[00:07:06] <trygvis> hm, adjusting the linux server actually worked
[00:07:22] <trygvis> now I just have to figure out why the directory is empty on the client side
[00:07:34] <trygvis> (I've loop mounted the b67 DVD
[00:15:38] *** bcrair has joined #opensolaris
[00:16:02] *** ed has left #opensolaris
[00:18:16] * alanc is waiting for the calls of sacrilege for throwing away type5 keyboards in our lab cleanup
[00:19:17] <dlg> wheres theh capslock and control keys on them?
[00:20:57] <alanc> mostly in the right places, though a couple had the PC layout
[00:21:25] <dlg> sacrilege
[00:21:43] *** devnull62 has joined #opensolaris
[00:22:02] <wesolows> who cares?  both should be mapped to control anyway
[00:22:11] <dlg> wesolows: YES
[00:22:15] <wesolows> the things that actually matter are Esc and Backspace
[00:22:37] <wesolows> (well, and |, which is related to backspace)
[00:22:48] *** USSJoin has joined #opensolaris
[00:23:25] <REllin1> let's put a [reset] key above backspace, like the old Apple ][... that was always fun...
[00:24:33] *** reidms has joined #opensolaris
[00:27:56] <coffman> how is the type 7 ?
[00:28:40] <wesolows> it's fine
[00:28:43] <wesolows> using one right now
[00:28:50] <richlowe> better than the 6, certainly.
[00:29:21] <wesolows> The keys don't have the great feel of the 5, nor does the whole thing have that solid construction, but it's quite nice - way better than the average PC keyboard, certainly
[00:30:45] <coffman> hard to get without a machine :(
[00:35:44] <wesolows> Hmm, maybe
[00:35:45] <richlowe> X3738A
[00:35:47] <richlowe> I think.
[00:35:53] <wesolows> I ordered them without, but that was using Sun's internal thingy
[00:36:16] <wesolows> richlowe: Yep
[00:36:21] <richlowe> coffman: hit sun.com, search country kit, hit products tab, scroll through, click.
[00:36:33] <wesolows> richlowe: Keyboard, mouse, and power cables (why power cables?  sigh)
[00:39:22] *** polk__ has quit IRC
[00:39:22] *** sparvu has quit IRC
[00:39:22] *** LeftWing has quit IRC
[00:39:29] *** sparvu has joined #opensolaris
[00:39:43] *** LeftWing has joined #OpenSolaris
[00:40:06] *** polk_ has joined #opensolaris
[00:40:09] <alanc> wesolows: because that's the part of the computer specific to each country - you buy one computer and a box with the appropriate local keyboard & power cord
[00:40:39] <wesolows> alanc: Yeah, I know the answer, it just makes little sense to me.
[00:41:12] <wesolows> alanc: But maybe having 1 P/N instead of 3 saves us a couple of pennies a year.
[00:41:17] <alanc> which is why last time I did a lab cleanup, I threw out tons of bizarre power cords, since we ordered international country kits to get the localized keyboards for testing
[00:42:22] <wesolows> ...and there went the savings
[00:43:51] <alanc> but the group responsible for producing them still saved money, so they met their goals
[00:44:14] <REllin1> the cables cost much, much less than the processing of the order
[00:44:40] *** trs81 has quit IRC
[00:44:45] <LeftWing> There's an ecological cost, though.
[00:45:00] *** trs81 has joined #opensolaris
[00:45:29] <REllin1> copper recycles quite nicely
[00:45:48] <LeftWing> Provided you actually recycle it.
[00:45:49] <sommerfeld> too easily, or else it wouldn't be stolen quite so often.
[00:46:13] <REllin1> just put the word out on the street and the tweekers will clean it up for you
[00:47:09] <LeftWing> Besides, there's still an ecological cost to the packaging, shipping and then disposal and recycling of unwanted power cables even if you can get 100% of the raw materials back in recycling.,
[00:47:14] <wesolows> Along with your catalytic converter
[00:47:26] <wesolows> LeftWing: Yep.
[00:47:35] <LeftWing> Shipping, like Friction, is a non-conservative field.
[00:47:52] <wesolows> LeftWing: There's probably a "green computing feedback" link somewhere on our web site.  I'd encourage you to offer that observation.
[00:48:30] <wesolows> Somehow every shop I've ever worked in ends up with a bin of 500 unwanted power cables even though every machine is fully cabled.
[00:48:37] <sommerfeld> ah, but in the common case where you do need kb, mouse, and power cord, i'd think that boxing them separatly would have greater overall cost
[00:49:23] <LeftWing> sommerfeld: When I replace a Desktop PC with a Sun Ray there are already two perfectly good IEC cables hanging out of the wall and pre-routed behind the desk that I certainly can't be arsed moving.  So the new ones land in a giant box of similar new ones.
[00:50:04] <Auralis> ntf ist gleich am anfang auf rm tanker losgegangen
[00:50:08] <LeftWing> Not to mention the operational cost of having to test & tag every single new cable that comes through prior to its use, so you can't just unbox and plug in anyway.
[00:50:33] <wesolows> LeftWing: That, and what percentage of shops actually use the 2-meter or 3-meter IEC13 cables that come with every box?  My guess is very few - we all have our own IEC14-IEC13 or 0.5m cables or whatever.
[00:50:35] <REllin1> on  a tangent, 2-way radio manufacturers are just now starting to make radios with wireless microphones (Bluetooth)... imagine that, a wireless wireless!
[00:50:44] <LeftWing> wesolows: Indeed.
[00:51:32] <wesolows> 2m cables are a recipe for a thoroughly unmanageable data center
[00:51:36] <LeftWing> Another non-conservative force is, it would appear, searching for Green feedback icons on sun.com.  =P
[00:53:11] <wesolows> or searching for anything else there, really.
[00:53:57] *** EchoBinary has quit IRC
[00:55:03] <LeftWing> It's OK provided your particular internal filing and sorting processes align with those of whichever committee is responsible for the thing. =P
[00:55:32] *** Fullmoon has quit IRC
[00:58:49] *** pfa3rh has quit IRC
[01:01:29] <boyd> ... this week
[01:01:32] <boyd> (Morning, all)
[01:01:44] <LeftWing> Morning boyd
[01:04:11] *** stevel has quit IRC
[01:04:37] <boyd> Hmm... FWARC 2007/417 is the first time I think I've seen a case denied (I know, I'm a n00b)
[01:05:19] <LeftWing> What is it?
[01:05:20] <alanc> it happens, but it's rare that something is so broken and the team so unwilling to fix it that a case gets denied instead of modified
[01:05:27] *** Plaidrab has joined #opensolaris
[01:06:00] <boyd> LeftWing: "Openboot NVRAM configuration variable to" That's all I can see
[01:06:08] *** nachox has joined #opensolaris
[01:06:09] <wesolows> or, possibly (though even less likely), the ARC is broken and unwilling to fix itself
[01:06:14] <nachox> evening people
[01:06:20] <boyd> Evening nachox
[01:06:57] <LeftWing> boyd: Heh.  *LOMs, perhaps?
[01:07:14] <LeftWing> oh, "variable to".  I read that as "available to".
[01:07:16] * boyd shrugs... could be coffee cup warmer for all I know
[01:07:35] <LeftWing> Coffee! Brilliant.
[01:07:38] * LeftWing leaves for Uni.
[01:07:39] <alanc> it's a closed case, since it involves a 3rd party's hardware product that they haven't granted permission to publish on OpenSolaris
[01:08:14] <boyd> Does that mean the third party had a broken idea. I suppose so.
[01:09:03] <wesolows> it's unclear whether this would even be visible to OpenSolaris anyway.
[01:12:07] *** SirFunk has quit IRC
[01:13:49] *** nemesis has quit IRC
[01:14:48] *** SirFunk has joined #opensolaris
[01:14:54] <sommerfeld> boyd: it concerns a tunable to work around broken hardware.  there's a right place to put such tunables, and the project team didn't want to put it there
[01:16:29] <nachox> /etc/system?
[01:16:51] <wesolows> the hardware is apparently much, much more broken than that
[01:17:14] <nachox> or the conf file of the hardware?
[01:17:25] <wesolows> (which is to say, it's a FWARC case - so whatever needs to happen needs to happen at a lower level than /etc/system)
[01:19:11] <Plaidrab> alanc: Are the IPLT ( Directory Server 5.2 on, etc ) packages named differently now? ( Says someone who was finally able to get boxes patched up to 5.1 SP 4 at work today....)
[01:19:37] <alanc> Plaidrab: I don't know - I never install JES
[01:19:58] <Plaidrab> heh
[01:20:04] <jbk> IPLT was the ones bundled with solaris 9
[01:20:05] <alanc> have no need for any of that stuff
[01:20:16] <jbk> JES packaged them as SUNW packages
[01:24:04] <Plaidrab> Heh. Hell, I'm still running on a site that installs them from scripties and no packages.
[01:25:16] <nachox> how do you patch that?
[01:25:25] <sommerfeld> it's the sort of thing that on a PC would be fixed with a new bios
[01:27:47] <Plaidrab> : catches up on two weeks of Indiana mail
[01:28:58] <nachox> ouch
[01:29:14] <tomww> Plaidrab: hope you'll survive
[01:29:30] *** jpdrawneek has quit IRC
[01:29:33] <Plaidrab> I have little emotional investment. It's mostly amusing.
[01:30:06] *** RobG has joined #opensolaris
[01:30:16] <tomww> Plaidrab: why?
[01:31:16] <RobG> Has anyone managed to get a build of GCC 4.2 or above working?
[01:32:30] <RobG> SS12 has a bug (with no useful workaround) and I need OpenMP support hence 4.2
[01:32:31] <Plaidrab> I find the conflicting viewpoints and the paradigm clash sometimes entertaining in ways they aren't meant to be.
[01:32:47] <nachox> use an older sun studio?
[01:32:59] <RobG> same problem
[01:33:08] <nachox> 10?
[01:33:21] <nachox> ohh, nm :)
[01:33:22] <Plaidrab> The prstat vs top conversation was most enlightening, for example.
[01:33:52] <boyd> I missed that onw
[01:33:53] <boyd> one
[01:33:55] <nachox> Plaidrab, where did that end?
[01:34:05] <Plaidrab> I didn't see that it did.
[01:34:32] <Plaidrab> They narrowed it down to about 8 possible solutions though. :)
[01:35:05] <alanc> Indiana discussions don't come to a decision, they just peter out as people tire of arguing and realize there is no mechanism for them to make a decision one way or the other
[01:35:30] <Plaidrab> A more specific mandate might solve some of that. :)
[01:35:33] <RobG> I would be happy with something that just says "Sorry no TOP, use PRSTAT instead"
[01:35:43] <wesolows> RobG: That was my answer.
[01:36:06] <RobG> wesolwos: Are you Bay area?
[01:36:07] <wesolows> RobG: Then a bunch of second- and third-raters chimed in with random whinging which I proceeded to ignore.
[01:36:13] <wesolows> RobG: Of course.
[01:36:22] <Plaidrab> I mostly agree. But there are some things in top prstat doesn't - to my knowedleg do. Those things should be added, IMNSHO.
[01:36:39] <RobG> I'll be there Aug 21/22... be nice to meet face 2 face... my shout ;-)
[01:36:46] <wesolows> Plaidrab: sure.  File RFE, write code, putback.
[01:37:10] <wesolows> RobG: Le me know where you'll be; if it's nearby I'll join you for a beer.
[01:37:15] <Plaidrab> Sadly, I'm not that level of hacker, wesolows.
[01:37:24] <wesolows> RobG: wesolows at eng dot sun dot com
[01:37:30] <Plaidrab> St.Johns and a trip to Fry's YAY!
[01:38:54] <Plaidrab> Sorry. Flashbacks. :)
[01:40:42] <Plaidrab> Anyway. Yeah. prstat is generally better. A symlink or set of default aliases "top='echo You really want to use prstat.'" heh
[01:40:59] <Samy> Hi guys
[01:41:08] <Samy> I have a question about Solaris's NUMA support
[01:41:19] <Samy> Does it have user-space numa management facilities?
[01:41:25] <Samy> In Linux, I can use numastat, for example
[01:41:32] <Samy> To look at number of remote fetches, local fetches
[01:41:35] <wesolows> Plaidrab: ahh, but how is that message i18n'd? :-)
[01:41:36] <Samy> Pages interleaved, etc...
[01:41:48] <Samy> I know Solaris supports some simple NUMA policies, but I would like to monitor some of its effects.
[01:42:06] <Samy> and I rather work on Solaris than Linux since I'm more familiar with it at this point ;]
[01:42:29] <alanc> wesolows: don't ask unless you want nrubsig to explain why ksh93's built in translation support is so much better than gettext(1)
[01:42:46] <wesolows> Samy: Interesting question - I'm looking for an answer for you.
[01:42:52] <Samy> wesolows, I'm looking with you :-)
[01:43:00] <Samy> wesolows, using http://blogs.sun.com/jjc/ as a starting point
[01:43:22] *** Ignacio_ has joined #opensolaris
[01:43:24] <wesolows> I'm wondering if there are kstats for this.
[01:43:35] <Plaidrab> I dunno. I won't use ksh though.
[01:43:42] <Samy> wesolows, hmmm, let's see
[01:43:43] <asyd> 2A/coin
[01:43:44] <asyd> oups
[01:43:45] <asyd> \_o<
[01:44:23] <wesolows> Plaidrab: The point is that a complete solution is much harder than it seems.  Which makes adding features to prstat and just not shipping top seem like a winner.
[01:44:57] <richlowe> I don't see anything in the obvious place.
[01:45:29] <Plaidrab> Would it be that much harder to add the features, have it look at arg0 for some beavhior options, and then ship it with a hardlink as both?
[01:45:42] *** nachox has quit IRC
[01:45:45] <sommerfeld> Samy: pmap -L  will show where memory is
[01:45:50] *** Ignacio_ is now known as nachox
[01:46:09] <richlowe> Yeah, numa observability add a lot of basic stuff
[01:46:14] <sommerfeld> lgrpinfo ("latency group") will show memory structure
[01:46:16] <richlowe> but it doesn't seem to go into the detail mentioned above.
[01:46:28] <wesolows> Plaidrab: That's another fairly easy solution, but it doesn't really solve the problem - that people are used to typing "top" and that the people who make the other "top"s they'll use may change it at any time and then ours will seem "weird" again.
[01:46:34] *** EchoBinary has joined #opensolaris
[01:46:54] <sahafeez> http://sandiego.craigslist.org/sys/376902382.html
[01:46:54] <Samy> sommerfeld, hrm
[01:46:54] <wesolows> hmm, my inability to type lgrp correctly explains my lack of success
[01:46:59] <tomww> RobG: I comipled gcc 4.2.0 with the CBE (SFEgcc.spec)
[01:47:04] <Samy> richlowe, I am pretty sure I can use dtrace though
[01:47:21] <Samy> richlowe, I'll just hook into the appropriate function for remote memory fetch
[01:47:42] <Samy> richlowe, though a majority of that is actually done by HT. Nevertheless, the address map is in kernel space somewhere for sure
[01:47:49] <Samy> screw it, I'll install solaris anyways ;]
[01:47:54] <Samy> Better off for this research anyways
[01:48:02] <Samy> We already have linux performance results (which blows)
[01:48:16] <Samy> the only thing that'll suck is Solaris doesn't have a notion of vsyscall/shared pages for things like gettimeofday
[01:49:50] *** reid_ has joined #opensolaris
[01:50:29] <Samy> Ok, thank you.
[01:50:44] <Samy> wesolows, if you do find additional information, it'll be appreciated. If I find out how, I'll also do the same :-)
[01:50:45] <Plaidrab> True. But you have to spend a few cycles thinking about how a potential convert will react to the "lack" of perceieved missing "unix" tools.
[01:50:46] * Samy poofs
[01:51:01] <nachox> aparently sch was actually working in package management
[01:51:23] *** karrotx_ has joined #opensolaris
[01:51:57] <RobG> tomww: excuse my ignorance but where is the spec file stored?
[01:52:12] <Samy> hi RobG
[01:52:20] <RobG> Yo Samy
[01:52:40] <tomww> RobG: a svn repo on sourceforge, let me look for a link
[01:52:43] <RobG> Hows the project going... or how's the blogging going ;-)
[01:53:03] <Samy> my goal is to write a numa-aware memory allocator (patch)
[01:53:08] <tomww> RobG: the build-tools are pkgbuild/pkgtool from the GNOME team/and others possibly
[01:53:09] <Samy> I think I'll have to port jemalloc
[01:53:22] <asyd> hmm, just for my info, is there a simple way to show blocks sectors numbers mark as unusable by ufs?
[01:53:24] <Samy> RobG, doing interesting things
[01:53:30] <Samy> RobG, I will start the project officially next week
[01:53:51] <Samy> RobG, I am sort of secure with $ for rent now, so I know I won't be homeless and can afford to do work that doesn't yield $ :-)
[01:55:31] <RobG> Samy: I remember being a student with no money... now I'm a husband and I still have no money thanks to 'she who must be obeyed'
[01:55:50] <tomww> RobG: if you follow paragraphs 3) and 4) from this link: http://sarine.nl/gmpc-solaris-build   --- a direct link into the sourceforgce-repo follows
[01:55:55] <Samy> RobG, why "blogging"? I usually blog just to post plots and/or patches. I closed kerneled.org to the public now due to some NDA-covered plots for now.
[01:56:09] <Samy> But I have it hosted on blogger.com now.
[01:56:15] *** Fish has quit IRC
[01:56:27] <Samy> RobG, well, you're not at the risk of being homeless ;-P
[01:56:42] * Samy was living the ghetto life just recently
[01:56:42] <RobG> Samy, So we can share in you GSoC journey
[01:57:07] <Samy> RobG, the NDA results are related to Sun products :-P
[01:57:21] *** sahafeez has quit IRC
[01:57:36] <Samy> RobG, I'll give you + others access since this is the case.
[01:57:42] <RobG> tomww: thx will try it
[01:57:43] <tomww> RobG: http://pkgbuild.svn.sourceforge.net/viewvc/pkgbuild/spec-files-extra/trunk/
[01:57:55] <Samy> Blogger won't allow me to open up some "tags", either the whole thing is open or closed
[01:58:16] <RobG> Samy, yeah I've had teh same problem
[01:58:37] *** karrotx_ has quit IRC
[01:59:07] *** reidms has quit IRC
[01:59:20] *** perlmonk has joined #opensolaris
[01:59:44] *** dgtl__ has quit IRC
[02:00:16] <Samy> Solaris is really nice.
[02:00:58] <nachox> and just wait till indiana is released <wink>
[02:01:54] *** karrotx_ has joined #opensolaris
[02:04:56] *** Gman has joined #opensolaris
[02:06:34] <Samy> wesolows, http://opensolaris.org/os/community/performance/numa/observability/dtrace/
[02:07:17] *** darehanl has joined #OpenSolaris
[02:07:17] *** stevel has joined #opensolaris
[02:07:18] *** ChanServ sets mode: +o stevel
[02:07:25] <Gman> hi stevel
[02:07:30] <stevel> hey glynn
[02:08:22] <RobG> Samy, I don't know if this is of any use to you http://developers.sun.com/solaris/articles/libumem_library.html
[02:08:48] <wesolows> Samy: Yeah you can certainly do whatever you want with fbt, but it won't be a stable interface like a hypothetical lgrpstat command might be.
[02:09:05] <stevel> lgrpstat?
[02:09:10] *** bcrair has quit IRC
[02:09:29] <wesolows> stevel: The nonexistent command that tracks lgrp-related activity over time.
[02:09:40] <wesolows> Possibly implemented using kstats and/or DTrace
[02:09:44] <stevel> interesting.
[02:09:53] <stevel> you should mention that to the observability community
[02:09:54] *** alanc is now known as alanc-away
[02:10:02] <stevel> i know they've put a lot of effort into NUMA/lgrp observability
[02:10:22] <Samy> wesolows, yes.
[02:10:31] <Samy> wesolows, I don't care for stability though.
[02:10:42] <wesolows> Samy: sounds like you've found a winner then :-)
[02:10:48] <Samy> :-D
[02:10:59] <Samy> wesolows, groveling through lgrp API now.
[02:11:13] <Samy> For this project to succeed, I'll need the ability to associate memory to a node.
[02:12:25] <Samy> Yes.
[02:12:28] <Samy> Very nice API :-)
[02:12:45] <Samy> I really expected Linux do me further ahead in ccNUMA.
[02:12:52] <Samy> * do me = to be
[02:12:53] <Samy> heh
[02:13:13] <nachox> ahead of what?
[02:13:38] <elektronkind> my god this array is kicking my butt
[02:13:47] <elektronkind> how hard is it to set it's management IP
[02:13:56] <Tempt> That's quite an array to be kicking.
[02:13:57] <stevel> samy: madvise()
[02:14:40] <Samy> stevel, yes. That is rather trivial though.
[02:14:56] <Samy> stevel, the lgrp API however, is very powerful compared to Linux's libnuma and cpuset stuff.
[02:15:03] <Samy> nachox, ahead of Solaris.
[02:15:06] <stevel> i know :)
[02:15:16] <stevel> i was saying madvise() was your API to associate memory to a node
[02:15:25] <stevel> sorry, i missed earlier conversation - so i might be missing context here
[02:16:25] <nachox> Samy, only hurd is ahead of solaris ;)
[02:16:56] <Samy> Solaris is pretty under-rated IMHO.
[02:17:10] <Samy> Most people I've met think FreeBSD and/or Linux almost ways > Solaris.
[02:17:15] <Samy> I've only seen this in specific cases :-)
[02:17:30] <Samy> Anyhow, I'll be rerunning the suite of tests I used for Linux. It'll be interesting to see lgrp effect on this.
[02:17:54] <Samy> Solaris x86 community download = x64 + NUMA, BTW? :-/
[02:18:16] *** karrotx has quit IRC
[02:18:20] <Samy> It did x64 on my EM64T box.
[02:18:21] *** karrotx_ has quit IRC
[02:18:27] <Samy> Though that was a while ago.
[02:18:30] *** karrotx has joined #opensolaris
[02:18:31] <stevel> you'll get NUMA support if you have a NUMA box, yeah
[02:18:44] *** karrotx_ has joined #opensolaris
[02:19:08] <nachox> stevel, i thought madvice was usefull to tell the os whether to cache, and readahead or not
[02:19:23] <elektronkind> no need to recompile the kernel to get it, or apply some odd patches to enable it... :)
[02:19:26] <stevel> nachox: that to, but you can apply NUMA specific advice as well
[02:20:28] <nachox> such as memory locality? that is interesting
[02:20:37] <stevel> or rather, you can apply advice that will affect NUMA-ness
[02:20:40] <Samy> nachox, you can use it to define various memory policies. Even things like page migration, etc...
[02:20:45] <stevel> actual affinity can be done with lgrp_affinity_set()
[02:20:59] <Samy> nachox, look at the "Performance Guidelines for AMD Athlon 64 and AMD Opteron ccNUMA Multiprocessor Systems"
[02:21:14] <Samy> stevel, madvise page migration can be done with lgrp?
[02:21:36] <stevel> nachox: as in MADV_NORMAL vs. MADV_RANDOM vs. MADV_SEQUENTIAL will affect which lgroups memory is allocated from
[02:21:37] <nachox> Samy, you dont happen to have the link handy do you?
[02:21:41] *** vmlemon has quit IRC
[02:21:42] <Samy> nachox, sure
[02:21:52] <stevel> you can imagine MADV_SEQUENTIAL and MADV_RANDOM would result in very different allocations from lgroups
[02:21:53] <Samy> nachox, http://www.google.com/url?sa=t&ct=res&cd=1&url=http%3A%2F%2Fwww.amd.com%2Fus-en%2Fassets%2Fcontent_type%2Fwhite_papers_and_tech_docs%2F40555.pdf&ei=k_-fRoHHDJLEeYGlpNQB&usg=AFQjCNEvOaX5JZIA566vzE20NdalBWM5qw&sig2=Ja0C0E7tw7gCqjsvyPwhZg
[02:21:56] <Samy> nachox, eeek
[02:21:59] <nachox> hehe
[02:22:20] <Samy> Ugly URL from AMD, too :-)
[02:22:25] <stevel> samy: it can be affected - but if you want more deterministic page migration, i think lgrp_affinity_set() is what you want
[02:22:34] <Samy> stevel, ok, cool.
[02:22:42] *** nrubsig has joined #opensolaris
[02:22:43] *** ChanServ sets mode: +o nrubsig
[02:22:54] <Samy> stevel, do you have gmail, BTW? or a google account?
[02:22:54] <jbk> what are you trying to do?
[02:23:04] <Samy> stevel, I think your feedback would be appreciated for this stuff.
[02:23:06] <nachox> stevel, i vagely remember some references in the numa chapter in solaris internals
[02:23:17] <stevel> pmadvise(1) and plgrp(1) are useful from the command line too
[02:23:24] * nrubsig curses Solaris's NUMA support.
[02:23:32] <Samy> nrubsig, ?!!!!
[02:23:39] * nachox stabs nrubsig and hides him in a closet
[02:23:46] <nrubsig> Samy: long story.
[02:23:51] <elektronkind> oh do not mind nrubsig... he will curse just about anything
[02:23:51] <Samy> I'm here
[02:23:52] <stevel> samy: nope, don't have one.  feedback for what?
[02:23:54] <Samy> Install will take a while
[02:23:57] * nrubsig is already dead.
[02:24:00] <Samy> stevel, ccNUMA "stuff".
[02:24:06] <stevel> stuff in what sense
[02:24:07] <stevel> ?
[02:24:11] <Samy> stevel, meta-project for ccNUMA, bunch of sub-projects.
[02:24:34] <Samy> stevel, well, one is developing a formal model, one is a QoS model for memory placement (+ NUMA aware memory allocators)
[02:24:39] <nrubsig> !seen sch
[02:24:43] <Samy> stevel, 2 others are Linux-specific
[02:24:54] <stevel> interesting. you should solicit feedback from the Observability & Performance communities
[02:25:00] <Samy> stevel, and the last one is more focusing on the AMD64 ccNUMA architecture itself
[02:25:17] <stevel> that's what i worked on :)
[02:25:18] * nrubsig curses the malloc()-happy Solaris kernel
[02:25:25] <Samy> stevel, I can't "open" it up, just yet. Atleast not until the end of the month.
[02:25:32] <Samy> Next month, that is.
[02:25:36] <Samy> End of August ;-p
[02:25:42] <Samy> stevel, what did you work on?
[02:25:43] <elektronkind> nrubsig: memory exists to be used!
[02:25:49] <stevel> seriously - check with the performance community, they will be interested
[02:25:54] <stevel> amd64 numa support in solaris
[02:25:59] <Samy> Hacker
[02:26:03] <stevel> i was the NUMA test guy
[02:26:06] <nrubsig> elektronkind: yes, but not the way how the Solaris kernel abuses the allocator.
[02:26:09] <Samy> stevel, haha
[02:26:18] <Samy> stevel, and you can't talk about this, I assume, NDA?
[02:26:26] *** RobG has quit IRC
[02:26:31] <stevel> no, all the source code is public.
[02:26:38] <stevel> open source, rather
[02:26:40] <Drone> I've never seen sch talk in #opensolaris.
[02:26:54] <stevel> I've never seen a bot lag so much in #opensolaris.
[02:26:57] <stevel> :-P
[02:26:58] <nrubsig> elektronkind: the mmeory allocation scheme used in Solaris assumes that memory allocations within the kernel are "free", but in reality they are quite heavywheight calls which in turn cause scalability problems.
[02:27:08] <Samy> stevel, well, the situation is I can't "open" the results myself :-) I can open them through a gateway, aka: you. End of August I can approach a community at large myself.
[02:27:24] <stevel> ??
[02:27:27] * stevel doesn't understand
[02:27:36] <nrubsig> elektronkind: almost every damn syscall in Solaris triggers a memory allocation for temporary memory and this is IMO _perverse_
[02:27:42] <nrubsig> (sorry)
[02:28:03] <Samy> stevel, the sponsor's contract ends next month, they don't want information related to them or their "big picture" of the work to be public I guess.
[02:28:13] <Samy> stevel, we are working on the general ccNUMA problem now.
[02:28:22] * richlowe is now more puzzled
[02:28:23] *** Plouj has quit IRC
[02:28:27] <stevel> i'm completely confused
[02:28:32] <stevel> what are you talking about?
[02:28:33] <stevel> what sponsor?
[02:28:36] <stevel> what work?
[02:28:37] <Samy> haha
[02:28:52] <Samy> The ccNUMA stuff my research group is working on :-)
[02:28:59] <richlowe> but you haven't said what that is.
[02:29:02] <Samy> Sponsor = the organization that is funding the work
[02:29:03] <richlowe> or is that what you can't say?
[02:29:12] <richlowe> in which case, how is this supposed to make sense? :)
[02:29:18] <Samy> richlowe, I did say, above. Exactly what the sub-projects are, that are Solaris related.
[02:29:31] <nachox> all but two
[02:29:54] <nachox> who is the sponsor? now i really want to know :)
[02:30:09] <Samy> ya ok
[02:30:18] <Samy> I'll go watch the install now :-P
[02:30:19] <Samy> Bye
[02:30:22] <jbk> setec astronomy :)
[02:30:41] <nrubsig> Samy: You're Bezroukov's pupil, right ?
[02:31:37] *** simford has joined #opensolaris
[02:32:32] <Samy> nrubsig, no
[02:32:34] *** Yamazaki-kun_ has joined #opensolaris
[02:33:11] <Samy> I'm a GRA, but not a graduate student. So the head of the group is not really my adviser (I actually work for another department).
[02:35:05] <nrubsig> wesolows: ping!
[02:36:21] *** danv12 has joined #opensolaris
[02:37:05] <nrubsig> Is there anyone else who knows the kernel details behind |mmap()| ?
[02:37:58] *** perlmonk has quit IRC
[02:38:07] 
[02:39:52] <nrubsig> Is there a public holiday in the US today ?
[02:40:19] <asyd> nrubsig: the source ? :p
[02:40:32] <jbk> nope
[02:40:47] <nrubsig> asyd: Erm, the source won't help.
[02:41:16] <Samy> nrubsig, what is your question?
[02:41:46] <nrubsig> asyd: I am trying to figure out whether it would be usefull to add a new flag to |mmap()| which says "return zeros for memory instead of a SIGSEGV signal if the page is inaccessible, e.g. file size changes etc."
[02:42:40] <nrubsig> s/the page/a page/
[02:42:49] <nrubsig> Samy: see above.
[02:42:52] <Samy> nrubsig, hrm. That sort of ruins the semantics of mmap IMHO. Maybe a flag to deliver a non-fatal signal.
[02:43:20] <nrubsig> Samy: yes, that may be good enougth.
[02:43:35] <Samy> nrubsig, how would you return zeros? Common zero-filled page, for example? (on phys)
[02:43:52] <Samy> nrubsig, because I don't think that'll be secure...again, ignoring how semantics will be changed.
[02:44:18] <nrubsig> Samy: problem is that some code like the libc locale functions would benefit from using |mmap()| but they can't because there is the SIGSEGV "gap" which may result in serious problems in some cases.
[02:44:39] <nrubsig> Samy: yes, zero-filled page.
[02:44:40] <Samy> nrubsig, other applications need to know of this situation, 0-filled page just bonks things.
[02:44:58] <nrubsig> Samy: SIGSEGV bonks things even worse.
[02:45:10] <Samy> nrubsig, atleast for my I/O library, for example, I do aggressive buffering, I rather segfault if another source changes the file
[02:45:17] <Samy> nrubsig, think of parallel I/O :-)
[02:45:24] <Samy> That's worse. It just ruins consistency completely.
[02:45:47] <nrubsig> Samy: for example if an application is currently running and using l10n catalogs and another thing (like patchadd) changes the catalog the underlying shell should not crash
[02:45:52] <Samy> SIGSEGV can be caught nrubsig, I can look at previous frame(s) to see the source.
[02:45:53] <nrubsig> with SIGSEGV
[02:46:03] <nrubsig> Samy: I know.
[02:46:03] <Samy> Something other than SIGSEGV though, would make such a job much easier.
[02:46:10] <nrubsig> Samy: but recovery is not simple
[02:46:12] <Samy> yes
[02:46:32] <Samy> So, better a non-fatal signal, no?
[02:46:37] <nrubsig> yes
[02:46:38] <nrubsig> maybe
[02:47:04] <nrubsig> Samy: problem is that it should not result in a signal mass-storm for each single page access
[02:47:45] <nrubsig> Samy: my idea is to make |mmap()| safe to use in situatoins like described above.
[02:47:50] <Samy> Yes, that would suck. However, atleast for my code, I rather this behavior than zero-filled.
[02:48:04] <Samy> In my case, for parallel I/O, this would be bad.
[02:48:16] *** syndrome71 has joined #opensolaris
[02:48:19] *** hile_ has joined #opensolaris
[02:48:26] <nrubsig> Samy: I discussed the issue with Sun's i18n engineers and they said they can't use |mmap()| because this would generate problems for patchadd&co.
[02:48:35] *** syndrome71 has left #opensolaris
[02:48:36] <Samy> SIGSEGV is very good there (internal state can track growth of buffer across threads though, so that isn't the problem, problem is a system error).
[02:48:47] <Samy> nrubsig, so add both? :-P
[02:48:48] *** syndrome71 has joined #opensolaris
[02:49:05] * nrubsig scratches his head
[02:49:13] <Samy> nrubsig, but think, again.
[02:49:23] <Samy> nrubsig, so, then, every process will have this zero-filled page.
[02:49:38] <nrubsig> Samy: this won't hurt for l10n catalog files.
[02:50:28] <nrubsig> Samy: my problem is that the "we can't use |mmap()| because it may fail with SIGSEGV" kills a huge performance improvment and would solve some file descriptor issues with open catalog files, too.
[02:50:52] <Samy> ic
[02:51:08] <nrubsig> s/it may fail with SIGSEGV/it may trigger a SIGSEGV later/
[02:51:15] 
[02:51:20] <Samy> I see.
[02:51:21] 
[02:51:38] <Samy> Beta in no context is as good as ? :-P
[02:51:40] *** syndrome71 has left #opensolaris
[02:52:06] <Samy> nrubsig, aka: ic = I see
[02:52:08] <nrubsig> Samy: and |mmap()| would allow that the pages for catalog files could be shared between processes instead of swallowing-up lots of memory
[02:52:20] <Samy> nrubsig, yes.
[02:52:25] <Samy> nrubsig, it is really nice.
[02:52:54] <Samy> nrubsig, so, obviously, that would be your only choice then? If that flag is used, you also allocate a zero-filled page for those cases (and can map a complete range to this).
[02:52:55] *** blueandwhiteg3 has joined #opensolaris
[02:52:58] <sommerfeld> if I'm not mistaken that's not a beta
[02:53:01] <nrubsig> Samy: it's nice but it may mean I would have to declare war against the kernel folks
[02:53:13] <sommerfeld> it's a letter that the germans have retired
[02:53:23] <nrubsig> sommerfeld: ?!
[02:53:48] <richlowe> no more double-s?
[02:53:50] <Samy> nrubsig, why?
[02:53:55] *** darehanl has quit IRC
[02:53:59] <Samy> nrubsig, if it is an optional non-default flag.
[02:54:58] <nrubsig> Samy: the "better choice" in such a case would be to return zero-filled pages instead of carpet-bombing the calling process with SIGSEGV
[02:55:11] *** Yamazaki-kun_ has quit IRC
[02:55:21] <nrubsig> Samy: just a feeling.
[02:55:22] <Samy> Zero-filled page
[02:55:26] <Samy> Rather than zero-filled pages
[02:55:44] <Samy> It is semantically...fucked up :-) However, it'll open a door for other applications to make use of mmap.
[02:55:51] <Samy> So that's good imho.
[02:56:23] <Samy> nrubsig, a shared page per-process rather than pages is nicer if 0 values = ignored.
[02:57:08] *** karrotx has quit IRC
[02:57:12] *** karrotx_ is now known as karrotx
[02:57:18] <jbk> well if i understand the problem, it's that if the message catalogs are modified while in use, it could cause applications to segfault (assuming it used mmap)??
[02:57:19] <nrubsig> Samy: Did you ever read the GNU grep manual page ? It has the following section:
[02:57:22] <nrubsig>        --mmap If possible, use the mmap(2) system call to read input, instead of the default read(2)  system  call.   In  some  situations,  --mmap
[02:57:23] <nrubsig>               yields  better  performance.  However, --mmap can cause undefined behavior (including core dumps) if an input file shrinks while grep
[02:57:24] *** karrot-x has joined #opensolaris
[02:57:26] <nrubsig>               is operating, or if an I/O error occurs.
[02:57:38] <nrubsig> jbk: yes
[02:57:52] <jbk> if so, what is the behavior if someone modifies a .so that's currently in use?
[02:58:06] <nrubsig> jbk: that's because the code changes.
[02:58:22] <Samy> nrubsig, I know this.
[02:58:31] <Samy> jbk, those are COW'd.
[02:58:33] <nrubsig> jbk: if you use /usr/bin/tar it will delete the file and creates a new inode... but there is still a race
[02:58:50] <Samy> jbk, atleast in FreeBSD, you have a shadow chain. If the executable is modified, then that vm object is chained off from.
[02:59:17] <nrubsig> Solaris is AFAIK not that forgiving.
[02:59:18] <jbk> cause to me, if you can match the behavior of .so's, i think that'd at least have a decent argument
[02:59:26] <Samy> jbk, so [original vnode] <- [vnode with changes]
[02:59:34] <Samy> and dependents chain off the vm object they had
[02:59:51] <nrubsig> jbk: catalog files are normal files and the kernel cannot detect it's special use.
[03:00:22] <nrubsig> jbk: and it won't help for cases like GNU grep (see above, --mmap switch)
[03:02:07] *** blueandwhiteg3 has quit IRC
[03:02:59] <jbk> though
[03:03:12] <jbk> well never mind, not relevant
[03:10:02] *** seanmcg has quit IRC
[03:10:21] <boyd> Anyone want to buy this for me? http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=270146164488
[03:10:21] *** seanmcg has joined #opensolaris
[03:11:17] <Plaidrab> mmm. no
[03:16:13] *** xuewei has joined #opensolaris
[03:28:01] *** Jondice has quit IRC
[03:28:46] <steleman> i need a new video card, and i found one: http://www2.pny.com/NVIDIA-Quadro-FX-5500-SDI-PCIE--P1818C38.aspx
[03:29:17] <steleman> it is only USD $6,500.00
[03:29:24] <steleman> which is a bargain if you think about it
[03:29:28] <steleman> :-P
[03:30:05] <hile_> wtf?
[03:30:11] <steleman> ROFLMAO
[03:30:15] <jamesd_> steleman, buy two, save twice the money, making it twice the bargain and give one to me, my b-day is coming up and you will get a nice big mention in my blog.
[03:30:26] <steleman> hile_: you can buy two of them if you want to use SLI :-P
[03:31:17] <hile_> I still don't see a need for your box to have a framebuffer... that's what Rays are for.
[03:31:35] <steleman> these cards perform very well for 3D interactive pr0n
[03:31:49] <hile_> heh
[03:32:10] <steleman> hile_: i don't need a Ray either. vt100 is da bomb.
[03:33:04] *** rpedro_ has joined #opensolaris
[03:33:17] <rpedro_> hello
[03:34:09] <jamesd_> hi
[03:34:42] <rpedro_> I just download solaris from the sun site and installed it
[03:34:45] <rpedro_> :)
[03:35:27] *** yongsun has joined #opensolaris
[03:35:41] <FrostCS> hang on, I'll get your cookie
[03:35:51] <rpedro_> it's not installed very well on vmware
[03:35:53] <rpedro_> hehe
[03:36:02] <rpedro_> thanks =)
[03:36:15] <FrostCS> oh on vmware? well, you still get half a cookie..
[03:36:37] <rpedro_> vmware complains, I think it's on first boot, says it needs 64bit support
[03:36:51] <rpedro_> but I downloaded the 32bit version
[03:37:05] <FrostCS> Solaris is both 32 and 64 bit..
[03:37:06] <rpedro_> and running a 32bit linux os
[03:37:09] *** hile_ has quit IRC
[03:37:16] <rpedro_> on same install?
[03:37:21] <FrostCS> yes
[03:37:27] <FrostCS> is your host 32 bit?
[03:37:30] <rpedro_> oh, interesting..
[03:37:31] <rpedro_> yes
[03:37:42] <FrostCS> and your on a 64bit processor?
[03:38:01] <rpedro_> yes, amd 2600+
[03:38:12] <FrostCS> I am no vmware expert, but I would say that would be the complaint
[03:38:12] <rpedro_> a few years old pc :)
[03:38:27] <rpedro_> what?
[03:38:33] <rpedro_> that the cpu is old?
[03:38:49] <FrostCS> no, if you were running a 32 bit host on a 64 bit cpu?
[03:39:02] <rpedro_> I mean, I selected 32bit guest os for the virtual machine
[03:39:19] <FrostCS> I'd google for your same problem, perhaps other users had experienced it..
[03:39:28] <FrostCS> if no one else in here speak up on it
[03:39:31] <jamesd_> runs  solaris in 64 bit mode, on a 64 bit cpu  opteron 252, inside a  32 bit OS XP
[03:39:44] <FrostCS> do you know the exact complaint it gives?
[03:40:22] <FrostCS> jamesd_ any idea's why his complains then?
[03:40:47] <jamesd_> maybe its not on the approved  list of 64 bit cpus..  check vmware's site
[03:41:19] <rpedro_> oh, it's vmware that complains that the solaris kernel is trying to get to 64bit mode, and that I should switch the vm to 64bit mode
[03:41:42] <rpedro_> but I do boot in 64 mode, I get another error while booting
[03:41:47] <FrostCS> can you type the complaint? so it would be googleable?
[03:41:48] <rpedro_> but when*
[03:42:09] <rpedro_> one min
[03:43:35] <jamesd_> rpedro_, try and using the newest version of  vmware server
[03:43:38] <jamesd_> its free
[03:44:37] *** hile_ has joined #opensolaris
[03:44:41] <hile_> well wasn't that fun..
[03:44:56] <FrostCS> a rollercoaster?
[03:49:43] *** devnull62 has quit IRC
[03:52:12] *** rpedro_ has quit IRC
[03:54:46] *** stevel has quit IRC
[04:00:33] <CIA-26> mh27603: 6582507 ddi_prop_free() error in acpica_get_handle_cpu()
[04:01:40] *** Atomdrache has quit IRC
[04:07:21] * hile_ bumped the power cable on the storage array here (personal lab)... insta-panic
[04:09:33] <nachox> hile_, ups?
[04:09:46] <hile_> no...
[04:09:54] <hile_> bumped the lead from the array to the UPS
[04:09:55] <nachox> lost data?
[04:10:00] <nachox> crashed os?
[04:10:05] <hile_> yeah, insta-panic
[04:10:10] *** EchoBinary has quit IRC
[04:10:11] <hile_> lose your boot LUN, you go bye bye
[04:10:25] <FrostCS> hard wire everything in :-)
[04:10:35] <hile_> Don't I wish I could.
[04:11:50] <nachox> use svm :)
[04:12:00] <hile_> eh?
[04:12:03] <hile_> lost no data
[04:12:06] <hile_> i just caused a panic
[04:12:20] <sommerfeld> d'oh
[04:13:00] <nachox> then just an ups
[04:13:16] <hile_> UPS woudln't have saved me
[04:13:23] <hile_> I bumped the lead that goes into the device
[04:13:32] <FrostCS> not if he hit the power from the ups..
[04:13:39] <nachox> buy a battery backed device?
[04:14:02] <nachox> or the less expensive alternative, dont bump into the cable :P
[04:17:33] *** danv12 has quit IRC
[04:19:04] *** danv12 has joined #opensolaris
[04:20:46] *** tombhad-AC has joined #opensolaris
[04:22:07] <sommerfeld> dual power grid might save you
[04:22:12] <sommerfeld> two ups's, two power cords
[04:26:09] <richlowe> oddly, I've found that adding to the level of coincidence necessary to make an accident catastrophic rarely makes it less likely.
[04:26:28] <richlowe> in fact, the probability of the accident doesn't change, and thus the probability of the coincidence markedly increases.
[04:26:51] *** elijahwright has joined #opensolaris
[04:32:25] <nachox> richlowe, what's the point of redundancy if that is the case?
[04:32:26] <CSFrost> I've started channeling my chi to power my equipment
[04:32:28] <CSFrost> so far it hasn't failed
[04:32:34] <CSFrost> but then again.. it hasn't powered up either.
[04:34:00] <richlowe> nachox: which is the point that leads me to suspect I just have rotten luck :)
[04:34:01] *** tombhadAC has quit IRC
[04:35:31] *** sommerfeld has quit IRC
[04:35:50] *** halton has joined #opensolaris
[04:35:59] *** hile_ has quit IRC
[04:36:19] <nachox> richlowe, hehe
[04:36:24] *** reid_ has quit IRC
[04:36:48] *** rbrown_ has joined #opensolaris
[04:41:48] <CSFrost> rbrown, your ip just reminds me I have to start threatening verizon :-P
[04:41:59] <CSFrost> or rbrown_ almost forgot the tail :-P
[04:47:36] <nachox> CSFrost, why?
[04:48:08] <elijahwright> no fios where you live either?  :)
[04:48:12] *** Dink has quit IRC
[04:48:57] <CSFrost> there is fios where I live
[04:49:10] <CSFrost> 100 feet in any direction
[04:49:16] <CSFrost> just not here..
[04:51:17] <CSFrost> and 4,000 calls to verizon never seem to have an effect :-P
[04:51:47] <elijahwright> that really, really, really, really stinks.
[04:51:54] <CSFrost> yet my apartment in tokyo, I had to use a childish technique to choose a fiber provider, because I had 4 to choose from
[04:52:19] <CSFrost> all 10-20 times faster then fios anyhow :-)
[04:55:35] <rbrown_> CS how so ? :-)
[04:55:44] <rbrown_> vzn is awesome as long as you dont use them for phone or dsl :)
[04:55:49] <rbrown_> fios >*
[04:56:16] <jbk> all phone companies are awesome if you don't have to use them :)
[04:57:58] * dlg hug sun some more
[05:01:59] <nachox> dlg, ?
[05:05:39] <Tempt> What's this @LongLink crap?
[05:05:46] <richlowe> gtarism, I think.
[05:05:47] <Tempt> Is this some sort of shite GNU tar thing?
[05:05:56] <Tempt> ffs.
[05:06:03] <Tempt> Schily is right.
[05:06:19] <richlowe> Yeah, that always annoys me, too. ;)
[05:06:46] *** Teknomancer has joined #opensolaris
[05:09:51] <elijahwright> ugh, gnutar.
[05:11:52] <nachox> Tempt, come on, linux userland is so much better!
[05:12:38] <CSFrost> rbrown_, yea I had just been trying to get fios instead of crappy timewarner service since the service came into the area this winter, with 0 luck :-P
[05:18:44] *** comay has quit IRC
[05:19:53] *** baijiutung has joined #opensolaris
[05:23:24] *** yongsun has quit IRC
[05:23:39] *** yongsun has joined #opensolaris
[05:25:01] *** whaq has quit IRC
[05:26:46] *** whaq has joined #opensolaris
[05:31:11] *** baijiutong has quit IRC
[05:59:27] *** bashusr has quit IRC
[06:00:33] <CIA-26> tomee: 6581855 add ::kmem_slabs dcmd to report slab usage per cache
[06:03:55] *** nachox has quit IRC
[06:08:23] <Teknomancer> anyone knows how to export an environment variable from a script?  (i tried PATH=<whatever>\nexport PATH) doesn't seem to export it in solaris
[06:10:01] *** swa has joined #opensolaris
[06:10:30] <Teknomancer> export from command-line exports it, but i need to export from a script ....
[06:11:22] <jcsmith> Teknomancer, source the script
[06:11:39] *** Drone has quit IRC
[06:11:42] <Teknomancer> jcsmith: source <script> ?
[06:11:47] <Teknomancer> k let me try
[06:11:48] <jcsmith> yes
[06:12:01] *** Drone has joined #opensolaris
[06:12:33] <Teknomancer> jcsmith: works :) thanks.
[06:13:28] <jcsmith> no problem
[06:13:34] <jcsmith> you can also do . script
[06:13:42] <jcsmith> ^ that is a period :-D
[06:14:09] <Teknomancer> yeah i wondered what that was
[06:14:24] <Teknomancer> source is cool , used it before while doing MakeMe, guess i forgot
[06:14:29] <Teknomancer> not much of a bash player :P
[06:15:02] <jcsmith> i'm more of a KSH scripter than bash, but they're pretty similar when it comes to things like that
[06:15:17] <jcsmith> hell any shell that is semi posix compliant is
[06:15:52] <Teknomancer> well bash comes into a play a lot when u're dealing with highly cross platform (configure etc) when in my previous job it was very specific platform
[06:16:02] <rbrown_> bash > *
[06:16:08] <Teknomancer> and it was very simple, we generate a very straightforward makefile and just call make...
[06:17:07] <jcsmith> i do like bash better, however at work the only shell that is installed on all of our systems is ksh so i end up using it more
[06:17:53] <rbrown_> that's pretty gay
[06:17:56] <rbrown_> why remove the other shells?
[06:18:12] <jcsmith> they aren't removed, just the default install of AIX does not include bash
[06:18:15] <Teknomancer> bash is more than enough, i fail to see the reason for anyother shell
[06:18:20] <jcsmith> i agree
[06:18:23] <Teknomancer> i mean honestly u can do so much with bash
[06:18:26] <Teknomancer> sed/grep whatever
[06:18:34] <dlg> has anyone hear heard of a sun chip called cheerio?
[06:18:41] <Teknomancer> its the same with competing DVD formats
[06:18:46] <jcsmith> don't forget awk Teknomancer
[06:19:03] <Teknomancer> jcsmith: never used awk :) (or sed for that matter) grep yes but not the other two
[06:19:07] <Teknomancer> sed syntax just freaks me out
[06:19:08] <Teknomancer> hehe
[06:19:28] <jcsmith> sed can get a little weird sometimes
[06:19:40] <jcsmith> but it sure comes in handy sometimes
[06:19:55] <jcsmith> but i believe i'm going to head to bed, g'night
[06:19:56] *** sartek has quit IRC
[06:19:58] <rbrown_> dlg, erio interface is using a cheerio chp
[06:20:02] <rbrown_> chip
[06:20:03] <Teknomancer> nite jcsmith, and thx
[06:20:24] <rbrown_> sed rocks
[06:22:13] <dlg> rbrown_: erio?
[06:22:31] <rbrown_> yes
[06:22:40] <rbrown_> I ment eri
[06:22:59] <rbrown_> found on like the older sun blades
[06:24:27] *** wickedwicky has joined #opensolaris
[06:24:36] <wickedwicky> good morning
[06:24:40] <Teknomancer> morning wickedwicky
[06:25:01] <wickedwicky> Hey :) how are you doing?
[06:26:54] <Teknomancer> fine :)
[06:37:24] *** Atomdrache has joined #opensolaris
[06:38:40] <Atomdrache> I've got a GIMP question, if anybody in here's familiar with it.  Basically everything involving using a mouse stopped working, right after I saved a file and opened another.  The coordinates box doesn't even work right, and stops at the edge of the image.  So I figure it might be wise to reinstall GIMP and GTK.  But it's not a package.  It came with Solaris.  So how do I reinstall something that's in /usr/bin?
[06:38:45] <Atomdrache> (let me know if that got cut off)
[06:39:02] <jmcp> it is a package, but you don't know what to look for
[06:39:28] <Teknomancer> the grep in opensolaris doesn't give me any --recursive options ... any other command that can recursively grep ?
[06:39:28] <boyd> I'd try nuking your personal config files first
[06:39:30] <Atomdrache> I was about to correct that as "It's not a package I installed".  How do I find out which package it is?
[06:39:39] <Atomdrache> Eh, I'll take a look.
[06:39:42] <boyd> Teknix: /usr/sfw/bin/ggrep
[06:39:46] <rbrown_> Teknomancer, use egrep
[06:39:48] <jmcp> SUNWgnome-img-editor, SUNWgnome-img-editor-root, SUNWgnome-img-editor-devel
[06:39:49] <rbrown_> egrep -r
[06:39:59] <Teknomancer> okay thanks boy and rbrown_
[06:40:02] <dlg> jmcp: mail4u
[06:40:24] <jmcp> ta
[06:40:38] <boyd> rbrown_: when did -r go onto egrep?
[06:40:41] <Teknomancer> hmm egrep -r doesn't work, i think its ggrep
[06:41:02] <boyd> Ah :)
[06:41:46] <Atomdrache> Aye.  I'll try getting rid of the personal configuration files first.
[06:45:11] <rbrown_> how does -r not work?
[06:45:14] <Atomdrache> Ah.  That did fix it.
[06:45:24] <Atomdrache> Just removed the entire configuration directory.
[06:48:55] <wickedwicky> egrep: illegal option -- help usage: egrep [ -bchilnsv ] [ -e exp ] [ -f file ] [ strings ] [ file ] ...
[06:49:11] <wickedwicky> rbrown: cuz egrep doesnt have the -r option :s
[06:49:34] <wickedwicky> I think GNU grep, like mentioned is the way to go
[06:50:21] <wickedwicky> or find /path/to/search -type f -exec grep string {} \;
[06:50:41] <wickedwicky> bit expensive search method though
[06:51:28] <wickedwicky> ow aaaaaaaaand, on69 didnt compile. Same error as without __SSNEXT
[06:51:54] <wickedwicky> I am trying to compile on a 64bit machine at work now, see if that changes anything
[06:54:31] *** karrotx has quit IRC
[06:58:05] <elijahwright> dlg: have heard of cheerio before, but not recently.
[07:05:18] <dlg> kay
[07:19:32] *** tg has quit IRC
[07:19:32] *** sniffy has quit IRC
[07:19:32] *** het_ has quit IRC
[07:19:32] *** LuckyLuke has quit IRC
[07:19:32] *** mustang has quit IRC
[07:19:32] *** yatesy has quit IRC
[07:19:33] *** Tiger^ has quit IRC
[07:19:33] *** FastJack has quit IRC
[07:19:33] *** sabor has quit IRC
[07:19:33] *** e^ipi has quit IRC
[07:19:35] *** Kernel86_ has quit IRC
[07:19:35] *** jwit has quit IRC
[07:19:35] *** mikaeld has quit IRC
[07:19:35] *** Bart_M has quit IRC
[07:19:36] *** TBCOOL has quit IRC
[07:28:01] <wickedwicky> allrighty
[07:28:10] <wickedwicky> going to the bus station, it's friday
[07:30:40] <wickedwicky> which means, last work day of the week \o/
[07:30:44] <wickedwicky> see ya all later
[07:31:02] *** wickedwicky has quit IRC
[07:39:26] *** st3fan has quit IRC
[07:47:40] *** st3fan has joined #opensolaris
[07:49:48] *** ciph3r has quit IRC
[07:53:22] *** monzie has joined #opensolaris
[07:53:32] *** cyril_plisko has joined #opensolaris
[07:58:44] *** Atomdrache has quit IRC
[08:00:09] *** estibi has quit IRC
[08:07:59] *** tsoome has quit IRC
[08:09:00] <Tempt>
[08:09:00] <Tempt>
[08:09:04] <Tempt>                                         ||        ||
[08:09:04] <Tempt>                                ||        ||
[08:09:04] <Tempt>                               ||||      ||||
[08:09:04] <Tempt>                           ..:||||||:..:||||||:..
[08:09:04] <Tempt>                          c i s c o S y s t e m s
[08:09:39] <g4lt-U60> Tempt, you fail at ascii art
[08:09:54] <purserj> and here I thought he was playing towers
[08:12:05] <Tempt> pfft.
[08:12:08] <Tempt> mis-paste.
[08:12:23] <Tempt> Also known as tossing a pile of stuff on top of the mouse having pasted it earlier.
[08:12:27] <Tempt> I'm having a very long day.
[08:12:39] *** FastJack has joined #opensolaris
[08:12:39] *** mustang has joined #opensolaris
[08:12:39] *** Tiger^ has joined #opensolaris
[08:12:40] *** sabor has joined #opensolaris
[08:12:40] *** sniffy has joined #opensolaris
[08:12:40] *** het_ has joined #opensolaris
[08:12:40] *** e^ipi has joined #opensolaris
[08:12:40] *** tg has joined #opensolaris
[08:12:40] *** yatesy has joined #opensolaris
[08:12:40] *** LuckyLuke has joined #opensolaris
[08:12:40] *** Bart_M has joined #opensolaris
[08:12:40] *** Kernel86_ has joined #opensolaris
[08:12:40] *** jwit has joined #opensolaris
[08:12:40] *** TBCOOL has joined #opensolaris
[08:12:40] *** mikaeld has joined #opensolaris
[08:14:17] *** bigjohnto has quit IRC
[08:14:45] <jmcp> Tempt: you.need.more.beer (tm)
[08:15:53] <purserj> jmcp: is that a callable function?
[08:16:40] *** tg has quit IRC
[08:16:40] *** LuckyLuke has quit IRC
[08:16:40] *** het_ has quit IRC
[08:16:40] *** sniffy has quit IRC
[08:16:40] *** mustang has quit IRC
[08:16:40] *** Tiger^ has quit IRC
[08:16:41] *** FastJack has quit IRC
[08:16:41] *** yatesy has quit IRC
[08:16:41] *** e^ipi has quit IRC
[08:16:41] *** sabor has quit IRC
[08:16:41] *** jwit has quit IRC
[08:16:42] *** Kernel86_ has quit IRC
[08:16:42] *** mikaeld has quit IRC
[08:16:43] *** Bart_M has quit IRC
[08:16:43] *** TBCOOL has quit IRC
[08:16:52] *** FastJack has joined #opensolaris
[08:16:52] *** mustang has joined #opensolaris
[08:16:52] *** Tiger^ has joined #opensolaris
[08:16:53] *** sabor has joined #opensolaris
[08:16:53] *** sniffy has joined #opensolaris
[08:16:53] *** het_ has joined #opensolaris
[08:16:53] *** e^ipi has joined #opensolaris
[08:16:53] *** tg has joined #opensolaris
[08:16:53] *** yatesy has joined #opensolaris
[08:16:53] *** LuckyLuke has joined #opensolaris
[08:16:53] *** Bart_M has joined #opensolaris
[08:16:53] *** Kernel86_ has joined #opensolaris
[08:16:53] *** jwit has joined #opensolaris
[08:16:53] *** TBCOOL has joined #opensolaris
[08:16:53] *** mikaeld has joined #opensolaris
[08:17:12] <g4lt-U60> purserj, if it is, I'm definitely permanently installing that library
[08:17:44] * purserj installs the free-beer library
[08:19:21] *** Gropi has joined #opensolaris
[08:23:24] <jmcp> purserj: I sure hope so
[08:23:57] * purserj calls you.need.more.beer(TRUE);
[08:31:14] *** tsoome has joined #opensolaris
[08:34:16] *** tg has quit IRC
[08:34:16] *** LuckyLuke has quit IRC
[08:34:16] *** het_ has quit IRC
[08:34:16] *** sniffy has quit IRC
[08:34:16] *** mustang has quit IRC
[08:34:16] *** Tiger^ has quit IRC
[08:34:17] *** FastJack has quit IRC
[08:34:17] *** yatesy has quit IRC
[08:34:17] *** e^ipi has quit IRC
[08:34:17] *** sabor has quit IRC
[08:34:17] *** jwit has quit IRC
[08:34:18] *** Kernel86_ has quit IRC
[08:34:18] *** mikaeld has quit IRC
[08:34:19] *** Bart_M has quit IRC
[08:34:19] *** TBCOOL has quit IRC
[08:34:21] *** phimic has joined #opensolaris
[08:36:35] * jmcp heads off to fill up the bbq gas bottle
[08:40:58] <Tempt> boyd: ping
[08:41:30] *** FastJack has joined #opensolaris
[08:41:30] *** mustang has joined #opensolaris
[08:41:30] *** Tiger^ has joined #opensolaris
[08:41:30] *** sabor has joined #opensolaris
[08:41:30] *** sniffy has joined #opensolaris
[08:41:30] *** het_ has joined #opensolaris
[08:41:30] *** e^ipi has joined #opensolaris
[08:41:30] *** tg has joined #opensolaris
[08:41:30] *** yatesy has joined #opensolaris
[08:41:30] *** LuckyLuke has joined #opensolaris
[08:41:30] *** Bart_M has joined #opensolaris
[08:41:30] *** Kernel86_ has joined #opensolaris
[08:41:30] *** jwit has joined #opensolaris
[08:41:30] *** TBCOOL has joined #opensolaris
[08:41:30] *** mikaeld has joined #opensolaris
[08:45:50] *** stranger has joined #opensolaris
[08:50:46] *** NikolaVeber has joined #opensolaris
[08:55:23] *** yongsun has left #opensolaris
[08:56:47] *** jpipkin has quit IRC
[08:57:52] <stranger> man, after using gnu/linux and occasionally bsd systems for so long, i feel so powerless with opensolaris. :( it's like learning unix all over again in some instances
[08:58:26] <trygvis> I know that feeling
[08:59:04] <tsoome> unix again? linux is not unix:P
[08:59:15] <stranger> bsd is unix though
[08:59:42] <tsoome> :)
[09:00:10] <stranger> well, the system i'm on, seems to be an old solaris, which i find even worse because i don't find dtrace or zfs utilities :(
[09:00:12] <tsoome> anyhow, unix is still unix, same commands, a bit different options...
[09:00:44] <stranger> tsoome: it's the options i'd come to take for granted. it'll take some getting used to, no big deal, but it's devastating at first when the command spews a sort of RTFM at you
[09:01:02] <tsoome> :D
[09:02:44] *** tsp has joined #opensolaris
[09:02:46] <stranger> anyone have any idea what /afs could be. i'm on a ssh'd on a university computer from home that has everything from *.gov to *.edu covered (i'd figured it would just be a university specific mount?)
[09:03:04] <tsoome> afs is another nf file service
[09:03:05] <tsp> stranger: afs is some weird type of filesystem
[09:03:06] <tsp> heh
[09:03:10] <seanmcg> Andrew filesystem
[09:03:12] <tsoome> nw*
[09:03:25] <tsoome> man google:D
[09:03:34] <stranger> i know that much, i mean how a particular AFS share can house so much stuff -- it has to be massive...
[09:03:38] <tsp> hmm, anyone know how to turn off the weird literal-after-\ thing that plagues my terminal?
[09:03:41] <seanmcg> network, distributed fs, like nfs
[09:03:46] *** cydork has joined #opensolaris
[09:03:55] <_mary_kate_> stranger: afs is distributed, those subdirectories are different hosts
[09:04:19] <tsp> I have a bunch of ed sessions open, and when I hit \ and then backspace, it puts a ^H in my file :)
[09:04:22] <stranger> _mary_kate_: but why would i be entrusted to my university's stuff plus other universities, etc. etc.?
[09:04:23] <tsoome> also they are using single access dir, /afs, for it
[09:04:26] *** slowhog has quit IRC
[09:04:51] <_mary_kate_> stranger: probably because they're public shares.
[09:04:54] <seanmcg> tsp: seems your shell is escaping the backspace for you as if you wanted the ^H...
[09:05:08] <tsp> seanmcg: I thought of that, but I think its the terminal driver
[09:05:24] <seanmcg> whats your TERM set to ?
[09:05:26] <tsp> I know its not ed, or more accurately edbrowse
[09:05:29] <stranger> _mary_kate_: hmm, i'll have to read up on afs. i'll go revive my high school LUG group and properly setup the network shares then :)
[09:05:31] <tsp> hmm, no idea
[09:05:38] <tsp> screen
[09:06:29] <tsp> I'm stuck doing mostly everything on my solaris server under vmware since my U5 keeps kicking the bucket at odd times
[09:06:30] *** jcea has quit IRC
[09:08:18] <tsp> it stays on anywhere from 3 seconds to 97 days, then dies
[09:08:29] <tsp> if I turn the power supply off for 10 seconds, I can restart it
[09:08:39] *** cyril_plisko has quit IRC
[09:08:44] <seanmcg> any errors from console ?
[09:08:49] <libkeiser> stranger: you're not entrusted to see anything in afs unless you're in an acl, and have kerberos credentials for that cell, or for a cross-cell trusted cell
[09:08:52] <tsp> nope, no system log either
[09:09:09] <tsp> the speech synthesizer I ahve hooked up says a random character or two as it goes off, but that's it
[09:09:50] <tsp> right now it's been on for almost 3 hours, but it'll probably go off again tonight
[09:11:20] <seanmcg> U5s are usually pretty good boxes.  Though I remember one years ago where I had to take the disk out and drop it to unsitick the head :)
[09:11:25] <tsp> heh
[09:11:48] <tsp> if I ever used something more than ed, the thing would probably fall over
[09:12:07] *** Cyrille has joined #opensolaris
[09:13:26] <tsp> but this ancient unix box can outperform my windows xp box depending on what I do with it
[09:13:40] <seanmcg> heh :)
[09:14:08] <tsp> I think my max was somewhere around 30 edbrowse sessions all opened to different files/websites
[09:14:15] <tsp> and it just worked
[09:17:08] <g4lt-U60> you also have to remember, that SunOS/Solaris can happily function at a loadavg of ~100 with no appreciable slowness.  try that with certain other oses
[09:17:51] <stranger> g4lt-U60: that's of course, reliant on hardware, yes?
[09:18:35] <g4lt-U60> with a misconfigured JVM and azureus my SB100 peaked in the 200s and I didn't notice until someone asked me why was I eating up most of a T1 on my own
[09:19:00] <tsoome> stranger: it is dependent on hw
[09:19:28] <tsoome> and workload details as well ofc
[09:19:41] <tsp> the biggest load averige I had was ~580, and 10 minutes later my os was back up
[09:19:52] <g4lt-U60> NFW would that happen on a X86, true, but even X86 can go to loadavgs that other OSen would die a screaming death at
[09:20:35] <tsp> I ran 500 conversion scripts in parallel instead of in sequence... big mistake
[09:20:42] <bda> I had a X2100 M1 hit 2500 loadavg while getting UDP flooded.
[09:20:47] <bda> (I blamed ipmon, heh)
[09:20:51] <tsp> ouch
[09:21:12] <tsp> heh ipfilter is a joke
[09:21:17] <bda> Yeah, I hate it.
[09:21:27] <tsp> bda: just put a linux box in its way
[09:21:33] <tsp> or a bsd box with pf
[09:21:36] <bda> OPenBSD. :)
[09:21:39] <tsp> yeah
[09:21:40] <bda> It's on the list.
[09:22:05] <tsp> although pf didn't like my network, my wimpy little iptables openwrt router I don't have to touch works far better
[09:22:23] <tsp> no ftp proxies :)
[09:22:30] <bda> I've only ever had problems with pf when doing weird QoS stuff.
[09:22:55] *** jpipkin has joined #opensolaris
[09:22:57] <tsp> hmm
[09:22:58] <bda> Well, you're still proxying ftp connex. Just doing it in kernel land instead of user, no? :P
[09:23:02] <tsp> I guess
[09:23:05] <tsp> but its transparent
[09:23:07] <bda> ipf's lack of macros makes me really sad.
[09:23:12] <tsp> I can connect to an ftp on any port, it'll work
[09:23:45] <tsp> the only good thing about ftp is the clients. else rsync/nfs would be the best
[09:26:10] *** rbrown_ has quit IRC
[09:26:16] <tsp> I'm waiting for the day when someone writes a whole net set of tools. If this box died, it would have created a history file on all the other boxes so I could launch a utility to bring everything back up
[09:26:21] <tsp> s/net/new
[09:30:01] *** Snake007uk has joined #opensolaris
[09:30:53] <bda> Linux LVM is so cute. Hardware RAID creating sda. 100G pv on sda8. 1TB drbd chewing on sda9. vgdisplay reports +1TB free for sd8.
[09:30:57] * bda shakes fist.
[09:31:26] <tsp> heh, lvm mostly works
[09:34:17] *** WickedWicky has joined #opensolaris
[09:34:26] <WickedWicky> hellows
[09:34:52] <seanmcg> tsp: some folks call what you waiting for a HA cluster :)
[09:38:35] *** stranger has quit IRC
[09:38:45] *** yongsun has joined #opensolaris
[09:40:22] <tsp> ah
[09:41:06] <tsp> if I'm write in the middle of reading a story, I don't want to lose it when the box dies :)
[09:41:35] *** razrX has joined #opensolaris
[09:44:15] <seanmcg> tsp: ya, a HA cluster works.  One node dies, other node sees that and takes over, no loss in services or descernable switchover from clients.  Used in telcos a good bit.
[09:49:55] *** WickedWicky has quit IRC
[09:52:33] <quasi> not just telcos
[10:00:08] *** dmarker has quit IRC
[10:00:19] *** dmarker has joined #opensolaris
[10:00:43] <CIA-26> ql147931: PSARC 2007/312 RealTek 8180L 802.11b Wireless Driver, 6523215 a wifi driver is needed to support card with realtek chipset
[10:02:27] <quasi> 11b drivers - I thought noone bothered with those any more
[10:07:11] *** vmlemon has joined #opensolaris
[10:10:02] *** Teknomancer has quit IRC
[10:10:26] *** calumb has joined #opensolaris
[10:11:05] *** IvanR_ has quit IRC
[10:12:08] *** WickedWicky has joined #opensolaris
[10:16:21] *** KermitTheFragger has joined #opensolaris
[10:16:26] *** Teknomancer has joined #opensolaris
[10:17:22] *** timsf has joined #opensolaris
[10:17:34] <timsf> Morning everyone
[10:17:58] <renihs> morning mammal
[10:18:09] <renihs> its friday :p
[10:18:52] <timsf> Yeah, beautiful thundery day here in Dublin - lots of rain. Aah Irish summers...
[10:20:28] <quasi> friday, beer, good!
[10:20:54] *** mikefut has joined #opensolaris
[10:22:21] <phips> timsf: it's no better in london. It's like autumn here today. What a great summer
[10:23:14] <timsf> Global warming -
[10:23:16] <Cyrille> same in Paris (though the week so far had been summerish)
[10:23:42] *** mikefut has quit IRC
[10:23:53] <Cyrille> haven't you seen the scientific documentary "the day after tomorrow"? ;-)
[10:24:07] <phips> Cyrille: :-)
[10:24:14] <timsf> Not yet - been raining too much every time I want to get to the video shop.
[10:24:23] <quasi> Sunshine in Copenhagen
[10:24:34] <phips> I'm off to Vienna for the weekend, where today it's supposedly 38 degress C - woot!
[10:24:35] <timsf> s'uppose I could drive there in my v12 6l SUV though ?
[10:24:57] <Cyrille> nah, just use the helicopter.
[10:25:12] *** Pietro_S has joined #opensolaris
[10:26:31] *** tombhad-AC has quit IRC
[10:27:20] <trochej> There goes sadist. He meets masochist on his way. Masochist says: Hit me. Sadist says: Nope.
[10:29:47] <WickedWicky> Amsterdam: same weather :P
[10:30:34] <WickedWicky> Cyrille: I've seen the day before the day after tomorrow (South Park)
[10:30:43] *** calumb has quit IRC
[10:34:33] <sbahra> Cyrille, what is it about?
[10:35:40] *** Teknomancer has quit IRC
[10:36:21] <Cyrille> it's one of those hollywood catastrophe movies, whose basic story is that due to global warming and the resulting weather disruption, the USA get covered in Arctic like weather.
[10:36:51] *** Fish has joined #opensolaris
[10:37:19] <Cyrille> obviously brave men risk their lives to save others, some die heroically in the process, fathers and sons reconnect, north meets south, and Everybody Learns Valuable Lessons (tm).
[10:37:57] <Fish> hello
[10:38:46] *** kumamoto has joined #opensolaris
[10:40:07] <trochej> Cyrille: Not USA, but whole world. But it's typical to Hollywood movies to just forget that the world is quite bigger than the USA itself
[10:41:26] *** calumb has joined #opensolaris
[10:42:02] <Cyrille> there are so few mentions of the rest of the world that I don't think it was worth mentioning. The only other place referred to is Mexico, which is then enjoying moderate weather, and becomes the refuge for USA citizens (which I assume is something screenwriters thought of as deep and thought provoking... yawn).
[10:46:02] <Pietro_S> Cyrille: that's that movie witch has nice effect with quicky frozen flag? (actually I remember only this effect from whole movie, but very cool effect ;-)
[10:46:55] <Cyrille> could be, quite frankly my memory of it is a bit hazy, it's not like I watch it every week ;-)
[10:47:33] *** bnitz has joined #opensolaris
[10:47:44] <trochej> :)
[10:51:07] *** calum_ has joined #opensolaris
[10:51:20] <seanmcg> the british military helicopter being frozen in a few seconds while flying was good (though somewhat above belief)
[10:52:00] <WickedWicky> morning sean
[10:52:14] <Pietro_S> hat effect left in my memory, because it's nice to watch, but very inatural - because in real world flag would explode (or implode) to lot's of pieces ...
[10:58:48] <tomww> Pietro_S: hi. tried pmpkg. the first package startet compiling after 3 minutes .-)
[11:02:25] *** Atomdrache has joined #opensolaris
[11:03:37] <seanmcg> morn' WickedWicky
[11:05:29] *** CSFrost has quit IRC
[11:05:40] *** calumb has quit IRC
[11:07:02] *** calum__ has joined #opensolaris
[11:08:34] *** CSFrost has joined #opensolaris
[11:09:24] <Pietro_S> tomww: hello, but I'm a bit confused - I don't have any connection to pmpkg (oxygen is in it) ...
[11:10:09] *** noyb has joined #opensolaris
[11:14:30] *** calum__ is now known as calumb
[11:16:15] *** tsp has quit IRC
[11:20:30] *** calum_ has quit IRC
[11:21:56] *** polk_ has quit IRC
[11:23:40] *** derchris has quit IRC
[11:23:56] *** derchris has joined #opensolaris
[11:29:44] *** CSFrost has quit IRC
[11:33:16] *** CSFrost has joined #opensolaris
[11:38:06] *** timeless has quit IRC
[11:38:36] <UberDuper> Any of you know what it takes to get netatalk to compile on CE?
[11:39:19] <WickedWicky> coffee?
[11:40:12] <trochej> A lot of
[11:44:24] <UberDuper> Meh. Almost 3am. I'll figure it out tomorrow when I have some coffee.
[11:46:28] *** halton has left #opensolaris
[11:59:24] *** simford has quit IRC
[12:00:23] <CSFrost> if there was a job to crash software, I would be raking in millions :-)
[12:01:07] *** calumb has quit IRC
[12:01:08] <seanmcg> CSFrost: millions of what ?  A job like that would pay peanuts :)
[12:01:27] <CSFrost> millions of peanuts then !
[12:02:10] <CSFrost> I seem to be able to take down any application in less than 4 minutes of use, time proven techniques.
[12:04:05] *** MattMan has joined #opensolaris
[12:04:10] <e^ipi> peanuts would probably be a decent thing to get paid
[12:04:21] <e^ipi> play the peanuts market
[12:04:55] <CSFrost> probably one of the most stable currencies on the market eh?
[12:04:58] <timsf> CSFrost, there is a job to crash software, I've got it!
[12:05:01] <timsf> (well, sort of)
[12:05:16] <CSFrost> woot, two more errors, on two seperate apps
[12:05:21] <CSFrost> that's another 2 peanuts right?
[12:05:48] <CSFrost> timsf, really? :-P
[12:05:53] <timsf> You need to be in QE!
[12:06:05] <timsf> Spend your time writing code that tests other code,
[12:06:22] <Cyrille> who tests your code then?
[12:06:32] <timsf> My code /never/ has bugs  ;-)
[12:06:36] <Cyrille> AH!
[12:06:41] * timsf looks nervous
[12:06:41] <Cyrille> Famous last words...
[12:06:44] <CSFrost> time might be better well spent just strangling developers who refuse to follow simple rules
[12:06:59] <timsf> We get ninja training for just such occurrences
[12:08:22] <CSFrost> lol
[12:08:52] <CSFrost> I enjoyed that 101 ways to make sure your code doesn't suceed thing or whatever it was :-P
[12:09:23] <timsf> That was a good document (how to write unmaintainable code, wasn't it?)
[12:09:28] <Cyrille> I didn't know there were so many ways, I only ever use a couple ;-)
[12:09:36] <timsf> I quite like the "Works on my machine certification program"
[12:10:30] <CSFrost> I'll dig it up in a second timsf, as soon as I stop crashing both opera and firefox that is.
[12:10:42] *** Gman has quit IRC
[12:10:51] *** dlewis has joined #opensolaris
[12:11:04] <timsf> http://www.codinghorror.com/blog/archives/000818.html (for the latter, at least)
[12:11:21] <CSFrost> If there is a hand to turn stuff to gold, and it's called Midas Touch or whatever.. what is it when you crash everything you touch? :-P
[12:12:11] *** delewis has quit IRC
[12:14:57] <CSFrost> timsf, http://www.codesqueeze.com/101-ways-to-know-your-software-project-is-doomed/
[12:17:39] <timsf> true
[12:19:33] *** cmihai has joined #OpenSolaris
[12:23:43] *** monzie has quit IRC
[12:26:59] <sickness> I'm back
[12:28:52] <seanmcg> CSFrost: ever find bugs in TeX then ?-)
[12:29:34] <CSFrost> are you suggesting I should look seanmcg? :-P
[12:30:22] <seanmcg> find a bug in TeX and make monies from Dr Knuth :)
[12:30:25] <CSFrost> Unless your referring to tex-mex, then the answer would be most definately.
[12:30:49] <seanmcg> Nope, the pure TeX, no LaTeX, no AMSTeX...
[12:31:15] *** ShanghaiScott has joined #opensolaris
[12:31:38] <ShanghaiScott> anyone know how to get the jabber protocol back in pidgin with SXCE 68?
[12:34:06] *** calumb has joined #opensolaris
[12:35:16] *** MattMan has quit IRC
[12:37:24] *** Gman has joined #opensolaris
[12:38:11] *** danv12 has quit IRC
[12:38:36] <CSFrost> "Debian-izing the open source counterpart to the Solaris operating system"
[12:38:43] <dme> House!
[12:38:45] *** Dar has joined #opensolaris
[12:40:43] *** MattMan has joined #opensolaris
[12:46:57] <WickedWicky> damn, I ate too much
[12:49:51] <Pietro_S> ShanghaiScott: maybe install latest vermillion, but I didn't try it
[12:50:14] <cmihai> Bloody hell, nuts to UPS, they're all crap.
[12:50:53] <cmihai> Got a new one, still won't hold.
[12:51:37] <WickedWicky> UPS as in United Postal Service or Unified Power Supply?
[12:52:24] <WickedWicky> n/m, guess you dont hold the UPS guys... unless the guy is a girl and tanned and long dark hair
[12:52:30] <cmihai> That's it, I'm gettin a 70Amp/h lead acid bus battery for it...
[12:52:47] <WickedWicky> buy a bike with a phat dynamo, hire someone to ride the bike all day
[12:54:58] <cmihai> Man, I got ANOTHER 2kVA UPS, and it won't even hold a 400 W PSU peecee (real power usage about 200W)
[12:57:03] *** SymmWork has joined #opensolaris
[12:57:14] <quasi> cmihai: sounds like you need new batteries ;)
[12:58:42] *** xisxon has joined #opensolaris
[12:59:16] <cmihai> quasi: both were brand new ;-)
[13:01:07] *** xisxon has quit IRC
[13:02:07] *** deather_ has joined #opensolaris
[13:02:12] *** dunc has quit IRC
[13:02:34] *** bubbva has quit IRC
[13:02:53] *** bubbva has joined #opensolaris
[13:04:35] <quasi> cmihai: and only good for keeping a flashlight alive for a couple of moments
[13:05:27] * WickedWicky just ordered a new gucci watch
[13:05:43] <WickedWicky> nv69 is still compiling btw
[13:05:54] <WickedWicky> on both the 64bit desktop and the 32 bit laptop
[13:06:22] *** Gman has quit IRC
[13:07:53] *** Gman has joined #opensolaris
[13:07:58] *** rmorokut has joined #opensolaris
[13:08:08] *** RobG has joined #opensolaris
[13:08:27] <rmorokut> Hi
[13:08:30] <RobG> Is there an easy way to tell what process has a file open?
[13:08:49] <WickedWicky> fuser?
[13:08:54] <PerterB> lsof, pfiles, fuser
[13:09:03] <RobG> thx
[13:09:29] <rmorokut> any experience with ON on nx6325 Compaq?
[13:09:46] *** MattMan has quit IRC
[13:11:33] <cmihai> quasi: heh, I've just realized the car battery setup will hold out 400W easily for at least an hour ;-)
[13:11:33] <cmihai> If my math is right anyway.
[13:11:51] <cmihai> And with second hand dead-battery APC UPS to charge it, I'll do it for half the money of this garbage.
[13:12:02] <coffman> cmihai: you should stop buying crp
[13:12:06] <coffman> *crap
[13:12:29] <cmihai> coffman: most UPS products on the market are crap.
[13:12:38] *** Fullmoon has joined #opensolaris
[13:12:57] <cmihai> With a puny 12V 7Amp/hour gel cell, it won't deliver..
[13:13:05] <cmihai> Doesn't matter who makes them.
[13:13:45] <coffman> cmihai: buy mge, power or liebert
[13:13:50] *** rmorokut is now known as solar-star
[13:14:19] *** RobG has quit IRC
[13:14:25] <cmihai> coffman: what about APC?
[13:14:33] <quasi> cmihai: I know someone who runs a couple of soekris on car battery backup
[13:14:44] <cmihai> quasi: neat.
[13:14:57] <quasi> but they use a lot less power than the average pc
[13:15:02] <coffman> cmihai: yust dont
[13:15:11] <cmihai> coffman: what?
[13:15:24] <quasi> I think he said something about easily running 24 hours off the batteries
[13:15:32] <coffman> cmihai: and dont even try to replace the batteries with something else
[13:15:34] *** Dar has quit IRC
[13:15:50] <cmihai> coffman: what do you mean?
[13:15:51] <coffman> cmihai: just dont - sry
[13:16:21] <cmihai> I don't understand what you mean mate. What are you talking about, the car battery setup?
[13:16:28] <coffman> yes
[13:16:31] <cmihai> Why not?
[13:17:05] <coffman> they will not get charged and your ups might get damaged
[13:17:58] <coffman> our ups battery problems always came from overheating, bad cooling of the ups
[13:18:38] *** deather has quit IRC
[13:18:44] *** deather_ is now known as deather
[13:18:51] <cmihai> coffman: car batteries wouldn't suffer such problems mate.
[13:19:02] <cmihai> Besides, all I need is a standard 12V PSU that any UPS has and will work.
[13:19:15] <cmihai> And a continuous sinewave inverter.. and a UPS has that too.
[13:19:45] <coffman> cmihai: sure, but they will kill your ups - the car batteries
[13:20:10] <cmihai> Take into account a Power Factor of 0.7 or so, and the fact I need about 400W, a 600W inverter will do just fine.
[13:20:31] <cmihai> coffman: I could get a desulfator for the car battery.
[13:20:36] *** Gman has quit IRC
[13:22:13] <cmihai> coffman: lead-acid batteries usually die because of sulfation when you take them down to half capacity, this thing will always be online, so I don't think it will be an issue.
[13:22:58] <cmihai> It's basically a regular UPS, but instead of the garbage 12V 7Ah lead-gely brick I put in a 45Ah or 60Ah lead-acid car battery.
[13:24:15] <cmihai> And with maintenence free batteries it won't bubble over and leak.
[13:24:59] <cmihai> I'll keep it in half open space should it leak hydrogen (though that won't usually happen)
[13:25:12] <libkeiser> why not deep cycle marine batteries?
[13:25:32] <solar-star> unable to configure date with Solaris Developer Edition 05/07
[13:28:07] <cmihai> libkeiser: too expensive
[13:30:02] *** bnitz has quit IRC
[13:30:51] <cmihai> libkeiser: if I find some cheap though, they'd be better than a cranking car battery though...
[13:31:20] *** calumb has quit IRC
[13:31:31] *** cypromis has joined #opensolaris
[13:31:43] <cmihai> 12 Volt 74 Ah Gel Cell Sealed Lead Acid Battery Your Price: $119.95
[13:31:57] <cmihai> libkeiser: actually, they're not THAT expensive. Just found one for 100Euro :-)
[13:32:31] <CSFrost> http://www.dansdata.com/diyups.htm
[13:32:36] <CSFrost> first hit ala google
[13:33:09] <cmihai> CSFrost: heh, cool, so now I have precedent :-)
[13:34:23] <CSFrost> that is the make it yourself type.. then as you said.. the upgrade the UPS type.. http://www.dansdata.com/upsupgrade.htm
[13:34:58] <cmihai> CSFrost: oh, he even used a UPS
[13:35:06] <cmihai> CSFrost: yeah, that's what I'm going for here ;-)
[13:36:29] <coffman> *sigh*
[13:37:35] <cmihai> coffman: bah, where's your hacking spirit mate?
[13:38:18] <CSFrost> I have some old telecom batteries in the basement, they are actually quite similar to truck batteries :-P
[13:38:26] <cmihai> hehe
[13:40:41] *** Gman has joined #opensolaris
[13:45:19] <Atomdrache> I'm attempting to install an application, and it seems to absolutely demand either ALSA (not happening) or OSS.  Like, it won't compile without one of them.  The Solaris sound system works fine on this computer, but it won't use that.  My question, then, is this:  Is it possible for me to install OSS for this application without hosing the existing Solaris sound system?
[13:45:49] <Atomdrache> And, if so, how likely?
[13:47:18] <Atomdrache> (I was *really* hoping that I wouldn't ever see OSS again after getting rid of the Intel box.)
[13:48:10] <coffman> Atomdrache: if you are lucky, oss will replace sun audio :P
[13:48:20] <Pietro_S> latest oss is quite nice
[13:48:35] <Atomdrache> Does it let more than one application access the same sound device at once?
[13:48:43] <coffman> yes
[13:48:50] <Atomdrache> Awesome, so that means it's not crap anymore.
[13:49:00] <Pietro_S> lot's of, you can use 3d audio and other stuff
[13:49:24] <Atomdrache> However, after the fun I had trying to get that to work in Nexenta, all I will ask is that it not destroy my existing sound system, which, well...works.
[13:49:27] <coffman> Atomdrache: i think there is no other way the using oss
[13:49:45] <coffman> Atomdrache: maybe libao if the app does it
[13:49:54] <Atomdrache> I know I have to use it to make the applicatiom compile.  But, like...can I install it without making anything other than this app use it?
[13:50:25] <Pietro_S> what app is it?
[13:50:30] <Atomdrache> Sabbu.
[13:50:43] *** bnitz has joined #opensolaris
[13:52:09] <Pietro_S> never heard about it, but you can try to install latest oss, it won't break anything on your computer and I heard that it has also some alsa compatibility layer
[13:52:27] <Atomdrache> As long as it doesn't destroy sound in SXCE.
[13:53:07] <Atomdrache> Are there any good binary OSS packages for SPARC, or will I need to compile it?
[13:54:17] <Atomdrache> (looking)
[13:54:20] <g4lt-U60> how did you get a app that requires OSS to freaking compile on sparc in the first place?
[13:54:54] <Atomdrache> It hasn't compiled yet because I don't have OSS :3
[13:55:20] <Atomdrache> I've had to move some libraries to /usr/lib, but besides this picky little detail it's been compiling fine.
[14:00:33] <CIA-26> narayan: 6528212 Programming error in ldc module: ldc handle gets kmem freed and then re-used, 6572885 ldc_init does not properly compute queue length from mtu, 6572891 ldc reliable mode does not process ACK packets properly, 6573492 ldc_rx_hdlr always sends CTRL/NACK on seqID mismatch, 6575608 i_ldc_send_pkt() uses seqID without grabbing Tx lock
[14:04:00] *** ShanghaiScott has quit IRC
[14:04:23] <Atomdrache> I'd better go to bed.  Not even going to try this until I've got some time to stay awake.
[14:06:28] *** triplah_ has joined #opensolaris
[14:07:29] *** master_o1_master has joined #opensolaris
[14:13:12] *** polk__ has joined #opensolaris
[14:19:16] *** master_of_master has quit IRC
[14:29:53] <coffman> g4lt-U60: well, any sound app that is not writen with solaris in mind ...
[14:30:11] <coffman> you ll be lucky if they do OSS
[14:30:29] <g4lt-U60> does OSS even compile on sparc?
[14:31:33] <coffman> g4lt-U60: there are OSS packages for sparc
[14:31:40] <Atomdrache> I found a .pkg for --he said it first.
[14:31:52] <g4lt-U60> most of the craptacular stuff OSS is designed for won't even work on sparc
[14:32:43] *** farak has joined #opensolaris
[14:32:48] <Atomdrache> Out of curiosity I gave another try at the other program that I was trying to compile before trying to compile Sabbu.  I got through configure but make won't work.  Is "macro assignment on dependency line" indicative of any common problem?
[14:33:05] *** calumb has joined #opensolaris
[14:35:24] *** MattMan has joined #opensolaris
[14:37:39] * g4lt-U60 imagines disabling the onboard CS4231 with a OPL3SA
[14:37:53] * g4lt-U60 wheezes from the guffawing
[14:40:14] * Atomdrache tries again with gmake.  Apparently gmake is less likely to barf when dealing with that.
[14:40:44] <Atomdrache> Awesome!  It didn't barf like instantly!
[14:40:51] <Atomdrache> Let's see if it finishes.
[14:41:58] <Atomdrache> Eh, took a few lines.  Different error.  Something I can probably fix later.
[14:42:05] <Atomdrache> For now?
[14:42:18] <Atomdrache> Fuck it, I need sleep :D
[14:42:30] <Atomdrache> (yay progress!)
[14:42:34] <Cyrille> I'm sorry, are you actually talking to someone here, or are you just having a conversation with your keyboard? ;-)
[14:42:51] <Atomdrache> I, ah...don't really know.
[14:42:56] <Atomdrache> Which is why I should go to sleep.
[14:43:10] <FrostCS> Cyrille, you mean you /ignore'ed all three of the people he is talking to?
[14:43:45] <Cyrille> oh dear, I knew I shouldn't have done that...
[14:44:16] <Cyrille> then again I don't recall /ignoring any makefile expert recently...
[14:44:42] *** triplah_ has quit IRC
[14:45:49] *** czar^ has left #opensolaris
[14:46:36] *** nostoi has joined #opensolaris
[14:48:42] *** Teknomancer has joined #opensolaris
[14:49:23] *** yongsun has quit IRC
[14:52:53] *** SirFunk has quit IRC
[14:57:29] *** nrubsig has quit IRC
[15:06:54] *** jmcp has quit IRC
[15:07:19] *** jmcp has joined #opensolaris
[15:09:41] <WickedWicky> mja, so now what am I doing wrong?
[15:09:51] <WickedWicky> the nv69 compile seems to be okay now
[15:10:05] <WickedWicky> but when I do bfu /export/snvbuild/proto/root_i386 /
[15:10:10] <Pietro_S> wtf, why xorg need 433 MB of *my* RAM?
[15:10:19] <WickedWicky> I get the error: Executing /tmp/bfu.4068 /opt/snvbuild/proto/root_i386  generic.root missing or in unknown compression format
[15:10:43] *** Plaidrab has quit IRC
[15:13:23] *** nachox has joined #opensolaris
[15:13:30] <WickedWicky> prolly cause I'm an idiot
[15:13:31] <nachox> morning all
[15:13:31] <WickedWicky> n/m
[15:13:57] *** Plaidrab has joined #opensolaris
[15:14:16] <Teknomancer> does opensolaris use /proc/self/cmdline or /proc/curproc/cmdline ?
[15:15:25] <jmcp> Teknomancer: use it for what?
[15:15:26] <sickness> good afternoon
[15:15:57] <seanmcg> WickedWicky: that proto area holds the tree of built stuff, not the bfu.  The bfu would be in /export/snvbuild/archives/...
[15:16:00] *** hohum has joined #OpenSolaris
[15:16:05] <Teknomancer> jmcp:  for data for the current process' argc, argv...
[15:16:33] <jmcp> why would you open something in procfs when you could use a system call instead?
[15:17:00] <jmcp> or, indeed, use standard features of libc
[15:18:15] <Teknomancer> i'm not sure why, but i'm porting code that does that
[15:18:30] <jmcp> so re-write it to use standardised libc functions and system calls that adhere to posix
[15:20:28] <Teknomancer> well
[15:21:53] <WickedWicky> sean: seems to work now!
[15:22:14] <WickedWicky> I had to set the variable GATEPATH before running bfu though, or it'd try to use /net/ovnbuild
[15:22:30] <WickedWicky> Extracting ufs modules for boot block ... 2120 blocks Extracting generic.usr ...
[15:22:32] <seanmcg> aye, thats a known anoying bug at times.
[15:22:45] <seanmcg> be sure to run ACR afterwards.
[15:22:47] <WickedWicky> fortunatly I can read shell scripts till an extend :D
[15:23:04] *** phimic has quit IRC
[15:23:32] <WickedWicky> okay, will do. It's extracting generic.usr now
[15:23:50] <timsf> Pietro_S is that the RSS value of your Xorg server ?
[15:24:03] <WickedWicky> so yeah, __SSNEXT=""; export __SSNEXT
[15:24:11] <WickedWicky> seems to make SS12 compile nicely
[15:24:12] <Pietro_S> timsf: yes
[15:24:24] <timsf> Leaky X clients ?
[15:24:27] <Pietro_S> size is 460
[15:24:32] <seanmcg> WickedWicky: wait till you reboot and hopfully all works :)
[15:25:24] <WickedWicky> yep! if not, redo from start, it rains like hell here so its not like I have anywhere to go
[15:26:38] <seanmcg> Rain here too... all of europe it seems
[15:26:48] <WickedWicky> yeps
[15:27:00] <Pietro_S> if I logout and log in does xorg restart itself or at least release resources?
[15:27:30] <seanmcg> Pietro_S: yes, xorg restarts
[15:28:09] * timsf seems to remember some way to determine where memory is being used in X
[15:28:14] <WickedWicky> updating /platform/i86pc/amd64/boot_archive...this may take a minute updating /platform/i86pc/boot_archive...this may take a minute Finished.  See /tmp/acr.v0aakz/allresults for complete log.
[15:28:17] <WickedWicky> so
[15:28:20] <WickedWicky> time to reboot I guess
[15:28:23] <Pietro_S> at least I run important things in screen (sometimes I forget it) ;-)
[15:28:26] <WickedWicky> see you all in some minutes, I hope :D
[15:28:36] <timsf> guessing pmap on the Xorg process will just show loads of heap...
[15:29:51] <timsf> alanc would probably know..
[15:31:45] <jmcp> timsf: xrestop
[15:31:53] <timsf> That's the one!
[15:31:56] <Teknomancer> is there any hex-editor (possible with GUI) in opensolaris that i can download ?
[15:32:02] <Teknomancer> need to look at some bin. files
[15:32:05] <jmcp> timsf: www.freedesktop.org/wiki/Software/xrestop
[15:32:18] *** WickedWicky has quit IRC
[15:32:21] <timsf> timf@cranialtrauma[504] grep xrestop /var/sadm/install/contents
[15:32:22] <timsf> damn.
[15:33:46] <Pietro_S> timsf, jmcp: too late already restarted xorg (now eating 37MB, but not for long time)
[15:33:51] * timsf compiles
[15:34:01] <coffman> Teknomancer: you could go by jedit, it has hex plugin
[15:34:19] *** WickedWicky has joined #opensolaris
[15:34:36] <WickedWicky> \\o o// \o/ <o/ \o>
[15:34:47] <WickedWicky> it boots
[15:35:03] <Pietro_S> Teknomancer: ghex you should have it in gnome menu
[15:35:12] <WickedWicky> and I see new things while booting
[15:35:13] <Pietro_S> under development entry
[15:35:42] <WickedWicky> NOTICE: cpu_acpi: _PSS package not found
[15:36:04] <WickedWicky> WARNING cpu_acpi: error parsing _PSS for CPU instance 0
[15:36:08] <WickedWicky> and the same for instance 1
[15:36:23] * timsf glares at firefox - thanks for the pointer jmcp
[15:36:23] <seanmcg> means your box doesn't have power setting feature, on laptops usually..
[15:36:41] <WickedWicky> this is a laptop, so that makes sense
[15:36:42] <Teknomancer> Pietro_S:  ghex? hm... let me look
[15:36:44] <WickedWicky> I can ignore it?
[15:37:25] <Teknomancer> Pietro_S: ok its called ghex2 (its not in the menu though, found it in command line)
[15:37:26] <Teknomancer> thanks
[15:37:31] <Teknomancer> thx also coffman
[15:37:55] *** cmihai has quit IRC
[15:38:08] <Pietro_S> Teknomancer: I use xfce and I have it in menu so I thought it's also in gnome ;-)
[15:38:30] <Teknomancer> Pietro_S:  ah no probs :) the name was good enough :)
[15:39:30] *** calumb has quit IRC
[15:45:00] *** cypromis has quit IRC
[15:48:57] <seanmcg> WickedWicky: ya, should be safe to ignore
[15:49:58] <seanmcg> though you're box may run hot.. and have low battery life, look into getting the frkit on it, google for frkit :)
[15:49:58] <Teknomancer> is it just me or is forums.java.sun.com quite slow :(
[15:50:55] *** IceGuest_7 has joined #opensolaris
[15:51:31] *** cypromis has joined #opensolaris
[15:52:11] <Pietro_S> WickedWicky: what do you have as cpu? only few are supported by frkit ....
[15:53:28] <seanmcg> WickedWicky looks to be having a dual core laptop.. so the current frkit power stuff won't work, unless you disable one of the 'cpus' :-}
[15:55:06] *** SirFunk has joined #opensolaris
[15:57:37] *** EchoBinary has joined #opensolaris
[15:58:51] <WickedWicky> I have a dual core t2300
[15:59:44] <WickedWicky> eratta
[15:59:45] <WickedWicky> 2400
[15:59:49] <WickedWicky> Genuine Intel(R) CPU T2400       U1
[16:00:11] <WickedWicky> bash-3.00$ /usr/sbin/prtdiag System Configuration: Hewlett-Packard HP Pavilion dv8000 (RG021EA#ABH)   BIOS Configuration: Hewlett-Packard F.15 06/13/2006
[16:04:47] <Teknomancer> bye all, must go out for a while,  thx for all the help
[16:04:49] *** Teknomancer has left #opensolaris
[16:05:56] <Gman> jmcp, there?
[16:12:58] *** pfa3rh has joined #opensolaris
[16:14:28] <g4lt-U60> am I the only one that just said "I fucking hope that's a joke" at osol-discuss?
[16:14:35] <IceGuest_7> hi all. I'm using sun download mgr to get the sol-nv-b68-sparc-vN-iso.zip set and it got 1-5 without problem, but disk 6 (v6) fails right away with "Download failed, maximum retries exceeded"
[16:14:44] <IceGuest_7> repeatedly, that is.
[16:15:11] <g4lt-U60> it's six parts?  ewww
[16:15:20] *** IceGuest_7 is now known as umodjm
[16:15:46] <movement> g4lt-U60: it's a bit of a weird joke
[16:15:52] *** obsethryl has joined #opensolaris
[16:16:06] <movement> umodjm: try again later. the SDLC tends to be ... unreliable
[16:16:14] <Pietro_S> iceq: cd version?
[16:16:33] <umodjm> movement, ok, thanks.  Do I actually need all 6?
[16:16:58] *** SymmWork has quit IRC
[16:16:59] <umodjm> it's going on a U10
[16:17:14] <g4lt-U60> umodjm, the short answer to "do I need all the CDs?" is "if you have to ask, yes"
[16:17:30] <coffman> umodjm: yes you need all cds
[16:17:39] <coffman> if you unziped the iso
[16:17:44] <coffman> it is valid
[16:18:08] <coffman> more likely that the cd is bork
[16:18:17] <coffman> just burn another with lower speed
[16:18:51] <dlg> gman?
[16:20:38] <Gman> dlg, hey
[16:21:15] <dlg> hi
[16:30:02] <Wez> anyone here used tcov successfully for multi-threaded apps?
[16:30:25] <pizdec> umodjm, you might want to paste new download link into SDLC
[16:30:40] <pizdec> as in refresh download page and paste new link
[16:30:41] <seanmcg> Gman: hows b-ham ?
[16:30:48] <Wez> the man page (from 1997) says that it likely won't work; I'm wondering if that note is stale, or if there is a preferred way to get coverage analysis for multi-threaded apps
[16:31:11] <Gman> seanmcg, it's been good
[16:31:28] *** m0le has joined #opensolaris
[16:31:36] <m0le> hi all
[16:32:05] <seanmcg> gnome folks any more interested in taking the patches ?-)
[16:32:18] <m0le> I have a question about X11 forwarding over ssh to a windows box. The reason I ask as I have never tried it. Anyone got any ideas of what to do?
[16:32:33] <seanmcg> ssh -X
[16:32:41] <seanmcg> don't set display or the like.
[16:32:54] <m0le> will that only do CDE or will it do gnome as well?
[16:32:55] <WickedWicky> enable Xforwarding in sshd_config, server side. Install Xming on your windows desktop and run it
[16:32:58] <Cyrille> to a windows box?
[16:33:04] <seanmcg> oh, wait to/from a windows box ?
[16:33:20] *** Snake007uk has quit IRC
[16:33:28] <m0le> to a solaris machine from a windowws box
[16:33:29] <WickedWicky> oh, to a windows box.. why x forwarding? why not RDP?
[16:33:40] <m0le> no from a windows machine to solaris
[16:33:42] <WickedWicky> ah
[16:33:48] <WickedWicky> install xming on your windows box
[16:34:40] <WickedWicky> start xming, and then tell your ssh client (putty/securecrt) to enable X tunneling
[16:34:52] <WickedWicky> and make sure Xforwarding is enabled in your sshd_config, server side
[16:35:12] <m0le> Alrighty i just wanted to see if it was possible
[16:35:24] <WickedWicky> it is, I use it on regular basis
[16:35:38] <m0le> i just told them to use one of the linux boxes to do it
[16:35:51] <WickedWicky> that's also possible
[16:35:56] <WickedWicky> xming is freeware btw
[16:36:15] <m0le> its a friend of mine trying to do it
[16:36:27] <m0le> he does not want to dedicate straight to open solaris or linux
[16:36:43] <m0le> he wants to use windows for OFFICE of all things
[16:37:25] <m0le> will that do forwarding for just CDE or could it do gnome as well?
[16:38:22] <WickedWicky> it will do forwarding for whatever X oriented process you start
[16:38:38] <WickedWicky> forwarding an entire desktop environment is not something I'd recommend
[16:38:52] <m0le> yeah true that
[16:38:56] <WickedWicky> I mostly use it to run Oracle installer remotely, or IBM WebSpehre administrator
[16:39:17] <m0le> I appreciiate the help and thank you so much. Now i can get more rest.
[16:39:28] <WickedWicky> ohhh, monthy python on tv tonight \o/
[16:39:36] <WickedWicky> rest well :)
[16:39:46] <m0le> oh i will be back in about an hour or two
[16:39:47] <m0le> thanks again
[16:39:56] <WickedWicky> yw :)
[16:40:16] *** reflect_ has joined #opensolaris
[16:40:17] *** reflect has quit IRC
[16:40:50] <umodjm> http://sourceforge.net/project/showfiles.php?group_id=156984    this shows 5 packages, is one of them all inclusive?
[16:40:56] <umodjm> (xming)
[16:41:26] <WickedWicky> I always download xming and xming-fonts
[16:41:51] <WickedWicky> xming-fonts are required to install oracle via the Oracle Installer
[16:42:21] <m0le> WickedWicky working like a champ.
[16:42:24] *** dlynes_laptop has joined #opensolaris
[16:42:46] <m0le> now i can enjoy my day off and kick my buddy in the shins for boycotting *nix at his house lol
[16:42:57] <WickedWicky> m0le: he might wanna try cygwin
[16:43:07] <WickedWicky> just to start off
[16:43:11] <m0le> yeah
[16:43:19] <dlynes_laptop> oin #asterisk
[16:43:21] <m0le> i told him google will be his friend as well hehe
[16:43:36] <WickedWicky> google is everybody's friend
[16:44:20] <m0le> ./agreed
[16:44:55] *** sparvu has quit IRC
[16:44:55] *** karrot-x is now known as karrotx
[16:46:58] <Pietro_S> m0le: open shell and type "yes" ;-)
[16:47:18] *** sartek has joined #opensolaris
[16:48:44] <m0le> ;)
[16:49:10] *** Netwolf_ is now known as Netwolf
[16:50:53] *** Juan has joined #opensolaris
[16:50:58] *** WickedWicky has quit IRC
[16:52:11] *** Gropi has quit IRC
[16:53:02] *** MattMan has quit IRC
[16:53:03] <Juan> hi guys :) I got an answer... Does anyone know if the sunray270 got BIOS ? I've been surfin' but I've not found anything. Thanks!
[16:53:54] <g4lt-U60> why do you need bios on a sunray?  it's a TERMINAL
[16:54:12] <g4lt-U60> that's like asking for the bios on a VT100
[16:54:41] *** morteng has joined #opensolaris
[16:55:18] <Juan> hehe yeah I know it g4lt-U60 :) But I just wondered how it can do a network boot without BIOS
[16:55:25] <morteng> Hi, Could  someone please take a look at my   Disk usage analyzer ?   http://666kb.com/i/aq82tywkxw72xc6kt.png
[16:55:59] <Juan> g4lt-U60: I know some terminal (thin clients, i.e.) like NICs that has BIOS
[16:56:51] <Juan> that was why I'm asking bout that... But it's like you're saying sunrays 270 don't have BIOS :)
[16:56:58] <Juan> Thank you g4lt-U60.
[16:58:43] <Stric> Juan: it depends on your definition of BIOS.. it does have some software that runs on powerup.. but it does not have the "IBM compatible" x86 BIOS thingie.
[16:59:20] <Juan> Stric: got it! Thanks :) I understand now.
[17:00:00] <Stric> and fyi, that software is about 300kB
[17:02:53] <Juan> Stric: does these sunrays 270 work properly with LTSP? You know that?
[17:03:22] *** arinel has joined #opensolaris
[17:03:23] *** jamesd_ has quit IRC
[17:03:23] <Stric> Sunrays works with Sunray software.
[17:03:34] *** jamesd__ has joined #opensolaris
[17:03:51] <Juan> hehe ok Stric.
[17:04:49] <Stric> Unless someone has written a compatibility layer that re-implements the sunray server part
[17:05:15] <Stric> sunrays are mostly just a network attached graphics card with keyboard/mouse
[17:06:20] *** umodjm has quit IRC
[17:06:22] <arinel> Can anyone please help me understand what is the correct command to do something like: "sudo -u postgres <command>" while logged in as an unprivileged user? I tried pfexec but I was unsuccessful :/
[17:06:51] *** tsoome has quit IRC
[17:07:24] <seanmcg> arinel: you'd need to give the user 'Postgres Administration' privs
[17:08:24] <Juan> ok Stric. Thanks.
[17:08:27] <arinel> seanmcg and how do I do that?
[17:09:47] <g4lt-U60> arinel, use SMC is the easiest way to start with roles
[17:10:16] <g4lt-U60> eventually, you can do it from the commandline, but it's much easier to learn it in SMC
[17:11:03] <arinel> g4lt-U60: thanks!
[17:11:32] <g4lt-U60> that's the first time I've ever been thanked for recommending SMC XD
[17:15:37] <seanmcg> that may come back to haunt you :)
[17:16:09] <arinel> ok, I managed to start the SMC up ;)
[17:16:36] *** solar-star has quit IRC
[17:16:36] <arinel> no worries, I need to use the shell, as I have been requested to make my script compatible with Solaris ..
[17:17:42] <arinel> 1 more question and I am out of your hair for a while: how do I install "stuff"? PostgreSQL server for example? Is there a non-GUI package manager or smth similar?
[17:18:07] <bda> pkgadd/pkgdelete/pkginfo/pkgchk, etc
[17:18:12] <arinel> thanx
[17:18:40] <bda> er. pkgrm.
[17:18:48] <bda> Bloody pkgsrc. Heh.
[17:19:48] *** m0le has quit IRC
[17:20:11] <arinel> wow
[17:20:13] <arinel> :O
[17:20:15] <arinel> # mkdir /home/user
[17:20:16] <arinel> mkdir: Failed to make directory "/home/user"; Operation not applicable
[17:20:28] <bda> arinel: Right. man vold.
[17:21:07] <bda> arinel: If you are just starting with Solaris, it would be really advisable to read Sun's SysAdmin books (docs.sun.com). They are very good.
[17:21:14] <arinel> ok, thank you
[17:21:27] <bda> It took me about a week to find my feet last year. :)
[17:21:34] <arinel> ;)
[17:23:35] <coffman> arinel: you should read this about the automounter http://uadmin.blogspot.com/2005/02/youre-never-far-from-home.html
[17:24:12] *** xuewei has quit IRC
[17:24:49] <g4lt-U60> of course, smc takes care of that when adding a user ;P
[17:25:25] *** timeless has joined #opensolaris
[17:26:12] <arinel> then i won't use it :D
[17:27:01] *** tsoome has joined #opensolaris
[17:27:18] *** Gman_ has joined #opensolaris
[17:28:26] <bda> I don't used the automounter, either, but stick with the /export/home model.
[17:28:34] <arinel> coffman thank you, that guide added the missing part. I already figured out that /export/home looks like a place to store home directories, but I wasn't aware of the automounter.
[17:29:29] *** Gman has quit IRC
[17:37:36] *** Juan has quit IRC
[17:38:40] <FrostCS> hrm, frkit tells me "You must reboot now" .. must I?
[17:39:34] <FrostCS> I'd really rather not.. lol
[17:40:56] *** sartek has quit IRC
[17:42:39] <g4lt-U60> what bonehead made frkit force a reboot?
[17:43:11] *** arinel has quit IRC
[17:43:16] <sickness> well, I suppose it's not meant for busy production servers...
[17:43:17] <FrostCS> I don't know, but I'd rather not reboot 'at all' if possible :-)
[17:43:18] *** Juan has joined #opensolaris
[17:43:42] <sickness> (I use it on my home fileserver too, instead, so I spare some Watts ;)
[17:44:15] <FrostCS> well my new laptop will run it also, once amd gets their act together :-P
[17:44:18] <g4lt-U60> furthermore, what bonehead wrote that error message? there is NOTHING in userland that should force an immediate reboot.  there are things in userland that may need a reboot to function properly, but they should not force a reboot
[17:44:29] <Pietro_S> sickness: what cppu do you have on that home servers?
[17:44:39] <FrostCS> it didn't force the reboot, but it told me I must.. lol
[17:44:47] <g4lt-U60> same diff
[17:44:50] <FrostCS> I thought it was going to force it
[17:45:10] <FrostCS> because it kept ticking along till it updated my boot archive
[17:45:42] <g4lt-U60> hell, even smpatch doesn't force a reboot, and it patches the kernel
[17:46:45] *** emilianOS has joined #opensolaris
[17:47:04] <FrostCS> last time I rebooted was when I updated to 67 (which I was late on).. so I didn't want to reboot now :-)
[17:47:08] <emilianOS> hi everybody
[17:47:17] <sickness> Pietro_S: athlon64, one is 939, the other is AM2 (both single core), they go as low as 1ghz, wich results in great watt and heat reduction :)
[17:47:22] <emilianOS> I've a question
[17:47:23] *** bondolo has quit IRC
[17:48:16] *** bondolo has joined #opensolaris
[17:48:40] <emilianOS> Have you tried to move form webmin 1.170 to webmin 1.350?
[17:49:09] <g4lt-U60> grrr
[17:49:14] <g4lt-U60> SunOS nightengale 5.11 snv_19 sun4u sparc SUNW,Ultra-60
[17:49:15] *** Yamazaki-kun has joined #opensolaris
[17:49:24] <g4lt-U60> FrostCS, ;P
[17:50:39] *** kloczek has quit IRC
[17:50:42] <FrostCS> lol, why 19??
[17:50:58] <FrostCS> what's your uptime?
[17:51:42] <g4lt-U60> not so hot, power outage :(
[17:52:20] <g4lt-U60> got so dmaned hot out that the wires basically plasticized and broke
[17:53:26] <FrostCS> ouch
[17:53:31] <FrostCS> you should have updated then
[17:54:37] <g4lt-U60> true, but as soon as the power got on, I was back on IRC bitching about the power outage ;)
[17:54:56] <sickness> lol
[17:55:43] <FrostCS> heh
[17:56:13] <FrostCS> I think my u60's are still on 65..
[17:56:26] <FrostCS> I've been slacking lately
[17:57:01] <emilianOS> Have you tried to move form webmin 1.170 to webmin 1.350?
[17:57:32] *** Cyrille has quit IRC
[17:57:35] <FrostCS> why?
[17:57:44] <g4lt-U60> eww, webmin?  you have a much stronger stomach than me sir
[17:58:24] <emilianOS> jaja
[17:58:32] *** halto1 has joined #opensolaris
[17:58:46] <FrostCS> mv webmin to trash? heh
[17:59:05] <emilianOS> I cannot log in anymore, I can install it, took the old config, but I cannot log in anymore
[17:59:06] <Pietro_S> sickness: does it also lower cpu fan speen? if yes I would install it on my desktop as well ;-)
[17:59:07] <emilianOS> jaja
[17:59:50] <axisys> is there a sol pkg for python 2.5 available?
[17:59:58] <sickness> Pietro_S: my two motherboards are an asus and a gigabyte, and with their respective slow fan systems enabled in bios, they already had quiet fans even before frkit
[18:00:05] <sickness> Pietro_S: don't know about other systems...
[18:00:51] <FrostCS> you can manually control fan speed with a controller also..
[18:00:55] <FrostCS> rheostat or whatever it's called
[18:00:56] <Pietro_S> I have alsoo asus mb, it lower fan speed but I still heard them
[18:01:10] <FrostCS> or, change fans :-P
[18:01:33] <sickness> Pietro_S: well... I've a system specifically build for noise reduction...
[18:01:40] <sickness> FrostCS: rheobus :)
[18:01:40] <bda> axisys: SUNWPython*
[18:01:53] *** jcea has joined #opensolaris
[18:01:55] <axisys> bda: for python 2.5?
[18:01:58] <FrostCS> sickness, rheobus is software controller right?
[18:02:23] <sickness> FrostCS: nope, hardware, it's a simple resistence with wires and a trimmer
[18:02:26] *** jcea has left #opensolaris
[18:02:29] <Pietro_S> I was thinking about buy big passive for cpu ..., right now I use box cooler witch comes with cpu
[18:02:34] *** jcea has joined #opensolaris
[18:02:41] <bda> axisys: ah, no. Sol10u3 has 2.3.
[18:02:56] <bda> axisys: You could use blastwave or pkgsrc, though.
[18:02:57] <FrostCS> sickness, ah, well atleast I had the rheo right :-)
[18:03:08] * FrostCS shutters
[18:03:17] <axisys> bda: blastwave has 2.3
[18:03:19] <axisys> :-(
[18:03:33] <sickness> Pietro_S: I've Scythe Katana2, virtually inaudible :)
[18:03:35] <axisys> bda: pkgsrc? u have a link? let me google it
[18:03:47] <FrostCS> Pietro_S, last time I was in tokyo I picked up about 40 fans, to replace a lot in my systems, cut noise in half :-P
[18:03:56] <g4lt-U60> axisys, there's always SUNWspro ;P
[18:03:58] <bda> axisys: pkgsrc.org. But it looks like 2007Q1 only has 2.4. :<
[18:04:05] <FrostCS> Scythe has some great fans.. especially I think it's the green or blue boxed ones (reduced noise)
[18:04:17] <bda> You could always graft your own on there or see what the wip branch has, but, eh.
[18:04:24] <FrostCS> was also some other makes too.. which I'd have to look up
[18:04:27] <bda> (building your own pkg is not an awful idea, really)
[18:04:55] *** MattMan has joined #opensolaris
[18:05:10] *** emilianOS has left #opensolaris
[18:05:31] *** Dar has joined #opensolaris
[18:05:32] <sickness> I only used the scythe for the cpu coolers, but yeah, I have to say they did a really good work on noise reduction
[18:05:48] <sickness> for case fans I used Noctua, theyre also top of the notch
[18:06:23] <axisys> g4lt-U60: absolutely
[18:06:52] <axisys> bda: i never used pkgsrc.. are they any good for sol platform? also what is there default prefix?
[18:06:57] <bda> axisys: There is also openpkg.[org,com], but I have not tried their stuff at all.
[18:07:06] <bda> axisys: /usr/pkg. I use it on all my prod systems.
[18:07:22] <bda> It is not 100%, but I don't usually run into major problems.
[18:07:44] <bda> You can tie PREFIX wherever you want, though.
[18:07:54] <bda> (/usr/pkg is annoying, if you use sparse zones)
[18:08:07] *** plutonas has joined #opensolaris
[18:08:29] <plutonas> i just got my opensolaris starter kit, and was not able to run any of the 3 live cd's provided,
[18:08:41] <axisys> bda: does it give u an option to change prefix?
[18:08:53] <bda> axisys: Yup, during bootstrap.
[18:08:54] <plutonas> If I try directly on the hw, it freezes, and using qemu, i get a segmentation fault
[18:08:55] <plutonas> :P
[18:09:08] <axisys> bda: i found python 2.5.1 on openpkg
[18:09:15] <axisys> bda: neat site
[18:09:23] <bda> axisys: If you use openpkg, let me know how it works out. :)
[18:09:34] <axisys> g4lt-U60: i wonder if they uses sun studio to build their pkgs ;-)
[18:09:51] <bda> Lots of Free software wants gcc. :\
[18:09:55] <axisys> bda: need to find out where they prefix and if i can change it
[18:10:08] <bda> axisys: I would be pretty shocked if you couldn't.
[18:10:21] <axisys> bda: well can u do that for CSW?
[18:10:53] <bda> Unsure, but presumably.
[18:11:34] <Pietro_S> what is default/recomended library for rendering svg?
[18:12:30] <FrostCS> plutonas, which hardware?
[18:13:14] <plutonas> FrostCS: pentium m 1.8
[18:13:25] <plutonas> but in qemu hw shouldn't matter
[18:13:29] <FrostCS> you need a little bit more information then that
[18:13:43] <plutonas> i managed to emulate a lot of different oses through qemu...
[18:13:48] <plutonas> FrostCS: what more do you want?
[18:14:23] <FrostCS> your system manufacturer, model? if you built it, what board did you use.. video card.. etc
[18:14:54] <plutonas> ok, just a minute, i think the best is to get a link
[18:15:05] *** morteng has quit IRC
[18:15:31] <plutonas> FrostCS: here you go, http://www.thinkwiki.org/wiki/Category:R50e
[18:16:02] <plutonas> though there are some minor differences, i have a pentium m 1.8 which is not listed there, 512M memory, 60G hd
[18:16:09] <plutonas> the rest should be as in the link
[18:18:11] <FrostCS> okay, now explain what happens?
[18:18:26] <plutonas> on real hw or in qemu?
[18:18:32] <FrostCS> your laptop is reported to work on SXCE atleast.
[18:18:39] <FrostCS> real hw, I could care less about qemu
[18:19:10] <plutonas> ok, the i select the first live cd
[18:19:15] <plutonas> don't remember the name
[18:19:24] <FrostCS> well this is important
[18:19:40] <FrostCS> because I only use Solaris Express Community Edition
[18:19:53] <plutonas> so i have to retry you say, and write down the name?
[18:20:08] <g4lt-U60> how much memory do you have allocated to the qemu session?  your laptop's entire memory isn't enough to install SXDE technically
[18:20:11] <FrostCS> mainly, that is the support you'd get in this room, Developer edition also, but it's just an older version of community..
[18:20:35] <plutonas> g4lt-U60: i don't allocate any, i don't know how much is allocated by default
[18:20:44] <FrostCS> I was hoping the media included wouldn't have been sxde, g4lt-u60 :-(
[18:20:45] <g4lt-U60> FrostCS, SXDE's min memory is 768M IIRC
[18:21:17] <g4lt-U60> min to install SXCE is still 128M, right?
[18:21:24] <FrostCS> g4lt-u60, you can still do a normal install without the gui right?
[18:21:28] <FrostCS> 256
[18:21:32] <Pietro_S> g4lt-U60: for text install? I doubt it I installed it on 512MB notebook
[18:21:33] <plutonas> g4lt-U60: 768M !!!!!
[18:21:46] <FrostCS> this laptop has 256mb.
[18:21:51] <plutonas> why would opensolaris need that much?
[18:21:57] <g4lt-U60> plutonas, yup, which is why I haven't installed SXDE on my craptop
[18:22:15] <plutonas> omg, i think i am not interested anymore in trying opensolaris
[18:22:21] <FrostCS> SunOS p1510d 5.11 snv_66 i86pc i386 i86pc
[18:22:24] <plutonas> that is almost as much as vistas requirements
[18:22:43] <FrostCS> I am running 256mb..
[18:22:51] <FrostCS> so you can either be serious, or cry
[18:22:54] <plutonas> but i still don't understand, why it needs that much for the gui installation
[18:22:57] <timsf> Me too (slowly)
[18:22:57] <g4lt-U60> plutonas, the initial ramdisk is a bitch, after you get over the install hump, you can run fine with 128M
[18:23:20] <plutonas> hm, i see... strange though
[18:23:33] *** mihaic has joined #OpenSolaris
[18:23:35] <trs81> sxde is 768 atm because it has two copies of java or something
[18:23:36] <timsf> 768 is only for running it - if you use the Solaris Express Community edition, it'll install with 256 (at least, last time I checked)
[18:23:37] *** mihaic is now known as cmihai
[18:23:55] <timsf> (there's different grub menu options for dev vs. community)
[18:23:56] <cmihai> timsf: what?
[18:24:17] <timsf> cmihai, what what ?
[18:24:21] <g4lt-U60> timeless, other way around, you need the 768 for ramdisk for the GUI install
[18:24:24] <FrostCS> "Memory size: 247 Megabytes" is what it tells me :-P
[18:24:46] <cmihai> The developer entry in Grub needs 768 to install the devel tools, but you can skip that and install them later... it's fine.
[18:24:55] <timsf> Right.
[18:25:01] <cmihai> FrostCS: 256RAM with 8MB video?
[18:25:14] <FrostCS> cmihai, oh yea.. forgot it's integrated
[18:25:35] <FrostCS> let's just be sane, and pretend SXDE doesn't exist, alright?
[18:25:44] <FrostCS> that will make my day go much smoother
[18:25:44] <cmihai> I love it when I get things right in the middle of a conversation :-)
[18:26:12] <cmihai> Well, since SXDE is pretty much just another SXCE, you can ignore it ;-)
[18:26:39] <g4lt-U60> basically SXDE during the SUNWspro installaation has a point where the JVM is running for the installation and it builds a SECOND JVM for running the IDE
[18:27:15] <g4lt-U60> ON TOP OF THE ramdisk
[18:28:18] <g4lt-U60> (or by that time is the ramdisk gone?)
[18:28:25] <cmihai> Nope
[18:28:28] <FrostCS> my skull is burning
[18:28:38] <cmihai> g4lt-U60: or wait, it does that after the first reboot
[18:29:03] <g4lt-U60> that's what I thought as well, but ICBW
[18:29:22] * g4lt-U60 hasn't used the GUI installer in months
[18:29:34] <cmihai> JumpStart ftw ;-)
[18:29:41] <g4lt-U60> ayup
[18:29:45] <cmihai> LiveUpgrade too. You hardly ever need to install.
[18:30:35] <g4lt-U60> now if only zfs root could be done so I could set the second BE on a zpool
[18:30:48] *** MattMan has quit IRC
[18:31:35] * FrostCS hasn't used the GUI installer in years
[18:32:12] <FrostCS> zfs root still isn't in sparc yet is it?
[18:32:21] <g4lt-U60> on that laptop, one can see why ;P
[18:32:48] <cmihai> FrostCS: nah, see onnv flag days
[18:33:07] <FrostCS> I try not to hurt myself both physically and mentally at the same time
[18:33:18] <g4lt-U60> FrostCS, nope, otherwise I'd force myself to update this and luupgrade at will
[18:33:19] <oxygene> FrostCS: they probably still design grub-for-sparc because they don't want to support zfs in OBP
[18:33:54] <richlowe> that's not (exactly) what the zfsboot spec suggested.
[18:33:57] <g4lt-U60> my proposed BE is a 711 full of cheap disks
[18:34:03] <FrostCS> g4lt-U60, I only use this for mail usually.. and I have pidgin running on it, I usually irc on csfrost box hehe
[18:34:04] <coffman> oxygene: hurts hurts hurts
[18:34:13] *** seanmcg has quit IRC
[18:34:36] <FrostCS> pfft, I'd rather have OBP
[18:34:40] <g4lt-U60> richlowe, you mean no lu BE on my 711? :(
[18:34:43] <FrostCS> rather have it on x86 also..
[18:35:40] <CSFrost> which reminds me.. gonna save my eyes some
[18:35:44] *** stevel has joined #opensolaris
[18:35:45] *** ChanServ sets mode: +o stevel
[18:35:55] <richlowe> oxygene: http://www.opensolaris.org/os/community/arc/caselog/2006/370/
[18:36:24] <g4lt-U60> oh, you weren't talking about lusetup on a zpool.  whew
[18:37:41] *** Dar has quit IRC
[18:38:18] *** morten1 has joined #opensolaris
[18:38:40] <g4lt-U60> don't SCARE me like that
[18:38:48] <morten1> After installing opensolaris, can I as multiboot also append win XP and fedora  Please?   http://multiboot.solaris-x86.org/v/2.html  says I  should always start with win NT
[18:39:33] <g4lt-U60> I'd assume so, but I'm not an X86 jock
[18:39:58] <coffman> morten1: yes and no - sxce/de will most likely take care about your windows, but not about your linux...
[18:40:24] <CSFrost> I'd leave windows completely out.
[18:41:36] <CSFrost> you can add grub entries into the solaris express boot loader for fedora
[18:42:15] <CSFrost> s/boot loader/grub
[18:54:38] *** sommerfeld has joined #opensolaris
[18:54:39] *** ChanServ sets mode: +o sommerfeld
[18:58:09] <nachox> 6522811 Implement backdoor for NSA to allow remote access to kernel data ??
[18:58:33] *** KermitTheFragger has quit IRC
[18:59:21] *** slowhog has joined #opensolaris
[18:59:42] *** Yamazaki-kun has quit IRC
[18:59:51] <sommerfeld> nachox: uh, wha?
[19:00:17] <g4lt-U60> nachox, already been discussed here, but movement and myself
[19:00:28] <CIA-26> kcpoon: 6530536 SCTP failed to initialize conn_t states in re-used conn_t blocks, 6578483 sctp_stack_init() can panic under low memory
[19:00:29] *** baijiutung has quit IRC
[19:00:48] <g4lt-U60> sommerfeld, osol-discuss mailinglist
[19:02:06] *** Gman_ has quit IRC
[19:05:33] <nachox> g4lt-U60: where? :)
[19:06:07] *** Juan has quit IRC
[19:06:08] *** swa has quit IRC
[19:06:11] *** morten1 has quit IRC
[19:07:13] <g4lt-U60> [0814MDT] g4lt-U60 am I the only one that just said "I fucking hope that's a joke" at osol-discuss?
[19:07:22] <g4lt-U60> [0815MDT] movement g4lt-U60: it's a bit of a weird joke
[19:08:53] *** halto1 has left #opensolaris
[19:10:20] *** bigjohnto has joined #opensolaris
[19:12:30] <bda> g4lt-U60: ?
[19:14:17] *** bnitz has left #opensolaris
[19:17:53] *** alanc-away is now known as alanc
[19:18:28] *** reflect__ has joined #opensolaris
[19:18:29] *** reflect_ has quit IRC
[19:19:53] <CSFrost> holy mother of.. I guess that's what I get for flipping through irc channels
[19:24:32] *** WickySolaris has joined #opensolaris
[19:30:21] <nachox> g4lt-U60: jokes like that scare people
[19:31:34] *** plutonas has left #opensolaris
[19:32:14] <alanc> sending the bogus notify to osol-discuss instead of onnv-notify may have got a bigger audience, but made it clearer it was a spoff
[19:32:17] <alanc> spoof
[19:32:30] <richlowe> and more clearly demonstrated that someone is a dumbass.
[19:33:19] <alanc> unlike Casper's April 1 putback notice one year that went out to the right mailing list, and listed all the files in ON changed since the last release with a putback message of something like "Too many new features confuse customers, reverting to Solaris.previous"
[19:33:56] *** m0le has joined #opensolaris
[19:34:15] <m0le> Hi everyone
[19:34:40] <nachox> alanc: :)
[19:35:11] <m0le> XDMCP is annoying lol
[19:35:13] <alanc> the real 6522811 is a bug against an internal tool that seems to have something to do with sales people entering orders into our db
[19:35:22] * alanc flees
[19:35:50] <stevel> alanc: he tried to send it to onnv-notify
[19:36:34] <m0le> Anyone here familiar with XDMCP logins? I just had a few quick questions about it.
[19:36:41] <stevel> too late
[19:36:42] <stevel> he fled
[19:36:53] <richlowe> stevel: -notify actually rejected something?
[19:37:05] <stevel> richlowe: yeah
[19:37:17] <richlowe> stevel: so how the hell does it get spammed (via something other than jive no less) now and then?
[19:37:40] <stevel> it's email - it's easy to spoof email
[19:37:51] <nachox> stevel: putting it in usr/closed is so not opensource, nobody reads the code anyway :P
[19:37:52] <stevel> and pretend you are someone else, who is allowed to post
[19:37:57] <alanc> m0le: I might know a little about XDMCP
[19:38:06] <richlowe> stevel: is this where you force me to pull out examples?
[19:38:34] <stevel> at least that's what happened with the most recent one earlier this week
[19:38:41] <stevel> which came from 'onnv-notify at opensolaris dot org'
[19:39:04] <m0le> alanc: I was just wondering when I actually logged in it kills the session. I am using this Xming to connect for testing for a friend to connect to my solaris box.
[19:39:15] <richlowe> stevel: I thought one of the earlier ones came through without such things.
[19:39:42] <m0le> alanc: I connect to it, once user name and pw is in and I hit enter it stops thinks and then closes
[19:39:46] <stevel> possibly - do you remember when it was?
[19:39:51] *** nostoi has quit IRC
[19:39:56] <stevel> ah
[19:39:59] <stevel> i see a couple in may
[19:40:04] <stevel> also from onnv-notify at opensolaris dot org
[19:40:13] <alanc> m0le: hmm, could be your session is crashing
[19:40:30] <alanc> look in /var/dt/Xerrors to see if any errors got logged
[19:41:51] <m0le> yeah there are some fatal errors
[19:42:14] <m0le> but I am wondering if the fact the user is not being created in the /home folder or the /export/home
[19:42:39] <m0le> one sec I will post
[19:42:42] <m0le> in paste
[19:43:23] <reflect__> lasseoe: you awake?
[19:45:57] <alanc> if the user has a valid home directory somewhere that he owns and can write to, the path shouldn't matter
[19:46:47] <m0le> that is the problem it does not
[19:47:16] <m0le> it can ssh to it and such but no folder to write to or anything but copying from this windows box is not working
[19:47:41] *** estibi has joined #opensolaris
[19:48:47] <g4lt-U60> ahh, time to trot out uadmin
[19:49:14] <g4lt-U60> http://uadmin.blogspot.com/2005/02/youre-never-far-from-home.html
[19:49:15] <m0le> uadmin? sorry not too familiar with that
[19:49:38] <g4lt-U60> jamesd's blog, where he explained why you don't have a homedir
[19:50:51] <m0le> going to read it now
[19:52:22] <estibi> hi
[19:53:40] *** dclarke has joined #opensolaris
[19:55:26] *** cydork has quit IRC
[19:56:47] <nachox> g4lt-U60: he has a real job now he doesnt actually need the hits :P
[19:57:17] <g4lt-U60> nachox, show me a better reference and I'll use it
[19:58:53] <nachox> there is non that i know of
[19:59:13] <g4lt-U60> so he gets the traffic
[19:59:34] *** sparvu has joined #opensolaris
[19:59:39] <dclarke> anyone know where I can get information about the ZFS versions currently releasd ?
[19:59:45] <timsf> Yes.
[19:59:49] <dclarke> I think S10U3 uses ZFS ver 3 .. but I am not sure
[20:00:00] <timsf> Oh, that. Hmm..
[20:00:18] * timsf digs
[20:00:37] <estibi> dclarke: s10u3 uses v3
[20:00:39] <timsf> http://opensolaris.org/os/community/zfs/version/3/
[20:01:34] <timsf> 4,5,6 are in nv_62
[20:01:45] <estibi> dclarke: the latest is v7
[20:01:48] <timsf> 7,8 are in nv_68
[20:01:57] <dclarke> thank you .. I have snv_68 here also
[20:02:04] <timsf> sorry, 8's in 69.
[20:02:09] <dclarke> but am currently seeing very strange behavior on S10u3
[20:02:13] <dclarke> its .. weird
[20:02:37] <timsf> do tell
[20:03:06] <dclarke> ls -lapin PHX*
[20:03:26] <dclarke> PHX-2001-001_23APR07_TEMPORARY.x_t: Value too large for defined data type
[20:03:28] <dclarke> 7428745 -rw-rw-r-- 1 3227 3000 446464 May 23 10:39 PHX-1001-008-TEMP.prt
[20:03:29] <dclarke> 7429302 -rw-rw-r-- 1 3227 3000 442880 May 23 12:06 PHX-1001-015-TEMP.prt
[20:03:41] <dclarke> Value too large for defined data type ??
[20:03:51] <nachox> cool error
[20:03:59] <dclarke> yeah .. scary too
[20:05:09] <timsf> How large is the filesystem ?
[20:05:18] <dclarke> is there a way to get the current ZFS version on this machine ?
[20:05:40] <estibi> dclarke: zpool upgrade -v
[20:05:41] <sommerfeld> dclarke: there's 64-bit ls
[20:05:52] <timsf> yeah, what I was thinking..
[20:06:02] <timsf> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6419989 explains a bit
[20:06:22] <dclarke> 1.39 TB filesystem
[20:07:04] *** Tpenta has quit IRC
[20:08:43] *** freakazoid0223 has quit IRC
[20:09:10] <dclarke> damn .. this is just a regular ZFS ver 3 on S10u3 and its nothing fancy
[20:09:21] <dclarke>  /usr/xpg4/bin/ls -lapin
[20:09:26] <dclarke> results in the same error
[20:09:27] <CSFrost> course not, fancy started at 5
[20:09:36] <timsf> Yeah, run the 64 bit version of ls.
[20:11:25] *** freakazoid0223 has joined #opensolaris
[20:12:01] <CSFrost> it's  /usr/bin/sparcv9/ls   right?
[20:12:28] <dclarke> thats not what I used
[20:12:37] <dclarke> and isaexec should resolve that for me anyways
[20:13:23] <richlowe> ls isn't set up to isaexec, I don't think.
[20:13:51] *** SirFunk has quit IRC
[20:14:07] <dclarke> damn
[20:14:32] <dclarke> so this is a PATH issue and I need to set /usr/sbin/sparcv9 and /usr/bin/sparcv9 in the head of my PATH .. thats just *wrong*
[20:14:35] <tomww> value too large i've seen in past times with brandz. booted in 32bit no probs, with 64bit ls gave value too large...
[20:15:57] <dclarke> feels like a bug that is messing up my NFS clients
[20:16:34] <dclarke> with CATIA running on NFS clients they blow up when they try to access the file on the NFS server .. which is S10u3 and ZFS based of course
[20:16:43] *** sfire||mouse has joined #opensolaris
[20:17:06] <dclarke> this is sooo *wrong*
[20:17:12] <dclarke> # /usr/bin/sparcv9/ls -lapin PHX*
[20:17:14] <dclarke> 7428745 -rw-rw-r-- 1 3227 3000 446464 May 23 10:39 PHX-1001-008-TEMP.prt
[20:17:16] <dclarke> 7429302 -rw-rw-r-- 1 3227 3000 442880 May 23 12:06 PHX-1001-015-TEMP.prt
[20:17:18] <dclarke> 6920106 -rw-rw-r-- 1 3227 3000 8667044 Nov 29 2076
[20:17:19] <dclarke> PHX-2001-001_23APR07_TEMPORARY.x_t
[20:17:29] <dclarke> wow ...
[20:17:39] <dclarke> Nov 29 2076 ?
[20:17:48] <dclarke> that's not right ..  it was created today
[20:17:58] <jbk> perhaps you live near a wormhole
[20:18:30] <dclarke> no .. something is seriouly wrong with a file created today that causes ls to blow up
[20:19:17] <nachox> dclarke: copy it to an ufs filesystem and try again, does that still happen?
[20:19:26] <nachox> or cp is also broken?
[20:19:43] <dclarke> maybe .. I'll try /usr/xpg4/bin/cp -p
[20:20:45] <CSFrost> This has been discussed in the past, unfortunately I can't remember the explainations for it (too old), but the same exact instances occured.
[20:22:12] <sommerfeld> the timestamp is out of range for a 32-bit time_t
[20:22:44] <sommerfeld> that's why the stat failed.
[20:22:48] <dclarke> I figured as much while looking at it
[20:22:54] <sommerfeld> the mystery is what caused the timestamp to be screwed up
[20:23:02] <dclarke> I have to figure out how some user created that thing
[20:23:21] <nachox> were you able to copy it?
[20:23:26] <dclarke> the client was running Solaris also
[20:23:31] <jbk> touch?
[20:23:44] <jbk> or perhaps buggy software?
[20:23:51] <dclarke> ummm .. well a user would need to set their date into the future .. that far .. and then create the file over NFS
[20:24:09] <dclarke> I think that a number of things would blow up if the time_t datatype is exceeded
[20:24:13] <jbk> i thought you could specify an arbitrary date w/ touch?
[20:24:21] <dclarke> never tried that
[20:24:24] <jbk> though dunnot if it uses 64-bit interfaces for that
[20:24:35] <dclarke> settime [-f ref_file] [date_time] file...
[20:25:44] <WickySolaris> hey guys? I am reading the zfs administrator pdf and I am a bit confused. Let's say I add c4d0s4, which is a 150GB slice, and I want to create a ZFS filesystem of 30GB and mounted it on /export/blerg, am I forced to use volumes then or can I settle with the quota property?
[20:26:32] *** Fetch_ has joined #opensolaris
[20:27:00] <jbk> zvols are for when you need a block device
[20:27:18] <jbk> quotas are how you cap the size of a filesystem in a zpool
[20:27:24] <jbk> (or reservations)
[20:27:32] <WickySolaris> ok, quotas is what I need then
[20:27:56] <WickySolaris> thank you!
[20:30:02] *** Fetch has quit IRC
[20:31:16] <dclarke> snv_68 has super cool up to date ZFS : http://rafb.net/p/uNxJEO37.html
[20:31:50] <coffman> WickySolaris: the freedom of zfs is that the space of the zpool is available to every fs that you create, future more you can make hard and soft quotas
[20:32:10] *** SirFunk has joined #opensolaris
[20:32:56] <coffman> when does sxce b69 come out ? next week or this?
[20:33:28] <sommerfeld> 69 will have even more up-to-date zfs (delegation)
[20:33:32] <WickySolaris> coffman: that's exactly what I am looking for. The hassle with LVM under Linux is that when you create a 50GB volume and they only use 60MB of it you kinda waste heaps of space
[20:33:42] <dclarke> sommerfeld:  thanks for the reality check there .. I figured that the date was the issue
[20:34:03] <dclarke> sommerfeld:  am also surprised that the 64-bit ls can render some result
[20:34:40] <alanc> coffman: 69 came out internally this week, so external should be next week
[20:34:53] <sommerfeld> the 64-bit ABI has a 64-bit time_t
[20:35:24] <dclarke> beautiful ..
[20:35:30] <coffman> alanc: ty
[20:35:37] <dclarke> now I need to figure out how some user created this
[20:35:39] <coffman> im just having trouble to keep up
[20:35:43] <dclarke> and then kill the user
[20:37:01] <coffman> WickySolaris: have a look at the demos http://www.opensolaris.org/os/community/zfs/demos/
[20:37:28] <coffman> hmm, i dont get it, since when is there more then one paste buffer?
[20:37:45] *** Murmuria has joined #opensolaris
[20:37:50] <coffman> i have difficults to paste to gnome terminal via hotkey
[20:38:04] <Murmuria> sommerfeld: hi!
[20:38:10] <WickySolaris> i just created my very first ZFS pool and fs
[20:38:17] * WickySolaris is not a ZFS virgin anymore \o/ :P
[20:38:35] <coffman> the hotkey buffer contains something different then right klick paste one
[20:38:56] <coffman> WickySolaris: you can make a couple more fs, like 100 or so :P
[20:39:16] <WickySolaris> now I can one by one mount my ext2fs partitions read-only, copy the stuff to the zfs pool, swipe the disks, and add them to the pool and slowely migrate from linux/ext2fs to solaris/zfs
[20:39:44] <WickySolaris> coffman: the problem is that I am migrating my OS and data from linux, started yesterday, and I definately dont wanna lose my mp3s and all
[20:39:51] <WickySolaris> so I do it disk by disk
[20:39:58] <coffman> WickySolaris: uhm, so you dont do raid ?
[20:40:15] <WickySolaris> hm? not at the moment no
[20:40:21] <WickySolaris> I broke the raid under linux
[20:40:50] <coffman> WickySolaris: you do with zfs, zpool does out stripe, means you will have a big raid 0
[20:41:26] <WickySolaris> I was more thinking about doing a: zpool add zpool_lilith mirror c0d0 c1d0
[20:41:31] <WickySolaris> is this possible?
[20:41:36] <richlowe> coffman: I suspect the key is pasting form the clipboard.
[20:41:47] <richlowe> coffman: whereas middle-click traditionally pastes the primary selection.
[20:42:00] <richlowe> (or possibly the cut-buffer, if you're in a world of suck)
[20:42:42] <sommerfeld> Murmuria: I owe you a bunch of replies...
[20:42:52] <coffman> how was that again with the multiple copies flag? if i have a zpool with a bunch of disk its assured that does are not go to one disk?
[20:43:13] *** REllin1 has quit IRC
[20:43:30] <sommerfeld> coffman: not assured, but if space is available it will avoid placing multiple copies on the same top-level device
[20:45:00] <coffman> so if i do 3 copies i can be quite sure if i have 2 disks and the space ?
[20:45:24] <coffman> WickySolaris: would be the better option yes
[20:45:40] *** m0le has left #opensolaris
[20:45:49] <WickySolaris> coolies
[20:45:59] <Murmuria> sommerfeld: and I owe GSoC program a lot of hard work...:(
[20:46:36] <Murmuria> sommerfeld: BTW what did you evaluate me on?
[20:46:40] <coffman> richlowe: pasting was from ctrl+insert and right mouse contex menu - so it should come from the same buffer or im wrong?
[20:46:59] <Murmuria> and what "grade" did I get?
[20:47:30] <richlowe> coffman: Hm, I thought they came from different places.
[20:47:57] <richlowe> shift+insert from the primary, right-click-and-paste from the clipboard.
[20:48:02] <richlowe> no clue about ctrl+insert, I avoid gnome-terminal ;)
[20:48:07] <sommerfeld> Murmuria: the form was about a dozen multiple-choice questions.  no single "grade"
[20:48:32] <coffman> richlowe: i mostly got thos between firefox and gnome terminal..
[20:48:44] <coffman> i would use xterm but its ugly...
[20:49:12] <coffman> and xfce terminal got the same bugs of gnome terminal and sucks more memory
[20:49:20] <coffman> *bitch*
[20:49:30] <Abe_Froman> rxvt
[20:49:34] <tomww> coffman: rxvt :-)
[20:49:37] <Abe_Froman> who needs unicode or tabs?
[20:50:00] <Abe_Froman> shirt idea: export LC_ALL=C
[20:50:11] <tomww> screnn solves tabs much better for *me*
[20:50:17] <WickySolaris> nv68's gnome-terminal is giving me artifacts some times
[20:51:10] <richlowe> 6566332, I'd assume.
[20:53:07] *** nachox has quit IRC
[20:55:21] * bubbva is away: lunch
[20:55:40] *** SirFunk has quit IRC
[20:56:05] <coffman> WickySolaris: im with you on that
[20:56:35] <WickySolaris> awww, this is even cooler. I can add a disk, non mirrored. buy a disk tomorrow, equal in size, and attach it to the disk added today and voila, it will be a mirror
[20:56:37] * WickySolaris sheds a tear
[20:56:50] <sommerfeld> well, after a pool traverse it will be..
[20:57:15] * WickySolaris is reading page 60 of the ZFS administration guide
[20:58:49] <WickySolaris> I am seriously going to make a case for the company where i work to ditch all Linux machines and migrate to opensolaris, I am using this for nearly a week now and I already see 0954098433094 advantages
[20:59:13] *** cmihai has quit IRC
[20:59:25] * coffman hands WickySolaris a gun
[20:59:37] <coffman> you will need to defend yourself
[21:00:42] <WickySolaris> I'll just put my Ministry of Sound cds very loud over the speakers till they agree with me
[21:00:48] <CIA-26> sjelinek: 6557849 Remove comma from "Solaris Express, Developer Edition", 6581317 Need internal pkg to deliver libdiskmgt.h for install consolidation CBE
[21:01:35] *** farak has quit IRC
[21:02:45] <WickySolaris> I think I'll have a struggle with the argument that Oracle and such don't deliver support on their software running on opensolaris though :(
[21:03:23] <stevel> for that matter, Sun won't deliver support on OpenSolaris ;-)
[21:03:28] <stevel> (yet, anyway)
[21:03:57] <WickySolaris> yea, true. Not that we have support on our linux now but hey.
[21:04:14] <WickySolaris> And I guess on our sparcs we keep running Solaris anyway, contract wise and all
[21:04:15] <sommerfeld> WickySolaris: anyhow, adding a mirror later is something you can also do with SVM.
[21:04:16] *** reflect_ has joined #opensolaris
[21:04:40] <reflect_> tell me something, if I want to use zfs in production..   what would be my best option?
[21:05:02] <WickySolaris> sommerfeld: I know :)
[21:05:05] <kito> option for what?
[21:05:18] <kito> I would say the best option is to use it!
[21:05:24] <WickySolaris> I was more comparing the advantages over Linux/mdadm/lvm with those of opensolaris/zfs
[21:05:31] *** EchoBinary has quit IRC
[21:05:35] <sommerfeld> reflect: it depends on what you're trying to do, what hardware you have available, etc.,
[21:05:47] *** cypromis has quit IRC
[21:05:50] <WickySolaris> at work we use SVM a lot, since v440 and v240s dont come with a hardware raid controller
[21:06:20] <WickySolaris> we also tried veritas but didnt really see the added value for paying $$$$$$
[21:06:20] *** reflect__ has quit IRC
[21:06:53] <reflect_> sommerfeld: then let me rephrase.. is ZFS stable enough to be used in vital production data?
[21:07:27] <reflect_> s/in/with
[21:09:29] *** koolniczka has joined #opensolaris
[21:09:35] <sommerfeld> reflect: absolutely, especially if your primary requirement is avoiding data loss
[21:10:21] <reflect_> that's one of the very top priorities, yes..  ofcourse "normal throughput" is on that list aswell
[21:10:41] <kito> a semi-related question, is zfs boot stable enough to use in production?
[21:12:34] <g4lt-U60> there ain't no xfs boot
[21:12:38] <g4lt-U60> zfs even
[21:12:43] <sickness> maybe, but AFAIK is not upgradeable
[21:12:47] <sickness> yet
[21:13:06] <g4lt-U60> until I can boot it on sparc, there ain't no zfs boot
[21:13:07] <sommerfeld> reflect: so, there are going to be particular areas where you'll find pathological performance.  they're working on those as they turn up...
[21:13:19] <sommerfeld> in general the performance is very good
[21:13:40] <sommerfeld> and stuff like the NVRAM-based-ZIL looks very promising
[21:14:01] <g4lt-U60> andmore importantly, until I can safely lucreate on zfs, there ain't no zfs boot
[21:14:08] <sickness> g4lt-U60: I thought it already could, with updated proms, isn't it this way?
[21:14:39] <g4lt-U60> that wwould be nice if the updated prom images existed for my two cases, the SB100 and the U60
[21:14:56] <coffman> WickySolaris: well, you can run oracle at solaris x86 with support, but i think oracle does not like zfs atm
[21:15:03] <sommerfeld> sickness: no OBP upgrade required
[21:15:13] <WickySolaris> sod oracle
[21:15:16] <jbk> there's a lot of things oracle doesn't like :)
[21:15:18] <sommerfeld> sickness: (at least for booting from a pool-in-a-slice)
[21:15:25] <jbk> it wants you to use asm for everything
[21:15:42] <jbk> and is trying to pressure everyone to move to it by refusing to bless anything else new
[21:15:58] <coffman> g4lt-U60: funny thing is, there is no need foor lucreate in zfs
[21:16:10] <g4lt-U60> EXPN
[21:16:11] <reflect_> sommerfeld: "pathological performance" - what does that mean? not so good?
[21:16:36] <g4lt-U60> you have my interest sir, please follow up
[21:17:00] <coffman> g4lt-U60: what does lu-create do?
[21:17:21] <coffman> g4lt-U60: a ccopy of your root or?
[21:17:22] <quasi> coffman: creates lu boot env
[21:17:23] <sommerfeld> right.  certain patterns of I/O that cause current versions of ZFS to perform poorly
[21:17:35] <g4lt-U60> it makes a bootable image so that you can do what the linuxheads call two-kernel monte
[21:17:51] <quasi> lucreate - create a new boot environment
[21:18:00] <sommerfeld> so, with zfs boot/root, "lucreate" boils down to "zfs snapshot" + "zfs clone" plus a fixup to your boot menus
[21:18:23] <coffman> yeah, thats what i was talking about sommerfeld
[21:18:26] <g4lt-U60> it still needs to be run though
[21:18:26] <jbk> yeah, you can do it by hand if you want :)
[21:18:43] <coffman> but ithink the problem is that there is no "new boot" at sparc
[21:18:53] <jbk> yet
[21:19:07] <g4lt-U60> and is this already in SUNWzfs?
[21:19:11] <coffman> so do they realy go for grub for sparc?
[21:19:29] <quasi> eventually
[21:19:58] <g4lt-U60> so I stilll deny that zfs boot exists ;P
[21:20:06] <coffman> g4lt-U60: i dont think it is a big problem to create a new entry in bootmenu - in grub x86 that is
[21:20:09] <WickySolaris> coffman: these are some of the toys we have in the datacenter http://jake.ruivo.org/~patrick/picies/20072007/20072007-001.html
[21:20:14] <sommerfeld> coffman: no.  current plan will use boot archive but not grub
[21:20:18] <coffman> g4lt-U60: on sparc, but on x86 there is
[21:20:26] <g4lt-U60> coffman, that would be fine if I were g4lt-U40
[21:20:46] <coffman> heh
[21:20:58] <coffman> you should realy change that
[21:21:11] <sommerfeld> g4lt-U60 -= 20
[21:21:21] <quasi> WickySolaris: small and boring place?
[21:21:25] <WickySolaris> no
[21:21:33] <WickySolaris> this is just 1/100th of what we have
[21:21:36] <sommerfeld> heh
[21:21:39] <WickySolaris> we have four data centers
[21:21:40] <coffman> WickySolaris: nice stuff
[21:21:45] <WickySolaris> spread over the entire country
[21:22:05] <coffman> WickySolaris: which?
[21:22:08] <WickySolaris> KPN Telecom
[21:22:17] <g4lt-U60> speaking of which....
[21:22:22] <wesolows> g4lt-U60: U40 too fast for you? ;-)
[21:22:42] <coffman> WickySolaris: thats netherlands?
[21:22:45] <WickySolaris> yes
[21:23:02] <quasi> WickySolaris: I shouldn't be teasing - since I'm about to leave the 500+ sparc box customer for a smaller joint
[21:23:14] <g4lt-U60> wesolows, no, it has the thrice-damned X86 instruction set
[21:23:27] *** tomww has quit IRC
[21:23:29] <WickySolaris> well I am about to move to a semi 3rd world country if things go well, so :P
[21:23:58] <wesolows> g4lt-U60: yeah, I hate it too.  Sure would be nice if I could get equivalent price/performance out of SPARC hardware, wouldn't it?
[21:23:59] <quasi> .nl being one of the few places in the world where you really want to put your datacenters on 3rd floor or higher ;)
[21:23:59] <WickySolaris> and my general opinion about big companies is that there is too much politics
[21:24:06] <wesolows> Or even competitive price/performance.
[21:24:07] <WickySolaris> I used to work for ING before going to KPN
[21:24:16] <wesolows> Or even *adequate* performance, period.
[21:24:19] <WickySolaris> quasi: aham!
[21:24:22] <g4lt-U60> or even 1:1 tradeins :D
[21:24:26] <coffman> kpn is a big isp
[21:24:39] <quasi> WickySolaris: I worked in .nl last year
[21:24:48] <WickySolaris> yea it is.. lots of management overhead as well :P
[21:24:54] <WickySolaris> really? you liked it here?
[21:24:58] <quasi> coffman: telco isp - monopoly
[21:25:17] <noyb> what vnc server is the best for opensolaris?
[21:25:25] <coffman> tight
[21:25:26] <quasi> coffman: they deliver at a rate that much reminded me of poland 5 years ago ;)
[21:25:35] <g4lt-U60> I'll stack a coolthreads version of a ultra 60 against any comers if they make one and trade me for it
[21:25:36] <coffman> noyb: tight vnc should do
[21:25:40] <WickySolaris> I dont think KPN has a monopoly at the moment.... there are at least two other competitors and we have cable providers servicing more or less the same stuff
[21:25:49] *** leal has joined #opensolaris
[21:25:50] * noyb has second thoughts about googling for "tight"   :-)
[21:26:09] <coffman> noyb: try "tight vnc"
[21:26:20] <noyb> thanks
[21:26:24] <quasi> WickySolaris: they're still shite at delivering dsls - took well over 4 months for us to get ours and we were in the middle of leiden
[21:26:35] <WickySolaris> leiden rox
[21:26:42] <quasi> WickySolaris: leiden sux
[21:26:43] <WickySolaris> and yeah, I know what you mean
[21:27:01] <leal> I did read that a disk in a zfs pool can be added to a quorum device... but zfs panics when have "two" writers... how sc 3.2 uses a quorum device?
[21:27:17] <WickySolaris> I am on the IPTV project at the moment, technicaly we're ready for it, but the order and administration desk is a big mess till today
[21:27:44] <quasi> WickySolaris: I used to work for an iptv project in leiden ;)
[21:27:54] <WickySolaris> coolies :D
[21:28:00] <quasi> WickySolaris: joost
[21:28:12] <WickySolaris> joost!
[21:28:17] <WickySolaris> joost is getting big!
[21:28:24] <WickySolaris> you're talking about the p2p tv thingy, right?
[21:28:35] <quasi> yeah
[21:28:44] <WickySolaris> that's just an awesome project, I use it a lot
[21:28:52] <WickySolaris> streaming M.O.S. channel mostly
[21:29:05] <wesolows> leal: That's a really great question.  If you find out, let me know - we speculate that they must be using libzfs to open and close the pool when doing takeover, but for quorum I don't know.  PSARC/2006/533 may be a useful starting point.
[21:29:07] <sommerfeld> quasi: Yike.
[21:29:19] <coffman> quasi: lucky that you are out in time Oo
[21:29:36] * quasi might have worked for them, but they never had a version running on an os I had
[21:30:01] <quasi> coffman: yeah, 7 months in .nl was more than enough
[21:30:04] *** thezerox has joined #opensolaris
[21:30:21] <quasi> (and we dropped the solaris bit, so not as fun either ;)
[21:30:32] <WickySolaris> what's wrong with the netherlands eh?
[21:30:39] <coffman> quasi: they droped solaris?
[21:30:45] <coffman> quasi: in there backend?
[21:30:54] <sommerfeld> they seem much more adept at keeping the sea at bay than new orleans..
[21:31:09] <quasi> coffman: in the somethingorother - can;r
[21:31:12] <WickySolaris> sommerfeld: after 1954, definately
[21:31:21] <quasi> can't really say ;)
[21:31:41] <quasi> WickySolaris: it is bloody crowded and has very lousy food
[21:32:26] <WickySolaris> well, I studied for being a cook, so.. I think my food rules :P
[21:32:46] <coffman> who does spec-file extra repro?
[21:32:46] <WickySolaris> then again, if i look to the food I had in croatia and brazil, yeah, our food sucks
[21:33:07] <WickySolaris> feijoada , I miss it :s
[21:33:08] <quasi> WickySolaris: maybe - but lunch at $work being white bread with hagelslag didn't work well for me
[21:33:15] <WickySolaris> LOLOL
[21:33:28] <WickySolaris> that doesnt work for me either, that's why I go out for lunch
[21:33:46] <quasi> WickySolaris: it got a bit expensive to go out every day
[21:33:47] *** rendero has joined #opensolaris
[21:33:52] <WickySolaris> true :(
[21:33:59] <WickySolaris> so where are you now?
[21:34:22] <quasi> WickySolaris: and being very interested in cooking myself, I was pretty appaled at the selection available in supermarkets
[21:35:13] <WickySolaris> the biggest problem in this country is that a) service is non existant, b) people are way too stressed
[21:36:21] <leal> wesolows: I did read something right know that could explain that... "Quorum information is written on private cylinders and it is not visiable by Solaris OS"
[21:36:52] <WickySolaris> leal: that's correct. the keys of the SC nodes is written raw to the quorum device
[21:36:55] <quasi> c) there's like 3 times as many people in the same area as here in .dk d) people may think scots are tight with their money, but they look genrous compared to the dutch ;)
[21:37:19] <WickySolaris> quasi: I totally agree, hence why I lived in three different countries already :P
[21:37:22] <wesolows> leal: which sounds to me like it's not actually using ZFS at all, just opening the pool to find out the devices it contains and then splatting stuff onto it.
[21:37:26] *** nostoi has joined #opensolaris
[21:38:15] <WickySolaris> wesolows: it might be it is trying to open the pool in raw mode, like it'd access a SAN disk
[21:38:25] <reflect_> sommerfeld: one last question.. are you affiliated with Sun in any way?
[21:38:45] <sommerfeld> yes, i am..
[21:38:46] <WickySolaris> like, totaly ignoring the fact its a ZFS pool and see it as a real device
[21:38:51] <noyb> coffman: build problems....   too close for gcc...  switching to blastwave...     :)
[21:39:27] <coffman> noyb: dunno, i think its in spec file extra but that one is down
[21:39:34] <wesolows> WickySolaris: there's nothing I know of called opening a pool in raw mode - I guess you mean just opening the underlying devices rather than using the pool.
[21:39:43] <coffman> i was just about to add it trough pmpkg
[21:39:48] <WickySolaris> wesolows: that
[21:40:24] <WickySolaris> I am not even sure if SC 3.2 is ZFS aware, quorum wise
[21:40:26] <noyb> coffman: no prob.  I have what I need now.   I was just making a lame Top Gun analogy
[21:41:00] <leal> If it is using "private" cylinders", there  is no relation to ZFS...
[21:41:33] <WickySolaris> yea, which is what i mean, it's treating the pool name as an actual device name and treating it as a device
[21:41:49] <leal> wickySolaris: sc 3.2 is ZFS aware.
[21:42:02] <WickySolaris> yea, for file systems
[21:42:10] <WickySolaris> quorum isnt a filesystem
[21:42:56] <leal> So, i think i can use a zfs disc as a quorum device without problems... :)
[21:43:13] <WickySolaris> maybe a ZFS volume
[21:43:19] <WickySolaris> not sure about a ZFS file system
[21:43:48] <leal> I think that is a good practice, because i did read some problems with quorum devices in a unused disc... you know, somebody can catch what is unused.
[21:44:16] <leal> WickySolaris: a disc.
[21:45:16] *** Gman has joined #opensolaris
[21:45:19] <WickySolaris> leal: if you find something, let me know pls :)
[21:45:56] <WickySolaris> I will let you know when I find something of course, maybe our sun engineer at work knows something, I'll try to ask him
[21:46:05] <richlowe> afternoon Gman.
[21:46:24] <leal> WickySolaris: i will do it right know, i will change the quorum device to a disc that is part of a zfs pool.
[21:46:37] 
[21:46:57] <Gman> hey richlowe
[21:47:24] *** IvanR_ has joined #opensolaris
[21:48:42] * Gman is looking forward to a 12 hour plane flight soon
[21:48:58] <Gman> lack of internet is a nice side effect
[21:49:10] <Pietro_S> ;-)
[21:49:21] *** cypromis has joined #opensolaris
[21:49:46] <coffman> is thre any mirror for the spec-files-extra svn?
[21:50:16] <Pietro_S> hmm, when I try to add new contact in thunderbird that window which open is impossible to move to another desktop :-( , I'm using xfce, can someone try it under gnome?
[21:50:27] <Pietro_S> coffman: why?
[21:50:45] <coffman> Pietro_S: cause sf is down? the svn i mean
[21:51:36] <reflect_> I offered this yesterday, but got no takers.. if anyone would like to write a opensolaris driver for the 3ware cards, I have one ready to ship
[21:51:41] <Pietro_S> coffman: it's online - right now I tried svn up
[21:52:12] <sickness> sommerfeld: oh really? how did that get solved? I mean, a sparc bootloader was introduced?
[21:52:53] <sommerfeld> sickness: my understanding is that the sparc zfs boot folks have written a mini zfs reader in the OBP forth dialect
[21:52:55] <Pietro_S> coffman: did you run svn --relocate on that repository? sf changed path for svn access
[21:53:14] <sommerfeld> that goes into the boot block area of the slice containing the zfs pool
[21:53:29] <sickness> sommerfeld: ok, so why that doesn-t need to be put back in the obp to work?
[21:53:53] <coffman> Pietro_S: i dont have the new path
[21:54:14] <sommerfeld> sickness: OBP knows how to read certain blocks out of a slice and execute the contents
[21:54:18] *** noyb has quit IRC
[21:54:38] <sommerfeld> sickness: so the existing OBP will do that, but instead of getting a UFS filesystem reader, it will get a ZFS reader
[21:55:06] *** karrotx_ has joined #opensolaris
[21:55:10] <g4lt-U60> you see, this was worked out in a proper manner in the design phase
[21:55:39] <g4lt-U60> unlike *cough*ivebeenmauled*cough*
[21:56:01] <sommerfeld> forth UFS reader is in usr/src/psm/stand/bootblks/ufs/common/boot_obp.fth
[21:56:37] <Pietro_S> coffman: try this: svn switch --relocate https://svn.sourceforge.net/svnroot/pkgbuild/spec-files-extra/trunk https://pkgbuild.svn.sourceforge.net/svnroot/pkgbuild/spec-files-extra/trunk
[21:57:53] <Pietro_S> by the way does anyone use edge's card with opensolaris?
[21:58:09] <sickness> interesting
[22:00:07] *** Atomdrache has quit IRC
[22:01:17] <bigjohnto> ok anyone in here know C, as the people in channel #C are not very helpful... I am using solaris... and for some reason keep getting a bus error FAULT 5 when executing the program... I use truss but it doesn't give me a specific error reason... any way to debug it further?
[22:02:27] <g4lt-U60> dtreace FTW
[22:03:35] <bigjohnto> is that suppossed to be the command?
[22:03:38] <bigjohnto> solaris 8
[22:03:44] *** tomww has joined #opensolaris
[22:07:35] <Pietro_S> it's dtrace, but it was firstly in solais 10 ...
[22:08:16] <bigjohnto> so yea it won't work for me lol
[22:08:20] <alanc> bigjohnto: you'ld want to use a debugger such as dbx to see where it crashed
[22:08:34] <alanc> if you built the program with -g, it should even know what line
[22:08:59] * bubbva is back (gone 01:13:38)
[22:09:06] <alanc> most common cause of bus errors is accessing a misaligned pointer (like a pointer to a 4-byte word using an odd address)
[22:09:09] <bigjohnto> ok let me see whati can do
[22:10:01] *** rendero_ has joined #opensolaris
[22:10:03] *** rendero has quit IRC
[22:10:04] *** sparc-kly_ has joined #opensolaris
[22:10:19] *** cypromis has quit IRC
[22:10:35] <bigjohnto> well compilation does not error out... just when the program is executing... it grabs args from a conf file
[22:13:02] <tomww> dclarke: zfs problem solved?
[22:15:36] *** karrotx_ has quit IRC
[22:23:04] *** RobG has joined #opensolaris
[22:23:19] <RobG> tomww: U there?
[22:23:32] *** Atomdrache has joined #opensolaris
[22:24:48] <asyd> \_o<
[22:25:34] <tomww> RobG: yes :-)
[22:27:53] <bigjohnto> hey alanc> i used dbx it claims it is crashing in my home folder... doesn't make too much sense
[22:28:18] <alanc> what do you mean crashing in your home folder?
[22:29:08] <bigjohnto> shows 25266 Bus Error
[22:29:14] <bigjohnto> and then the path to the program
[22:29:17] <bigjohnto> which is my home folder
[22:29:31] <bigjohnto> well /home/user/path/to/program/program
[22:30:02] *** obsethryl has quit IRC
[22:30:03] <alanc> yeah, that's not the interesting part - that's just telling you what program crashed
[22:30:14] <bigjohnto> ok how do i get it to give me more info?
[22:30:19] <bigjohnto> i just typed dbx
[22:30:22] <bigjohnto> and ran the program
[22:30:27] <RobG> tomww: I follow the instructions for a pkg build yesterday but ran into some problems... to be honest I just need to get gcc 4.2 installed so was wondering if you have the pkg available and could send it to me?
[22:30:46] <alanc> exit, and restart it as "dbx program"
[22:30:54] <alanc> then type "run" at the dbx prompt
[22:31:17] <alanc> (where "program" is the name of your program"
[22:31:18] <alanc> )
[22:31:34] *** Kernel86_ has quit IRC
[22:31:45] <bigjohnto> ok i see it gives me somewhat of a better error
[22:31:51] <bigjohnto> llappend issue
[22:32:27] <alanc> you can type "where" to see the stack of calls that got you there
[22:32:40] <alanc> and "print var" to print the value of the variable named var
[22:33:14] <bigjohnto> ok perfect thanks
[22:34:01] <alanc> and of course "help" will give you lots more help on the available commands
[22:34:22] <bigjohnto> thanks i really appreciate that bit of info... it gave me quite a good start
[22:34:59] *** dclarke has left #opensolaris
[22:38:30] <bigjohnto> alanc, you ever used llappend before?
[22:38:31] *** WickySolaris has quit IRC
[22:38:34] *** sparc-kly__ has quit IRC
[22:39:29] <leal> hello there...
[22:39:58] <leal> i did a "clnode evacute node1" and after that a shutdown...
[22:40:41] <leal> when the machine was up again, my nfs share was not responding (but the services was running on node2)...
[22:41:20] *** vmlemon has quit IRC
[22:41:21] <leal> so i did a "clnode evacuate node2".. and everything was ok again. It's a known behavior or i did something wrong?
[22:43:09] *** jcea has left #opensolaris
[22:44:24] *** Kernel86 has joined #OpenSolaris
[22:45:35] *** zooko_ has joined #opensolaris
[22:47:02] <RobG> tomww: I follow the instructions for a pkg build yesterday but ran into some problems... to be honest I just need to get gcc 4.2 installed so was wondering if you have the pkg available and could just send it to me?
[22:47:14] <Pietro_S> RobG: try SFEgcc, it's 4.2 version
[22:47:54] <RobG> sorry fat fingers... didn;t mean to send again
[22:50:24] <Pietro_S> RobG: I would try that, most of time packages from SFE work well ;-)
[22:50:54] <alanc> bigjohnto: sorry, never heard of "llappend" before
[22:51:05] *** leal has quit IRC
[22:52:20] <zooko_> Dear people of #opensolaris: I am trying to port a package that I wrote, which includes a Python extension module, and I get a compatibility error.
[22:52:23] <zooko_> So when I compile this here Python extension module (that I wrote), I get: #error "Compiler or options invalid for pre-UNIX 03 X/Open applications  and pre-2001 POSIX applications".
[22:52:40] <zooko_> nvestigating, I see that __XOPEN_OR_POSIX is set, and XPG6 is not set.
[22:52:40] <zooko_> [2:41pm] zooko_: Now the Python header -- pyconfig.h -- that the Python configure script produced defines _POSIX_C_SOURCE 200112L  and  _XOPEN_SOURCE 600  and  _XOPEN_SOURCE_EXTENDED 1.
[22:52:41] <zooko_> [2:42pm] zooko_: So, where is the problem?  Is it that solaris's features_tests.h should be tweaked to recognize that code with _XOPEN_SOURCE 600 has XPG6?  Or is it that Python's pyconfig.h should #define something else in addition to _XOPEN_SOURCE 600 in order to communicate that fact?
[22:53:41] *** WickySolaris has joined #opensolaris
[22:54:36] *** zooko_ has quit IRC
[22:54:44] *** Samy has quit IRC
[22:55:03] *** zooko has joined #opensolaris
[23:00:31] <CIA-26> tm144005: 6546188 system panics in nxge:crc32c_tab+b149c98 during hotplug/dr testing with Atlas cards.
[23:00:58] <zooko> Ah, this seems to be related: http://osdir.com/ml/python.matplotlib.general/2005-10/msg00030.html
[23:03:09] *** reflect__ has joined #opensolaris
[23:03:10] <movement> zooko: I'm sadly familiar with this problem
[23:03:34] <zooko> Here's something else that looks related, and looks sadly unfixed: https://sourceforge.net/tracker/?func=detail&atid=105470&aid=1306253&group_id=5470
[23:03:44] <zooko> (That's a bug report to Python.)
[23:03:48] *** CSFrost has quit IRC
[23:04:12] *** obsethryl has joined #opensolaris
[23:04:27] <movement> zooko: are you using your own Python, or solaris's?
[23:05:01] <zooko> I built it myself.
[23:05:02] *** reflect_ has quit IRC
[23:05:20] <movement> (btw, I *really* wish python would use a decent bug tracker)
[23:05:44] <movement> zooko: my recommendation is to start from Sun's package sources. we have this stuff significantly tweaked
[23:06:00] <zooko> movement: What's the tweak?  :-)
[23:06:19] <movement> zooko: I'm somewhat forgetting the details now unfortunately. but basically python was originally asking for contradictory standards
[23:06:20] <movement> essentially
[23:06:33] <movement> on linux, they take it to mean "give me everything" so it's not a problem.
[23:06:34] <zooko> Well, I can fix that.
[23:06:41] <zooko> As soon as I understand what it is doing wrong.  :-)
[23:06:52] <movement> solaris's headers are much more picky, purportedly to help with portability problems
[23:06:55] *** nostoi has quit IRC
[23:07:18] <movement> zooko: OK. the way I worked out this stuff was by carefully reading standards(5)
[23:07:57] <movement> zooko: I hit a lot of problems related to these changes: http://blogs.sun.com/levon/entry/64-bit_python_in_nevada_build
[23:08:03] *** CSFrost has joined #opensolaris
[23:09:26] <movement> zooko: you can also look at the installed python's pyconfig.h of course
[23:11:56] <zooko> Oh, actually I'm not going to be that good at fixing this, because I really don't understand the autoconf/configure language very well.
[23:11:57] <coffman> i dont think zooko uses (open)solaris, he uses nexenta
[23:11:57] <zooko> :-(
[23:12:20] <movement> ah.
[23:12:33] <movement> in that case, I suggest perusing, and applying, our python patches
[23:12:38] <movement> then using pycc
[23:12:44] <zooko> What's pycc?
[23:12:58] *** comay has joined #opensolaris
[23:12:59] <zooko> So I would really like to contribute at least this one patch back to python upstream.
[23:13:08] <zooko> If I can figure out which one it is.
[23:13:23] <coffman> can i tell people that search for help for solaris and have nexenta installed to go away?
[23:13:24] <zooko> Let's see...  How do I get your python patches?
[23:13:26] <movement> pycc is a script we use to get stuff right fr both compilers
[23:13:47] <movement> I don't believe our patches are really ready to go upstream. but I would certainly be interesting in working with upstream
[23:14:13] <coffman> i would not do that for belenix or martux or something...
[23:14:22] <movement> zooko: http://src.opensolaris.org/source/xref/jds/spec-files/trunk/patches/
[23:14:54] <zooko> movement: thanks!
[23:15:15] <movement> Python-05-isalibs.diff is the critical one
[23:15:48] <Pietro_S> coffman: people are comming here, because here are gurus willing to help ;-)
[23:16:33] <zooko> movement: thanks, again!
[23:16:39] <coffman> Pietro_S: sure, they should say what they are using , he did that an get kicked in #solaris, but they are dicks anyways
[23:16:55] <coffman> zooko: next time tell the people that you use nexenta
[23:16:56] <movement> zooko: actually that doesn't seem right. hold on
[23:16:59] <zooko> They were, indeed, pretty rude.
[23:17:13] <zooko> coffman: in this case it was irrelevant.
[23:17:23] <coffman> i dont think so
[23:18:13] <coffman> ill vote for a channel faq in the topic
[23:18:37] <coffman> i would even write it if some one gives hosting for it
[23:18:40] <movement> zooko: OK look at http://src.opensolaris.org/source/xref/jds/spec-files/trunk/SUNWPython.spec
[23:19:02] <zooko> Cool!  Thanks a lot for the help, movement.
[23:19:03] <movement> line 121. it's a hack rather than a good solution.
[23:19:10] <movement> email me if you follow up on this.
[23:22:23] <zooko> Wow.  advogato is really alive again?
[23:22:38] <zooko> </offtopic>
[23:22:48] <reflect__> zooko: were you actually using nexenta, or did you just try any channel that was available to solve your problem?
[23:23:46] *** derchris has quit IRC
[23:23:53] *** derchris has joined #opensolaris
[23:25:01] <coffman> reflect__: 22:42  zooko_> I just installed nexenta a few days ago on a new server. - thats from #solaris
[23:25:38] <coffman> not to be rude, isnt there a nexenta channel?
[23:25:45] <reflect__> there is
[23:26:11] <reflect__> coffman: but python is fairly OS-agnostic
[23:26:33] <movement> zooko: yup.
[23:26:44] <reflect__> your statements made me blink.
[23:26:54] <coffman> reflect__: the build of it etc is
[23:26:58] <movement> coffman: but this has very little to do with nexenta, so how about you stop giving him a hard time?
[23:27:32] <zooko> For what it is worth, I am to some degree trying to help solve other people's problems in addition to getting help solving my problems.
[23:27:51] <Pietro_S> isn't nexenta packages build by gcc instead of studio?
[23:28:07] <coffman> thats my point
[23:28:14] <zooko> In particular there are plenty of ways that I get could my module to compile on my box, and there are even plenty of ways that I could make it compile on anybody else's solaris system, but I'm trying to (a) learn about the sun build/configure system and its ideas about xpg
[23:28:19] <reflect__> coffman: well you made it sound like "if people are using opensolaris, or belenix, or martux" then they would get help
[23:28:23] <reflect__> but not nexenta
[23:28:38] <zooko> (b) maybe make it so that the upstream python sources compile more easily even with c99 extension modules on solaris, etc.
[23:29:11] <zooko> Pietro_S: I compiled the python interpreter using gcc instead of studio.
[23:29:47] <coffman> reflect__: not being a nazi, but t people tend to come in and dont say where they , so it makes it kind of hard since nexenta differs much to solaris/sxce/belenix etc
[23:30:13] <zooko> End of thread?  ;-)
[23:30:14] <Pietro_S> jds is compiled by studio, so take care about importing patches from them, because they should be done for studio
[23:30:42] <reflect__> coffman: I agree with you there, people SHOULD state what they are using. however, they way you stated it almost seemed like you singled out nexenta and wanted nothing to do with it.
[23:31:40] <reflect__> and I just hope that isn't this channels official position, cause if it is, I will leave this instant.
[23:31:49] <coffman> reflect__: yeah, sometimes im a bit to rude, amd yes i single it out (for me) - many people try to escape the linuxisch stuff
[23:32:11] <palowoda> It will be interesting when Indiana/OpenSolaris distribution encounters these issues.
[23:32:15] *** gm152 has joined #opensolaris
[23:32:39] <zooko> movement: something I haven't quite figured out yet, is why the feature_tests.h don't think that XPG6 is true, even though
[23:32:55] <zooko> _XOPEN_SOURCE is 600
[23:33:07] <zooko> This message from matplotlib mailing list: http://osdir.com/ml/python.matplotlib.general/2005-10/msg00030.html
[23:33:21] <zooko> suggests that this is because gcc isn't c99 compliant.
[23:33:25] <Abe_Froman> wow.  there's a sun office right off the orange line
[23:33:53] <zooko> If that's true, then this means ... Um.. somewhere in the sun header files it is checking if a gcc define is present and if so undefining XPG6 ??
[23:34:04] <zooko> Let's see...  How can I go about answering this question for myself...
[23:34:09] <wesolows> gcc -std=c99 or -std=gnu99 should be sufficient
[23:34:10] <coffman> palowoda: while you dont want 300 distros, it may produce other distros/forks
[23:34:40] <coffman> which is not that bad
[23:34:43] <wesolows> this was 2 years ago, so there's hope
[23:34:57] <wesolows> and I remember we fixed something to do with wchar_t in our headers in the build 14 timeframe
[23:35:41] <wesolows> the message also implies that SOS10 worked, so maybe Studio 11 does too.  Another option, since it can now be had for $0.
[23:36:16] <coffman> reflect__: and yes, i would like if we could split the discusions into specific channels to lower the traffic
[23:37:11] <reflect__> well, is that the official stance of this channel?
[23:37:17] <stevel> reflect__: no
[23:37:19] <richlowe> we have stances?
[23:37:28] <reflect__> channels always have rules.
[23:37:30] * stevel assumes a judo stance
[23:37:35] *** obsethryl has quit IRC
[23:37:38] <richlowe> at this point, I'd go for the captain morgan thing.
[23:37:52] <stevel> nexenta is an OpenSolaris distribution, and as far as i'm concerned, is welcome in #opensolaris
[23:37:58] <reflect__> stevel: thanks for that clarification
[23:38:03] <stevel> speaking for myself, not for the channel or anyone else
[23:38:18] <reflect__> hehe
[23:38:30] <reflect__> so who can speak for the channel?
[23:38:31] <zooko> I was chatting with Jonathan Swartz the other day, and he mentioned "Oh by the way, please feel free to chat about Nexenta in #opensolaris on freenode.".
[23:38:33] <sommerfeld> the channel doesn't have a stance
[23:38:34] <zooko> Just kidding.
[23:38:44] *** obsethryl has joined #opensolaris
[23:38:44] <zooko> Anyway, what I want to know is how XPG6 gets defined.
[23:38:51] <zooko> I think maybe a recursive grep would be in order...
[23:39:29] <coffman> zooko: im not bad with that, as long as you say what you use
[23:40:02] <zooko> Okay.  I'll try to avoid confusing people by using unexpected tools and not mentioning that I'm using them.
[23:40:03] <stevel> reflect__: nobody speaks for the channel
[23:40:04] <coffman> kind of anoying if people try to help and it does not work cause the target is on a different system
[23:40:20] <reflect__> stevel: and you have no rules, either?
[23:40:30] <stevel> reflect__: not as far as i'm aware of
[23:40:35] <e^ipi> be respectful is the only one I know of
[23:40:36] <reflect__> I love it
[23:41:23] *** RobG is now known as RobG_AFK
[23:41:24] <Pietro_S> reflect__: did you ever saw anyone got kicked from here? I don't remember it
[23:41:25] <reflect__> if I may, then..   anything using opensolaris is welcome, and it is appreciated if you would state which distribution and/or build you're currently using is stated along with your problem?
[23:41:34] *** m0le has joined #opensolaris
[23:41:52] <reflect__> Pietro_S: kicked..  I don't much care about that.. I do care about the tone of voice people use.
[23:41:54] <bda> When the channel starts speaking for itself, it's time to lay off the Absinthe.
[23:41:55] <stevel> pietro_s: i believe i kicked richlowe yesterday
[23:42:35] <stevel> (15:23:36) richlowe left the room (Kicked by stevel (punk)).
[23:42:36] <stevel> yup
[23:43:00] <CSFrost> Are we having flame wars over *distros* (I know how you guys love that word) ?
[23:43:05] <richlowe> I don't think that was yesterday.
[23:43:11] <coffman> CSFrost: nah
[23:43:16] <stevel> sorry, 3 days ago
[23:43:26] <richlowe> Oh, right, I know why that happened.
[23:43:27] <richlowe> ok.
[23:43:28] <reflect__> CSFrost: no, it's about what is and what isn't allowed.
[23:43:35] <Pietro_S> oh, I know why I didn't noticed it - I use ignore for pats, joins, and kicks and other crap
[23:43:39] <stevel> i forget why
[23:43:41] <CSFrost> I need a little help installing vista on an E15k, anyone around for a challenge?
[23:43:49] <zooko> Thanks for the help, folks.  See you around.
[23:44:02] <wesolows> I've kicked people, but always for spamming.
[23:44:13] <reflect__> good luck with your stuff, zooko
[23:44:16] <wesolows> And never without warning them first.
[23:44:21] <Pietro_S> by the way why gnome-chess were removed from sxce?
[23:44:30] <reflect__> sounds like a good policy to me, wesolows
[23:44:44] <CSFrost> cause the ogre's always squashed the gnomes
[23:44:46] <coffman> stevel: i would like to make a faq that could be included in the channel header
[23:45:06] <coffman> CSFrost: you might could try that with bochs
[23:45:11] <stevel> do we have frequently asked questions?
[23:45:14] <Pietro_S> coffman: you can add it to wiki
[23:45:23] <CSFrost> I would like my mail to be categorized by color of the stamp, neatly placed in my mailbox..
[23:45:28] <richlowe> stevel: only, really, "When will the SXCE based on build $N be available"
[23:45:28] <coffman> stevel: yes
[23:45:38] <richlowe> stevel: and that has a non-constant answer.
[23:45:41] <reflect__> hehe
[23:45:46] <coffman> richlowe: thats mine
[23:45:55] <alanc> richlowe: don't forget "Where can I download OpenSolaris?"
[23:46:09] <reflect__> not sure if it's funny or scary that the ops knows of no rules, nor faqs.. ;)
[23:46:11] *** rendero_ has left #opensolaris
[23:46:12] <wesolows> and "Can someone help me install OpenSolaris?" A: No.
[23:46:32] <alanc> who needs rules?
[23:46:33] <richlowe> 'bfu nightly-nd'
[23:46:35] <wesolows> reflect__: Ops are just here to maintain basic order, kick spammers and runaway scripts.
[23:46:40] <coffman> automounter <- such things
[23:46:58] <Pietro_S> another often question "is anyone here?" A:"nope"
[23:47:05] <Gman> reflect__, sounds good, you should totally write it, and if it's useful, i'm sure it'll get added
[23:47:06] <CSFrost> "My ipod's filesystem doesn't work with OpenSolaris, is there anyone that can help?"
[23:47:12] <m0le> <---n00b, so sorry for le spam ;)
[23:47:31] <wesolows> reflect__: Otherwise there aren't really any rules.  We seem to prefer /ignore to /kick for dealing with people we don't like.
[23:47:48] <coffman> CSFrost: hey, my mp3 play still not works :(
[23:47:51] <reflect__> well, as long as it works for you, all the more power, I guess
[23:48:00] <stevel> it seems to work
[23:48:07] <CSFrost> telekinesis isn't working on my box yet either
[23:48:13] *** ichigo has joined #opensolaris
[23:48:14] <coffman> i like modless channels
[23:48:28] <e^ipi> CSFrost: in fairness I said "my ipod's filesystem doesn't work, this annoys me so I'll port it from darwin when i've got some free time"
[23:48:43] <reflect__> coffman: in the case of you and me..  it worked out well
[23:49:03] <reflect__> but it only works as long as the parties are able to come to some sort of agreement
[23:49:32] <coffman> you can easiely tell some to leave and he will
[23:49:47] <reflect__> if you would have taken the stance of "fuck you, nexenta sucks and should not be allowed to come here" then what could I do? :)
[23:49:49] <CSFrost> e^ipi, ipozfs? :-P
[23:49:55] <e^ipi> HFS+
[23:50:43] <coffman> did i said that pcfs sucks? :P
[23:51:09] <coffman> still searching some java thing that does it
[23:53:14] <CSFrost> we should be fearful of demons
[23:53:52] <CSFrost> appologies, I am just going through a 10 foot stack of mail and causing trouble :-P
[23:54:02] *** nrubsig has joined #opensolaris
[23:54:03] *** ChanServ sets mode: +o nrubsig
[23:54:23] <coffman> CSFrost: kaiwai isnt around :P
[23:54:38] <coffman> but there is nrubsig :)
[23:55:00] <oxygene> reflect__: "/ignore" helps in such cases
[23:55:03] <nrubsig> coffman: ?!
[23:55:04] <coffman> (speaking of demons)
[23:55:21] <nrubsig> coffman: I am not a daemon
[23:55:25] <nrubsig> er
[23:55:27] <nrubsig> demon
[23:55:41] <oxygene> nrubsig: aren't you the komodo dragon gate keeper?
[23:55:47] <nrubsig> I am sick (headaches, sore throat), half dead and tired.
[23:55:57] *** dlg has quit IRC
[23:56:07] <nrubsig> oxygene: not gate keeper., just keeper. And this week someone else is doing that job.
[23:56:13] * coffman hands nrubsig some painkillers
[23:58:45] 
[23:59:20] <ichigo> nrubsig: 30C here and it is midnight
[23:59:30] <coffman> ugh
[23:59:46] <coffman> its like 16C now here
[23:59:47] <ichigo> ~40C in the afternoon. hardly can breathe the air





top