd go with 10kVA
[11:40:36] <cmihai> That's gonna cost a lot of money hehe
[11:41:19] <Berny_> yeah
[11:41:54] <Berny_> apcc's online thingie suggests a 12kVA which comes to 12kEuro (with N+1 redundancy)
[11:42:11] <cmihai> w00t
[11:42:18] *** CosmicDJ has joined #OpenSolaris
[11:42:20] <cmihai> Hm.. yeah, I guess it's best you play it safe with 12kVA
[11:42:45] *** DebolazX has joined #opensolaris
[11:42:50] <Berny_> might just be they don't have a 10kVA model
[11:43:35] <cmihai> That too.
[11:43:49] <cap_> Berny_, that's probably a good approximation, but it's important to understand that it will depend on the power factor of the machines connected to the UPS
[11:44:43] <cap_> Berny_, most modern machines today have good power factors (say, ~0.9)
[11:44:56] <ofu> but i thought, active pfc is standard nowadays
[11:45:03] <Berny_> cap_: that would be a v880z, an ultra 5, 2 x2200M2, one W2100z, 3 3510, a L100 and a L25 :-)
[11:45:16] <cmihai> Heh, fire up MathCAD dude :P
[11:45:29] <cmihai> That's a pretty complex load ;]
[11:45:30] <cap_> Berny_, If you have them running today without ups you can measure the power factor
[11:45:33] <dlg> Berny_: give me some
[11:45:46] <cap_> ofu, quite so, yes, but there are exceptions..
[11:46:16] <Berny_> cap_: there now running on 2 smaller ups, but i want/have to get some redundancy, as i'm f*cked if one of the ups' fails
[11:46:21] <ofu> in germany, active pfc is required since 2005 or so
[11:46:39] <Berny_> dlg: sorry dude, all in heavy use ;-)
[11:47:14] <cap_> Berny_, then buy the new ups as VA(oldups1 + oldups2) and you're safe =)
[11:47:18] <dlg> booerns
[11:48:09] <cap_> ofu, oh, either way it doesn't give you power factor 1.0
[11:49:58] *** calumb has joined #opensolaris
[11:50:00] <Berny_> cap_: there's just one problem: the x2200M2s are not yet on the ups ;-)
[11:50:03] <ofu> but 0.90-0.95 is a realistic value
[11:50:32] <Berny_> and the old ups run with 99% load 8-)
[11:50:33] <quasi> Berny_: hopefully you've got nothing important wired to them? ;)
[11:50:38] <dlg> Berny_: theyre not real computers so it doesnt matter
[11:50:41] <Berny_> so a little extra would help
[11:50:50] <Berny_> quasi: no there still being installed/tested
[11:53:16] <Berny_> right... let's call some folks who sell that stuff...
[11:53:33] <cmihai> APC ftw
[11:54:09] <Berny_> apc, powerware, mge
[11:54:21] <Berny_> though i must admit i don't like apc that much
[11:54:48] <Berny_> i used to have apc's smart ups for some time and had to rplace them like every 6 months :-\
[11:57:11] <Berny_> hmm, damn lunchtime over here... pretty little chance to get someone on the phone
[11:57:52] * cap_ has a riello 220KVA UPS, it's nice, especially the part that when it fails it never fails to auto-by-pass itself
[11:58:58] *** benignbala has quit IRC
[11:58:59] <Berny_> 220kva? what the heck are you running?
[11:59:24] <PerterB> a bunch of E450s maybe :)
[12:00:20] <quasi> a few racks of dell would eat that easily ;)
[12:02:30] <dlg> altix
[12:02:33] <dlg> they love power
[12:03:13] <cap_> Berny_, try www. + the domain from my whois...
[12:04:37] *** nostoi has joined #opensolaris
[12:04:44] <Berny_> oh well...
[12:04:59] * Berny_ is just running small biz then ;-)
[12:06:20] <cmihai> 220kVA?!
[12:06:21] <cmihai> Dude, WTF
[12:06:34] <cmihai> Oh, wait, that's not big ;)
[12:07:01] <cmihai> We got something... we have something of a few MVA's :)
[12:07:11] <Berny_> pfft
[12:07:30] <cmihai> Huge room full of batteries keeps the whole building running for like an hour :)
[12:07:47] <cmihai> Something around 2000 PCs + servers and lights
[12:08:38] * Berny_ doesn't need light as long as the crt got power ;-)
[12:08:46] <cmihai> Made that mistake before.
[12:08:53] <cmihai> Put everything in the UPS sockets
[12:08:57] <cmihai> Apart from the monitors :)
[12:09:27] <Berny_> :>
[12:09:47] <cmihai> Had some of those serial CRT consoles.. vt420, vt220 and the likes.
[12:09:54] <cmihai> No power though (DUH)
[12:10:17] <Berny_> yeah i got a nice amber vt320 at home... a real dec thingie :-)
[12:10:28] <cmihai> Yup ;]
[12:10:40] <Berny_> should hook up that bugger again :-)
[12:10:51] *** nostoi has quit IRC
[12:24:12] *** lplatypus has quit IRC
[12:24:47] *** Kmays has quit IRC
[12:24:55] <PerterB> oh bugger...
[12:25:11] <PerterB> looks like the machine I liveupgraded remotely is stuck in single user mode
[12:25:28] <cmihai> PerterB: no it's not :P
[12:25:39] <cmihai> 10$ says it's asking you for "Language" and then starting sshd :D
[12:25:48] <cmihai> PerterB: just happened to the ones I was LU :)
[12:26:13] <cmihai> It's just _not fun_
[12:26:21] <PerterB> language?
[12:26:47] <cmihai> Yeah, upgraded 54 to 55b via ssh / LU... and when i did init 6 I could ping but not ssh
[12:27:01] <cmihai> So when I came to work this morning... the thing was asking me for "Select language and press F2"
[12:27:08] <cmihai> Did that, wam, ssh was up
[12:27:43] <tsoome> sounds like the machine dropped into single user prompt....
[12:28:46] <cmihai> tsoome: still looking for solutions to this thing.
[12:28:58] <cmihai> I have plenty more machines to LU remotly... don't want this happening to all of them
[12:29:40] <Berny_> hmm riello was the first to answer
[12:30:15] <tsoome> I had a wierd experience - did a lucreate with split mirrors, new BE spec was to use cancat devices from mirror - and it dropped to single user mode with complaint that metadevwas were missing
[12:30:36] <tsoome> but they did actually exist
[12:30:54] <PerterB> this was solaris 9 to 10 so I suspect it isn't that, isn't that press F2 thing specific to one of the x86 X config scripts?
[12:31:26] <cmihai> PerterB: yeah, it was similar to kdmconfig thing, but... this only happened with svn_55b here... so if it's 9->10.. could be something else
[12:31:33] <tsoome> so I fixed vfstab to use plain  slices, booted, did set up a mirror on top of existing md's and after reboot it was just ok
[12:33:57] <PerterB> ah well... it's my own machine but colocated, and I have a friend coming in from out of town this evening so I won't be able to go down their and sort it till next week...
[12:34:40] <cmihai> That sucks...
[12:35:23] <cmihai> Doing upgrades of any kind without a serial console (BIOS console/iLO/ILOM even) is just a scary thought.
[12:35:38] <tsoome> true
[12:36:03] <cmihai> Especially when you find yourself muttering the words: Well, this never happened before...
[12:36:31] <quasi> sounds like me trying remote bios upgrades of an x2200 ;)
[12:37:02] <Berny_> hehe
[12:37:32] *** swa__ has joined #opensolaris
[12:37:38] <quasi> all it took was for someone to press the power button afterwards, but still annoying as hell
[12:37:41] <cmihai> quasi: heard a really funny story of someone typing "halt" instead of "reboot" :)
[12:37:42] <PerterB> cmihai: I know, I was just being over ambitious :)
[12:37:44] <Berny_> cap_: does the riello software support x86 solaris?
[12:37:57] <cmihai> So the server went: Cool, press any key to reboot :))
[12:38:04] *** nostoi has joined #opensolaris
[12:38:36] <quasi> oh, you mean init 5 instead of 6? neeeeever ;)
[12:38:56] <PerterB> I've had not-so-funny stories of typing init 5 instead of init 6 and having to make a trip to the data centre
[12:39:35] <PerterB> if only I could type, I'd be so much better at this job ;)
[12:39:50] *** swa_wtf has quit IRC
[12:39:57] <ofu> start /SP *g*
[12:40:04] <ofu> start /SYS i mean
[12:40:16] <quasi> another classic (that's not happened to me, yet) is last| reboot
[12:40:41] <ofu> i once pasted last-output onto the command line
[12:40:50] <ofu> there were lines beginning with "reboot"
[12:42:29] * twincest has seen "ssh otherhost; reboot" and ctrl-alt-del while in ssh
[12:42:31] <cmihai> quasi: heheheh
[12:42:56] <cmihai> That's not fun heh
[12:43:44] <cmihai> last | reboot is safe as regular user :)
[12:44:28] <quasi> cmihai: no, 2 of my former co workers have done that (and a couple of others at an even further back job discovered that killall does what the command says ;)
[12:44:29] <cmihai> Ah great, my Solaris started writing in German lol
[12:44:39] <cmihai> quasi: lol, good old Linux admins :P
[12:44:46] <PerterB> wunderbar!
[12:45:19] <quasi> cmihai: indeed
[12:45:21] <cmihai> "Why does kill all.. kill... all?!" hahaha :)
[12:46:41] <cmihai> Too many horror stories with UNIX administration. Stuff like rm -rf ?as8#d dir and such.
[12:50:37] *** klocze2 has joined #opensolaris
[12:51:14] <razrX> cmihai: eeprom keyboard-layout=US-English ??
[12:55:45] *** wamty has joined #opensolaris
[12:56:36] *** uebayasi has joined #opensolaris
[13:00:55] *** wamty has left #opensolaris
[13:02:24] *** tsoome_ has joined #opensolaris
[13:03:16] *** tsoome_ has left #opensolaris
[13:03:37] *** dunc has quit IRC
[13:05:51] <twincest> other than support, what reasons are there not to change root's shell?
[13:06:20] <Cyrille> because it's an indication you're using the root account too much ;-)
[13:06:21] *** svoboda has quit IRC
[13:08:59] *** tsoome has quit IRC
[13:11:11] *** tsoome has joined #opensolaris
[13:14:23] *** calumb has quit IRC
[13:14:59] *** tsoome has quit IRC
[13:15:12] *** tsoome has joined #opensolaris
[13:15:50] <quasi> twincest: on older versions, you could end up with your root shell being outside the dirs being mounted in single user or linking to files there
[13:16:26] <quasi> twincest: and choosing a shell with tab completion is just asking for trouble...
[13:16:31] <mv> twincest: sh is staticaly linked, others are not
[13:16:37] <twincest> mv: no it's not
[13:16:49] *** calumb has joined #opensolaris
[13:17:07] <quasi> it isn't an issue on s10, but was on older versions
[13:17:27] <mv> well, you're right
[13:17:36] <mv> but it was some time ago
[13:18:25] *** MikeTLive has joined #opensolaris
[13:20:53] *** dunc has joined #opensolaris
[13:21:02] *** svoboda has joined #opensolaris
[13:22:57] <sickness> good lunch all
[13:26:27] *** pogma has joined #opensolaris
[13:28:42] *** gm152 has joined #opensolaris
[13:31:27] <twincest> how odd, i can label a disk with currently mounted partitions, but not one with unmounted partitions in vfstab
[13:31:56] <lasseoe> http://unixzone.dk/homeoffice.jpg
[13:32:08] <lasseoe> oops
[13:32:23] <cmihai> Ooops?
[13:32:34] <quasi> lasseoe: damn, that's looking tidy ;)
[13:32:53] <cmihai> Damn, nice place
[13:32:56] <cmihai> It's so.. clean.
[13:33:05] <lasseoe> I like it tiy, less to distract me :)
[13:33:07] * Stric gets a green alien in pink dress
[13:33:08] <lasseoe> tidy
[13:33:47] <cmihai> lasseoe: I don't know, I like it more when I have to be on the lookout for stuff crashing down on me
[13:33:50] <cmihai> It
[13:34:07] <cmihai> It's more exciting when you know there's stuff around that can kill you :)
[13:34:25] <cmihai> Nothing like getting crushed by ancient VAXen :)
[13:34:32] <lasseoe> cmihai: do you have a "Hazardous material" sign on your door? :)
[13:34:35] <lasseoe> haha
[13:37:52] <sickness> error 404?
[13:41:42] <Vratha> oh, nice.  apple release a 5-license "family pack" edition of OS X for only 199.00 compared to their single seat 129.99 price
[13:41:45] <Vratha> er, 129.00
[13:41:56] <cmihai> Vratha: that's nice
[13:42:02] <Vratha> yeah!
[13:42:54] <cmihai> Beats paying 400Euro for Vista Ultimate lol
[13:42:59] <Vratha> haha, yeah
[13:43:16] <Vratha> though i learned on dl.tv that newegg is selling OEM copies of vista right now for 50% off the retail price
[13:43:49] <hali> Beats paying 9000Euro for AIX5L lol
[13:43:56] <Vratha> man, their 10-client server is only 499.00 as well
[13:44:12] <Vratha> and the unlimited-client edition is 999.00
[13:44:18] <Vratha> that sound cheap if you ask me
[13:45:06] <cmihai> Haven't seen many MacOS servers yet ;)
[13:45:29] <Vratha> me either
[13:45:58] <Vratha> but i've been totally digging my OS X laptop, and they're pulling in great features from all the OS out there, including ZFS and DTrace
[13:46:09] <cmihai> Yeah, guess so.
[13:46:18] <cmihai> Best of all worlds I guess.
[13:46:47] <cmihai> You can run some MS stuff (Office), you can run some BSD stuff (pkgsrc && the likes), you can run MacOS stuff, you can run commercial apps.. makes a rather neat desktop.
[13:46:47] <Vratha> yes
[13:46:54] <Vratha> yeah :)
[13:46:57] <cmihai> Really looking forward to ZFS and DTrace.
[13:47:01] <Vratha> and it's powerful as a supercomputing platform too
[13:47:14] <cmihai> Though I wouldn't bother with it as a server myself.
[13:47:31] <cmihai> Neither do the big companies it seems (Eg: Oracle)
[13:47:54] <Vratha> yeah, the big companies go with linux it appears now
[13:48:01] <cmihai> Sadly, yes.
[13:48:30] <cmihai> Seen something.. interesting I'd like to checkout.
[13:48:31] <Vratha> i'd use some mac servers they were better supported; they're sexy
[13:48:42] <Vratha> s/they/if they/
[13:48:52] <Vratha> i really can't type tonight for some reason, and i don't know why
[13:49:05] <cmihai> A Windows Server without Explorer, Iexplore and the likes.. running Xen and command prompt / serial console.
[13:49:10] <Vanuatoo> Vratha: full moon maybe :)
[13:49:19] <Vratha> hehe
[13:50:29] <cmihai> Vratha: rip that out of your BOFH excuse calendar?
[13:50:44] <cmihai> "Fool moon - gravity affecting keyboards"
[13:51:45] <Vratha> :)
[13:51:50] <Cyrille> "fool" moon? Is that when people turn into idiots?
[13:52:01] <jmcp> Cyrille: probably
[13:52:04] <jmcp> hence "lunatic"
[13:52:29] <Vratha> oh hell.. OS X even fully supports windows joining a domain created on the mac, alnog with all the win 2k3 ACLs
[13:52:34] <Vratha> that is just sexy right there
[13:53:07] <tsoome> domain on mac? thats the regular samba function....
[13:53:21] <tsoome> every unix with samba can do
[13:53:26] <Vratha> no, this is different
[13:53:27] <Vratha> http://www.apple.com/server/macosx/windowsservices.html
[13:53:33] <Vratha> they've enhanced samba afaik
[13:53:56] <Vratha> they also support all the win 2k3 ACLs for windows clients
[13:54:01] <cmihai> Open Source Reliability
[13:54:02] <cmihai> Samba 3 delivers high-performance, reliable file and print services to Windows clients using the native
[13:54:05] <cmihai> bla bla bla, it's Samba.
[13:54:10] <tsoome> exactly
[13:54:27] <cmihai> And OpenLDAP etc.
[13:54:45] *** dunc has quit IRC
[13:54:46] <Vratha> i'm sure that it's a better samba.. they'd need kerberos working too
[13:54:46] <cmihai> Just gave it a pretty GUI.
[13:54:47] <tsoome> my solaris is joined to AD and I have no users in passwd (except system users)
[13:55:00] <Vratha> otherwise you have to send cleartext passwords to the samba machine
[13:55:15] <cmihai> It's integrated with MIT Kerberos.
[13:55:24] <Vratha> samba?  not that i could see
[13:55:27] <cmihai> Nothing you can't get on any other UNIX machine.
[13:55:40] <tsoome> Vratha: have you ever read samba source ./configure --help output?
[13:55:42] <Vratha> yeah, but you can get full compatibility with win 2k3 ACLs on OS X server :)
[13:55:55] <Vratha> tsoome: i generally install a debian package
[13:56:28] <tsoome> kerberos is needed for AD support  - and samba supports heimdal and mit
[13:56:56] <Vratha> i know it's needed for AD support; i didn't find anything in the docs talking about how to get samba working with ldap and kerberos simultaneously
[13:57:03] <tsoome> ful compatibility with samba means it's compatible till the next windows service pack
[13:57:06] <Vratha> could've missed them though
[13:57:28] <tsoome> unless you upgrade your samba
[13:57:31] <Vratha> tsoome: no, apple was talking about this being something you can't get elsewhere
[13:57:55] <cmihai> Every vendor thinks they're special.
[13:58:08] <Vratha> New in Mac OS X Server v10.4 is support for the native access control list (ACL) file permissions of Windows Server 2003 and Windows XP clients ? a capability unprecedented on any UNIX- or Linux-based platform.
[13:58:29] <Vratha> i've seen ACLs work in samba, but they were always kinda screwy
[13:58:32] *** Kernel86 has quit IRC
[13:59:15] <cmihai> Still think MacOS makes a lousy server.
[13:59:30] <Vratha> why?  looks fine to me
[13:59:30] <tsoome> maybe maybe not
[13:59:53] <tsoome> if linux can be used as server, why not macos
[14:00:04] <cmihai> Costs I guess.
[14:00:11] <Vratha> yeah, and OS X is getting ZFS and dtrace!
[14:00:23] <Vratha> and it'll work... unlike linux where it's in alpha i think
[14:00:24] <cmihai> So? Solaris is free.
[14:01:01] <jteo_> eventually, every IRC conversation degenerates to comparisons.
[14:01:02] <Vratha> no company is going to use an OS with no support
[14:01:05] *** jteo_ is now known as jteo
[14:01:09] <Vratha> so it's not REALLY free
[14:01:24] <tsoome> Vratha: you don't wwanna know;)
[14:01:31] <Vratha> yeah, i don't wanna know
[14:01:38] <Vratha> our uni. uses it with no support, but that's different
[14:01:39] <twincest> i've never used macos, but it seems like it would be a reasonable choice for smaller environments where you perhaps don't have many experts on hand
[14:01:54] <tsoome> I have to fight with morons using debian....
[14:02:10] <Vratha> tsoome: moron users or moron admins?
[14:02:21] <cmihai> Moron users who think they're admins :D
[14:02:24] <tsoome> admins+managers
[14:02:25] <Vratha> hehe
[14:02:27] <Cyrille> he said "morons using", not "morons admining" ;-)
[14:02:31] <Vratha> i use debian, bitch!  >:)
[14:02:42] <Vratha> but that's now only because opensolaris doesn't run on my current hardware
[14:03:07] <Cyrille> that's a handy excuse...
[14:03:08] <jmcp> Vratha: what are the gaps in hw support?
[14:03:20] <tsoome> debian to serve some 100k users on top of some SAN storage?
[14:03:27] <Vratha> jmcp: it doesn't support my promise IDE raid controller
[14:03:34] <jmcp> ah
[14:03:42] <cmihai> Your fakeraid controller?
[14:03:43] <Vratha> it's my cheap ass home server it doesn't support
[14:03:44] <cmihai> :]
[14:03:47] <Vratha> cmihai: pretty much
[14:03:53] <Vratha> i won't lie; i know it's cheap
[14:04:02] <Vratha> then again it's for my home
[14:04:02] <cmihai> Throw that junk away.
[14:04:09] <Vratha> when i have a job with cash, i will
[14:04:11] <cmihai> Just use ZFS for RAID.
[14:04:13] *** gm152 has quit IRC
[14:04:29] <Vratha> cmihai: well, the thing is that that card also gives me ATA-133 support
[14:04:32] *** Kernel86 has joined #OpenSolaris
[14:04:36] <Vratha> i don't really want to degrade to ATA-33 just to run opensolaris
[14:04:42] <cmihai> Ugh...
[14:04:49] <cmihai> This is on your 486, right?
[14:04:58] <Vratha> 486 doesn't have ata-133
[14:05:09] <sickness> pio mode 4 anyone? ;P
[14:05:09] <Vratha> it's a tbird 1.34 GHz
[14:05:42] <Vratha> cmihai: but if you want to spend money on my system, i'm willing to accept donations :)
[14:05:59] <Vratha> however, my new mini-itx server coming in works with opensolaris... except for the freakin' vt6103 NIC
[14:06:04] <sickness> maybe if you could share bandwidth... ;P
[14:06:08] <Vratha> unless that somehow magically works, which i doubt it will
[14:06:11] <razrX> the moment ipfilter is capable of filtering IPv6 traffic will prolly be the moment i'd replace my freebsd 6.1 gateway install with a opensolaris install
[14:06:12] <cmihai> It's always the same thing. People having problems usually want to run an Enterprise grade Operating System on 5-year old sub-optimal desktop hardware.
[14:06:23] <sickness> Vratha: doesn't it simply need the murayama's driver/
[14:06:24] <sickness> ?
[14:06:33] <Vratha> cmihai: yeah, and that problem with that is?  not all of us have money to spend
[14:06:36] <Vratha> stop being elitist
[14:06:49] <Vratha> sickness: maybe; i will check.. i've seen reference to it
[14:06:56] <sickness> Vratha: i've used it :)
[14:06:59] <dme> i use the murayama driver on a mini-itx machine happily.
[14:07:05] <sickness> Vratha: use the via-rhine one ;)
[14:07:10] *** calumb is now known as calLNCH
[14:07:13] <twincest> Vratha: obviously solaris is only allowed to be used in the enterprise, if you can't afford the hardware you have no right to use it
[14:07:20] <Vratha> sickness: is yours the VT6103 (VIA Rhine-III)?
[14:07:46] <Vratha> twincest: yeah, i guess no more new solaris admins unless they have rich pappies and mommies :-/
[14:08:09] <cmihai> Bah, I'm saying a 500$ new PC or something from eBay will do
[14:08:09] <sickness> Vratha: I don't remember because It was at my previous workplace, but it was an epia-6000 fanless, with 1 ethernet onboard, and it was rhine II or III, anyway, same driver
[14:08:30] <dme> pci bus 0x0000 cardnum 0x12 function 0x00: vendor 0x1106 device 0x3065 VIA Technologies, Inc. VT6102 [Rhine-II]
[14:08:39] <Vratha> sickness: oh, well if the rhine 2 and 3 both work with the same driver, then i think it should work out of box when i get it
[14:08:43] <dme> (this is a 1Ghz nehemia)
[14:08:44] <PerterB> interesting viewpoint... but I don't think the majority of new solaris admins come from people who've learned at home (on sub-optimal hardware etc etc)
[14:08:49] <Vratha> because i've seen opensolaris has support for vt6102
[14:09:00] <sickness> Vratha: put on 2 of them, and run 2 mysql nodes fanless replicating between them for more than a year, no problems whatsover :)
[14:09:08] <Vratha> PerterB: yeah, well, i'm sure some did start that way
[14:09:18] <PerterB> sure, some
[14:09:21] <Vratha> anyway, that's an entirely elistist attitude
[14:09:28] <sickness> Vratha: I even managed to do network install, putting that driver inside the bootarchive/miniroot
[14:09:49] <PerterB> no, it's a realistic attitude regarding how the IT industry works
[14:10:10] <Vratha> no, it's elitist.. some of us just want to use something w/o paying for top-of-the-line equipment
[14:10:19] <Vratha> but you're welcome to your opinion
[14:10:26] <cmihai> It's really for you own good, if you can't afford a simple 500$ machine of eBay or something.. I'm not talking TOP of the line, I'm talking entry level
[14:10:40] <cmihai> You won't be able to enjoy any of the features of Solaris anyway
[14:10:55] <Vratha> cmihai: like my new $500 mini-itx server i just bought?  we shall see if the vt6103 even works w/o much difficulty
[14:10:58] <cmihai> ZFS, Containers, etc. all chew up memory and stuff. So why bother?
[14:11:04] <Vratha> i think what should be done is more drivers ported to solaris rather than you guys bitching
[14:11:16] <cmihai> Well, start porting.
[14:12:19] <Vratha> actually
[14:12:25] <Vratha> i'm going to port my userspace device driver framework
[14:12:31] <sickness> Vratha: use an opensolaris distro, it already has all those drivers included right from install time, belenix, nexenta, and so on, try them
[14:12:53] <sickness> Vratha: sxcr unfortunately does not include those 3rd party drivers :/
[14:13:00] <Vratha> sickness: i was going to use SX:CR (right initials?)
[14:13:05] <Vratha> hmm
[14:13:20] <Vratha> well i'll try to bring them in from wherever if i have to
[14:13:23] <sickness> sx:cr does not have 3rd party network drivers, I'm sure
[14:13:40] <cmihai> You can get driveres for most stuff like realtek or the likes from opensolaris.org
[14:13:44] <sickness> (granted, you can always install them :)
[14:13:46] <cmihai> Plenty of 3rd party stuff that works.
[14:14:19] <Vratha> sickness: yeah, if it doesn't work, i'll just burn a CD with them and take them over
[14:14:31] <sickness> :)
[14:16:08] *** Mazon is now known as mazon
[14:16:23] <Vratha> but this is an honest question: do many places still even use sun products anymore?  i only ever hear talk about "we switched to linux" and crap
[14:16:40] <Vratha> i never really hear about sun systems being purchased
[14:16:43] <cmihai> Get the facts...
[14:16:52] <Vratha> cmihai: isn't that why i asked?
[14:17:15] <cmihai> All this "We switched to Linux thing" is a "Get the facts" style campaign by RedHat and the likes.
[14:17:20] <jmcp> Vratha: uh .... if "nobody ever uses Sun" then why did they have non-zero earnings for last quarter, let alone earnings which exceeded Wall St's expectations?
[14:17:22] <cmihai> Yes, Sun is still used and in business.
[14:17:29] *** mazon is now known as Mazon
[14:17:31] <tsoome> actually quite many of our clients are switching from linux to solaris
[14:17:55] <Cyrille> I think he knows Sun is still used, he would like to know where.
[14:18:13] <Vratha> yes, thanks Cyrille
[14:18:15] <cmihai> Army? :)
[14:18:20] <Vratha> and i'm kinda curious about the purposes of use
[14:18:30] <tsoome> everywhere:)
[14:18:33] <jmcp> all the places you expect - governments, banks, hospitals, universities, businesses in general
[14:18:54] <Vratha> is it still used much for grid computing or supercomputing?
[14:19:09] <Cyrille> Vratha: http://www.sun.com/customers/index.xml
[14:19:11] <Cyrille> have fun.
[14:19:25] <cmihai> Let's just say it's got more users that all BSD's and most Linux distros put together.
[14:19:36] <Vratha> thanks Cyrille :)
[14:20:04] <Vratha> cmihai: hmmm.. perhaps my university and the place where i'll be working (morgan stanley) are not a good representative sample then
[14:20:27] <Vratha> i mean sun is used, just not much it seems
[14:20:31] <Vratha> at my place
[14:21:37] <cmihai> And I'd still say it has more usage then AIX, HP-UX, Tru64 and the likes put together.
[14:22:06] <cmihai> So it's pretty much alive and kicking.
[14:22:20] <Vratha> yeah
[14:22:25] <Vratha> could be quite true
[14:23:35] <cmihai> Doesn't have that big of a market share since the days when UNIX ruled the world and mammoths walked the server rooms (the days of big bad IRON) and commercial unices... but since then Linux happened. And Windows Servers started popping up everywhere.
[14:23:40] *** flowtech has quit IRC
[14:23:45] <Vratha> anyway, it wasn't a trolling comment; i was really curious
[14:23:50] <jmcp> ah, I love my 24" monitor ... I can have 2 bugster windows open side by side :-)
[14:23:55] <cmihai> Somehow Linux and Windows Servers got a big chunk of the market share.
[14:24:14] <Cyrille> jmcp: and that's good?
[14:24:16] *** flowtech has joined #opensolaris
[14:24:26] <Vratha> cmihai: yeah :-/
[14:24:41] <jmcp> Cyrille: darned right it is :)
[14:25:05] <Vratha> i'm really not sure how windows got up in there... i can only say it's probably due to AD lock-in and devs not wanting to code in anything other than ASP and now ASP.NET
[14:25:20] <cmihai> Mostly AD lock-in.
[14:25:23] <cmihai> Netware lost.
[14:25:40] <cmihai> Netscape Directory Server lost. AD really won out there. Must've been the cool name I guess.
[14:25:51] <cmihai> AD, Exchange, etc.
[14:26:22] <Vratha> yeah, and that's why i wonder where sun is even used
[14:26:25] <cmihai> And the x86-thing. x86 "servers" really started to compete with entry level enterprise stuff and 1/4 of the price.
[14:26:28] <Vratha> since windows and linux make up such a large share
[14:26:30] <jmcp> netware still lingers in some environments....
[14:26:42] <jmcp> much to my disgust
[14:26:45] <cmihai> jmcp: hell, we still have Netscape Directory Server tucked somewhere around here.
[14:26:51] * jmcp shudders
[14:26:52] <cmihai> And a HP-UX of all things.
[14:27:26] <Vratha> cmihai: so you're saying windows and linux make up such a large share because they're easy to use on x86, which are cheap and good for entry level systems?
[14:27:30] <cmihai> Vratha: well, enterprise databases (think Oracle) and CAD workstations (think CATIA) are still places for commercial UNIX. Storage, etc.
[14:27:43] <cmihai> Vratha: pretty much.
[14:27:46] <Vratha> though, linux can run on heavy iron now too
[14:27:54] <cmihai> Sure, but it's not where it got it's market share from.
[14:28:09] <Vratha> and as far as i can tell, morgan stanley uses linux on big iron (IBM) machines for intensive computing
[14:28:20] <tsoome> every moron thinks he/she can manage linux, thats why
[14:28:24] <cmihai> Hell, look at Tru64 or AIX, they're stuck on their platforms and there's a handfull of people who know how to use it (or even heard of it)
[14:28:50] <cmihai> Now Linux and Windows.. well.. almost anyone can manage that... they've all used it, know of it's existence, etc. And it runs on their hardware.
[14:28:56] <Vratha> yeah
[14:29:02] <cmihai> Most people aren't even aware of something beyond x86 and Windows/Linux.
[14:29:04] <Vratha> well i can only say that linux has mindshare going for it
[14:29:11] <cmihai> They really don't know what a SPARC or PA-RISC or Alpha is.
[14:29:15] <Vratha> i could see it making up for its problems in storage someday
[14:29:16] *** nachox has joined #opensolaris
[14:29:17] <Vratha> databases too
[14:29:54] <cmihai> Just go ask your regular Linux admin about what Tru64 is and doe.s
[14:30:15] <Vratha> i don't really care about their knowledge on the subject actually
[14:30:24] <nachox> he would go to wikipedia and answer :P
[14:30:35] <cmihai> Well, if they have no clue, imagine what the management knows about it
[14:30:42] <cmihai> The ones that actually buy the stuff...
[14:30:57] <cmihai> Hell, MS salesdroid pops over, and has them buy Windows on Xeon.
[14:31:18] <Vratha> then i can only say sun and IBM and people need better marketing...
[14:31:41] <jmcp> Vratha: IBM has serious marketing muscle. Sun tends to work in a less direct fashion
[14:31:42] <tsoome> management knows that there is new fancy vista and linux and all software with prefix "open" is in and without it is not
[14:32:04] <Vratha> jmcp: i guess that's true; that's likely why morgan stanley runs all linux on the servers
[14:32:41] <Vratha> tsoome: well then maybe opensolaris will be used more if that is indeed true :)
[14:32:52] <tsoome> it is already
[14:33:19] <jmcp> Vratha: I await the arrival of "openvista" with eagerness
[14:33:24] <Vratha> haha
[14:33:43] <tsoome> they are telling stories like solaris sucks but opensolaris is really cool...
[14:34:01] <Vratha> tsoome: who is
[14:34:26] <tsoome> some kind of morons, I guess:)
[14:34:33] <Vratha> :)
[14:35:19] <tsoome> usually one can just skip this kind of people, but unfortunately sometimes you can't:)
[14:35:27] <Vratha> does the opensolaris license allow products to be sold with opensolaris installed on them without any compensation to sun?
[14:35:54] <jmcp> Vratha: I believe so
[14:35:59] <Vratha> nice
[14:36:04] <nachox> Vratha: check the cddl, but if that wasnt possible, there wouldnt be opensolaris distributions
[14:36:07] <twincest> you can't really install opensolaris
[14:36:23] <twincest> but yes, you can do that with an opensolaris-based distribution, the binary bits are under a fairly open license
[14:36:24] <Vratha> nachox: well, the distributions aren't necessarily installed on machines by default
[14:37:15] <Vratha> as i get more familiar with opensolaris, perhaps i'll try to nudge morgan stanley into trying a couple machines out to see what they think... maybe have some Sun marketers over
[14:37:19] <oxygene> Vratha: the cddl definitely allows it - but read the license for the binary bits, I'm not sure about that one
[14:37:20] <Vratha> but we'll see
[14:37:28] <Vratha> oxygene: will do
[14:38:21] *** nbkk6fo_ has joined #opensolaris
[14:38:22] *** nbkk6fo__ has quit IRC
[14:38:37] <nbkk6fo_> Vratha you better than most that MSDW isnt moving off Linux
[14:38:45] *** nbkk6fo_ is now known as phusnikn
[14:38:52] *** phusnikn is now known as rodrik-brown
[14:38:57] <Vratha> rodrik-brown: huh?
[14:39:16] <Vratha> yeah, as far as i know, they just recently got into linux with IBM
[14:39:58] <Vratha> but whatever; i kinda have this feeling that IBM is going to improve linux's storage capabilities and what not
[14:40:23] <rodrik-brown> MS pretty much brought Linux to Wall St.
[14:40:32] <Vratha> yeah
[14:40:37] <Vratha> rodrik-brown: what were they using before that?
[14:40:40] <rodrik-brown> the years of engineering they have put into Linux I doubt they will ever give it up
[14:40:45] <sickness> didn't they put windows2000 in the new york stock exchange? =)
[14:40:58] <Vratha> god, maybe that's why the NYSE is so slow right now :)
[14:41:08] <Vratha> though i hear the NYSE is getting a big upgrade to improve latencies
[14:41:09] <rodrik-brown> Vratha Solaris
[14:41:18] <Vratha> rodrik-brown: ah
[14:41:25] <rodrik-brown> NYSE is also heavily invested in Linux
[14:41:29] <nachox> since they already invested in ibm to have them using linux i doubt they will re-invest into switching to opensolaris, and they will certainly want support which leaves only solaris
[14:41:41] <Vratha> well the nice thing for MSDW switching to linux is that now they have more mindshare
[14:41:44] <sickness> so mycrosoft was telling bullshit? :)
[14:42:16] <Vratha> rodrik-brown: yeah, i'm sure they are
[14:42:25] <Vratha> i know all the client machines are windows though, heh
[14:42:51] <Vratha> i'm actually completely fine with that; i like XP
[14:45:15] <sickness> I was meaning the server machines, indeed...
[14:47:30] *** cmihai has quit IRC
[14:50:56] *** delewis has joined #opensolaris
[14:51:37] * jmcp sleeps
[14:53:30] * nachox wonders, has anyone tried solaris in one of those via x86 compatible procs?
[14:54:05] <quasi> nachox: yes
[14:54:29] <quasi> epia M10000 works ok
[14:55:01] <nachox> they do not even need a cooler right?
[14:55:39] <dlg> does solaris use the crypto ops on it?
[14:55:46] <quasi> the m10000 does
[14:55:59] <quasi> dlg: I don't think it does
[14:56:25] <quasi> nachox: http://guses.org/home/projets/dedibox.en is another example
[14:57:07] <quasi> dlg: http://opensolaris.org/os/project/crypto/Accelerators/ - says there's a driver in development
[14:57:34] <dlg> development implies the future
[14:57:39] <dlg> so maybe you meant "it will"
[14:58:00] <quasi> it may
[14:59:13] <dlg> :)
[14:59:27] <nachox> nice, those would make great firewalls once the crossbow project gets integrated in solaris :)
[14:59:59] <tomww> quasi: doe the M10000 work with two disks?
[15:00:29] *** triplah_ has quit IRC
[15:03:20] <ofu> hmmm, this dedibox-datacenter looks much like ours...
[15:04:07] *** rodrik-brown has quit IRC
[15:05:44] *** sparc-kly has quit IRC
[15:05:59] *** sparc-kly has joined #opensolaris
[15:12:23] * nachox yawns
[15:14:38] *** tsoome has quit IRC
[15:14:58] *** tsoome has joined #opensolaris
[15:19:53] *** calLNCH is now known as calumb
[15:33:35] <Vratha> well, looks like CDDL is pretty nice in terms of executable (binary) distributions of software :)
[15:34:21] <Auralis> it is nice for source as well
[15:34:38] <Vratha> yes
[15:34:41] *** Hunger- has quit IRC
[15:35:05] <Vratha> but i was curious about the binary part right now since i was wondering if it was even possible to distribute a product with opensolaris installed
[15:35:21] *** tsoome has quit IRC
[15:35:36] *** tsoome has joined #opensolaris
[15:36:03] <Auralis> ah
[15:36:52] *** logic has quit IRC
[15:36:55] *** Hunger- has joined #opensolaris
[15:37:04] *** logic_ has joined #opensolaris
[15:39:19] *** MikeTLive has quit IRC
[15:40:32] *** cmihai has joined #OpenSolaris
[15:45:33] *** tsoome has quit IRC
[15:45:50] *** mrdeviant has joined #opensolaris
[15:45:50] *** tsoome has joined #opensolaris
[15:52:14] <axisys> what is the dst patch number for sol 8? does it require a reboot or run it in single user mode?
[15:53:36] <axisys> got it.. http://sunsolve.sun.com/search/document.do?assetkey=1-26-102775-1
[15:55:29] <nachox> solaris 2.6 is still supported? wow
[15:57:12] <Doc> no, it's not
[15:57:19] <Doc> went EOSL 23 July 2006
[15:58:44] <nachox> but the patch axisys mentioned was released for 2.6 as well
[16:00:01] <nachox> it was also released for 2.5.1
[16:00:36] <Doc> yup. and it'll cost you $$ per machine to get your hands on it
[16:01:03] <axisys> nachox: i think they have patch only for sol 8 to 10
[16:01:50] <axisys> but for others you basically copy the timezone files from sol 8 to sol 7 or 6 or 5 and recompile.. that should work
[16:02:55] <Doc> nope. there are libc changes as well for some areas
[16:03:18] <Doc> there are patches for 2.5.1 and 2.6 (costs $$ per host), and solaris 8, 9 and 10 (needs a contract)
[16:04:15] *** RaD|Tz has joined #opensolaris
[16:04:32] <nachox> only the security patches for sol 10 are free right?
[16:05:33] *** derek has joined #opensolaris
[16:05:38] *** derek is now known as hell`
[16:05:41] *** hell` is now known as hell``
[16:10:03] *** Mazon is now known as mazon
[16:10:15] *** sommerfeld has joined #opensolaris
[16:10:15] *** ChanServ sets mode: +o sommerfeld
[16:11:43] *** gallium has quit IRC
[16:11:51] *** dunc has joined #opensolaris
[16:12:17] *** LeftWing has quit IRC
[16:12:19] *** LeftWing_ has joined #OpenSolaris
[16:15:04] <quasi> tomww: I can't see why it shouldn't - it is just plain old ide
[16:15:30] *** bnitz has joined #opensolaris
[16:15:39] <jbk> morning
[16:15:55] *** laca has joined #opensolaris
[16:16:25] *** tsoome has quit IRC
[16:16:41] *** tsoome has joined #opensolaris
[16:20:56] *** tsoome has quit IRC
[16:21:11] *** tsoome has joined #opensolaris
[16:23:30] <axisys> Doc: we changed it for host running sol 7
[16:24:20] *** uebayasi has quit IRC
[16:25:33] *** Fish- has joined #opensolaris
[16:26:03] *** Fish- is now known as Fish
[16:26:39] <Fish> hello
[16:28:10] *** Peanut has joined #opensolaris
[16:28:13] *** logic__ has joined #opensolaris
[16:34:53] <Vratha> why are sun desktops overpriced?
[16:35:23] *** axxl has joined #opensolaris
[16:35:53] <quasi> Vratha: so that you can't get one ;)
[16:36:09] <Vratha> i could get one, but i'd spend my money at apple first
[16:36:27] <Vratha> better price/performance there
[16:36:40] <Doc> best one is http://flickr.com/photos/docbert/369909527/
[16:36:45] <Doc> blah.. wrong channel
[16:37:12] <axisys> how do find out the active tcp window size of a remote system? apparently that window size is smaller than my side. I read an articlet .. do not remember if it were snoop or tcpdump that has enough info to calculate that.. anyone knows?
[16:37:34] <Stric> netstat -a   can say stuff about active connections
[16:38:10] <quasi> Doc: pretty nice though (beats my friends at http://www.flickr.com/photos/zzathras777/369612745/ ;)
[16:38:18] *** tsoome has quit IRC
[16:38:27] <axisys> Stric: how about remote side that i have no access to except ftp
[16:38:34] <Doc> quasi: yes, but he took that from a plane
[16:38:38] *** tsoome has joined #opensolaris
[16:39:15] *** tsoome has quit IRC
[16:39:36] <quasi> Doc: yeah - he usually just sticks the flying pictures in his office window category
[16:39:40] *** tsoome has joined #opensolaris
[16:39:56] <gdamore> good morning *!
[16:41:00] <Doc> yah.. i was talking to him about that one a few hours ago
[16:41:22] *** logic_ has quit IRC
[16:41:32] <Stric> wtf.. "NVIDIA Quadro FX 5500 uber high-end 3D graphics card, RoHS-6 Compliant" .. first entry, then midrange, ultra and uber..
[16:41:39] *** alanc_away is now known as alanc
[16:43:40] <LeftWing_> Doc: Nice photo.
[16:43:56] *** LeftWing_ is now known as LeftWing
[16:47:34] *** nostoi has quit IRC
[16:50:13] *** tsoome has quit IRC
[16:50:34] <sommerfeld> axisys: both snoop and tcpdump can display the tcp window
[16:50:46] <sommerfeld> (you need to do snoop -v to get that level of detail IIRC)
[16:52:01] *** nachox has quit IRC
[16:52:16] <sommerfeld> excuse me, it's in the -V output, too (-V is denser than -v)
[16:52:20] *** LordKing has joined #opensolaris
[16:54:00] *** LordKing has quit IRC
[16:57:41] <axisys> sommerfeld: so this is the window size correct Win=49232 ? from snoop -V
[16:58:44] *** Vratha has quit IRC
[17:01:00] <sommerfeld> yes
[17:01:15] <sommerfeld> the currently offered window
[17:01:36] <sommerfeld> (if the receiving app doesn't read data, it will shrink...)
[17:02:25] <axisys> sommerfeld: so will it alwasy show the smallest window size of the two?
[17:03:14] <sommerfeld> it's the amount of in-sequence data that the receiver is prepared to buffer at the time it sent the ack.
[17:04:47] <axisys> so to increase the throughput i need to make sure both hosts has the high tcp_xmit_hiwat or tcp_recv_hiwat?
[17:12:27] *** piefra has quit IRC
[17:16:25] <sommerfeld> axisys: the relationship between throughput and window size can be subtle.
[17:16:44] <sommerfeld> axisys: in high-loss networks you can sometimes increase throughput by dropping window size.
[17:18:27] <sommerfeld> axisys: boost both on both ends if the flow is bidirectional.
[17:22:29] *** klocze2 has quit IRC
[17:22:46] *** klocze2 has joined #opensolaris
[17:26:13] *** nachox has joined #opensolaris
[17:27:21] <nachox> hehe, that just made my day, i read a mail to osol-discuss that just said "+1 for killing jive" :)
[17:27:24] *** fgd has quit IRC
[17:27:33] *** fgd has joined #opensolaris
[17:33:25] <axisys> but do i play with only tcp_xmit_hiwat?
[17:33:32] <axisys> as far as the OS
[17:33:57] <axisys> sommerfeld: but do i play with only tcp_xmit_hiwat? or both
[17:34:47] <sommerfeld> both.
[17:35:35] *** polk__ has quit IRC
[17:38:43] *** bunker_work has quit IRC
[17:38:54] *** bengtf has quit IRC
[17:39:17] *** polk_ has joined #opensolaris
[17:42:21] *** bondolo has joined #opensolaris
[17:43:33] <axisys> sommerfeld: thnx
[17:44:09] <axisys> sommerfeld: is there any other parameter /dev/tcp values that need to play w/?
[17:44:54] *** polk_ has quit IRC
[17:45:30] <kimc> greetings..
[17:46:28] *** stevel has joined #opensolaris
[17:46:28] *** ChanServ sets mode: +o stevel
[17:47:03] <kimc> slice 8 on a disk is 'boot' and fsck complains the magic number is wrong
[17:47:09] <richlowe> mornin' stevel.
[17:47:14] <kimc> its a disk copied with dd
[17:47:30] <kimc> any thoughts on this ?
[17:47:46] <kimc> the disk will boot the machine
[17:49:20] <nachox> i thought dd didnt work when used with solaris disks
[17:49:45] <kimc> it made is disk which boots and appears for run normally..
[17:50:37] <kimc> what i'm trying to do is replace the boot disk with a raid array.. the process was moving along fine to this point
[17:50:55] <kimc> just one more partition to mount and ufsrestore to the raid array
[17:51:25] <Doc> many raid arrays use the last X Mb on the disk to store their config, so...
[17:51:26] <kimc> maybe i should replace the dd-copy with the original installed boot disk
[17:51:39] <kimc> yes Doc
[17:52:26] <kimc> problem is that slice 8 'boot' cannot mount due to complaint about the magic number being wrong -says 'run fsck'
[17:52:51] <Doc> right, so is slice 8 at the end of the disk?
[17:53:02] <kimc> i'll shut it down and replace the boot disk with the original..
[17:54:04] <kimc> yes slice 8 is at the end
[17:54:19] <kimc> but its the source disk its not happy with
[17:54:27] *** polk__ has joined #opensolaris
[17:56:33] <stevel> morning rich
[17:59:17] <kimc> the original disk mounts fine -great
[18:02:49] <sommerfeld> axisys: there are many tcp tunables.  play with them at your peril.
[18:05:03] <axisys> sommerfeld: u mean in the context of bandwidth?
[18:05:09] <kimc> looks like maybe the 'boot' partition doesn't need/want to be restored with ufsrestore
[18:05:28] *** sommerfeld has quit IRC
[18:05:41] *** sommerfeld has joined #opensolaris
[18:05:41] *** ChanServ sets mode: +o sommerfeld
[18:06:56] *** polk__ has quit IRC
[18:08:11] <kimc> that last remark it wrong..
[18:11:37] <delewis> hmm, sweet. seems I've acquired a Tadpole SPARCbook 6500 with 4GB of memory.
[18:11:59] <jbk> nice
[18:12:32] <jteo> was the sale of an organ involved?
[18:12:35] *** mazon is now known as Mazon
[18:12:43] <delewis> jteo, actually, no. :-)
[18:12:44] <delewis> more like a finger.
[18:12:46] <Cyrille> and more importantly whose
[18:13:04] <delewis> if I told you guys the price I paid, you fall over :-)
[18:13:12] <delewis> you'd*
[18:13:13] <elektronkind> "it fell off the back of a truck"
[18:13:22] <jteo> so you're bragging now.
[18:13:28] <jteo> ;)
[18:13:31] <delewis> $1,045 :-)
[18:13:32] *** polk__ has joined #opensolaris
[18:14:20] <jteo> there is some bizarre need for a sparcbook in your life?
[18:15:09] <elektronkind> it would make a good wall-mounted terminal for the bathroom
[18:15:29] <LeftWing> heh
[18:15:38] <dme> or a good step, to reach higher things.
[18:15:43] <delewis> elektronkind, hardly. 2x60GB drives, 650MHz UltraSPARC-IIe, and 4GB of memory isn't exactly a wall-mounted terminal. :-)
[18:15:50] <jbk> when i used to work at an ISP many moons ago, one of the customers worked for plantinum technologies and had a sparc book
[18:16:09] <LeftWing> A SunRay 150 would make a better wall-mounted terminal. =P
[18:16:12] <jbk> but this was in the solaris 2.4 days, and the other admins didn't want to try to get ppp working on it :)
[18:18:16] <delewis> no modems on modern SPARCbooks. Apparently, they've got two NICs.
[18:18:34] <delewis> why you need two NICs on a laptop is beyond me, but who am I argue, otherwise.
[18:18:55] <LeftWing> delewis: To set up a NAT firewall in the field? =P
[18:19:17] <jbk> link aggregration!
[18:19:25] <delewis> jbk, :-)
[18:19:35] <LeftWing> Could you run Sun Cluster with only one public-facing and one aggregation link?
[18:19:45] <LeftWing> s/aggr../interconnect/g
[18:19:58] <delewis> LeftWing, nope.
[18:19:59] <LeftWing> Bloody 4.20am and I'm just typing what I see. =P
[18:20:20] <LeftWing> You'd need at least two for interconnect I guess?
[18:20:33] <jbk> hmm dunno
[18:20:42] <jbk> vcs allows you to run with one (though it complains)
[18:20:43] <delewis> http://www.tadpole.com/upload/specifications/sparcbook.pdf
[18:21:10] <delewis> Sun Cluster checks the number of interfaces upon installation
[18:21:42] <LeftWing> How many does it require to proceed?
[18:21:49] *** rcheli has joined #opensolaris
[18:22:04] <delewis> its either 2 or 4.
[18:22:09] <rcheli> What's the command in VPROM to list all aliases for disks ?
[18:22:15] <rcheli> printenv something
[18:22:18] <jbk> devalias
[18:22:45] <rcheli> thanks jbk
[18:22:51] <jbk> np
[18:23:00] <delewis> what's PGX64 performance compared to say, an Elite3D?
[18:23:17] <dme> pgx64 has no 3d acceleration.  okay for 2d.
[18:23:26] <delewis> dme, that should suffice.
[18:23:37] <delewis> Elite3D sucks for most of the modern 3D workloads nowadays :-)
[18:23:51] <dme> pgx64 is ATI Rage XL, as i recall.
[18:23:55] <delewis> right
[18:24:20] <delewis> Xine apparently has "accelerated" PGX64 video output.
[18:25:07] <dme> where "accelerated" means "almost watchable" probably :-)
[18:25:46] *** Cyrille has quit IRC
[18:25:49] <delewis> can't be worse than doing video output on an Elite 3D :-)
[18:25:53] <jteo> depends on what you're watching.
[18:26:33] <delewis> no video scaling :-(
[18:26:33] <delewis> hardware video scaling, that is.
[18:26:33] <delewis> jteo, nothing spectacular.
[18:26:38] <delewis> no H.264 1080p video, that is, which seems to be latest craze.
[18:27:25] <delewis> where apparently a 3GHz P4 can barely handle it, though I imagine the CoreAVC codec is far more optimal than the x264 code.
[18:27:54] *** alanc is now known as alanc_away
[18:28:13] *** reflect has quit IRC
[18:28:22] <PerterB> H.264 is pretty CPU intensive, especially at 1080p
[18:28:31] <delewis> yes
[18:28:52] *** qdk has quit IRC
[18:28:55] <PerterB> my laptop can just about handle it with a bit of jitter in the frame rate
[18:29:27] <jamesd> isn't everything cpu intensive at 1080p
[18:29:36] *** rcheli has quit IRC
[18:29:36] <PerterB> well yes :)
[18:30:00] <jteo> relative to what and how many cores.
[18:30:29] <delewis> regardless, most of my video is encoding in MPEG4 and a slightly less than DVD-quality resolution.
[18:30:36] <delewis> encoded, rather.
[18:30:56] <delewis> enough for a 750MHz UltraSPARC-III to encode and decode.
[18:32:43] *** mdj has quit IRC
[18:33:25] *** Netwolf has quit IRC
[18:34:10] <jteo> i suppose 2 sparcbooks would be more fun than buying 1 Modbook. alas i digress.
[18:36:19] <nachox> arent those sparcbooks more like sparclibraries? ie, impossible to carry
[18:44:30] <delewis> nachox, 7Lbs.
[18:44:37] <delewis> with a full load (two disks, etc.)
[18:44:51] <delewis> they actually use a standard laptop chassis that other vendors use.
[18:46:43] *** dunc has quit IRC
[18:47:04] *** dunc has joined #opensolaris
[18:53:10] *** Theoden-Nexenta has joined #opensolaris
[18:54:33] <gdamore> sparcbook and sparcle are in standard 15 and 14" cases.  the bullfrog is the 28lb bad boy for the military
[18:54:59] <delewis> which is equivalent to what? a v210?
[18:55:42] <gdamore> i think so.
[18:56:06] <gdamore> its a dual us3i system, with a full size pci slot, 17" screen, and all the trimmings. :-)
[18:56:27] <delewis> right. :-)
[18:57:15] *** bengtf_ has joined #opensolaris
[18:57:18] *** bengtf_ is now known as bengtf
[19:00:12] *** bnitz has left #opensolaris
[19:00:26] <nachox> heh, i thought the whole deal with laptops was being able to actually move them :)
[19:00:54] <twincest> my laptop is a desktop, i move it maybe once a month
[19:01:01] <twincest> but when i do, i still think i'd rather not lug around a bullfrog :)
[19:02:08] <gdamore> bullfrogs are "mobile servers", not laptops.  they mounted in humvees and such. :-)
[19:02:24] <nachox> i'm still thinking about the niagra chip as an appliance... but i have never heard of sun selling he actual chip, motherboard and stuff to make them
[19:02:29] <gdamore> but we started with a standard 17" laptop case like you'd find from acer.
[19:02:50] <gdamore> nachox: not yet.  i think you can license the designs, though.
[19:03:16] <gdamore> i want to see niagra2 in a laptop. :-)
[19:03:50] <gdamore> unfortuntely, most of our buyers for sparc insist on Solaris 8 compatibility right now.  (So even PCIe is a non-starter.)
[19:04:18] <delewis> well, if you can put a v210 in a box, you can certainly put a T1000 in one :-)
[19:04:33] <gdamore> and most of those have been saying they'll switch to x86 (possibly Windows) before moving away from S8.
[19:04:43] <delewis> gdamore, ugh.
[19:05:06] <delewis> I would ask why they need Solaris 8 compatibility, but I have a feeling you wouldn't be able to respond. :-)
[19:05:15] <gdamore> exactly.  we're going to have x86 products on our price sheet soon, I think. :-(
[19:05:33] <Stric> delewis: probably "we know it works on s8"
[19:05:36] <delewis> gdamore, pity.
[19:05:40] <delewis> Stric, my thinking precisely.
[19:05:41] <gdamore> Stric: exactly.
[19:05:45] <delewis> silly vendors :-(
[19:05:58] <gdamore> nobody wants to requal apps, and they don't "get" Solaris binary compatibility
[19:06:04] <Stric> and even if they try it on s10 and it "seems to work", that doesn't mean it does
[19:06:07] *** reflect has joined #opensolaris
[19:06:26] <delewis> gdamore, how long does Tadpole typically support a laptop (as far as driver maintenance, etc.)? (of course if the code gets in Nevada, that'll be irrelevant)
[19:06:39] <gdamore> delewis: 1) the code will get into nevada
[19:07:08] <gdamore> 2) we are still supporting SPARCbook and UltraBook IIe hardware
[19:07:13] <delewis> right.
[19:07:23] <gdamore> 3) our gov't customers require some minimum time (i think 5 years or so)
[19:08:09] <gdamore> what's interesting to me, is that i've never ever seen anyone from nature hanging around the os.o groups.
[19:08:27] <delewis> gdamore, nope.
[19:08:40] *** polk__ has quit IRC
[19:08:43] <gdamore> but then i don't see any x86 laptop vendors/mfgrs involved, either.
[19:08:46] <delewis> Naturetech also doesn't seem to have the tools that Tadpole distributes with their laptops, either.
[19:09:11] <gdamore> no, afaik nature has no software people on staff at all.
[19:09:12] <nachox> gdamore: i bet most here know, but where do you work? tadpole?
[19:09:19] <gdamore> yes.
[19:09:34] <gdamore> except that now we are a division of General Dynamics.
[19:09:49] <nachox> you guys are selling sun ray laptops right?
[19:09:49] *** polk__ has joined #opensolaris
[19:09:51] <delewis> and have now become Itronix?
[19:10:00] <gdamore> Itronix == GD.
[19:10:04] <delewis> ah.
[19:10:14] <gdamore> basically, GD bought both Tadpole and Itronix, and merged us together.
[19:10:27] <gdamore> (Itronix does x86 hardware, in case people are curious.)
[19:10:43] <delewis> x86 hardware designed for government applications and all that, right?
[19:11:01] *** axxl has quit IRC
[19:11:03] <gdamore> right.  but also commercial.  Sears is one of their biggest customers of handheld units, I believe.
[19:11:11] <delewis> interesting.
[19:11:24] <gdamore> apparently truck drivers for sears are much harder on their handhelds than the US marines are. :-)
[19:12:54] *** polk__ has quit IRC
[19:13:35] *** polk_ has joined #opensolaris
[19:13:52] <nachox> heeh
[19:17:22] *** bubbva has joined #opensolaris
[19:18:39] <e^ipi> likely has a lot to do with a field marine knows that his life might depend on it
[19:18:44] <e^ipi> & the guy at sears doesn't care
[19:19:05] <gdamore> i think you are exactly correct.
[19:19:25] <gdamore> even if his life won't depend on it, the marine has probably learned to treat his equipment with respect.
[19:25:49] *** alanc-away is now known as alanc
[19:26:10] <nachox> i think the truck driver's job is much more dangerous, he should be spared :)
[19:26:31] *** polk_ has quit IRC
[19:26:42] *** Theoden-Nexenta has left #opensolaris
[19:27:37] <e^ipi> driving in general is a fairly dangerous job ( I got in a really bad car wreck delivering pizza in highschool ), but i don't think it approaches the danger of being shot at for a living
[19:28:09] *** nachox has quit IRC
[19:28:43] *** slowhog has joined #opensolaris
[19:35:54] *** nostoi has joined #opensolaris
[19:36:43] *** estibi has quit IRC
[19:37:12] <alanc> e^ipi: actually, I think the odds of dying in a traffic accident are higher than those of dying in the military
[19:37:23] *** estibi has joined #opensolaris
[19:39:30] *** tsoome has joined #opensolaris
[19:40:39] *** polk__ has joined #opensolaris
[19:42:01] *** calumb has quit IRC
[19:42:21] <sommerfeld> gdamore: so, apparently, one of the most difficult environments for gear is wallstreet trading floors.  angry traders beat up on the gear.
[19:42:32] <gdamore> haha.
[19:42:33] <sickness> sommerfeld: cool :P
[19:42:48] <gdamore> good market for ruggedized keyboards, i guess. :-)
[19:42:50] <sommerfeld> or so i'm told
[19:43:20] <sommerfeld> yah.  but make sure that anything more expensive than a ruggedized keyboard can survive a hit from the ruggedized keyboard or phone handset or...
[19:43:43] <gdamore> plexiglass monitor screen covers? :-)
[19:44:01] <alanc> "I sold SUNW at 5.60 on Monday and today it's at 6.30! I'll go kick a computer!"
[19:44:06] <gdamore> hmm... plexiglass might scratch.  :-)
[19:44:19] <sommerfeld> lexan, maybe
[19:47:02] *** polk__ has quit IRC
[19:54:28] <LeftWing> alanc: Or, potentially, some switch gear as they'd read somewhere that the Network is the Computer.
[19:55:51] *** yofuh has joined #opensolaris
[19:57:31] *** tsoome_ has joined #opensolaris
[19:58:06] *** polk__ has joined #opensolaris
[19:59:45] *** comay has joined #opensolaris
[20:00:35] *** syllogism has joined #opensolaris
[20:00:38] <syllogism> hey guys
[20:00:50] <syllogism> does anyone know of my SiS900 NIC will work with solaris 10?
[20:00:53] <bubbva> hey (even if we aren't all guys... ;-)
[20:01:19] <syllogism> how's this
[20:01:37] <syllogism> howdy y'all - anyone know if my SiS900 NIC will work with solaris 10 before I go and install it?
[20:01:52] <sickness> syllogism: try masayuki murayama's drivers
[20:02:00] <sickness> will work with them
[20:02:33] *** lisppaste3 has quit IRC
[20:02:57] <syllogism> what about an intel 2200BG wireless card?
[20:03:13] <sickness> there's a page on opensolaris.org for that
[20:04:11] *** dfgas has joined #opensolaris
[20:04:35] <dfgas> how do i install my mic its either a 8139 or a tulip so i can get online?
[20:06:27] *** lisppaste3 has joined #opensolaris
[20:07:10] <syllogism> I'm downloading solaris 10 from sun's web site. . am I missing out on anything by not downloading from opensolaris.org?
[20:08:07] <sickness> syllogism: lots of new things
[20:08:28] <bubbva> Solaris Express is really builds of Nevada, the latest & greatest OS.
[20:08:30] <sickness> syllogism: but that's a stable release, opensolaris.org has sx:cr that's solaris11 beta
[20:08:47] <bubbva> Lots of things do get backported to the 10 updates, but not everything.
[20:09:25] <syllogism> so if some of my hardware doesn't work in 10, I should download Solaris Express "Nevada" and give it a try?
[20:09:41] <bubbva> plus there is significant delay for things to get to the s10 updates (related to testing, stability, bug fixing, and the time it just takes to get a full release out the door)
[20:10:03] <bubbva> syllogism: you may want to ask in here first to see if support has been added :-)
[20:10:14] <dfgas> anyone know how to get my network card installed?
[20:10:19] <syllogism> bubbva, roger that. . .
[20:10:24] *** Odin-LAP has joined #opensolaris
[20:10:45] <syllogism> I'm planning to install 10 on my laptop (unfortunately SiS chipset with an ati card)
[20:10:46] *** tsoome1 has joined #opensolaris
[20:11:05] *** tsoome1 has left #opensolaris
[20:11:07] <syllogism> anyone think i'll run into some trouble with that?
[20:11:12] <bubbva> unfortunately I'm more of a sparc person (always have been) , so I'm not very useful for helping with x86 issues
[20:11:39] <syllogism> I've never really gotten the chance to use a sparc system
[20:12:20] <bubbva> well, I've been at Sun for 10 years, and before that, used mostly sparc systems at Purdue (except my last year, where we had 2.5.1 on x86 boxes that had been donated to the school)
[20:12:51] <bubbva> I do have a laptop, though, that is in desperate need of an upgrade...perhaps next wekk...
[20:13:10] *** tsoome1 has joined #opensolaris
[20:13:20] <syllogism> sparc have better hardware?
[20:16:41] <comay> syllogism: http://www.sun.com/bigadmin/hcl/hcts/device_detect.html has a java program that runs under windows or linux and will identify which devices are supported by solaris nevada
[20:17:56] *** Fish- has joined #opensolaris
[20:18:05] <syllogism> comay, thank you - but my laptop is out of commision at the moment
[20:18:33] <syllogism> I was running netbsd on it(well, attempting to), and I soon realized that netbsd just isn't up to par with hardware support yet. .
[20:21:41] *** RaD|Tz has quit IRC
[20:23:25] *** tsoome_ has quit IRC
[20:27:00] *** tsoome has quit IRC
[20:27:17] <stevel> opensolaris.org about to go down for a minute or two
[20:27:54] <stevel> or longer if i screw up
[20:29:12] *** alfism has joined #opensolaris
[20:29:49] *** bengtf_ has joined #opensolaris
[20:31:03] *** logic__ has quit IRC
[20:31:52] <syllogism> is sparc architecture better designed then x86?
[20:33:18] <DerJoern> sometimes
[20:33:19] <mrdeviant> depends what you are doing with it
[20:34:18] *** Fish has quit IRC
[20:35:00] <sommerfeld> syllogism: questions like that do not have yes or no answers.
[20:36:18] <syllogism> I was really hoping to get some opinions wth backup links
[20:37:26] <bubbva> syllogism: for me, it's just everything (obviously) just works. I do a lot less fighting with my equipment than others who are getting the latest & greatest x86 boxes (but that's good that they fight, and hopefully things will be better for external folks)
[20:37:39] <sommerfeld> the answers are inherently subjective and depends on what you want to do with a system.
[20:41:34] <DerJoern> sparc boxes doesnt change every month that is an advantage
[20:42:25] * bubbva is away: lunch/gym
[20:43:57] <jbk> hmm anyone know if s10u2 supports usb flash drives on an ultra20? Simply inserting them doesn't appear to work
[20:44:28] <gdamore> jbk: i'm pretty sure it works -- I've used them on my u20 running s10 IIRC.
[20:44:48] *** jopi has quit IRC
[20:45:00] <jbk> hmm perhaps jit's was junk
[20:45:44] *** ka46 has joined #opensolaris
[20:46:33] <jbk> nope.. not working
[20:46:35] *** jopi has joined #opensolaris
[20:47:18] <jbk> i think i have a couple gb's worth of flash drives with teh sun logo on them by now :)
[20:48:31] *** logic_ has joined #opensolaris
[20:55:08] *** ka46 has quit IRC
[20:55:26] *** ka46 has joined #opensolaris
[20:57:26] *** phips has quit IRC
[20:57:59] * jbk curses vxvm
[20:58:45] * charlieS hugs zfs
[21:00:49] <jbk> i wish we were using zfs :(
[21:01:06] <jbk> but mgmt wants to wait until it's bootable and you can shrink pools
[21:04:25] *** mayskenn has joined #opensolaris
[21:04:51] *** zerovirus2002 has joined #opensolaris
[21:08:06] <zerovirus2002> hi
[21:08:26] *** mayskenn has quit IRC
[21:09:03] <jamesd> jbk, when was the last time you saw a file system shrink?   i never have enough disk space....
[21:10:04] <jbk> actually somewhat often here
[21:10:12] <jbk> because they tend to grossly overallocate
[21:10:25] <jbk> then to save money go back and reclaim what's not used on the SAN to reallocate elsewhere
[21:10:46] <jbk> since they decided to use all EMC, cost becomes a big deal :)
[21:13:32] *** twocarlo has joined #opensolaris
[21:13:55] <twocarlo> hi all
[21:14:13] <zerovirus2002> Hi twocarlo
[21:14:32] <twocarlo> hi zero
[21:14:37] <jbk> for example, their goal is to not purchase any new storage this year
[21:14:51] <jbk> (we all laugh violently at this, but still, they're gonna try)
[21:19:36] <jamesd> jbk, have the sun storage guys do a review on your site, and they will tell you how much new storage you really need... and the truth may come out is that you have enough storage unallocated or misallocated that you don't need new storage for the next 3 years.
[21:20:52] <jamesd> on average only  13% of storage on site is actually used, its either  allocated to projects that have died, or  overallocated to machines/projects that don't need it, but no one dares take it back.
[21:23:06] <zerovirus2002> I believe you are absolutly right jamesd
[21:23:37] <jbk> oh i know, but it's very political :)
[21:23:39] <zerovirus2002> sometimes even we keem clones of production systems on the same storage media, which wastes alot of space
[21:23:47] <jbk> we have a 'san' but we really just do port aggregration
[21:24:02] <jbk> because emc says it's very bad to have luns from multiple arrays visible to the same hba
[21:24:21] <sommerfeld> !
[21:24:22] *** qdk has joined #opensolaris
[21:24:38] <jbk> so, if you need luns from two different arrays visible
[21:24:52] <twocarlo> im a solaris newbie, i use a generic pc or ibm compatible computer can i install solaris on this?
[21:24:54] <jbk> you need 4 ports (as there's always two paths to a lun)
[21:25:00] <jamesd> or the other major use is "misuse"  joe blows special entertainment files need to stop using up 5TB...    of course  marco who is no longer with the company still has a mirror of joe's entertainment files...  that need to disapear as well..
[21:25:16] <jbk> actually it's more oversized oracle databases
[21:25:21] <jbk> than mp3s or videos or such
[21:25:48] <jbk> i've never actually seen anyone do that (though I can't speak for the windows world)
[21:25:55] <jbk> (here)
[21:26:47] <twocarlo> anyone would like to answer my question
[21:27:07] <sommerfeld> sure the oversized oracle databases aren't storing mp3's and videos? :-)
[21:27:08] <jbk> twocarlo: it's very possible
[21:27:22] <jbk> hopefully not
[21:27:38] <jbk> typically it's 'we need 435635463 copies of the 600gb database'
[21:28:07] <jamesd> then i could mention that  almost 0% of files needs to be accessed again once they are 90 days old, but they can't be deleted because you may need to get audited and what they really need is to be moved to tape... oh yeah and EMC doesn
[21:28:15] <jbk> i don't even flinch at allocating things a terabyte at a time anymore
[21:28:17] <jamesd> t sell tape so they will never tell you that
[21:28:27] <sommerfeld> twocarlo: in general, you can run solaris on x86 systems.
[21:28:50] <sommerfeld> but, as with everything-but-windows, whether it works depends on the precise hardware you have.
[21:28:51] <jbk> *sigh* and crs just booted this whole cluster..
[21:29:02] * jbk waits for hpost to complete...
[21:30:38] <jbk> but i just don't get the whole 1 array : fc-port thing
[21:30:40] * Teknix waits for zfs+fma
[21:30:54] <jbk> seems to kinda defeat the whole idea behind a san (at least as it's been presented to me)
[21:31:07] <Teknix> and using intelligent arrays in jbod mode with zfs is still sucky
[21:34:29] <Teknix> does anyone know of a vendor selling a 16bay FC<->SATAII dumb JBOD?
[21:35:01] <Teknix> there's plenty of straight FC JBODs, and SAS JBODs, but not FC/SATAII
[21:45:36] *** dlynes_laptop has quit IRC
[21:46:17] *** slowhog has left #opensolaris
[21:52:14] <syllogism> good god this thing has some serious compression on it. . .
[21:56:04] *** dunc has quit IRC
[21:56:07] *** zerovirus2002 has quit IRC
[21:59:32] <Teknix> i just talked to a vendor that swears they have one
[22:00:48] <twincest> Teknix: http://www.infortrend.com/2_product/a16f-r2431.asp <- it's a RAID device, but i believe ost infortrend devices can act as jbod as well
[22:01:04] <Teknix> i have infortrend gear.  This is not what I need
[22:01:26] <Teknix> even in JBOD mode it's not working the way I think it should
[22:01:59] <Teknix> the drive swap/fail detection doesn't work right when you export the disks as individual LUNs
[22:02:21] <Teknix> so you manually have to re-scan the drive and remap the LUN
[22:03:30] <Teknix> and zfs doesn't pick up on media errors unless a zpool scrub is run, and fma isn't reporting problems.. zfs and the OS just keeps trying to access disks even when they're pulled from the array
[22:11:18] *** Gr|ffous has quit IRC
[22:11:20] *** Griffous has joined #opensolaris
[22:16:38] *** broadcast has quit IRC
[22:16:47] *** broadcast has joined #opensolaris
[22:19:35] * bubbva is back (gone 01:37:11)
[22:20:07] *** mrdeviant has quit IRC
[22:35:17] *** hspaans has joined #opensolaris
[22:35:56] <hspaans> g'day all
[22:37:33] <quasi> hey hey
[22:38:05] *** pikapika has joined #opensolaris
[22:38:40] *** nostoi has quit IRC
[22:38:44] *** TrogL has joined #opensolaris
[22:39:01] <bubbva> hi!
[22:39:57] <axisys> how do i find out if any file is symlink to a file
[22:40:19] <pikapika> hello
[22:42:12] *** piefra has joined #opensolaris
[22:43:09] <quasi> axisys: you could always do a find on type s and see if any of them link to the file
[22:43:56] <jbk> or do you want to check if a particular file is a symlink?
[22:44:14] *** ka46 has quit IRC
[22:45:22] <axisys> quasi: :-(
[22:45:32] <axisys> jbk: nah
[22:46:51] <axisys> if linkfile --> origfile, how do i find out some one pointing to orig file
[22:47:38] <quasi> find / -type s |grep originalfile
[22:47:48] <quasi> how hard can it de?
[22:47:52] <quasi> be
[22:50:38] <axisys> quasi: well unless there is magic hook to `ls' or `gnu ls' that shows the attributes
[22:50:58] <axisys> quasi: i did not fix one yet
[22:50:59] <lasseoe> man ls   and ye shall find out
[22:51:23] <axisys> s/fix/find/
[22:51:43] <axisys> lasseoe: none as fas as i can tell
[22:52:41] <quasi> ah, you might want to stick some xargs ls -l in there as well
[22:53:16] <lasseoe> hard link or symbolic link?
[22:53:25] *** cmihai has quit IRC
[22:58:23] <CIA-22> Wyllys Ingersoll: 6507361 pktool help doesn't; 6514823 kmfcfg is misspelled in exec_attr; 6515109 Error message when using "pktool gencert" should use 'serial' instead of 'serno'
[23:15:07] *** deedaw has joined #opensolaris
[23:15:07] *** klocze2 has quit IRC
[23:15:22] *** dlynes_laptop has joined #opensolaris
[23:17:49] <TrogL> I still haven't been able to get Studio11 to isntall on OpenSolaris Nov/06.  No matter what I do, get same erorr message:
[23:18:15] <TrogL> Exception in thread "Thread-30" java.lang.NoClassDefFoundError: com/sun/install/ panels/ComponentSelectionListener
[23:18:51] *** Gman has joined #opensolaris
[23:19:01] <TrogL> as explained by others, it's a JAVA problem
[23:19:32] *** Fish has joined #opensolaris
[23:20:21] *** slowhog has joined #opensolaris
[23:20:47] <TrogL> ...and the solution (as explined to me) invovled reverting to an earlier version of the JAVA lbraries(?)
[23:20:55] <TrogL> which I was unable to properly implement.
[23:21:00] <TrogL> Can anybody give me a walkthrough?
[23:23:17] *** Tramsie has joined #opensolaris
[23:27:33] *** Fish- has quit IRC
[23:34:59] *** Tpenta has joined #opensolaris
[23:42:32] *** Qapf has quit IRC
[23:43:11] <doof> does some people use opensolaris or solaris on  workstation ?
[23:45:42] *** pseudoXh4 has joined #opensolaris
[23:46:07] <pseudoXh4> Anyone know if Belenix comes with drivers for all wireless devices supported by opensolaris thus far?
[23:46:08] <Auralis> sure, lots of people do
[23:46:24] <pseudoXh4> I tried it out and ifconfig -a said nothing about my ralink, although Ralink support is indicated on the website.
[23:46:24] <pseudoXh4> :\
[23:47:53] *** Fish has quit IRC
[23:48:42] <delewis> doof, workstation? I'll be using it on my laptop once it arrives.
[23:48:58] <hspaans> the laptop or solaris?
[23:49:30] <Gman> pseudoXh4, check the drivers, and if not, you can easily download them from opensolaris.org
[23:50:32] <pseudoXh4> Hm, thanks for the suggestion Gman.
[23:50:42] <doof> i dont know if i must stay with my debian or change to solaris10/opensolaris in order to learn more . the pb i dont know if solaris have good desktop, tools, multimedia soft
[23:50:48] <lasseoe> pseudoXh4: the interface needs to be plumbed before you can see it with ifconfig
[23:50:52] <pseudoXh4> Would anyone happen to know if DRI works with just Intel 915s, or are 945s/950s supported by now?
[23:51:04] <pseudoXh4> lasseoe, I looked through dmesg too though... and nothing.
[23:51:10] <lasseoe> ah right on
[23:51:13] <Gman> i believe so
[23:51:16] <delewis> doof, uh, you can build the vast majority of multimedia tools that build on Linux for Solaris.
[23:51:35] <delewis> MPlayer, FFmpeg, VLC, Xine, XMMS, mkvtoolsnix, etc.
[23:51:53] <pseudoXh4> Does OpenSolaris have multi-user/multiple terminal screens like the BSDs do, with CTRL+ALT+Fx?
[23:52:00] <delewis> pseudoXh4, no.
[23:52:12] <Gman> not yet
[23:52:12] <pseudoXh4> :(
[23:52:15] <Gman> soon!
[23:52:18] <pseudoXh4> :D
[23:52:27] <doof> delewis: you'll use solaris or opensolaris ?
[23:52:28] <Gman> check out the virtual consoles project on opensolaris.org
[23:52:36] <delewis> doof, you cannot use OpenSolaris.
[23:52:46] <doof> why ?
[23:53:27] <delewis> its just a source dump -- not an operating system, but merely the bits to construct an OpenSolaris-based distribution, like Solaris or Nexenta.
[23:53:37] *** robelkin has joined #opensolaris
[23:54:13] <delewis> its analogous to throwing the Linux kernel source, GNU coreutils, bash, glibc, etc. in a tarball.
[23:55:19] <CIA-22> Artem Kachitchkin: 6488051 media directory created with wrong permissions; 6496068 If statement terminated prematurely in hald-runner; 6496370 bug in media_findname() causes fdformat to fail; 6504803 HAL should support multisession/hybrid CD
[23:56:46] <hspaans> go Artem
[23:56:59] *** estibi has quit IRC
[23:58:02] <doof> nexenta seem to be nice
[23:59:46] *** bubbva has quit IRC

top