[00:00:23] <tsoome> surprise?
[00:00:24] <Error_e^ipi> bunch of jerks...
[00:01:40] <tsoome> acroreader would be good....
[00:02:18] <Error_e^ipi> I am surprised actually
[00:02:35] <Error_e^ipi> they support everything on everything else
[00:03:14] <Error_e^ipi> there's an IRIX/MIPS port still
[00:03:19] *** adlpaf has joined #opensolaris
[00:04:39] <quants> with opensolaris give it time
[00:04:56] <quants> it won't take long
[00:05:31] <Error_e^ipi> I got a response from a hu-man when i emailed... *shrug* maybe that'll help
[00:05:31] <sickness> can you advice me a software voip phone for opensolaris x86? (it doesn't have to be compatible with skype)
[00:06:00] <Triskelios> sickness: solaris express ships with ekiga
[00:06:20] <Error_e^ipi> neato
[00:06:25] <sickness> uhmmm
[00:06:28] <sickness> interesting...
[00:06:31] <Error_e^ipi> at least mathematica has a solaris/x86 port
[00:06:32] <quants> skype sux
[00:06:42] <sickness> infact I'm not interested in skype
[00:06:53] <sickness> but I need to find a multiplatform software phone
[00:06:56] <quants> we use FWD
[00:07:20] <quants> I use twinkle, ut we're looking at ekiga
[00:07:25] <sickness> all the people continues to advice me ser and asterisk, it seems that I'm not able to explain that I need a *client* :)
[00:07:33] <Triskelios> sickness: ekiga uses regular SIP
[00:07:46] <sickness> Triskelios: yeah
[00:08:00] <quants> freeworlddialup is a sip service
[00:09:17] <quants> getting something sip for windows is truly challenging
[00:09:47] <sickness> heh
[00:10:20] <quants> the best I find that support STUN is pulver, and ekiga
[00:10:50] <quants> we use kphone, and twinkle
[00:13:27] *** lloy0076 has joined #opensolaris
[00:15:51] *** Gman_ has joined #opensolaris
[00:16:02] *** estibi has quit IRC
[00:18:43] *** razrX has quit IRC
[00:19:45] *** CosmicDJ has joined #OpenSolaris
[00:20:30] *** razrX has joined #opensolaris
[00:27:00] *** Gman has quit IRC
[00:29:14] *** postwait__ has joined #opensolaris
[00:30:44] *** postwait_ has quit IRC
[00:31:39] *** Gman_ has quit IRC
[00:32:04] *** triplah_ has joined #opensolaris
[00:34:05] <onbot> commit by Aleksandr Guzovskiy:  6510640 ISM pagefaults are very slow
[00:34:28] *** MikeTLive has joined #opensolaris
[00:35:42] *** jsubl2 has joined #opensolaris
[00:38:35] *** MikeTLiv1 has joined #opensolaris
[00:40:03] *** MikeTLiv1 has left #opensolaris
[00:43:41] <mv> does it only seems to me, or osol x86 with 512MB ram or less su*ks...
[00:44:05] <Error_e^ipi> 512 is fine
[00:44:10] <Error_e^ipi> a bit swappy
[00:44:15] *** MikeTLive has quit IRC
[00:44:16] <Error_e^ipi> less, not so much
[00:44:32] <mv> well, i've had 512MB ram and something about 1300MB od swap
[00:44:40] <Error_e^ipi> I don't tend to use any less than a gig
[00:44:43] <mv> and i can't even build b55
[00:44:48] <Error_e^ipi> on any machine
[00:45:23] <mv> yeah, i put in that machine 2x512MB a it seems to be fine
[00:45:42] <mv> but my freebsd router is crying :) it was his
[00:47:10] <Error_e^ipi> if it's just a router you can strip *bsd down pretty well & give it 256 or 128 even if you get lucky
[00:47:59] <mv> well, it's not just router, but he'll be fine
[00:52:06] *** abackos has quit IRC
[00:54:18] *** uebayasi has joined #opensolaris
[01:00:35] *** coffman has joined #opensolaris
[01:01:07] *** knic has quit IRC
[01:06:21] *** tsoome has quit IRC
[01:08:48] <Auralis> http://blogs.sun.com/alanbur/entry/it_s_official_sun_is1  nice
[01:21:16] *** kb7sqi has quit IRC
[01:21:22] <richlowe> ooooh, beatnik spam on discuss.
[01:22:44] <Error_e^ipi> ?
[01:22:53] *** gm152 has quit IRC
[01:22:59] *** nbkk6fo__ has joined #opensolaris
[01:23:29] <richlowe> Error_e^ipi: oh, the filler text in the spam that just hits osol-discuss reads almost like beat poetry.... at least, if you're really really tired I guess.
[01:23:34] <richlowe> s/hits/hit/
[01:24:50] <agliodbs> richlowe: I got a great e-mail from a Japanese collaborator
[01:24:54] <agliodbs> it read like Haiku
[01:24:57] <agliodbs> hold on ...
[01:26:49] <agliodbs> http://rafb.net/p/PcXqVR57.html
[01:27:48] <agliodbs> that's like my favorite e-mail ever
[01:27:54] <agliodbs> I keep wanting to use that
[01:27:59] <agliodbs> "there was an expression of the will"
[01:28:28] *** Fish- has quit IRC
[01:28:48] <lloy0076> And just like a haiku, well
[01:28:54] <lloy0076> Written it is, but an ending
[01:28:57] <lloy0076> Gives it not...
[01:29:48] <agliodbs> including we were pgadmin upcoming in update4,
[01:30:05] <agliodbs> but 508 compliant it was not and there was and expression of the will
[01:34:14] <richlowe> oh, it's hitting all the lists now, via jive no doubt.
[01:34:16] <richlowe> great.
[01:34:41] <Triskelios> Auralis: but how many of those lines of code are just OO.o? =P
[01:34:51] <Auralis> lol
[01:36:47] *** coffman has quit IRC
[01:37:31] <alanc> I guess this is what we get for complaining about the spam on vconsole-discuss - they spread it to the rest of the lists instead
[01:38:02] <richlowe> alanc: the greater question is Why do we care about filtering the lists, then hand people a gaping hole to shove their stuff into (jive)?
[01:38:21] <alanc> I've never understood why we care about jive at all
[01:38:48] <Error_e^ipi> jive?
[01:39:11] <alanc> it's main purpose seems to be to make it easier to do drive-by participation instead of joining the list and becoming a long-term part of the community
[01:39:26] *** MikeTLive has joined #opensolaris
[01:39:34] <onbot> commit by Michen Chang:  6513238 autofs coredumps when a machine is setup as ldap client
[01:39:36] *** stevel has quit IRC
[01:39:37] <agliodbs> what's jive?
[01:39:40] <alanc> the god-awful web forum crap on opensolaris.org that's gatewayed with the mailing lists
[01:40:01] <agliodbs> well, webforums are good for a business audience
[01:40:14] <richlowe> Yeah, we're seeing.
[01:40:21] <agliodbs> and for people who just want help with one issue and don't know where to start
[01:40:21] <richlowe> as long as you want to buy... whatever the hell this is trying to sell.
[01:40:33] <agliodbs> I've wanted to gateway the postgresql lists for a while
[01:40:36] <Error_e^ipi> oh... I use the web forum on account of I don't want to fill my email box with mailing lists
[01:40:52] <alanc> though I suppose they're better than the "support" webforums on sun.com that go mostly ignored because they're not gatewayed to something like a mailing list to keep up on it
[01:40:53] <agliodbs> forums are more appealing to the casual/temporary user
[01:41:13] <richlowe> my larger complaint is that they're entirely unfiltered, seemingly.
[01:41:21] <richlowe> while they have other issues, at least they're mainly tolerable ones.
[01:41:34] <richlowe> (except the tendency for people to use forum-markup in mail, and render their text unreadable)
[01:42:26] <agliodbs> ah
[01:42:42] <agliodbs> yeah, any forum we had would be authenticated
[01:42:53] <agliodbs> actually, that's one of the things which held us up
[01:43:18] <alanc> websites that require you to go check regularly for discussion are not long-term-participation-friendly - I've never found any forum worth the effort of visiting often enough to take part in
[01:43:21] <agliodbs> the need to synch forum and e-mail subscriber lists
[01:43:44] <agliodbs> alanc: right, but not everybody is in it for the long term, or wants to get 80 mails a day
[01:44:08] <agliodbs> say for example that I'm a sysadmin and I have 11 servers only one of which is opensolaris
[01:44:12] <richlowe> 80? that must be pleasant.
[01:44:16] <agliodbs> and I want to know if there are any backup tools
[01:44:19] <agliodbs> built-in
[01:44:37] <agliodbs> I don't want to join osol-discuss and get 300 emails just to answer that one question
[01:45:02] <richlowe> yeah, especially since we all know you should use star anyway ;)
[01:45:02] *** nbkk6fo__ has quit IRC
[01:45:03] *** nbkk6fo_ has joined #opensolaris
[01:47:03] * steleman sings "Twinkle Twinkle Little Star"
[01:47:25] <Error_e^ipi> hey steleman
[01:47:42] <steleman> hey Error :-)
[01:48:26] <Error_e^ipi> how goes?
[01:48:33] <steleman> it goes ....
[01:49:59] *** mikaeld_ has joined #opensolaris
[01:51:44] *** mikaeld has quit IRC
[01:53:01] <agliodbs> star?
[01:53:09] *** ShadowHntr has joined #opensolaris
[01:53:54] *** Gman has joined #opensolaris
[01:54:19] <delewis> between
[01:54:22] <delewis> er
[01:54:31] * quants reinstalling sol
[01:54:38] <delewis> between "everyone should use star" and ksh93 domination
[01:54:43] <delewis> I don't know what this world is coming to :-(
[01:54:47] <quants> it needs a bigger hdd
[01:55:50] <lloy0076> What's so amazing about ksh?
[01:56:29] <quants> might as well use bash
[01:56:34] <delewis> lloy0076: it's a fairly complex programming language.
[01:56:43] <delewis> associative arrays, etc.
[01:56:56] <delewis> and it maintains POSIX compliance.
[01:57:05] <delewis> which is something Bash does not even do in POSIX-compatibility mode.
[01:57:27] <delewis> its also far more optimal in both memory footprint and performance.
[01:57:34] <quants> don't get me wrong I have a ksh reference
[01:57:39] <delewis> </gisburn>
[01:58:25] <quants> dt's shell is really interesting
[01:58:49] <quants> I refer sh for scripting, or perl
[01:58:54] <quants> prefer*
[01:59:49] <quants> too bad the gnu people can't do for bash what open group did for ksh
[02:00:02] *** peteh has quit IRC
[02:00:03] <quants> gtkbash?
[02:00:22] <delewis> quants: there is something similar to dtksh
[02:00:27] <delewis> I can't recall the name of it at the moment
[02:00:30] <delewis> Solaris ships it
[02:00:30] <quants> really?
[02:00:32] *** peteh has joined #opensolaris
[02:00:33] <delewis> ugh, what is it called..
[02:00:36] <delewis> zenity
[02:00:47] <richlowe> zenity and dtksh aren't even vaguely alike...
[02:00:48] <quants> nahunnawha?
[02:00:55] <delewis> richlowe: really?
[02:00:57] <delewis> never used it
[02:01:12] <delewis> I've just heard it being compared by the JDS folk on the mailing lists to dtksh.
[02:01:21] <quants> dtksh is genius
[02:01:30] <delewis> and zenity is?
[02:01:40] *** sahafeez has quit IRC
[02:01:48] <quants> no idea what zenity is
[02:02:04] *** adcsi1123 has joined #opensolaris
[02:02:05] <alanc> zenity allows displaying gtk dialogs from shell scripts
[02:02:07] * Gman never compared the two fwiw
[02:02:18] <Tpenta> so how was the keynote glynn?
[02:02:23] *** nachox has joined #opensolaris
[02:02:24] <delewis> Gman: sorry, might've been someone else :-)
[02:02:25] <alanc> ah there he is - was waiting for Gman to pop up
[02:02:33] <richlowe> hah, pop up.
[02:02:37] <nachox> hey guys
[02:02:42] <quants> holla
[02:02:51] <Gman> Tpenta, was pretty interesting - most of the crowd didn't like it
[02:03:06] <Tpenta> he was controvesioal?
[02:03:09] <quants> sux to be them huh?
[02:03:09] *** pitty has joined #opensolaris
[02:03:20] <pitty> anyone using pconsole on solaris 10?
[02:03:27] <Gman> well, he just described minix basically
[02:03:31] <Gman> and talked about reliability
[02:03:32] <alanc> "Zenity is a tool that allows you to display Gtk+ dialog boxes from the command line and through shell scripts. It is similar to gdialog, but is intended to be saner. It comes from the same family as dialog, Xdialog, and cdialog, but it surpasses those projects by having a cooler name."
[02:03:37] <delewis> Gman: not surprising :-)
[02:03:54] <Tpenta> so, since it was not linux, the crowd disapproved?
[02:04:05] <Gman> basically talked about separating things up to improve reliability
[02:04:12] <Gman> can't really fault him for that
[02:04:19] <Tpenta> agreed
[02:04:31] <Gman> but since it's not linux, don't think anyone agreed :)
[02:04:48] <Gman> http://www.linux-watch.com/news/NS8979755794.html
[02:04:50] <Gman> ah fun ;)
[02:05:00] <Gman> amusing reading the lwn comments too to that story
[02:05:32] <delewis> http://www.osnews.com/story.php/16973/Sun-To-License-OpenSolaris-Under-GPLv3
[02:05:43] <delewis> I was wondering when that would appear
[02:05:44] <Tpenta> sjvn is there?
[02:05:49] <Tpenta> oh that
[02:05:52] <Gman> yeah
[02:05:53] <quants> what was that quote about the best attack on great minds being the stupidity of small minds?
[02:05:56] *** pitty has left #opensolaris
[02:06:01] <Auralis> they can stick the gplv3 where the sun doesn't shines
[02:06:12] <Gman> just sat in on the kernel report from jonathan corbet
[02:06:20] <Gman> funny that they're facing very very similar problems
[02:06:25] <Gman> their release schedule is a little crazy
[02:06:33] <Gman> a major kernel every 3 months
[02:06:50] <Gman> and funnily enough, they're experiencing a slower buggier kernel
[02:06:51] <Gman> go figure
[02:06:58] *** axxl has joined #opensolaris
[02:07:19] <richlowe> Someone really should respond to the GPLv3 articles.
[02:07:33] <Gman> and say what?
[02:07:40] <richlowe> "If they are, nobody told us"
[02:07:48] <Gman> heh
[02:07:49] <Gman> exactly
[02:08:03] * Gman has absolutely no clue what might be being talked about in dark corridors
[02:08:04] <lloy0076> How about: You got it wrong. It's actually the Artistic License we were thinking of...
[02:08:05] <lloy0076> :P
[02:08:08] <Tpenta> the funny thing is that they are all saying we are doing it, but no-one has been able to get anyone from sun to comment
[02:08:09] *** postwait__ has quit IRC
[02:08:36] <nachox> i cant believe people still  read Steven J. Vaughan-Nichols, he is a joke as a journalist
[02:08:43] <delewis> and from what I've heard the CAB members don't know anything, either.
[02:08:50] * Gman imagines it's a somewhat proverbial hot potato ;)
[02:09:09] <delewis> nachox: we have a similar journalist by the name of John C. Dvorak.
[02:09:13] *** axxl has quit IRC
[02:09:22] * delewis wonders what it is the nonsense journalists and fancy names
[02:09:33] <delewis> with*
[02:10:38] *** axxl has joined #opensolaris
[02:10:51] <nachox> ahh, yes, that one too
[02:11:09] *** axxl has quit IRC
[02:11:10] <delewis> even Slashdot doesn't read Dvorak :-)
[02:11:13] <alanc> thought the powers-that-be had gotten the message when jonathan and rich did the "and what about solaris?" bit at the GPL announcement that this is not something to do unilaterally without talking to the community
[02:11:16] <lloy0076> delewis: Just like pr0n stars, when they become journos they change their names to become sexy^Wnewsworthy...
[02:11:41] <delewis> lloy0076: :-)
[02:11:45] <alanc> "Robert X. Cringely"
[02:11:50] *** abackos has joined #opensolaris
[02:12:14] <nachox> hehe, top 10 of the worst it journalists :P
[02:12:20] <alanc> (err, that was supposed to be "at the Java/GPL announcement" above)
[02:12:26] <jmcp> alanc: I've often wondered whether Cringely's middle name is Xavier, or he just can't spell it so uses an X isntead
[02:13:10] <alanc> I thought it stood for "I couldn't think of anything better when I made up this fake name"
[02:13:37] <Tpenta> grumble cringlye, grumble non-existant journalistic ethics, grumble
[02:13:45] <alanc> http://en.wikipedia.org/wiki/Robert_X._Cringely
[02:13:52] *** xusheng has joined #opensolaris
[02:14:31] <nachox> and people actually pay them to write
[02:15:28] <jmcp> I'd comment on that piece of SJVN drivel on lwn but I can't be bothered signing up
[02:15:49] <nachox> which makes me wonder, i mean, were talking about them, that means they are effective at what they do, it's not important if we say positive or negative things about them, we talk about them
[02:15:59] <lloy0076> We should start "Microsfot are going to license Vista under the CDDL rumours."
[02:16:16] <lloy0076> Microsoft too
[02:16:24] <nachox> jmcp: that stopped me the last n times too, hehe
[02:17:53] <jmcp> I'll blog instead
[02:18:03] <quants> MS licensed Vista under the CDDL, in other news paint peels...
[02:19:08] <quants> tux stuck his little pee pee into Vista, and left behind GPL code
[02:19:27] <quants> new outlets world wide shudder
[02:19:30] *** triplah_ has quit IRC
[02:19:39] <lloy0076> Weapons of Mass Destruction were found in the USA...
[02:19:40] <quants> news*
[02:19:43] * lloy0076 did I say that?
[02:20:10] *** simford has joined #opensolaris
[02:21:17] <Gman> http://planet.lca2007.linux.org.au/ - if anyone wants to follow conference blogs for linux.conf.au
[02:21:43] *** adcsi1123 has left #opensolaris
[02:21:48] <Error_e^ipi> lloy0076: you know nothing of american propaganda tactics
[02:21:51] <Error_e^ipi> try it this way:
[02:22:04] <Error_e^ipi> "Are there weapons of mass distruction in the USA?"
[02:22:06] <Error_e^ipi> ...
[02:22:15] *** triplah_ has joined #opensolaris
[02:22:34] <Error_e^ipi> if you ask a question, you're not lying :)
[02:23:06] <jmcp> so what's an appropriate blog post title? SJVN's head is stuck where the Sun doesn't shine   ??
[02:23:06] <lloy0076> We know they have nuclear warheads.
[02:23:11] <lloy0076> :P
[02:23:14] * lloy0076 shuts up
[02:23:31] <lloy0076> jmcp: Sun's Silence Blinded SJVN to the Truth
[02:23:59] <quants> lloy0076: right if you can call that lil thing a penis
[02:24:03] <delewis> you seriously don't think that even SJVN made up the GPLv3 do you?
[02:24:34] <delewis> are mis-reported it from the Java GPL announcement when Schwartz and the other guy were asked about Solaris being GPL'd
[02:24:37] <quants> lloy0076: who is our "warhead"
[02:24:38] <delewis> s/are/our/
[02:24:48] <delewis> er or*
[02:25:16] <jmcp> delewis: SJVN once again reports rumour as fact.
[02:25:25] <delewis> jmcp: yeah.
[02:25:26] <jmcp> delewis: eweek, once again, reporting rumour as fact
[02:25:32] <delewis> given the CAB and no one seems to know
[02:25:34] <delewis> that seems logical
[02:26:10] *** adcsi1123 has joined #opensolaris
[02:26:32] <nachox> anyway, gplv3 is a draft atm, we will have an ogb when the licence is ready
[02:26:51] *** steleman has quit IRC
[02:27:24] <delewis> yeah, I highly doubt Sun legal would decide to go with a license that's currently not past the draft-point.
[02:29:05] <quants> rumor mill: I hear Cingular is being killed by its rollover
[02:29:15] *** xusheng is now known as Steven_Hou
[02:29:31] <jmcp> quants: rollover to Apple for the iphone?
[02:29:47] <quants> jmcp: ?
[02:29:52] <dwc-> no, rollover minutes
[02:30:02] <dwc-> unused minutes "rollover" into the next month
[02:30:09] <dwc-> up to a max number of minutes
[02:30:49] <quants> they have yet to expire any of my minutes
[02:30:56] <dwc-> how many do you have
[02:30:57] <jamesd> if sun legal went for that, it will be time for that guy to turn in his crackerjack law degree
[02:31:21] <dwc-> and how long have you had them
[02:31:26] <quants> dwc-: I use them once they get too big
[02:31:26] <nachox> well, it would certainly be a bold move
[02:31:59] <jmcp> hey, does SJVN call himself a journalist?
[02:32:08] <dwc-> this page says rollover minutes expire after 12 billing periods
[02:32:15] <quants> dwc-: true
[02:32:28] <quants> dwc-: I use them once they get too big
[02:32:39] *** adcsi1123 has left #opensolaris
[02:35:29] *** Steven_Hou has quit IRC
[02:38:18] *** steleman has joined #opensolaris
[02:39:35] * jmcp posts
[02:39:48] *** adcsi1123 has joined #opensolaris
[02:40:07] *** Gman has quit IRC
[02:43:07] *** deedaw has quit IRC
[02:44:25] <Error_e^ipi> posts as in "check your memory & registers" ?
[02:44:43] *** CosmicDJ has quit IRC
[02:46:47] <kAv_> has anyone made bootable dos cd's with extra filesystem in solaris with mkisofs?
[02:48:57] *** rachel_ has joined #opensolaris
[02:55:55] *** adcsi1123 has left #opensolaris
[03:00:30] *** gm152 has joined #opensolaris
[03:01:13] *** Dar has joined #opensolaris
[03:08:02] *** abackos has quit IRC
[03:26:40] *** jteo has quit IRC
[03:28:22] *** comay has quit IRC
[03:34:44] *** ShadowHntr has quit IRC
[03:36:00] *** edwardocallaghan has joined #opensolaris
[03:36:07] <edwardocallaghan> Hi
[03:36:15] <jbk> evening
[03:36:20] <jbk> or morning
[03:36:47] <edwardocallaghan> s/morning/evening/k
[03:39:02] * boyd is having trouble with his vision today. Can anyone tell me what the "news" actually is on this page? http://www.sun.com/launch/2007-0116/feature.jsp?intcmp=hp2007jan16_launch
[03:39:50] <boyd> I can't see anything that isn't in the Sol 10 u3 release
[03:40:19] <edwardocallaghan> Hi boyd
[03:40:22] <richlowe> maybe it's a very very late bit about u3?
[03:40:22] <jsubl2> looks that way to me too boyd
[03:40:31] <edwardocallaghan> Nothing they are just summing things up
[03:40:39] <richlowe> either way it's marketing, so sod that.
[03:41:01] <edwardocallaghan> Maybe they plan to drop out another product soon ?
[03:41:08] <edwardocallaghan> lol
[03:41:21] <edwardocallaghan> boyd: How are you?
[03:41:56] <richlowe> not much that could cause that all to make sense other than an update release.
[03:41:59] <richlowe> so probably u3.
[03:42:01] <boyd> edwardocallaghan Yeah, ok... I've been a bit crook for about 3 weeks
[03:42:06] <richlowe> (and the LDOM firmware is *still* not around, is it?)
[03:42:40] <boyd> I think it's just an ad wrapped up as news. For people who missed u3 before the holidays
[03:43:06] <edwardocallaghan> boyd: I got my ticket
[03:43:18] <edwardocallaghan> I will be there in less then two weeks
[03:43:57] <boyd> Woo hoo.. Don't cook when you get here
[03:44:14] <edwardocallaghan> Why ?
[03:44:19] 
[03:44:26] <edwardocallaghan> Sorry?
[03:44:29] <edwardocallaghan> Oh yes, nice
[03:44:36] <boyd> I mean I hop you personally don't get cooked
[03:44:43] <boyd> hope
[03:44:47] <edwardocallaghan> lol I'll be find
[03:44:51] *** nbkk6fo_ has quit IRC
[03:44:54] <edwardocallaghan> s/find/fine/
[03:45:05] *** nbkk6fo_ has joined #opensolaris
[03:45:21] *** jwk404 has quit IRC
[03:45:21] *** Tpenta has quit IRC
[03:45:38] <edwardocallaghan> What do you think of Japan airlines ?
[03:45:58] <boyd> I've only flown them once. 5/10
[03:46:16] <boyd> Stric: Where are you?
[03:46:19] *** Tpenta has joined #opensolaris
[03:46:24] <boyd> Hi Tpenta
[03:46:25] <Stric> edwardocallaghan: you still not decided which airline? just pick one who isn't ryan air and is kinda cheap..
[03:46:38] <Stric> boyd: northern sweden.. close to the polar circle ;)
[03:46:41] <Tpenta> grrr, something just crashed my gaim client
[03:46:42] <boyd> I'm guessing that since he has the tickets he's chosen the airline
[03:47:01] <boyd> Tpenta: I thought it didn't need something else to crash it :)
[03:47:32] <edwardocallaghan> Yes boyd is right I have already got my tickets
[03:47:39] <edwardocallaghan> As I said ^
[03:47:40] <Tpenta> identify ah-alan
[03:47:47] *** ChanServ sets mode: +o Tpenta
[03:47:51] <Tpenta> arg
[03:47:56] <Error_e^ipi> heh
[03:48:01] <edwardocallaghan> I was just wondering if I am going to crash ?
[03:48:04] <Stric> good one.. now change your password ;)
[03:49:08] * Tpenta changes his nickserv passwd (again)
[03:49:17] <boyd> hehe
[03:49:22] <nachox> hehe
[03:50:33] * boyd wonders how people can say that a new license will be used when it doesn't exist yet
[03:50:34] <nachox> just use op rights when you need them, it is not only a good idea but the recommended freenode policy
[03:52:50] <g4lt-U60> nachox, and said policy doesn't do any good agains irc spammers
[03:53:05] <richlowe> it does, if services is there.
[03:53:12] <g4lt-U60> IF
[03:53:15] <richlowe> such thinking really didn't survive the period where services was down more often than not.
[03:53:26] <g4lt-U60> irc spammers usually try to DOS services
[03:53:53] *** abackos has joined #opensolaris
[03:54:09] <nachox> g4lt-U60: channels like ##slackware live by that rule and have very few problems
[03:54:34] <edwardocallaghan> Be back later...
[03:57:07] *** MikeTLive has quit IRC
[03:57:51] <edwardocallaghan> boyd:stick around as you do, I would like to have a chat later tonight
[03:58:40] <lloy0076> heh
[03:58:56] <lloy0076> Sometimes I think I should have a nick called "chanser" and the like...
[03:59:01] <lloy0076> Imagine all the mistells you'd get.
[03:59:34] <minerale> So sun organised a conference in atalnta today (and tomorrow) and a bald guy who's in charge of opensolaris gave a speech, among the things they mentioned something about a OpenSolaris Developer Edition (express) and a new package managing system
[03:59:44] <minerale> Atlanta*
[04:00:16] <minerale> Can't remember his name
[04:01:09] *** edwardocallaghan has left #opensolaris
[04:01:35] <nachox> glenn faden? :P
[04:01:54] *** adcsi1123 has joined #opensolaris
[04:06:45] <minerale> Otherwise they gave a pretty nice (day long) presentation complete with a 3000$ system giveaway, all kinds of features, they took notes too. After the new package management (and patach management) system was mentioned this one guy kept mentioning Gentoo's portage and why it's soo good, one of the fellows there took notes, actually and promised to look into it
[04:07:01] <Error_e^ipi> lol
[04:07:06] <Error_e^ipi> portage can't even remove packages
[04:07:12] <Error_e^ipi> give me a break
[04:07:29] *** peteh has quit IRC
[04:08:04] *** peteh has joined #opensolaris
[04:08:09] <lloy0076> portage is a pile of stinking...
[04:08:12] <lloy0076> I'd prefer apt
[04:08:40] <minerale> WellOh and another thing coming out: community build farms, zones assigned to various developer wanting to install solaris, play around and build their things on it
[04:08:52] <g4lt-U60> I'd prefer proper ports trees
[04:09:18] <lloy0076> g4lt-U60: I'm convinced there was a guy who was proposing a community to do something like that a while back on the OpenSolaris discuss list.
[04:09:48] <Error_e^ipi> minerale: solaris runs on junk hardware, who needs build farms?
[04:09:49] <nachox> i doubt there is a way out from sysv packages that mantains backwards compatibility
[04:10:15] *** edwardocallaghan has joined #opensolaris
[04:10:16] *** jimgris has joined #opensolaris
[04:10:17] <g4lt-U60> nachox, technically, ports trees build sysv packages and install them
[04:10:44] <edwardocallaghan> Think I may go to sleep, not as hardcore as I use to be
[04:10:56] <minerale> error_e^ipi: for developers who don't have access to sparc & higher end processors
[04:11:03] <lloy0076> edwardocallaghan: You'll be fine once you get to Australia.
[04:11:09] <g4lt-U60> and it wouldn't be too hard to make PORTfoo overrid SUNWfoo
[04:11:12] <lloy0076> edwardocallaghan: It's about 1:40 PM here.
[04:11:39] <Error_e^ipi> minerale: old ultra2's and 5's are cheap as well
[04:12:54] <edwardocallaghan> Yea, my body clock keeps shifting to Australia time :D
[04:13:17] <edwardocallaghan> I am trying to keep it with GMT at the moment :p
[04:14:28] <minerale> error_e^ipi: I personaly pay a monthly fee for a xen domU to run whatever on, I've found it easier to run everything remotely (~20 gnu screen sessions) anyhow, that's what they said they are bringing out (I believe sourceforge provides server farms too)
[04:15:01] *** abackos has quit IRC
[04:15:42] *** brs has quit IRC
[04:20:15] <edwardocallaghan> Edward switches to Run level -- 'Bed time'
[04:20:34] <alanc> minerale: bald guy in charge of opensolaris is probably Stephen Harpster, who I know is in Atlanta this week
[04:21:03] <alanc> while glenn faden is bald, he's just Trusted Extensions, not all of OpenSolaris
[04:21:25] <edwardocallaghan> Night
[04:21:26] <minerale> is glenn faden one of the fellows ?
[04:21:27] *** edwardocallaghan has left #opensolaris
[04:22:04] <minerale> Ah yes, Stephen Harpister (just googled his photo) that's him
[04:24:02] <minerale> if you work with him tell him good job (from one of the attendees)
[04:24:50] <lloy0076> Dilbert: How did you know that was Stephen Harpster and not Glen Faden? PHB: He has 73.24% baldness, Glen Faden has 73.10% baldness...
[04:24:59] *** alfism_ has joined #opensolaris
[04:25:27] <alanc> Harpster is my boss'es boss
[04:25:37] <minerale> Ok, there were two bald guys actually, one who's a fellow (sp?) and more on the technical side, and Harpster who's more of a business person
[04:25:38] <alanc> so I see him around
[04:26:05] <alanc> not sure who the fellow is
[04:26:36] *** linma has quit IRC
[04:28:53] <boyd> Does anyone know what bug state "Accepted' means? Does it imply progress? Allocation to an engineer?
[04:29:33] <Error_e^ipi> "outcast" -> "tolerated" -> "accepted" -> "king of all that is!"
[04:29:38] <Error_e^ipi> *nod*
[04:29:50] <alanc> not much - mostly that someone took a quick glance and decided it wasn't obviously NOTABUG or obviously assigned to the wrong place
[04:30:07] <alanc> it may be assigned to an engineer
[04:30:28] <boyd> 'kay, thanks
[04:32:59] *** jimgris has left #opensolaris
[04:34:09] <alanc> hmm, of the Sun Fellows I've seen pictures of or met in person, James Gosling is the only bald one I know
[04:34:26] <alanc> (had to search hard to find the list of Sun Fellows)
[04:35:06] <alanc> but of the 13, I can only recognize 4 by sight
[04:35:13] * boyd can't help but wonder why alanc went looking for a list of Sun Fellows so that he could count the bald ones
[04:35:18] <richlowe> 1-Dispatched - default state. 2-Incomplete - there's not enough information there, 3-Accepted - Yup, that's a problem, 4-Defer - ... I dunno, "not yet"(?), 5-Cause Known.- We know what the problem is, 6-Fix Understood - ... and we know how to fix it, 7-Fix In Progress ... and we are fixing it, 8-Fix Available - We have the fix, 9-Fix Failed - ...and it didn't work, 10-Fix Delivered - the fix is in a build, 11-Closed - closed.
[04:35:24] <richlowe> or something to that effect.
[04:35:34] <alanc> though the picture of the newest one is clearly not bald (Radia Perlman)
[04:35:59] <alanc> boyd: earlier discussion about someone talking at TechDays today who was bald and was a Sun Fellow
[04:36:00] <boyd> richlowe: thx
[04:36:06] <boyd> Ah
[04:36:12] <richlowe> boyd: well, wait to see if either of the Alan's correct me, first. ;)
[04:36:16] <boyd> :)
[04:36:33] <alanc> "Highly prestigious, the title of Sun Fellow is the most senior rank that can be attained within Sun's engineering organization. Sun Fellows help guide Sun's technical direction, identify new opportunities and advise the management team on technical issues. Radia's appointment adds her to the elite group of Fellows at Sun?Nick Aneshansley, Richard Dee, Whitfield Diffie, James Gosling, Jim Hughes, Tim Marsland, Jim Mitchell, Mike Splain, B
[04:36:33] <alanc> ob Sproull, Guy Steele, Ivan Sutherland and Marc Tremblay."
[04:37:11] <alanc> Tim Marsland is the one I'd most expect talking about Solaris, but he wasn't bald the last time I saw him
[04:37:33] <movement> xen would send most people bald... he's made of sterner stuff
[04:37:41] <boyd> lol
[04:38:07] <alanc> richlowe: that's pretty much it
[04:38:14] <movement> despite the easy fix, http://lists.xensource.com/archives/html/xen-devel/2006-11/msg00482.html is still my favourite so far
[04:38:56] <alanc> Defer is usually used for "we won't just close as WONTFIX, but aren't going to fix it for a while for some reason (lack of resources, conflict with other plans, waiting for an upcoming project, etc.)"
[04:40:47] <minerale> This guy used to work heavily into storage, carried around a mac and a little black notebook
[04:43:01] *** axxl has joined #opensolaris
[04:43:24] <alanc> not sure then
[04:45:16] *** abackos has joined #opensolaris
[04:50:19] *** minerale has quit IRC
[04:50:22] *** minerale has joined #opensolaris
[04:51:41] *** edwardocallaghan has joined #opensolaris
[04:51:44] <edwardocallaghan> Hi
[04:52:00] <edwardocallaghan> Can't get to sleep, quick question
[04:52:12] <edwardocallaghan> How does ZFS get around endianness ?
[04:52:18] *** UnixTitan has joined #opensolaris
[04:52:44] <boyd> It tags metadata blocks with the endianness they were written with. Reads either writes native
[04:52:47] <jmcp> edwardocallaghan: it was designed from the start to not be endian-specific
[04:52:54] <jmcp> it's in the design docs on opensolaris.org
[04:53:15] <jamesd> edwardocallaghan, each block had a bit in it that specifies the block if it matches the hardware it does nothing, if the endianess doesn't match zfs converts it...
[04:53:39] <jamesd> thus the more you write to a filesystem the more it matches the current system.
[04:54:42] <jbk> though, that does nothing for the contents of any particular file (veritas tries to imply that a similar feature for vxvm does this)
[04:55:57] <boyd> I can imagine that VxVM can do that for the disk group metadata. I can't imagine it works on FS data the same way
[04:57:01] <nachox> edwardocallaghan: http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/common/fs/zfs/zfs_byteswap.c
[04:57:16] <jbk> diskgroup metadata + fs metadata
[04:57:33] <jbk> but they try to imply it will magically do it for fs data too (which of course it cannot)
[04:58:09] <boyd> To even do FS metadata they would have to parse the whole FS. VxVM would not do that. VxFS may.
[04:59:29] <jmcp> boyd: not to my knowledge
[04:59:42] <edwardocallaghan> Thanks all
[05:01:26] *** Gman has joined #opensolaris
[05:02:04] <quants> cmihai: you about?
[05:02:10] <Error_e^ipi> hey gm152
[05:02:11] <Error_e^ipi> err,
[05:02:14] <Error_e^ipi> hey Gman
[05:02:17] <Error_e^ipi> stupid tab complete
[05:02:35] <edwardocallaghan> OK Night night
[05:02:37] *** edwardocallaghan has left #opensolaris
[05:02:41] <Gman> howdy Error_e^ipi
[05:02:54] <Triskelios> how's LCA going?
[05:03:00] *** alanc is now known as alanc-away
[05:06:03] <Gman> it's been good
[05:06:14] <Gman> just sitting in dave airlie's talk about nouveau
[05:06:24] <Gman> previously watched raster talk about e17
[05:06:44] <boyd> Ah... so there is some activity on the e17 front then?
[05:06:51] <Gman> sure
[05:07:01] <Gman> it's like at 0.16.999.37 or something :)
[05:07:07] <Gman> it's looking pretty sweet
[05:07:10] <Gman> and fast, obviously
[05:07:13] <boyd> No doubt.
[05:07:32] * boyd wonders if they'll ever get over adding cool stuff to make, you know, a release :)
[05:07:43] <Gman> bit easier to keep to those programming principles when you're basically the only one writing the code :)
[05:08:26] *** adcsi1123 has left #opensolaris
[05:08:50] *** freakazoid0223 has quit IRC
[05:09:11] <jsubl2> any url
[05:09:25] <jmcp> this dates me.... every time I see/hear about E17 I think of the boyband East17 from the early 90s
[05:09:32] <boyd> EGAD!
[05:09:37] <jsubl2> any url's on how  to get e17 going on the current sxcr
[05:09:55] * boyd feels the need to shower after that jmcp comment
[05:10:04] <jmcp> boyd: hehe ... J has their album :)
[05:10:33] <boyd> jsubl2: benr used to be the guy for that. I think he's too busy these days
[05:10:46] <jsubl2> cuddletech?
[05:10:51] <boyd> The same
[05:10:57] <jsubl2> i think i have that url somewhere
[05:11:43] *** gm152 has quit IRC
[05:11:44] *** freakazoid0223 has joined #opensolaris
[05:15:58] *** nachox has quit IRC
[05:22:38] *** salmandr has joined #opensolaris
[05:24:11] *** jsubl2 has quit IRC
[05:27:37] *** Gman has quit IRC
[05:29:15] *** MikeTLive has joined #opensolaris
[05:29:54] *** nbkk6fo_ has quit IRC
[05:30:55] *** MikeTLive has quit IRC
[05:31:30] *** MikeTLive has joined #opensolaris
[05:32:46] *** danmorg has joined #opensolaris
[05:32:58] <danmorg> hi
[05:34:30] <lloy0076> Howdy
[05:34:38] <danmorg> when will dhcp be supported out-of-the-box on opensolaris?  right now, i tried build 53 and it never asks.  it just ask for my default router, my ip address, my dns servers, etc... but nowhere did i see it asking for dhcp
[05:35:07] <jmcp> danmorg: it's been supported out of the box since, uh .... Solaris 7 iirc
[05:35:08] <Error_e^ipi> then you weren't paying attention
[05:35:33] <Error_e^ipi> *shrug*
[05:35:40] <Error_e^ipi> it asks me for DHCP or not
[05:35:45] <lloy0076> danmorg: What type of DHCP server are you using? My *crappy* ADSL2+ modem has a DHCP server but it's so crap that Solaris sometimes won't see it there...
[05:36:31] <jmcp> lloy0076: that's not the question that danmorg is asking
[05:36:32] <danmorg> linksys dsl/cable router
[05:37:08] <lloy0076> jmcp: I was under the impression that if it thought it wasn't there it would silently drop you to the questions danmorg says he's being asked.
[05:37:52] <boyd> It asks me every time and there's definitely not DHCP server here
[05:38:18] <lloy0076> I'll have to look again when I get home.
[05:40:13] *** laca has quit IRC
[05:43:47] <danmorg> i tried open solaris build 46 (using it now) and the gu installer ran and installed sort-of-okay.  but build 53 the gui installer fails, so it falls back to text mode installer.  was there a regression here or a dummy like me should just stick with stable releases?   i read a blog that build 53 included hal support for devices... i wanted to try listening to my cds from solaris.
[05:43:59] <Theoden-Nexenta> http://www.eweek.com/article2/0,1895,2084284,00.asp
[05:44:44] <Error_e^ipi> Theoden-Nexenta: and that's been passed around to death... is there any official confirmation of anything of the sort?
[05:45:41] <Tpenta> not that i'm aware of
[05:46:02] <Error_e^ipi> that's what I thought
[05:46:17] <Error_e^ipi> it's just the linux sharks wishing sun would let them take all the code they want on their terms
[05:46:36] <Theoden-Nexenta> Error_e^ipi: don't know yet.
[05:47:41] <Tpenta> though simon phipps HAS been saying the gplv3 looks to be heading to somewhere that we'd consider using it
[05:49:07] <danmorg> with the CDDL, can you take code and make it closed source and sell it if you want to?
[05:49:22] <jmcp> I don't believe so
[05:49:27] <boyd> Tpenta: Yes, I think that's as far as anything's officially got. But untill GPLv3 exists it's rather acedemic, surely
[05:49:50] <Tpenta> that's what i'd think boyd
[05:49:51] <jmcp> danmorg: if you license something under CDDL then it's open. I don't think you have to include everything that you want to deliver under CDDL though
[05:49:56] <Error_e^ipi> my main gripe with that move would be "linux compatible license" more than anything else.... solaris evangelism is hard enough without linux taking & halfassing the compelling features that solaris has
[05:49:58] <jmcp> danmorg: because CDDL is file based not project based
[05:50:33] <boyd> Error_e^ipi: But *is* it linux compatible? What with them being GPLv2?
[05:50:36] <danmorg> ok.  i prefer MIT/X11 and BSD licenses myself.... true freedom
[05:50:52] <jmcp> danmorg: don't quote me, and make sure you check out the FAQs @ opensolaris.org. I'm a software engineer and definitely not a lawyer
[05:50:57] <Error_e^ipi> "why should i use solaris" "zfs" "oh neat..." versus "why should i use solaris?" "more stable, faster" "i've never had problems" "nice anectdote"
[05:51:05] <jmcp> danmorg: please leave the politics of "free"/"freedom" somewhere else
[05:51:38] <jmcp> Error_e^ipi: "no, you've never been trying to run multiterabyte datawarehouses on parallel clusters either"
[05:51:55] * jmcp looks for the mop+bucket to clean up after the pissing contest
[05:52:02] <boyd> #endless_discussion_about_relative_merits_of_software_licenses may be a place to look
[05:52:13] <Error_e^ipi> meh, whatever
[05:52:16] <Error_e^ipi> more jobs for me then
[05:52:31] <danmorg> hehe
[05:54:00] <danmorg> any idea why the gui installer would fail trying to install solaris build 53 and resort to text mode installation?
[05:59:32] <lloy0076> danmorg: Because it can't decide if it fits the GPLv3 requirements...
[05:59:35] * lloy0076 ducks for cover
[06:00:00] <lloy0076> The most likely answer lies in that it can't detect your graphics card, mouse or can't determine it can safely run on your monitor.
[06:01:35] *** whaq has quit IRC
[06:03:10] <Error_e^ipi> that never stopped it from locking up my video card on my laptop
[06:06:26] <jmcp> boyd: you any good with sed?
[06:06:41] <boyd> So so (I'm not a guru)
[06:06:42] <Tpenta> james, what are you trying to do?
[06:06:54] <jmcp> PMs coming
[06:10:09] *** salmandr has quit IRC
[06:11:03] <danmorg> i guess i'll wait for another build.   both solaris 10 and SX build 46 were able to install via the gui installer.  :-(
[06:11:06] *** danmorg has quit IRC
[06:12:05] *** abackos has quit IRC
[06:14:24] *** Gman has joined #opensolaris
[06:24:54] *** yydzero has joined #opensolaris
[06:28:16] *** MikeTLiv1 has joined #opensolaris
[06:31:03] *** mritun_ has joined #opensolaris
[06:31:19] <mritun_> hello folks
[06:42:37] *** MikeTLive has quit IRC
[06:45:13] *** Gman has quit IRC
[06:49:51] *** lloy0076 has quit IRC
[06:55:50] *** lloy0076 has joined #opensolaris
[06:59:46] *** ShadowHntr has joined #opensolaris
[07:07:54] *** yongsun has joined #opensolaris
[07:08:19] *** bougie has joined #opensolaris
[07:08:30] *** ilyxa has joined #opensolaris
[07:14:07] *** Viking667 has joined #opensolaris
[07:14:18] <Viking667> hi all. Finally found out how to mount a cd (mount -F hsfs /dev/device /place/to/mount)
[07:15:38] <jmcp> uh .... didn't you think to ask here?
[07:18:54] *** lloy0076 has quit IRC
[07:19:19] <Viking667> Well, I asked SOMEwhere...
[07:20:09] <Viking667> someone in some solaris channel mentioned hsfs to me... at the time I thought they were talking about the Mac... then I found /usr/lib/fs/* and realised that hsfs stands for HighSierra File System...
[07:20:39] <Error_e^ipi> that's HFS+
[07:20:45] <Error_e^ipi> the mac filesystem
[07:20:50] <Viking667> yeah. So I realised.
[07:20:51] <Error_e^ipi> it doesn't work with solaris
[07:20:57] <Error_e^ipi> but i have an iPod that wants it to
[07:21:08] <Viking667> Neither does ext2/3,XFS,JFS,Reiser, etc etc.
[07:21:20] <Error_e^ipi> and XNU (osx's kernel) is a cddl compatible license
[07:21:42] <Error_e^ipi> well, I don't care about ext/ricerfs so that's someone elses' problem
[07:21:54] *** ilyxa has quit IRC
[07:21:56] <Viking667> XNU ?
[07:22:02] <Error_e^ipi> osx's kernel
[07:22:07] * Viking667 wonders what that is - he's only seen Darwin.
[07:22:28] <Error_e^ipi> darwin is XNU + apple's libc + the bsd userspace + some bsd api stuff
[07:22:35] <Error_e^ipi> xnu is just the kernel
[07:22:54] <sickness> Error_e^ipi: *less* all the useful kext's ;P
[07:22:56] <Viking667> ahh.
[07:23:11] <Error_e^ipi> sickness: doesn't matter, HFS is in xnu, not one of the kexts
[07:23:22] <Viking667> err, so what happened to the Mach kernel? Or am I mistaking that for GNU's Hurd?
[07:23:26] <sickness> yeah
[07:23:29] <dlg> maybe ipods will use zfs soon
[07:23:44] <Error_e^ipi> xnu is a crazy mach3 freebsd amalgam
[07:23:55] <Viking667> Anyhow, now I'm able to mount the disk, and update packages manually. Whew.
[07:24:06] * Viking667 happy happy happy.
[07:24:12] <Error_e^ipi> dlg: god, i could only wish
[07:25:06] <Viking667> only one more problem left. My monitor doesn't want to display the picture that X puts out through the video card (Radeon 7000, supported by Xorg)
[07:25:49] *** Theoden-Nexenta has left #opensolaris
[07:27:31] *** slowhog has joined #opensolaris
[07:27:54] <Error_e^ipi> it's kinna cute how apple's seeking UNIX03 certification for osx 10.5
[07:28:11] <Error_e^ipi> it's like a little girl dressed up in her mom's dress with a face covered in lipstick
[07:28:19] <Error_e^ipi> "aww... does little osx want to play unix too? "
[07:30:53] <Viking667> hm. something wedged.
[07:31:08] *** Viking667 has quit IRC
[07:39:32] *** alobbs has quit IRC
[07:46:06] *** ShadowHntr has quit IRC
[07:48:43] *** kb7sqi has joined #opensolaris
[07:53:00] *** bougie has quit IRC
[07:55:51] *** Gman has joined #opensolaris
[07:56:47] <richlowe> 'ello again Gman
[07:56:52] <Gman> howdy
[07:56:59] <Gman> sorry, ping ponging in and out while at the conf ;)
[08:04:22] * _syphilis_ wonders if there are any useful documents for optimising networking programs on solaris
[08:06:03] <jmcp> hi Gman
[08:06:12] <Gman> hey james
[08:07:29] <jmcp> Gman: did you have a good afternoon @ the conf?
[08:07:56] <Gman> yep, all good
[08:07:59] <Gman> looked at ubuntu's upstart
[08:08:06] <Gman> then keithp's hotplugging in x
[08:08:11] <Gman> both interestx
[08:08:27] <richlowe> the quick look I've taken at upstart, it seemed pretty well done, given the constraints.
[08:08:30] <richlowe> though it was a rather brief look.
[08:08:32] <jmcp> what's upstart?
[08:08:40] <Gman> yeah, seems to be cool
[08:08:43] <richlowe> jmcp: Ubuntu's init replacement.
[08:08:50] <Gman> some interesting challenges of making it event based
[08:09:07] <jmcp> ah
[08:09:10] <richlowe> yeah, and it's near impossible for him to be as invasive as SMF is, either.
[08:09:15] <richlowe> which is the main constraint I was referring to.
[08:09:24] <Gman> not necessarily
[08:09:37] <Gman> assuming it's only ubuntu
[08:09:38] <jmcp> hehe ... somebody on another channel just asked for LT to be given a nipplecripple :)
[08:09:50] <richlowe> Gman: SMF carries the contract(5) stuff, and various other changes with it.
[08:09:53] <Gman> they'll basically have to hack a bunch of packages
[08:10:05] <richlowe> Gman: retrofitting it all into upstream software that doesn't want it would be interesting to the degree of probably being more trouble than it's worth.
[08:10:10] <Gman> yeah
[08:10:25] <Gman> though there's not that much software that uses init is there?
[08:10:28] <Gman> anyway, brb
[08:10:48] <richlowe> adding manifests is one thing, adding the contract model to the kernel (and the other bits that would be bitten by it, like cron and sshd) is quite another.
[08:11:03] <jmcp> Gman: when do you head back across the pond?
[08:14:37] *** karrotx has joined #opensolaris
[08:18:17] <Gman> richlowe, yeah
[08:18:20] <Gman> jmcp, sunday probably
[08:18:30] <Gman> interesting to hear about the kernel releases
[08:18:39] <Gman> major one every 3 months - insane!
[08:18:56] * Gman has a feeling that he's repeated himself, shuts up
[08:19:03] * jmcp mutters about KU scheds ...... & looks @ Tpenta
[08:19:03] <richlowe> well, major isn't Major
[08:19:16] <richlowe> I think they could deal with it appropriately, given their model.
[08:19:43] <Gman> richlowe, i would say they're less equipped to deal with it
[08:22:08] <Error_e^ipi> what about the kernel schedule?
[08:22:10] <Error_e^ipi> which kernel?
[08:22:26] <richlowe> jmcp's reference to KU would be Solaris, Gman's reference would be Linux.
[08:22:57] <Error_e^ipi> 2.x.y : major update = x or major = y?
[08:23:27] <Gman> 2.6.x = where x is supposed to be 'major'
[08:23:45] <Gman> though i imagine it's more like solaris style features
[08:23:45] <Error_e^ipi> ahh
[08:23:46] <_syphilis_> where is port_notify_t documented?
[08:23:52] <Gman> except they might break a little more stuff :)
[08:26:51] <richlowe> damnit, I've found it once when postwait had the race in that stuff.
[08:27:25] <richlowe> _syphilis_: try port_associate(3C)
[08:27:59] <_syphilis_> mmh, i was hoping for a little more than an undocumented example :)
[08:28:13] <richlowe> well, grep's still running, give me time.
[08:28:14] *** karrotx has quit IRC
[08:28:46] <_syphilis_> signal.h(3head) just says to see port_associate(3c), so i'm not too helpful
[08:29:28] <richlowe> I suspect it's intented to be as opaque as it is.
[08:29:43] <richlowe> though that's not really made clear anywhere (and I swear I've found more documentation than this previously, as I said...)
[08:30:00] *** Gman is now known as GmanAFK
[08:30:05] <GmanAFK> later dudes
[08:30:11] <richlowe> later, Gman.
[08:30:34] <jmcp> ciao
[08:30:52] * _syphilis_ wonders if aio with ports is likely to be faster than recv/send
[08:31:12] <richlowe> just be damn careful if you're threaded, too.
[08:31:21] <richlowe> I'm not sure I ever saw a fix for postwait's race/deadlock go back.
[08:31:24] <_syphilis_> i have a separate port fd for every thread
[08:31:32] <richlowe> that's pretty much what he was doing, iirc.
[08:31:37] <richlowe> close them in the thread they started in.
[08:32:02] <_syphilis_> yes, clients never change thread.. i did it that way to reduce need for locking
[08:32:14] <richlowe> or you can end up with use in one thread, and the close in the other deadlocking you fairly deep in the kernel bits.
[08:32:25] <richlowe> (unless that got fixed and I just didn't notice)
[08:33:46] *** qdk has quit IRC
[08:35:36] <Doc> ok, so Linux has "unexpected resilience" according to Linus
[08:35:57] <asyd> \_o<
[08:36:12] <Doc> hard to know if that's a good thing or a bad thing
[08:37:47] *** yydzero has quit IRC
[08:37:48] * _syphilis_ wants signal delivery via ports
[08:40:58] *** alfism_ has quit IRC
[08:41:39] <Error_e^ipi> Doc: doesn't surprise me he'd say such a thing
[08:41:43] <Error_e^ipi> it ran away from him
[08:42:04] <Error_e^ipi> it was a school project & a dumping ground for ideas for a long time & then people started using it for some reason
[08:43:17] *** sahafeez has joined #opensolaris
[08:44:18] *** sahafeez has quit IRC
[08:44:36] *** sahafeez has joined #opensolaris
[08:48:16] *** mazon is now known as Mazon
[08:48:35] * jmcp wanders off
[08:48:51] <richlowe> Bugger.
[08:49:43] <richlowe> so, is knowingly duplicating a Closed:Not A Defect CR while providing much more information as to why it actually *is* a defect acceptable?
[08:50:55] <richlowe> (given the added and obvious context, that I don't know who closed the original, only have the barest idea of why, and can't do anything to it)
[08:55:27] *** Cyrille has joined #opensolaris
[09:00:04] *** sahafeez has quit IRC
[09:00:43] <asyd> hmm http://www.eweek.com/article2/0,1895,2084284,00.asp
[09:00:44] <asyd> is that a koke ?
[09:00:48] <asyd> joke even
[09:02:58] <Error_e^ipi> so long as gpl3 is gpl2 incompatible, i'm cool with it
[09:03:17] <Error_e^ipi> by which i mean, a gpl2 project can't take code from a v3 project
[09:03:24] <asyd> ok
[09:03:35] <asyd> it's true I'm far to be familiar with GPLv3
[09:03:37] <andersmo> asyd: the sun guys in the channel don't seem to know anything about it - at least as far as I can see in the backlog. =)
[09:03:51] <asyd> ok
[09:03:54] <asyd> thanks :)
[09:04:57] <andersmo> (but then again, they may be 1) held in the dark and fed shit by their bosses or 2) just maintaining confidentiality on corporate matters as their employment contracts probably tell them to. ;)
[09:09:53] *** Posixzombie has joined #opensolaris
[09:10:27] <Error_e^ipi> where on earth did i get richard stallman's PGP key in my ring ?
[09:10:30] <Error_e^ipi> *shrug*
[09:10:41] <Posixzombie> dladm show-dev show 100MB speed and I want to change it to 1000MB (it's conencted to 1000MB switch) ; how ?
[09:13:11] <cmihai> You sure you have Gigabit cards in that thing?
[09:13:36] <cmihai> And it's running in full duplex / 1000Mbit and not 100Mbit? See the switch config also.
[09:14:27] <richlowe> andersmo: #2 would be complex, given that while no doubt their contracts do, they also hold them to doing their jobs, which in part involves cooperating with us. :)
[09:14:48] <Posixzombie> chihai, yes, it is e1000
[09:15:26] <Posixzombie> and I saw that it is running 100Mb only by: dladm show-dev
[09:15:51] <Posixzombie> in linux there is ethtool to set the speed of the NIC
[09:16:02] <Posixzombie> what should I do in solaris to achieve this ?
[09:16:33] <andersmo> richlowe: uhm, I don't really think there is a problem there. Their jobs are software engineering, not public relations, and even marketroids maintain confidentiality even though their job may be to inform the rest of the world about the company. =)
[09:19:12] <Posixzombie> any idea, anybody?
[09:19:55] *** slowhog has left #opensolaris
[09:20:06] <loke_> Posixzombie: you use ndd -set something...
[09:23:08] <Posixzombie> loke_, there is a way to list all properties which can be set by bdd
[09:23:10] <Posixzombie> how
[09:23:32] *** Posixzombie has quit IRC
[09:23:41] <_syphilis_> posix: query the ? property
[09:25:50] *** alobbs has joined #opensolaris
[09:27:56] *** tsoome has joined #opensolaris
[09:33:26] *** Posixzombie has joined #opensolaris
[09:33:44] <Posixzombie> I try : ndd -set /dev/e1000g0 link_speed 1000
[09:33:46] <Posixzombie> and get
[09:33:54] <Posixzombie> operation failed: Permission denied
[09:34:03] <Posixzombie> an idea why ?
[09:34:06] <Posixzombie> I am root
[09:34:44] <Error_e^ipi> is it a zone?
[09:34:59] <Posixzombie> no
[09:35:01] <Posixzombie> and
[09:35:04] <Posixzombie> kstat -m e1000g | grep cap_1000fdx
[09:35:07] <Posixzombie> returns
[09:35:13] <Posixzombie> adv_cap_1000fdx                 1
[09:35:14] <Posixzombie>         cap_1000fdx                     1
[09:35:14] <Posixzombie>         lp_cap_1000fdx                  1
[09:35:21] <Posixzombie> adv_cap_1000fdx                 1
[09:35:21] <Posixzombie>         cap_1000fdx                     1
[09:35:21] <Posixzombie>         lp_cap_1000fdx                  0
[09:35:24] <Posixzombie> so it seems ok
[09:36:09] *** coffman has joined #opensolaris
[09:36:28] <tsoome> posixzombi what solaris?
[09:36:42] <tsoome> for older versions ndd support was quite limited
[09:37:01] <tsoome> now they are startint to improve...
[09:37:08] <Posixzombie> tsoome, sxcr snv_54
[09:37:15] <Posixzombie> which is really recent
[09:37:19] <tsoome> true
[09:37:20] <Posixzombie> on x86_64
[09:37:22] <coffman> hey there
[09:37:31] <coffman> b55 is overtime right?
[09:37:40] <richlowe> Yes.
[09:37:50] <richlowe> it's been respun at least twice, I think.
[09:38:00] <Posixzombie> tsoome, and I know for sure
[09:38:18] <tsoome> for -set you should use adv* vars, those are rw
[09:38:22] <rydis> posixzombie: "ndd /dev/something \?"
[09:38:32] <Posixzombie> that this nic on the same machine (but with linux) on the same switch port and LAN was set to 1000Mbits
[09:38:33] <rydis> ... lists parameters.
[09:38:52] <Posixzombie> rydis, thnks
[09:38:57] <tsoome> it should default to 1g via autoneg
[09:39:07] <coffman> ah greate, mobile internet :)
[09:39:10] <Posixzombie> I found the link_speed with kstata -m e1000g
[09:39:34] <tsoome> man e1000g will help as well;)
[09:40:24] <Posixzombie> ndd -get  /dev/e1000g0 autoneg_cap
[09:40:26] <tomww> what about the dladm thing to set speed?
[09:40:28] <Posixzombie> returns 1
[09:40:47] <Posixzombie> tomww, dladm with which option ?
[09:41:16] <coffman> i have this pcmcia card wich gives me umts/gprs, its an usb serial modem bridge and i connect via ppp - i tryed it under b53 but every time i put it in the system crashed
[09:41:31] <coffman> i wonder if any one tryed too?
[09:45:13] *** glick has joined #opensolaris
[09:48:23] <coffman> tomww: im not registered so now priv :/
[09:48:40] <coffman> tomww: cant try atm since im connectet via :)
[09:50:38] *** LordKing has joined #opensolaris
[09:51:34] *** LordKing_ has joined #opensolaris
[09:52:02] *** LordKing_ has joined #opensolaris
[09:52:02] *** linma has joined #opensolaris
[09:52:28] *** LordKing_ has quit IRC
[09:55:42] *** qdk has joined #opensolaris
[10:00:12] <tomww> coffman: http://es.opensolaris.org/os/project/wwan/  and the cardbus drivers from "frkit", this works for me with a E61
[10:02:17] *** gustav3d has joined #opensolaris
[10:10:06] *** yippi has quit IRC
[10:10:07] *** alanc-away has quit IRC
[10:10:22] *** yippi has joined #opensolaris
[10:10:39] *** alanc-away has joined #opensolaris
[10:10:56] *** timfoster has joined #opensolaris
[10:15:44] *** Mazon is now known as mazon
[10:16:50] *** glick has quit IRC
[10:24:15] *** bengtf has joined #opensolaris
[10:25:39] *** bengtf_ has joined #opensolaris
[10:27:48] *** bengtf_ has quit IRC
[10:28:09] *** mikaeld_ is now known as mikaeld
[10:29:33] *** _357181_ is now known as estibi
[10:37:03] *** Fish- has joined #opensolaris
[10:37:24] <Fish-> hello
[10:37:25] *** simford has quit IRC
[10:39:22] *** dunc has quit IRC
[10:46:57] <Gr|ffous> hi fish
[10:50:50] <raph_ael> hello
[10:56:55] *** uebayasi has quit IRC
[10:57:28] *** Posixzombie has quit IRC
[10:59:21] *** svin0 has joined #opensolaris
[11:01:04] *** mazon is now known as Mazon
[11:03:27] *** coffman has quit IRC
[11:04:50] *** MattMan has joined #opensolaris
[11:07:56] *** svin0 has quit IRC
[11:07:57] *** Tpenta has quit IRC
[11:07:58] *** timfoster has quit IRC
[11:08:21] *** timfoster has joined #opensolaris
[11:08:39] *** Fish- is now known as Fish
[11:09:06] <onbot> commit by vi117747:  4765561 tcp panic due to memory corruption
[11:10:53] * Doc wonders if onbot has an off switch?
[11:10:58] *** dunc has joined #opensolaris
[11:11:10] <cmihai>  /ignore ;]
[11:11:53] *** yongsun has quit IRC
[11:12:46] <Doc> onbot: a/s/l ?
[11:12:47] <onbot> Doc: Error: "a/s/l" is not a valid command.
[11:14:00] <Doc> onbot: do you know the the IRC RFC says that bots shouldnt answer to public messages with another public message?
[11:14:01] <onbot> Doc: Error: "do" is not a valid command.
[11:14:29] <Doc> i'll take that as a no
[11:16:22] *** hou has joined #opensolaris
[11:18:02] <sickness> lol
[11:18:39] *** Dar has quit IRC
[11:20:30] <richlowe> Doc: beyond /ignore onbot, I don't think so.
[11:21:27] <richlowe> though for it's actual use, I can't think why you'd want to, other than perhaps when the bridge backs up and it dumps a lot of stuff at once.
[11:24:56] *** yippi has quit IRC
[11:27:56] *** timfoste1 has joined #opensolaris
[11:28:08] *** timfoster has quit IRC
[11:29:05] *** mh_ has joined #opensolaris
[11:30:43] *** gallium has joined #opensolaris
[11:34:50] *** damienc has joined #opensolaris
[11:53:43] <Doc> rich: you appear to be assuming i'm being serious
[11:54:03] <Doc> onbot: i_love_you!
[11:54:04] <onbot> Doc: Error: "i_love_you!" is not a valid command.
[12:02:38] <silk> see Doc, no one loves you
[12:02:47] <silk> not even brainless bots
[12:02:51] <Doc> not unexpected
[12:03:22] *** trip_ has joined #opensolaris
[12:03:35] * Doc goes off to putback the fix for a CR entitled "onbot loves doc!"
[12:04:48] <silk> fix su instead
[12:04:57] <silk> more value I think
[12:04:59] *** thomsog has joined #opensolaris
[12:05:15] <Doc> did you open a call?
[12:05:20] <Doc> do you even remember the phone number?
[12:05:20] <silk> no
[12:05:27] <silk> 1800 555 SUN?
[12:05:37] <Doc> close enough
[12:05:40] <silk> (876) ?
[12:05:46] <silk> or was that the platinum one?
[12:05:55] <silk> 1800 Phiiiilllllll
[12:06:25] <silk> trying to figure out whether novating a car is worth it now
[12:06:37] *** jteo has joined #opensolaris
[12:06:46] <Doc> how many kms you drive a year?
[12:06:54] <silk> ~20K
[12:07:21] <Doc> hmm.. that's in the "probably worth it" category under the old rates, but unless you earn over 150k it's possibly not now
[12:12:43] *** trip_ has quit IRC
[12:13:26] <_syphilis_> is it safe to call port_alert() from a signal handler?
[12:13:58] <jteo> hmm. what's this noise about osol under GPL3. -mutters-
[12:15:35] <Doc> syphilis: one of the man pages tells you what's safe to call - it's a very small list
[12:15:43] *** triplah_ has quit IRC
[12:16:24] *** calumb has joined #opensolaris
[12:16:29] <_syphilis_> oh, MT-Level i guess
[12:16:35] <_syphilis_> and it's not async-signal-safe :-(
[12:17:08] <Doc> that's it
[12:18:16] *** rachel_ has quit IRC
[12:18:20] <_syphilis_> hmm, port_send isn't either.. i wonder what the usual way of handling signals with event ports is
[12:19:10] *** mh_ has quit IRC
[12:20:19] <Doc> the reason behind it is that a thread could already be running port_send, then get a single, which then calls port_send
[12:20:41] <Doc> so the one thread can be halfway through the same function twice - at the same time
[12:20:49] *** trip_ has joined #opensolaris
[12:21:01] <Doc> which leads to "Weird-Shit(tm)" occuring
[12:21:38] * tsoome is away - bbl - [AcidJazz - Log ON]
[12:22:53] <jteo> s/single/signal ?
[12:24:21] *** trip_ is now known as triplah
[12:24:43] <Doc> english is my 2nd language, so i'm not very good at it
[12:24:55] <Doc> unfortunately i dont have a first language
[12:25:58] <cap_> you have a 2nd 2nd language then?
[12:26:00] <cap_> ;-)
[12:33:19] *** calumb is now known as calAFK
[12:33:56] <jteo> Doc: excellent excuse.
[12:34:15] <silk> fucking clowns
[12:34:32] <silk> I wonder why the RFC says at least 2 nameservers on seperate networks
[12:35:02] <jteo> ?
[12:35:03] <silk> oh thats right.. cause its a good ideda
[12:35:10] *** sdrom has joined #opensolaris
[12:35:19] <silk> oh peckerheads and their dns configurations
[12:35:40] <silk> lets create 2 NS records, and 2 A records for them, but point them both to the same IP
[12:35:59] <gallium> stop f4cking clowns, silk ;)
[12:36:30] <silk> clowns can be cute
[12:36:45] <lasseoe> if they have big feet and a red nose sure
[12:36:46] <gallium> heh
[12:37:11] <Doc> be nice to silk, or he'll start sticking needles in you
[12:37:25] * lasseoe stops clowning around
[12:38:45] <mv> my 11hours build of b55 just ended with build errors
[12:41:45] *** trs81 has joined #opensolaris
[12:42:35] *** nrubsig has joined #opensolaris
[12:42:36] *** ChanServ sets mode: +o nrubsig
[12:43:15] <gallium> mv, bugger.
[12:44:45] *** CosmicDJ has joined #opensolaris
[12:46:33] <mv> well, i'm little bit sad...after 4 build with 512MB ram, which ended with sucking out all memory+swap and unusable system and after doubling the ram it ended with errors
[12:47:40] <Doc> hmm.. i put a CD in my PC, and now windows is trying ot generate a thumbnail of the 589 meg TIFF on it
[12:48:07] <Doc> stupid bloody OS
[12:48:59] <gallium> Doc, PEBKAC ;-)
[12:48:59] *** bougie has joined #opensolaris
[12:49:02] <LeftWing> lol
[12:49:21] <silk> Layer8 problem
[12:49:42] <bougie> hello :)
[12:49:53] <Doc> why, because i'm using windows in the first place?
[12:50:12] <Doc> talk adobe into writing photoshop for Solaris (again) and i'll consider changing
[12:50:22] <silk> use a mac
[12:50:33] <gallium> that and because you have a 589mb tiff
[12:50:51] <LeftWing> 589MB TIFFs are great.  They're the ideal size.
[12:50:54] <Doc> gallium: well the .psd wouldnt fit on a single CD
[12:51:29] <silk> Doc, whst this one of?
[12:51:35] <Doc> this is just sydney
[12:51:58] <Doc> MP is a 2Gb TIFF (13.5 Gb PSB)
[12:57:56] <Stric> Hm. Seems like I can't do a mirror(raidz(x,y,z),raidz(a,b,c)), can anyone verify that?
[12:58:11] <Doc> just do RAID-Z2
[12:59:08] <Stric> had some whacko idea about mirroring two disk arrays for some possible additional protection against weird hw faults..
[12:59:31] <Stric> that means going hw-raid5, since I can't mirror sw-raidz it seems..
[12:59:35] *** calAFK is now known as calumb
[12:59:40] <LeftWing> Surely double-parity and data checksumming will be enough?
[13:00:03] <Stric> unless the controller in the disk array blows up or something
[13:00:18] <andersmo> you can mirror x+a, y+b and c+z to be somewhat resilient against losing a whole array at once.
[13:00:48] <LeftWing> No don't do that.
[13:00:50] <Doc> yah.. you can do straight mirroring, which will actually give you more space that your mirror of raidz's
[13:01:05] <LeftWing> You only need to lose x & a to bring down the whol show, andersmo.
[13:01:20] <andersmo> LeftWing: yep. I know. =)
[13:01:28] <Doc> leftwing: yes, but the benefits of the cross-controller/array redundency outweights that
[13:01:39] <LeftWing> Doc: Arrr.
[13:01:42] <Doc> add in a hot-spare and a regular scrub if you're worried
[13:01:49] <LeftWing> fair enough
[13:01:56] <Stric> I'm gonna do double parity, and wasted disk space isn't a problem..
[13:02:18] <Doc> double parity still doesnt give you cross-array/controller/etc
[13:02:37] <Doc> if you've got the hardware, and you're worried about loss of an array, mirror
[13:02:50] <Stric> that's why I asked for mirror(raidz,raidz) ;)
[13:02:53] <Doc> all depends on what you're trying to achieve
[13:03:10] <Doc> loss of controller doesn't result in data loss, just data unavailibility
[13:03:18] <Stric> and with that setup, I can detach a mirror and do tests on one array if I want
[13:03:53] <andersmo> I think you can raid5 on top of mirrors with vxvm. Dunno how hairy you like your configuration, though... =)
[13:04:19] <andersmo> three-way-mirror (x, y, a) + three-way-mirror ( z, b, c) ? =)
[13:04:50] <andersmo> or just megamirror (x,y,z,a,b,c) ;)
[13:04:51] <Doc> yay.. now windows explorer has died
[13:05:01] <Stric> I'd like to leave vxvm out of this
[13:05:30] <andersmo> Stric: probably a good idea. =)
[13:13:30] <Stric> Or a butt ugly solution.. zpool create blah1 raidz2 x y z;zpool create blah2 raidz2 a b c;zfs create -V 100G blah1/mirror1;zfs create -V 100G blah2/mirror2;zpool create data mirror /dev/zvol/dsk/blah1/mirror1 /dev/zvol/dsk/blah2/mirror2
[13:13:31] <Stric> :P
[13:14:58] <Stric> (amount of disks and disk storage is highly fictional in this case)
[13:27:28] *** cyberops has joined #opensolaris
[13:34:28] *** nrubsig has quit IRC
[13:36:09] <andersmo> Hm. I just watched SMF restart an entire cyrus service due to one minor daemon (not the master daemon) segfaulting. I know I can turn off "restart on segfault" for the whole service, and I've modified it's manifest and tested that it works. However, if I used that, and the master daemon dies, SMF wouldn't notice anything wrong before all the other daemons have dies as well, and that could take a while... Is there any way to make SMF more protective
[13:36:27] <andersmo> s/have dies/have died/
[13:36:36] <andersmo> That was probably cut off...
[13:36:47] <LeftWing> "SMF more protectiv"
[13:36:48] *** edwardocallaghan has joined #opensolaris
[13:36:50] <edwardocallaghan> Hi
[13:37:23] <andersmo> thanks. Is there any way to make SMF more protective of the master process (and react on, say, segfaults), and letting it's subprocesses (who are watched by the master daemon) live and die without harming the master? =)
[13:38:30] <andersmo> I've considered making the master use ctrun to start it's subprocesses, but that would probably cause one "contract" for each of the subprocesses, of which there are a thousand or so, give or take a few hundreds.
[13:38:41] <edwardocallaghan> I got a question on ZFS
[13:38:42] <andersmo> Dunno if that's scalable or insane. =)
[13:39:22] <edwardocallaghan> If you have one disk in ZFS and add a new one can you add it to a mirror and if so, how is that done ?
[13:40:36] <CosmicDJ> http://opensolaris.org/os/community/zfs/docs/zfsadmin.pdf
[13:41:21] <edwardocallaghan> *Thanks
[13:44:33] <edwardocallaghan> So if you did zpool create -f newpool c1t2d0 then put data on that disk then did zpool create -f newpool mirror c1t2d0 c1t3d0
[13:44:44] *** triplah has quit IRC
[13:45:02] <edwardocallaghan> Would that lose the new data or would it mirror it keeping the new data on the old disk ?
[13:55:36] <tomww>
[13:58:17] *** cmihai has quit IRC
[13:58:42] <edwardocallaghan> Be back later tonight
[13:58:43] *** edwardocallaghan has left #opensolaris
[13:59:04] *** cmihai has joined #OpenSolaris
[14:01:44] <loke_> when (if ever) is there going to be a "zpool remove" command available? To be able to remove devices frm an activestorage pool.
[14:02:46] <andersmo> loke_: maybe sometime in 2007, is what I've seen on sun.com... But I don't have the URL reference handy.
[14:04:35] <loke_> andersmo: oh... so it'll be some time until it finds its way into the supported Solaris releases
[14:08:17] <jteo> loke_: post on zfs discuss. i would like to see an active discussion of the need for such a feature. :)
[14:09:08] <loke_> jteo: I was quite surprised it didn't exist... Esp. since I realised it when I was playing around with ZFS for the first time. That's the time you really want it :-)
[14:09:25] <loke_> But it would be reallyuseful when migrating hardware
[14:11:44] <jteo> true.
[14:12:37] *** calumb has quit IRC
[14:20:31] *** MasterShrek has joined #opensolaris
[14:20:58] *** nachox has joined #opensolaris
[14:25:16] *** bunker_work has joined #opensolaris
[14:26:08] <nachox> for those that care about bsds http://www.freebsd.org/news/status/report-oct-2006-dec-2006.html#Porting-ZFS-to-FreeBSD
[14:27:54] <cmihai> Already tried it.
[14:28:05] <cmihai> It actually works.
[14:28:10] <cmihai> Too bad it's on a perforce repo :P
[14:30:16] <mv> well, i've tried it to, but it didn't work for me
[14:30:42] <mv> it crash as soon as it tries to mount pool on /tank for example
[14:31:12] <cmihai> You must have gotten a different version from the cvs I guess :)
[14:31:40] <nachox> they also have a mostly working 3945abg driver
[14:31:41] <mv> well, it was current from date, when pjd released zfs patch
[14:34:52] <nachox> in any case, that is opensource in action, they are working in dtrace too iirc which is very nice
[14:35:35] <mv> yes, dtrace is actually quite working :)
[14:36:03] <delewis> mv: how many probes are they up to?
[14:36:22] <cmihai> Not all the probes
[14:36:24] <cmihai> About 70%
[14:36:29] <delewis> still impressive
[14:36:32] <cmihai> Indeed.
[14:36:45] <cmihai> They've been working on it ever since it came out
[14:37:11] <mv> delewis: i'm not sure
[14:37:34] <mv> i've heard that sun is helping them a lot
[14:37:36] <cmihai> Devon O'Dell was working on it at some point.. I think someone else took over now
[14:37:48] <mv> hw to test on, help from developers etc
[14:37:51] <delewis> I'm curious if ZFS on FreeBSD is going to have the same hardware requirements for decent performance as it does on Solaris
[14:37:56] <delewis> like 64-bit, 1GB+ of memory, etc.
[14:38:14] <cmihai> I'm sure it does.
[14:38:57] <delewis> mv: yes, and Sun has actually integrated patches submitted by the ZFS FreeBSD team to make the code even more portable than it already was.
[14:39:21] <mv> wow
[14:39:31] <mv> delewis: well, ZFS FreeBSD team is only 1 guy
[14:39:38] <delewis> proof that the CDDL actually works :-)
[14:39:41] <delewis> mv: true
[14:39:45] <mv> Pawel Jakub Dawidek
[14:40:25] *** MasterShrek has quit IRC
[14:40:39] <cmihai> mv: and he did it in 10 days
[14:40:44] <cmihai> Guy's a bleeding genius :)
[14:40:54] <nachox> that only shows how portable zfs is
[14:41:04] <mv> cmihai: he is coding machine :)
[14:41:11] <cmihai> Well, to be honest, since MacOS ported ZFS already...
[14:41:19] <cmihai> It might have been a lot simpler if Apple gave him a hand :P
[14:41:22] <delewis> nachox: well, Sun claims one of the design goals of ZFS was portability.
[14:41:22] <mv> they did?
[14:41:30] <cmihai> Yes, ZFS is in Leopard
[14:42:01] <zdzichuBG> ZFS became new FAT of data-exchnage
[14:42:16] <cmihai> And techincally, MacOS is a XNU (modified Mach 3 kernel) on FreeBSD userland (well, Darwin) with Aqua interface.
[14:42:17] <nachox> not quite
[14:42:37] <cmihai> Well, not that simple, but it's similar enough to allow porting of apps.
[14:42:59] <cmihai> Seen a few FreeBSD devels that also worked on Darwin.
[14:42:59] <nachox> zfs would have to be ported to lots of oss first, netbsd, windows...
[14:43:14] <cmihai> nachox: DragonFlyBSD was also working on a port.
[14:43:17] <cmihai> Didn't get anywhere though.
[14:43:27] <cmihai> Linux is working on a port too..
[14:43:34] <cmihai> But they're using FUSE and stuff.
[14:43:57] <nachox> cmihai: i havent seen any news other than intentions to actually port it
[14:44:09] <andersmo> nachox: http://zfs-on-fuse.blogspot.com/
[14:44:56] <nachox> that is for linux right? i think they are probably the biggest losers in this game, hehe
[14:45:14] <cmihai> Not really, that is for FUSE
[14:45:27] <cmihai> And FUSE runs on Linux, FreeBSD, NetBSD and Solaris.
[14:45:27] <zdzichuBG> I'm actually using ZFS on Linux via FUSE
[14:45:42] <zdzichuBG> cmihai: that's "Filesystem In Userspace"
[14:45:45] <cmihai> Yes.
[14:45:50] <zdzichuBG> out of kernel, free of licensing problems
[14:45:53] <cmihai> zdzichuBG: since GPL isn't compatible with CDDL
[14:46:00] <cmihai> And they can't add it to the kernel.
[14:46:12] <cmihai> FreeBSD can though.
[14:46:24] <cmihai> GPL is like a virus :P
[14:46:31] <GoodKarma> meh
[14:46:40] <cmihai> Steve Ballmer said that ;)
[14:46:40] <GoodKarma> is it?
[14:46:47] <zdzichuBG> it would be hard to go through linux developers with such "blatant layering violation" anyway :)
[14:46:48] <cmihai> He also said Linux is like a cancer :)
[14:46:57] <GoodKarma> good boy
[14:47:00] <cmihai> Yup
[14:47:42] <GoodKarma> where do we send pizzas for that guy again?
[14:47:45] <mv> intresting opinion
[14:47:48] <mv> mostly true
[14:48:43] <nachox> licencing is a design decision too, the funny thing is that they still have binary blobs
[14:48:46] <cmihai> GoodKarma: he's also know for throwing chairs and telling people "I'm going to fucking burry you"
[14:48:52] <cmihai> So I'd be carefull with that pizza if I were you
[14:49:07] <GoodKarma> cmihai: oh I like that attitude
[14:49:31] <cmihai> "I'm going to f---ing bury that guy, I have done it before, and I will do it again," the declaration quotes Ballmer.
[14:49:44] <mv> send him few bottles of wine
[14:49:49] <cmihai> Heheh
[14:54:06] <GoodKarma> yeah
[14:57:28] <sickness> can someone confirm to me that solaris10u3 is 11/06 ?
[14:58:11] <zdzichuBG> sickness: yes it is
[14:58:20] <GoodKarma> u2 was 06/06 I guess, so yea
[14:58:23] *** cmihai has quit IRC
[15:00:03] *** cmihai has joined #OpenSolaris
[15:02:02] *** bengtf has quit IRC
[15:02:03] *** syndrome71 has quit IRC
[15:03:17] * tsoome is away - bbl - [AcidJazz - Log ON]
[15:08:31] *** mrdeviant has joined #opensolaris
[15:14:25] *** merlin_ has joined #opensolaris
[15:19:41] <AbeFroman>                        Solaris 10 11/06 s10s_u3wos_10 SPARC
[15:22:37] <_syphilis_> SX has TCP_CORK... i suppose the chance of this being backported to S10 is fairly small?
[15:28:51] *** timfoste1 has left #opensolaris
[15:33:55] *** calumb has joined #opensolaris
[15:35:51] *** brs has joined #opensolaris
[15:37:00] *** mritun_ has quit IRC
[15:39:56] <quasi> sickness: yes, confirmed
[15:45:54] *** gallium has quit IRC
[15:46:40] *** MikeTLiv1 has quit IRC
[15:48:16] <rydis> I'm having a little problem with Solaris 10 crashing, rather hard. Is it OK to talk about that, here? (I can't experiment with it, right now, since I don't have console access to any machines, and most are in actual use.)
[15:48:35] *** CosmicDJ has quit IRC
[15:49:29] *** jcea has left #opensolaris
[15:49:36] <Auralis> sure
[15:49:37] <quasi> rydis: did it do a crashdump?
[15:50:08] *** gustav3d has quit IRC
[15:50:11] <rydis> quasi: I don't think so.
[15:50:27] <rydis> The machine I just crashed is one on which I'm not root.
[15:50:32] *** laca has joined #opensolaris
[15:50:41] <rydis> So I can't check /var/crash.
[15:51:10] <quasi> in that case, helping out is very hard
[15:51:14] <_syphilis_> rydis: if it's not something obvious (like eating all the memory) you should contact security-alert at sun dot com
[15:51:32] <rydis> The current snapshot (and some earlier versions) of CMU Common Lisp, running as a regular user, seems to, sometimes, crash the kernel. It doesn't necessarily crash it immediately, either, so it might be difficult to check.
[15:52:34] <rydis> It shouldn't be eating all memory. I don't think so, at least. It runs fine on Solaris 9, by the way.
[15:53:01] <jbk> actually causing a OS kernel crash?
[15:53:09] <rydis> Yes.
[15:53:49] <jbk> i mean, i've heard of certain lisp implementations crashing because they were using a register they shouldn't have, that with the thread changes in solaris 10, was previously unusued (but reserved) but is now used
[15:54:07] *** CosmicDJ has joined #opensolaris
[15:54:10] <rydis> (The snapshot I'm running is built on Solaris 8, if that helps.)
[15:54:37] <rydis> jbk: Yeah. No longer an issue. That just crashed the lisp, though.
[15:55:37] <rydis> The binary and sources are available here: <URL: http://common-lisp.net/project/cmucl/downloads/snapshots/2007/01/ >
[15:56:01] <trygvis> is it not possible to get prstat to sort by state?
[15:56:09] <trygvis> I want to see all running processes
[16:00:39] *** Theoden-Nexenta has joined #opensolaris
[16:01:34] <jbk> jesus.. i hate non-technical people who think they know technical issues
[16:01:46] <jbk> (sorry.. ranting about this call i'm on)
[16:02:08] *** hile_ has quit IRC
[16:02:53] *** kloczek has quit IRC
[16:03:49] <jbk> because we had an issue with oracle a few months ago (where it triggered a bug in libaio), it just _has_ to be the same problem we're seeing here
[16:04:03] <jbk> even though the stack trace doesn't agree
[16:04:06] <Berny>        morning
[16:04:13] <Berny> afternoon
[16:04:16] <Berny> whatever
[16:04:33] <rydis> I've mailed the admin, asking if there's a crash-dump. I don't think there is, but there might be. Would/could a user using a reserved register crash the kernel, if there isn't a Solaris bug?
[16:04:48] <jbk> i wouldn't think so
[16:04:53] <jbk> though the crash dump should help
[16:05:45] <rydis> There might be something weird in the memory usage, since it's slightly unorthodox, but I really don't think it'd run out.
[16:07:22] <Berny> any of you folks tried to run some xview programs in a lx branded zone?
[16:07:23] *** sdrom has quit IRC
[16:07:40] <Berny> i'm seeing some strange behaviour where cmdtool just hangs
[16:08:04] <Berny> my guess would be i tries to set some terminal stuff which is not (yet) supported?
[16:11:02] *** kloczek has joined #opensolaris
[16:14:05] *** calumb has quit IRC
[16:16:54] *** calumb has joined #opensolaris
[16:21:24] *** hile_ has joined #opensolaris
[16:25:36] <rydis> Seems the sysadmin has logged a case with Sun.
[16:25:58] <rydis> (There was an attempt at a crash dump, but it apparently failed.)
[16:26:38] <rydis> " 37% doneXIR failed - performing SPOR"
[16:27:26] *** osdummy has quit IRC
[16:27:51] *** nachox has quit IRC
[16:29:10] <lisppaste3> rydis pasted "Console log, on crash" at http://paste.lisp.org/display/35526
[16:41:13] *** nachox has joined #opensolaris
[16:45:35] *** jcea has joined #opensolaris
[16:45:40] <_syphilis_> when i build a trivial application (int main() {}) with gcc -m64 -pg, i get this error when running it: "No space for profiling buffer(s)".  a 32-bit binary is fine.  what would cause that?
[16:47:49] *** hou has quit IRC
[16:51:19] *** mrdeviant has quit IRC
[16:51:20] *** alanc has joined #opensolaris
[16:51:59] <nachox> _syphilis_:  try with -mcmodel=medlow
[16:52:30] <nachox> or turn off profiling
[16:52:39] <nachox> or use sun studio :P
[16:54:49] <_syphilis_> nachox: this is x86
[16:54:58] <_syphilis_> and none of the x86-specific memory models seem to change it
[16:56:02] <nachox> x86? it's x86-64 right?
[16:56:07] <_syphilis_> yes
[17:07:33] *** Mazon is now known as mazon
[17:09:14] *** ShadowHntr has joined #opensolaris
[17:14:46] *** ShadowHntr has quit IRC
[17:16:19] <sickness> quasi: tnx :)
[17:16:50] *** UnixTitan has quit IRC
[17:17:48] *** UnixTitan has joined #opensolaris
[17:19:36] <onbot> commit by Kacheong Poon:  6513446 ETSI TS 102 144 requires that SCTP delayed SACK frequency to be tunable
[17:19:39] <nachox> i have been googling for a while and all i can find is problems when profiling using gcc and 64 bit binaries in solaris, so i guess you should just use sun studio or not profiling
[17:20:08] <_syphilis_> well, i can compile in 32-bit mode
[17:21:04] *** mrdeviant has joined #opensolaris
[17:21:05] <nachox> that too
[17:23:21] *** qufk has joined #opensolaris
[17:23:50] *** stevel has joined #opensolaris
[17:23:51] *** ChanServ sets mode: +o stevel
[17:24:02] *** kFuQ has quit IRC
[17:24:35] *** salmandr has joined #opensolaris
[17:25:44] *** bunker_work has quit IRC
[17:31:43] *** alfism_ has joined #opensolaris
[17:32:33] *** alfism_ has quit IRC
[17:32:36] *** dunc has quit IRC
[17:32:39] *** alfism_ has joined #opensolaris
[17:32:50] *** dunc has joined #opensolaris
[17:35:23] *** dunc has quit IRC
[17:35:40] *** dunc has joined #opensolaris
[17:37:33] *** calumb is now known as calAFK
[17:37:39] *** axisys has joined #opensolaris
[17:40:47] <axisys> netservices limit keeps rpc port 111 open per nmap.. is it really open?
[17:41:27] <axisys> i can telnet to the port but i think someone said here that, that does not mean rpc port open
[17:41:44] <axisys> looking for a article w/ some explanantion
[17:42:43] *** alfism_ has quit IRC
[17:43:26] <_syphilis_> 111 is portmap, limited services makes it only accept connections from localhost, iirc
[17:43:39] *** dunc has quit IRC
[17:43:45] <_syphilis_> portmap on its own doesn't actually do anything anyway
[17:44:07] <onbot> commit by James D Carlson:  PSARC 2006/597 DHCPv6 Client; 4863327 dhcp client causes drag; 6369116 dhcpagent doesn't notice when logical interfaces disappear; 6386331 dhcpagent should implement RFC 4361 client identifier
[17:44:27] <axisys> _syphilis_: this is where i am little lost.. security team wanted to close that port down per their nmap scan :-)
[17:44:52] <_syphilis_> you can stop portmap, but then rpc won't work
[17:45:23] <axisys> _syphilis_: what does it need for? i use mainly ssh
[17:45:42] <_syphilis_> nfs uses it, CDE has some RPC service
[17:46:40] *** UnixTitan has quit IRC
[17:48:14] <andersmo> axisys: just firewall away their nmap scanning host. "Security by not really." ;)
[17:49:25] *** dunc has joined #opensolaris
[17:49:58] <axisys> andersmo: i wish i could.. its the policy :-(
[17:50:29] <axisys> so 111 is portmap port or rpc port?
[17:50:46] <axisys> _syphilis_: sorry u already answered that
[17:50:53] <_syphilis_> axisys: portmap is used to translate rpc server names to port numbers
[17:50:59] <_syphilis_> axisys: it's a service enumerator
[17:51:31] <_syphilis_> it's not an RPC service itself but RPC requires it
[17:51:50] <andersmo> "where do I connect to talk to $service"? - the rpc servers themselves may be pretty much anywhere. Makes RPC notoriously difficult for firewalls. =)
[17:54:20] *** calAFK is now known as calumb
[17:54:27] *** hile_ has quit IRC
[17:55:33] <axisys> how can i find out what will break on a system if i turnoff portmap?
[17:56:04] <_syphilis_> disable it and see ;-)
[17:56:11] <_syphilis_> 'svcs -a|grep rpc' might be a start
[17:56:37] <sniffy> _syphilis_, -i on that grep too I'd bet
[17:56:56] <axisys> rpcinfo -p gives me a whole list
[17:56:57] <_syphilis_> i thought all FMRIs were lowercase
[17:57:15] <sniffy> True.
[17:57:30] <axisys> here is a list for one host http://rafb.net/p/xzCvgD81.html
[17:57:58] <_syphilis_> axisys: are you an NFS server?
[17:58:25] <axisys> scratch that url.. here is a complete one for that host http://rafb.net/p/ZWlOsm82.html
[17:58:34] <axisys> _syphilis_: nawp.. not that host
[18:07:21] *** deather_ has joined #opensolaris
[18:08:46] <axisys> _syphilis_: that host actually nfs sharing one folder
[18:10:07] *** RaD|Tz has joined #opensolaris
[18:10:14] *** Netwolf has quit IRC
[18:10:41] *** hile_ has joined #opensolaris
[18:14:22] *** woffer has quit IRC
[18:16:09] *** alobbs has quit IRC
[18:18:59] *** UnixTitan has joined #opensolaris
[18:20:07] *** CosmicDJ has quit IRC
[18:20:30] *** woffer has joined #opensolaris
[18:21:04] *** Pro^^DiGy-PunK has joined #opensolaris
[18:21:17] <Pro^^DiGy-PunK> He regresado! (demasiado autismo -autoaway, 15 mins-) (Away: 1m23s) desde el 17 Jan 00:16:53
[18:21:18] <Pro^^DiGy-PunK> hellos
[18:24:31] *** sahafeez has joined #opensolaris
[18:25:15] <_syphilis_> does aio_read/aio_write work on sockets?
[18:25:23] <cmihai> What the fuck was that
[18:25:40] <cmihai> God, turn that crap off.
[18:25:57] *** deather has quit IRC
[18:26:06] <dme> _syphilis_: it should, but the asynchronicity is done using threads rather than any true async calls into the kernel, as i recall.
[18:26:15] <_syphilis_> oh.  that's a shame
[18:27:00] *** sahafeez has quit IRC
[18:27:09] *** sahafeez has joined #opensolaris
[18:27:40] *** Cyrille has quit IRC
[18:28:17] <_syphilis_> what about the solaris-native interfaces (aioread/aiowrite)?
[18:31:29] *** Pro^^DiGy-PunK has quit IRC
[18:34:22] *** comay has joined #opensolaris
[18:36:29] *** Netwolf has joined #opensolaris
[18:55:11] *** damienc has quit IRC
[18:58:13] *** alfism has quit IRC
[18:58:28] *** alfism has joined #opensolaris
[19:05:39] *** tsoome has quit IRC
[19:12:06] *** mazon is now known as Mazon
[19:17:56] *** MattMan has quit IRC
[19:19:51] *** LordKing_ has joined #opensolaris
[19:28:37] <nachox> hmm, this is quite off topic, but anyone knows where i can find info about identity management of web services? :)
[19:30:32] *** tsoome has joined #opensolaris
[19:32:23] *** tsoome1 has joined #opensolaris
[19:33:07] *** thomsog has quit IRC
[19:36:19] *** hile_ has quit IRC
[19:38:03] *** hile_ has joined #opensolaris
[19:41:29] *** alanc is now known as alanc_away
[19:41:36] *** calumb has quit IRC
[19:44:33] <nprice> zpool iostat is very awesome hehe
[19:46:36] *** silk has quit IRC
[19:47:42] *** axxl has quit IRC
[19:48:03] <quasi> nprice: I don't find it very useful compared to regular iostat
[19:48:30] *** tsoome has quit IRC
[19:59:55] *** dunc has quit IRC
[20:02:10] *** tsoome1 is now known as tsoome
[20:02:12] *** alanc-away is now known as alanc
[20:04:21] *** Hobz has joined #opensolaris
[20:04:57] *** silk has joined #opensolaris
[20:05:59] <delewis> oh, no.
[20:06:04] <Error_e^ipi> ?
[20:06:09] <delewis> the GPLv3 stuff hit Slashdot.
[20:07:01] <Error_e^ipi> yay rumour propogation
[20:08:41] <Error_e^ipi> it occurred to me that a v2 kernel wouldn't be able to steal code from a v3 kernel anyways
[20:09:10] *** delewis has quit IRC
[20:09:11] *** nachox has quit IRC
[20:09:17] <Error_e^ipi> in which case, i don't care... gpl it or not
[20:10:46] *** delewis has joined #opensolaris
[20:11:42] <AbeFroman> boycott slashdot
[20:11:48] <AbeFroman> you'll gain iq
[20:12:27] <g4lt-U60> yeah, read the inquirer, where you can read about blu-ray pr0n
[20:12:46] <AbeFroman> haha
[20:16:31] <Error_e^ipi> unless you can see in infinite clarity every single abraded ingrown pubic hair's shaving bumps, it's just not porn
[20:18:15] *** mrdeviant has left #opensolaris
[20:18:26] <AbeFroman> amen
[20:20:51] *** bondolo has quit IRC
[20:21:15] *** UnixTitan has quit IRC
[20:26:52] *** Fish- has joined #opensolaris
[20:27:56] *** sickness_ has joined #opensolaris
[20:28:51] *** sickness has quit IRC
[20:29:20] *** sickness_ is now known as sickness
[20:29:37] <sickness> i'm back
[20:33:31] *** slowhog has joined #opensolaris
[20:34:13] *** Fish has quit IRC
[20:43:21] *** estibi is now known as _357181_
[20:44:03] *** estibi has joined #opensolaris
[20:45:29] *** ShadowHntr has joined #opensolaris
[20:50:39] <estibi> hi
[20:52:37] *** benr has joined #opensolaris
[20:52:37] *** ChanServ sets mode: +o benr
[20:52:44] <sickness> hi benr
[20:52:48] *** b3stbuddy has joined #opensolaris
[20:56:36] <benr> howde.
[20:57:33] <b3stbuddy> hello
[20:57:55] <Error_e^ipi> hey bestbuddy rocko guy
[20:57:58] <Error_e^ipi> *slurp*
[20:58:07] <elektronkind> benr has internets again
[20:58:27] <b3stbuddy> does anyone know the basic differences between opensolaris and solaris 10
[20:58:38] <Error_e^ipi> the internet's full of dweebs and perverts
[20:58:43] <Error_e^ipi> hate that place...
[21:00:02] <Error_e^ipi> b3stbuddy: think of it this way... Solaris10 is an OpenSolaris distro that includes some non-free stuff
[21:00:15] <b3stbuddy> ok
[21:00:17] <Error_e^ipi> ken?
[21:01:01] <elektronkind> b3stbuddy: new coe is also first put into opensolaris, tested and matured. Once it is suitable, those new features or bug fixes are rolled back into Solaris 10.
[21:01:08] <elektronkind> that's another way to look at it
[21:01:12] *** RaD|Tz has quit IRC
[21:01:14] <elektronkind> s/coe/code
[21:01:25] <b3stbuddy> ic
[21:01:41] <b3stbuddy> sames as fedora > redhat
[21:01:46] <elektronkind> Error_e^ipi: "I beat the internet once. The end boss was hard"
[21:01:54] <andersmo> oversimplified: opensolaris is source code for much of an operating system. solaris is a complete operating system.
[21:01:57] <elektronkind> b3stbuddy: similar general concept
[21:02:01] <andersmo> or, not really oversimplified either.
[21:02:46] <b3stbuddy> do they both use zfs?
[21:03:08] <andersmo> Yes, ZFS was added to Solaris 10 update 2 (I think?)
[21:03:18] <b3stbuddy> right now i use reiserfs 3 on linux and was looking to go to something better
[21:03:19] <andersmo> ...but it was first released as source code as part of opensolaris.
[21:03:21] <Error_e^ipi> zfs is in s10u2+ and O/N some version or other
[21:03:22] <zdzichuBG> yes, u2 6/06
[21:04:06] <b3stbuddy> was going to go with xfs but then i started hearing good things about zfs
[21:04:26] <andersmo> b3stbuddy: are you familiar with the distinction between linux the kernel and the various distributions built around it?
[21:04:50] <andersmo> b3stbuddy: opensolaris and solaris are somewhat similar - except that opensolaris is more than just a kernel. =)
[21:05:57] <b3stbuddy> so more packages come with it ...
[21:06:14] <andersmo> I decided to try JFS on the last Linux machine I installed - just out of curiosity - and I'm quite pleased with the experience.
[21:06:41] <b3stbuddy> what were you using before that andersmo?
[21:06:49] <andersmo> b3stbuddy: http://www.opensolaris.org/os/about/distributions/
[21:08:12] <andersmo> Mostly ext{2,3} and some reiserfs (3). We tried reiserfs on some file servers where I worked, but that was about the worst thing we've ever had in production. fsck.reiserfs --rebuild-tree isn't funny when thousands of users are trying to get some work done. =)
[21:08:30] *** UnixTitan has joined #opensolaris
[21:09:49] <andersmo> ...so we've gone back to solaris-based file servers with vxfs, and have been reasonably happy since. No ZFS in production except for internal needs yet - but I consider it very likely that ZFS will replace vxfs here, eventually.
[21:11:03] *** bengtf2 has joined #opensolaris
[21:11:42] *** hile_ has quit IRC
[21:13:44] <AbeFroman> we've had problems with (in this order): ext3, jfs, xfs, reiserfs
[21:14:04] <cmihai> Ugh, RicerFS
[21:14:16] <AbeFroman> wasn't my idea
[21:14:19] <Error_e^ipi> it murders your data, like he murdered his wife
[21:14:22] <Error_e^ipi> "allegedly"
[21:15:38] <elektronkind> i like burritos
[21:16:06] <AbeFroman> oh no.  here comes a tacos vs burritos holy war
[21:16:25] <zdzichuBG>  it's time for kebab
[21:19:12] <axisys> how much amps T2000 draws when starts
[21:20:06] * tsoome is away - bbl - [AcidJazz - Log ON]
[21:22:56] <axisys> 2 amps
[21:22:58] <axisys> thnx
[21:30:57] *** Tpenta has joined #opensolaris
[21:31:19] *** hile_ has joined #opensolaris
[21:32:54] *** Tpenta has left #opensolaris
[21:33:05] *** GmanAFK has quit IRC
[21:33:17] *** Tpenta has joined #opensolaris
[21:33:17] *** ChanServ sets mode: +o Tpenta
[21:38:57] *** slowhog has quit IRC
[21:46:46] *** LordKing_ has quit IRC
[21:47:46] *** hspaans has joined #opensolaris
[21:48:14] <hspaans> g'day to all
[21:49:07] <Error_e^ipi> hey
[21:50:56] *** abackos has joined #opensolaris
[21:53:08] *** Gman has joined #opensolaris
[21:53:37] <hspaans> hmmm rumors about opensolaris going gplv3 as well
[21:54:02] <hile_> rumours are just that
[21:54:05] <hile_> morning Glynn
[21:54:30] <Gman> hey hile_
[21:54:41] <hspaans> hile_: thats why them "hmmm" ;-)
[21:55:39] <sahafeez> ahah! gpl3 is evil
[21:56:29] <hspaans> really?
[21:56:52] <AbeFroman> but it stays crunchy in milk
[21:56:57] *** adlpaf has quit IRC
[21:59:31] <Error_e^ipi> i thought about it, and v2 isn't compatible with v3
[21:59:38] <Error_e^ipi> a v2 project can't contain v3 code
[21:59:43] <Auralis> correct
[21:59:57] *** rodrickbrown has quit IRC
[22:00:00] <Auralis> for the same reasons gplv2 is incombatible with cddl
[22:00:03] <Error_e^ipi> and since linus is opposed to v3, It takes away my main opposition to the whole thing
[22:00:04] <hspaans> FSF is telling people otherwise
[22:00:30] *** Gman has quit IRC
[22:00:34] <movement> nope, fsf have admitted it now
[22:00:35] <Error_e^ipi> which is to say, linux half-implementing the "compelling cool features" that help push people towards opensol
[22:01:21] <hspaans> movement: when? a dutch fsf laywer told something else in december
[22:01:47] <hspaans> but then again gplv2 is a problem outside the US
[22:02:07] <g4lt-U60> w00t, our first post-slashdotting license troll
[22:02:27] <richlowe> woohoo, a licensing argument!
[22:02:30] <hspaans> you mean me?
[22:02:43] <g4lt-U60> ya think?
[22:02:53] <richlowe> I say we set everyone on fire, and let the people who burn blue redistribute.
[22:03:10] <richlowe> at least that way the only ones annoyed won't be able to argue about why... ;)
[22:03:43] <sahafeez> there is no argument. glpv2 was a nice idea in order to promote code sharing and development. gplv3 is a political statement by RMS
[22:06:22] <Error_e^ipi> richlowe: who's trolling?
[22:06:28] <richlowe> I said someone was trolling?
[22:06:37] <richlowe> Oh, no, I didn't, g4lt did.
[22:06:49] <Error_e^ipi> oh, sorry
[22:06:50] *** nrubsig has joined #opensolaris
[22:06:50] *** ChanServ sets mode: +o nrubsig
[22:07:14] <Error_e^ipi> i didn't mean to start a flame war... i was just making an innocent comment :(
[22:07:19] <alanc> the people at the top know they can't change the license without talking to the community first, so until you see someone posting a proposal to do so on an opensolaris list, it's just pure fantasy and rumor-mongering
[22:07:39] <nrubsig> alanc: GPLv3.mod-Sun ?
[22:07:40] <b3stbuddy> why even have any gpl?
[22:07:51] *** Netwolf has quit IRC
[22:08:00] <alanc> (and of course, until GPLv3 is finalized, no proposal could be taken seriously - no one will adopt a license that's still being rewritten/changed)
[22:08:05] <b3stbuddy> just stamp it open source and let that be it
[22:08:21] <nrubsig> alanc: So the rumor is not true that Sun makes it's own GPL version ?
[22:08:26] <hspaans> b3stbuddy: public domain? ;-)
[22:08:33] <zdzichuBG> what for? cddl is good
[22:08:37] <alanc> why would Sun do that?
[22:08:47] <nrubsig> alanc: I don't know.
[22:08:54] <nrubsig> alanc: IRC rumor.
[22:08:57] <Error_e^ipi> if sun does it & really wanted to run the Linux war machine in full gear... they could grep through LKML for the pro v3 guys, remove those who are in redhat or IBM's employ, and invite them for a relaxing weekend + primer on Solaris kernel technologies at sun's campus
[22:09:10] <alanc> if GPL was used, it would be to be license-compatible with other GPL software
[22:09:16] <quasi> nrubsig: the rumor I hear is that they will go gpl v3
[22:09:20] <alanc> making another GPL variant would be useless
[22:09:44] *** cyberops has quit IRC
[22:09:47] <Error_e^ipi> which would be quite funny.... if a good chunk of linux's core developers jumped ship & came to solaris
[22:09:52] <quasi> gpl is also a way of making anyone but the fsf miserable
[22:10:08] <richlowe> For what it's worth, a couple of us have asked Simon and others for a joint Sun/CAB response to all this.
[22:10:44] <quasi> richlowe: that would be nice
[22:11:06] * quasi makes note to self about bugging Roy for that
[22:15:34] *** dunc has joined #opensolaris
[22:16:14] <sahafeez> what would happen is a large chunk of solaris code would end up in linux
[22:16:56] <sahafeez> and the way i read the cddl was that sun can do anything it wants. so if they say, this is now gplX then can
[22:17:12] <Error_e^ipi> sahafeez: v3 is v2 incompatible
[22:17:27] <Error_e^ipi> linux (as a v2 project) couldn't use opensolaris code (as a v3 project)
[22:17:30] <sahafeez> yes i know that. i do not like v3
[22:17:38] <sahafeez> ah, ok.
[22:17:45] <sahafeez> did not see the connection
[22:17:57] <hspaans> sahafeez: and why don't you like v3 btw?
[22:18:06] <Error_e^ipi> opensolaris could use linux's code, but i don't think there's anything in there we'd want
[22:19:06] <sahafeez> i do not like the forced restriction that are placed on code (re DRM, etc) even tho i hate DRM as i am a libertarian and i do not think you can force the politics down everyones gut.
[22:19:27] <richlowe> There's bits in linux that are pretty interesting, but there's not exactly much in the way of shared interface.
[22:19:31] <richlowe> want, maybe.  Could use? no.
[22:19:40] *** UnixTitan has quit IRC
[22:19:45] <sahafeez> like i said, GPLv3 is more of a political statement then anything else.
[22:20:22] <Error_e^ipi> tbh, i'm pretty ambivalent eithe way
[22:20:33] <Error_e^ipi> CDDL is cool, GPL3 is cool too
[22:20:53] <Error_e^ipi> *shrug* sun can do what they want with it, so long as it's still solaris i'm happy
[22:21:06] <sahafeez> i was pretty much a BSD guy but i have seen that the GPL has done alot more..
[22:22:03] <alanc> sahafeez: the CDDL doesn't allow Sun to do that - ownership of the code allows Sun to do that
[22:22:12] <sahafeez> ok
[22:22:26] <hspaans> sahafeez: it only closes the tivo loophole for drm, but the advantage of v3 vs v2 is that it tries to be more international and not based on the US legal system. v2 causes some problems in the Netherlands for example
[22:23:10] <alanc> (which is why contributors to opensolaris have to sign an agreement giving co-ownership to Sun, so it can relicense as necessary without tracking down hundreds of people like Mozilla had to)
[22:23:31] <Error_e^ipi> is that what i had to sign & scan & email back?
[22:24:15] <alanc> the Contributor Agreement?   you would have had to do that before submitting code to go in
[22:24:20] <Error_e^ipi> it was a big pain in the ass too... i don't have a scanner
[22:29:17] *** jamesd has quit IRC
[22:29:18] <richlowe> you have to do it before the code goes back, I think the barrier to sponsorship is merely due to bookkeeping.
[22:29:59] <richlowe> and I still think it's somewhat silly that you end up *having* to post to request-sponsor, even if you already have the sponsor you're requesting. :)
[22:30:38] <onbot> commit by Cynthia McGuire:  6513804 fmtopo assertion hit when hostbridge and pcibus enumerators can't construct module property
[22:33:18] <Error_e^ipi> yeah, it's kinna a silly way about going about adding to the project
[22:34:12] * tsoome is away - bbl - [AcidJazz - Log ON]
[22:34:28] *** jsubl2 has joined #opensolaris
[22:34:43] <Error_e^ipi> the "patch dumping ground" that other projects have Isn't too bad IMO... subsystem maintainers pick up patches as they see fit & integrate them
[22:35:38] <alanc> that assumes all subsystems have maintainers, which isn't the way Solaris has always worked
[22:35:54] <richlowe> I don't like the fire and forget approach for other reasons, too.
[22:36:10] <alanc> very few parts of Solaris have dedicated people who take long-term responsibility for a set of code
[22:36:20] <richlowe> but for something that's meant to be shrink-to-fit, shrinking the parts that involve doing things twice for the sake of bookkeeping by management would seem pretty good.
[22:43:17] <comay> http://blogs.sun.com/richgreen/entry/all_the_news_that_s
[22:45:33] *** istappning has joined #opensolaris
[22:47:29] <steleman> bummer. if the GPLv3 rumor was true, then we could have upgraded the discussion to GPLv4. now we're stuck with v3 until it's final.
[22:48:06] <delewis> I highly doubt the osnews.com and Slashdot commenters regularly read Rich's blog :-)
[22:48:52] <hspaans> steleman: its seems to be started by Tom Goguen, but maybe Microsoft Shared Source is also an option ;-)
[22:48:54] <steleman> yeah but discussing GPLv4 should yield numerous slashdot posts dont you agree ? there isn't even a draft yet, so everyone can speculate ahoy :-)
[22:50:06] <delewis> I can't imagine the hysteria if were GPLv2...
[22:50:09] <alanc> but at the rate of new GPL releases, GPLv4 should be finalized around 2020
[22:50:22] *** Gman has joined #opensolaris
[22:51:20] * boyd writes a note to self. "The star option -stats does not print stats, it replaces the first occurrence of 'a' in each pathname with 's' as the files are added. You don't want that"
[22:51:34] <alanc> heh
[22:51:46] <Gman> howdy all
[22:51:49] <alanc> why would you ever want that?
[22:51:51] <boyd> Took me a while to work out what had happened
[22:51:54] <boyd> Gman: Hey
[22:52:28] <boyd> alanc: I don't know... it's part of the generic -s/foo/bar mechanism, with user-selectable /s
[22:52:32] <quants> cmihai: you about?
[22:52:37] <alanc> ah, like vi & sed
[22:52:40] <boyd> yeh
[22:53:36] 
[22:54:09] <boyd> e.g. if you mis-type -fifostat you get it writing to a file called "ifostat"
[22:54:11] <alanc> just make star2 with a sane command line and leave star for compatiblity
[22:54:25] <boyd> That's my choice
[22:54:47] <boyd> It could be called "blar" :)
[22:54:50] <Gman> heh
[22:55:01] <Gman> pretty much how i feel about most tar commands :)
[22:55:07] <boyd> heh
[22:55:20] <alanc> yarr!   the pirate archiver...
[22:55:31] <boyd> hehe "aarrgghhh
[22:56:02] <hspaans> alanc: thank you. now I need to clean my screen
[22:56:15] <boyd> :)
[22:56:22] <alanc> 8-)
[23:05:45] <richlowe> alanc: it Joergifies your work. :)
[23:10:34] *** bougie has quit IRC
[23:10:43] <steleman> omg
[23:11:06] <steleman> star2 ?
[23:12:27] <boyd> We're just suggesting that it needs a command line revamp
[23:12:52] <steleman> the only thing missing in star's command-line language is that it is not Turing-complete.
[23:13:34] <steleman> yet.
[23:14:13] <stevel> star:tng
[23:15:31] <lasseoe> heh
[23:15:37] *** axisys has quit IRC
[23:15:40] <delewis> steleman, stranger things have happened and will happen. I suspect an ARC case to rewrite the kernel in ksh93 before the day is over with.
[23:15:59] <Gman> hey stevel
[23:16:00] *** slowhog has joined #opensolaris
[23:16:10] * Tpenta treis really hard and manages not to kick dlewis for that ;-)
[23:16:21] <delewis> Tpenta, reading up on ksh93, yet? :-)
[23:16:22] * delewis ducks
[23:16:29] <cap_> who needs star when afio exists ;-)
[23:16:42] <Tpenta> right now, given the amount to read on, I am avoiding it as I have too much on my plate
[23:17:02] <richlowe> 2007/035 seems to be going well.
[23:17:08] <richlowe> I imagine Roland is out dancing in the streets.
[23:17:22] <delewis> richlowe, that or fondling the komodo dragons of his.
[23:18:10] *** jamesd has joined #opensolaris
[23:18:10] *** ChanServ sets mode: +o jamesd
[23:18:33] <Gman> richlowe, probably a realization that the world (and jek3) really isn't out to get him :)
[23:18:34] <delewis> jamesd, still schooling users on DTrace in the osnews.com discussions? :-)
[23:18:55] <delewis> its amazing how uninformed people tend to be of the technologies out there.
[23:20:46] *** mdj has joined #opensolaris
[23:24:59] <quants> http://feeds.feedburner.com/~r/arstechnica/BAaf/~3/76505051/20070117-8635.html
[23:27:03] * mv have just ordered free solaris dvd
[23:28:37] <jamesd> delewis, of course.. perhaps they will visit my website.. and accidently click on an ad.. or add me to there daily read list.. in any case its a big win...
[23:28:52] <delewis> jamesd, :-)
[23:29:10] <delewis> I enjoyed the fellow that stated SystemTap could handle any task and failed to see the point of userland probes :-)
[23:29:30] <delewis> clearly, someone's brain isn't operating a peak capacity.
[23:29:48] <delewis> at, rather.
[23:30:21] <jamesd> yeap..   systemtap can't even print out a userland stack trace..
[23:30:49] <jamesd> but damm they have now redesigned kprobes to use 10 different methods in the last 2 years.
[23:31:16] <sahafeez> to many cooks
[23:31:45] <delewis> jamesd, poor planning or "hey, let's just do it this way for the hell of it and bestow it upon everyone else"
[23:31:46] <jamesd> and none of the cooks are willing to commit userland probes..
[23:31:47] <delewis> take your pick.
[23:32:12] <cap_> the kprobes part is actually quite good now
[23:32:13] <delewis> you can actually burn a cd now by "cat foo.iso > /dev/hda"
[23:32:19] <delewis> in Linux, that is.
[23:32:30] <cap_> ..but as you say, the userspace part is really lacking =/
[23:33:11] <delewis> so now the kernel is a CD recording application.
[23:33:19] * delewis sighs
[23:33:32] <cap_> jamesd, IBM has tried for than a few times with various ideas on userspace probes (show down in flames by the kernel guys for being ugly every time)
[23:33:40] <delewis> next, they'll be stuffing an X11 server in the kernel -- remind you of anyone else?
[23:33:53] <quants> "But there's a catch: the drug isn't patented, and pharmaceutical companies may not be interested in funding further research if the treatment won't make them a profit. In findings that 'astounded' the researchers, the molecule known as DCA was shown to shrink lung, breast and brain tumors in both animal and human tissue experiments." it's a sad day
[23:34:36] <quants> what the hell is wrong with people?
[23:35:11] <jamesd> cap_, i follow the servertap mailing list... 1/2 of the time they don't even post it to the lkml.. its just another damm way to probe the kernel... but hardly anyone is using any of it..   and if you look at the systemtap scripts you will think how is anyone going to accomplish real world tasks with this besides kernel coders
[23:35:16] <delewis> cap_, doesn't surprise me. Oracle has had a bit of trouble getting OCFS2 into the mainline.
[23:35:30] <delewis> though, the Linux VFS guys tend to be somewhat generous enough to tell you what's wrong and how to fix it.
[23:37:07] <onbot> commit by Hai-May Chao:  6509342 Code licenses for KMF need to be cleaned up.
[23:37:55] *** triplah_ has joined #opensolaris
[23:38:56] <cap_> jamesd, I don't quite agree that systemtap is useless for real world tasks (I use it frequently since it's part of cento-4 (EL4 clone)). IMHO it's good enough to be useful but far far from DTrace-level
[23:39:24] <jamesd> cap_, how often do your crash your system?
[23:39:43] <cap_> never happend acatully (unless I count trying to kill it on my dev system)
[23:40:06] <cap_> damn it I spell bad, should have gone home a few hours ago *grmbl*
[23:40:57] * delewis has a 6PM class here in awhile
[23:40:58] <delewis> :-(
[23:41:04] <hile_> damn
[23:41:07] <delewis> I'll get home about 10PM tonight
[23:41:08] <jamesd> lucky you... i find it all too bloody easy to crash...  just probe every  function in a module works wonders, or probe every function call in the main kernel and it reboots or gives me a nice oops
[23:41:38] <stevel> 'afternoon gman
[23:41:51] <delewis> jamesd, I'm guessing you can't take a SystemTap application and run it on a system with a different kernel can you?
[23:42:02] <cap_> jamesd, I usually use it for very safe and easy stuff (enough to be useful as I said). like syscalls and some tried and tested probepoints
[23:42:12] <jamesd> nope... it will have to be rebuilt though they are working on it...
[23:42:44] <cap_> eh?
[23:42:45] *** abackos has left #opensolaris
[23:43:26] <cap_> stap scripts are typically portable, the kernel module that gets built is only a temporary file that is hidden from the normal user
[23:43:35] <jamesd> cap_, you can't move a compiled systemtap module to a new kernel...  and most people dont like having gcc and friends installed on production machines that is why its a valid question.
[23:44:28] <cap_> ah, different world then, we have gcc on all machines I think and I would never think of moving a stap module
[23:44:39] <jamesd> of course it gets much worse if you try and use the functionality that you can probe any line in the kernel, if the line moves.. you have to recompute it.
[23:45:22] <cap_> 'nuff said, there's a day tomorrow too
[23:45:29] * cap_ went home
[23:45:57] <delewis> jamesd, different gcc versions can do funky things, too.
[23:46:20] <delewis> if a kernel is compiled with gcc version foo it might pack, say structures differently than it would if it were compiled with gcc version bar.
[23:46:30] <jamesd> of course... and you need to have a full kernel with all debugging code availible... and those aren't easy to find either.
[23:46:31] <delewis> I imagine that would effect SystemTap quite a bit.
[23:46:35] <Gman> oooh
[23:46:38] <delewis> jamesd, ugh!
[23:46:39] * Gman reads the kde thread
[23:46:42] <delewis> say it ain't so!
[23:47:27] <delewis> I don't mind having compilers on production system, provided there is a valid reason; however, compilers should *not* be necessary for functionality.
[23:47:32] <jamesd> delewis, it gets worse, systemtap in normal mode  doesn't know how to read a value out of a struct.. so to access it, you either have to create a tapset  or  enable dangerous mode...
[23:47:37] <alanc> enjoying Roland's attempt to hijack the project to twist it to what he wants?
[23:47:51] <cap_> jamesd, not true
[23:48:46] <jamesd> well if you have debugging information, you can do it, but with out the debugging information its not...  so userland or reverse engineering is useless...
[23:49:12] *** rodrickbrown has joined #opensolaris
[23:49:13] <cap_> systemtap requires debuginfo and gcc, that's just the way it is
[23:49:29] <jamesd> btw  gcc has an awful habbit of screwing up debugging information
[23:50:00] <delewis> jamesd, does Red Hat ship kernels with symbols?
[23:50:27] <cap_> delewis, kernel debuginfo is available in the kernel-debuginfo rpm on redhat
[23:50:28] <jamesd> i think you have to download the debugging information its about 15MB IIRC
[23:50:38] <cap_> jamesd, alot more than 15M
[23:50:43] <cap_> more like 115M
[23:52:31] <cap_> 280 megs for EL4u4 actually
[23:58:27] <sahafeez> syncing your phone for the 1st time over bluetooth takes to damn long.





top