[00:00:27] <CIA-16> sommerfe: 6584986 *nightly* needs a hook for shared build machines
[00:00:28] <CIA-16> egillett: PSARC/2007/453 MSI-X interrupt limit override, 6592959 MSI-X interrupt limit override
[00:06:00] *** sommerfeld has quit IRC
[00:06:29] <elektronkind> man this mpt console boot spew is annoying
[00:07:18] <wesolows> yep
[00:07:27] <wesolows> wait until you get Winchester
[00:07:37] <wesolows> NOTICE: idmap: Registered door_handle_t ffffff033c55c5b0
[00:07:47] <wesolows> because I really needed to know that address
[00:08:04] <richlowe> I was annoyed enough by the coherent console 'os-io' bit.
[00:15:09] *** Fish has quit IRC
[00:23:20] *** g4lt-superbat is now known as g4lt-mordant
[00:23:35] *** hajma has joined #opensolaris
[00:25:35] *** alobbs has quit IRC
[00:27:09] *** yippi has joined #opensolaris
[00:28:25] <jbk> hmm.. now to find a mb that works well w/ nevada
[00:28:50] *** scarab1 has quit IRC
[00:29:35] <wesolows> the ultra 40s seem to work fine
[00:31:36] <jbk> a bit more than i want to spend :)
[00:31:39] *** pfa3rh has quit IRC
[00:31:46] <wesolows> ultra 20 it is then
[00:31:59] <jbk> or that, though the ultra 20 i had at my old job was nice
[00:32:15] <wesolows> of course, tbh although I haven't tried I have yet to find a board that won't boot nevada
[00:32:31] <wesolows> I'm sure there's such a thing but even the shitty asus boards I have work just fine
[00:32:49] <jbk> well not just boot, i'd like onboard sata + networking to work
[00:33:13] <wesolows> those seem to work ok for me
[00:33:25] <wesolows> even though the nic is (blecch) nge
[00:33:44] <wesolows> even sata + dvd writer, amazingly enough
[00:38:10] <gcj> is there a SUNW package for g++?
[00:38:21] <richlowe> SUNWgcc
[00:38:50] <jbk> but i need to get a desktop -- both as a fileserver, and as something i can use while i send in my laptop to be fixed
[00:40:37] <g4lt-mordant> SUNWfuckmehardergnu
[00:40:46] <gcj> hehe
[00:41:01] <gcj> better than SUNWfuckmeharderformore$$$pleasesun
[00:41:14] <g4lt-mordant> richlowe, I thought it was SFWgcc?
[00:41:24] <richlowe> I think this is where flyingparchment comes and explains everything that's wrong with with Studio's CC
[00:41:37] <flyingparchment> ooh someone mentioned CC?
[00:41:57] <e^ipi> gcj: studio is free
[00:42:00] <e^ipi> *shrug*
[00:42:03] <jbk> though then i'd also need a char to encourage me to actually use it -- i like to lean back, but my height makes that a bit unconfortable with most chairs
[00:42:25] <gcj> e^ipi, oh that changed? i thought sun used to charge $$$ for CC
[00:42:29] <masta> sbahra: yes I'm here these days.
[00:42:34] <g4lt-mordant> richlowe, right, how dare CC actually not overflow the stack when compiling {switch foo case 1: case 2: case3: case 4: case 65536:}
[00:42:35] <e^ipi> it's been this way for years
[00:42:37] <flyingparchment> gcj: until like two years ago
[00:42:40] <wesolows> The whole cc vs gcc thing doesn't need to happen if it's just an ideological argument; you have the choice and should use whatever you prefer
[00:42:53] *** hile__ has joined #opensolaris
[00:43:00] <gcj> ah ok, i'm new to solaris :-)
[00:43:15] <jbk> just try to avoid gcc-isms :)
[00:43:16] 
[00:43:46] <gcj> is there a SUNW package for studio which includes a C++ compiler?
[00:43:48] <flyingparchment> i like to point on the poor C++ language support in studio
[00:43:53] <g4lt-mordant> wesolows, what if my pont about gcc being broken by design because of a "make it more like microsoft cc" hack is correct, is it still ideological?
[00:44:26] <e^ipi> flyingparchment: which makes it even more a shame that GNU decided to ignore the solaris C++ abi
[00:44:29] <gcj> g4lt-mordant, i guess that depends whether brokenness is ideological or practical in this case :)
[00:44:30] <wesolows> g4lt-mordant: I must have missed that
[00:44:55] <jbk> does it actually do something incorrectly, or that encourages bad practices?
[00:45:01] <wesolows> though bug compatibility with an inferior and nonstandard compiler hardly seems like BAD; it may be a misfeature but BAD is going a bit far
[00:45:25] <flyingparchment> e^ipi: gcc chose to use the Intel ABI standard; i understand studio is looking at adding that as an option
[00:45:55] <gcj> some way to list files in a package?
[00:45:58] <wesolows> ahh, the endless C++ ABI war
[00:46:01] <flyingparchment> sadly, one of the main studio options to fix its language support (stlport4) also breaks the abi
[00:46:12] <g4lt-mordant> wesolows, with gcc4, [honestly, I don't know if it got fixed, nver treid after I proved it could be done] the compiler made a switch/case of [1,2,3,4,65536] the exact same way it made a case/switch of [1,2,3,4..65536]
[00:46:14] <wesolows> you'd think that by now people would realise that this is a fundamental reason C++ should never be used
[00:46:32] <wesolows> g4lt-mordant: well, if that's true, it would be a bug.
[00:46:48] <gcj> wesolows, doesn't seem that different to a platform war to me :-)
[00:46:49] <g4lt-mordant> and in the process, it overflowed the variable table and thus overflowed the stack during the compile
[00:46:54] <wesolows> of course, gcc4 isn't part of OpenSolaris
[00:47:08] <richlowe> neither is studio. :)
[00:47:16] <wesolows> very true
[00:47:19] <masta> sorta, in express
[00:47:22] <jbk> it would be neat to see sometihng that was C-ish (retained the abi), but supported classes (single interitance) + some sort of reasonable exception-like mechanism
[00:47:30] <wesolows> the OpenSolaris compiler is gcc 3.4.3 + fixes which are of course available to all in source form
[00:47:31] <e^ipi> wesolows: lack of a well-defined standard isn't  a reason a language shouldn't be used, it's a reason that standards should be defined
[00:47:46] <richlowe> they wouldn't be adhered to.
[00:47:53] <flyingparchment> they were defined and are adhered to
[00:48:02] <richlowe> like all after-the-fact standardization, it would be come an ill-specified clusterfuck of badly defined common practice.
[00:48:03] <wesolows> e^ipi: I agree.  Since C++ has been out there for over a decade and that hasn't happened, I can only conclude that the C++ community is full of morons.
[00:48:05] <richlowe> and benefit nobody.
[00:48:34] <jbk> c++ is needlessly complicated
[00:48:49] <richlowe> most likely leaving enough leeway that any existing offering with a voice of the committee met the standard without change.
[00:48:50] <gcj> anyone know a way to list files in a package?
[00:49:00] <jbk> pkgchk
[00:49:21] <gcj> thanks jbk
[00:49:52] <g4lt-mordant> wesolows, not so fast.  C never really got standardized until the 90s.  K&R 1E has a copyright of 1978
[00:50:13] <elektronkind> C89
[00:50:18] <elektronkind> and thank god for it
[00:50:50] <wesolows> g4lt-mordant: the language standardised slowly, yes, mainly in the late 80s.  C++ has gotten there too.  But the _architecture_ ABIs were around well before that.
[00:50:53] <g4lt-mordant> elektronkind ...which was 11 years after, and still wasn't inclusive.  borland wasn't fully C89 complaint IIRC
[00:51:09] <gcj> is sun studio shipped with opensolaris developer? and if so which package should i install for a C++ compiler?
[00:51:33] <g4lt-mordant> yes, and you don't have to if you do a full+OEM install of SXDE
[00:51:36] <elektronkind> gcj: yes, and it should be installed for you. SUNWspro is the package name
[00:51:46] <masta> gcj: with the express you cannot choose to not have gcc or studio
[00:51:55] <masta> you just have them
[00:52:24] <gcj> strange, doesn't seem to be installed for me :-)
[00:52:25] <e^ipi> gcj: /opt/SUNWspro/bin/CC or /usr/sfw/bin/g++
[00:52:28] <e^ipi> take your pick
[00:52:31] <gcj> pkgchk shows nothing
[00:52:47] <g4lt-mordant> masta, please, be precise.  with the SXCE defintion of "the express" you can have neither
[00:52:49] <masta> lemme go fire up my laptop and see the path
[00:53:04] <gcj> no SUNWspro package on the installer DVD either
[00:53:22] <e^ipi> fwiw, in case anyone's interested, the lions share of KDE4 compiles with studio now
[00:53:23] <masta> g4lt-mordant: no, sorry... no idea what you're talking about
[00:53:29] <g4lt-mordant> gcj do you have SXCE or SXDE?  what build do you have?
[00:53:41] <e^ipi> thanks to a couple kde/solaris types
[00:53:46] <gcj> i have sol-nv-b64a-x86-dvd.iso
[00:54:19] <g4lt-mordant> is b64 thee last SXDE I have is b57, can anyone confirm b64a is SXDE?
[00:54:28] <richlowe> 57 wasn't SXDE
[00:54:38] <masta> my developer version is 64
[00:54:43] <richlowe> 55a, 64b
[00:54:44] <gcj> it's called "Solaris Express Developer Edition 5/07 Developers Release" on their website
[00:54:54] <masta> right
[00:55:15] <masta> there is no option to choose packages on there... you get the full suite of tool-chain
[00:55:19] <gcj> i guess it doesn't come with studio?
[00:55:24] <gcj> cos i can't find it
[00:55:27] <masta> it does
[00:55:28] <richlowe> but b64a never existed in an appropriate context.
[00:55:29] <wesolows> how are you looking for it?
[00:55:32] <richlowe> contact the DE people, ask them to help you.
[00:55:39] <g4lt-mordant> richlowe, want to bet?  the set I got from get.opensolaris.org as soon as they opened up was snv_57
[00:55:40] <gcj> no support contract :-)
[00:55:54] <gcj> there's no /opt/SUNWspro on this box
[00:55:56] <wesolows> Studio is an unbundled product; it's not part of OpenSolaris nor of Solaris, although Sun has, in its wisdom, included it on the SXDE DVDs.
[00:55:59] <richlowe> g4lt-mordant: the fact there is a boot option that says "Developer Edition", doesn't make it a developer edition release.
[00:56:18] <richlowe> I cannot be held responsible for the people who put Solaris media together being loons.
[00:56:23] <richlowe> they can, however.
[00:56:27] <gcj> lol
[00:56:30] <richlowe> so if you find this confusing, non-productive, and stupid.
[00:56:31] <richlowe> tell them.
[00:56:34] <gcj> maybe that's what got me
[00:56:35] <richlowe> it won't help, but it'll make me feel better.
[00:56:38] <gcj> and me
[00:56:42] <g4lt-mordant> I thought g.o.o only shipped "official" releases of SXDE
[00:57:18] <gcj> do i have to reinstall then? :(
[00:57:25] <gcj> and download a new dvd?
[00:58:00] <g4lt-mordant> gcj no, you can get sunstudio 11 off the opensolaris.org website, or you can get [the still unsupported to build opensolaris] sunstudio 12 off sun.com
[00:58:07] <alanc> g4lt-mordant: get.opensolaris.org is for getting OpenSolaris releases, not SXDE releases - there was nother site to get SXDE official releases
[00:58:08] *** hile_ has quit IRC
[00:58:20] <gcj> ok thanks
[00:58:27] <g4lt-mordant> alanc, we might want to put that in /topic....
[00:58:32] <masta> ok yea
[00:58:37] <masta> sun studio is in the menu
[00:58:56] <alanc> why?  topic doesn't say to go there for SXDE, says to go there for OpenSolaris starter kits
[00:59:23] <masta>  /usr/bin/sunstudio
[00:59:30] <wesolows> wtf is /usr/bin/sunstudio?
[00:59:36] <richlowe> the crack addled GUI
[00:59:37] <wesolows> and why is it delivered to /usr?!
[00:59:42] <richlowe> they install symlinks in /usr/bin now.
[00:59:51] <masta> that is my path on my developer express of 64 nevada
[00:59:51] <richlowe> at least, they were planning to.
[00:59:54] <richlowe> I take this as indication they did.
[00:59:58] <richlowe> apparently LSARC have no standards.
[01:00:09] <g4lt-mordant> s/standards/taste/
[01:00:10] <wesolows> "all the benefits of being in the WOS with none of the responsibilities!" Yay for "growth targets"
[01:00:13] <richlowe> except for alanc, obviously. ;)
[01:00:14] <alanc> we have standards, they're just low
[01:00:21] <richlowe> Oh, including alanc then.
[01:00:53] *** pablomh has joined #OpenSolaris
[01:01:03] <alanc> it was just a way to get /usr/bin/cc back so people could stop asking where the compiler was
[01:01:11] <gcj> i can install studio on opensolaris 5.11, right?
[01:01:11] <wesolows> I keep hoping that someone will take OpenSolaris and make a really cracking distribution out of it.  Sun desperately needs some competition to remind it of its standards and give it some motivation.
[01:01:20] *** linux_user400354 has quit IRC
[01:01:26] <richlowe> wesolows: I was going to say, it sure as hell won't be Sun.
[01:02:20] <jbk> well perhaps indiana can help with that :)
[01:02:30] <richlowe> see my prior sentence.
[01:02:43] <wesolows> indiana is another Sun product, a sibling of Solaris
[01:02:44] * g4lt-mordant puts the final stake into the heart of man standards
[01:02:58] <richlowe> g4lt-mordant: Sorry, DevPro beat you to it.
[01:03:12] <richlowe> but I think you mean filesystem(5)
[01:03:21] <wesolows> made by some of the very people who moan about "stop energy" and seem dedicated to a technical race to the bottom with the GNU/Linux camp.
[01:03:32] <g4lt-mordant> richlowe, man standards also has paths
[01:03:46] <jbk> well one of the problems with /opt is that then people have to know to modify their path (or have some way of doing that as part of installation)
[01:03:55] <wesolows> But we shall see.  It's always possible that some good will come of it.
[01:04:21] <gcj> jbk, many linux systems have /etc/profile.d, seems like a good idea to me :-)
[01:04:22] 
[01:04:48] <richlowe> you're the kind of person that would say candyman 3 times to see what happened, aren't you?
[01:04:49] <jbk> gcj: i've also heard of issues with such an approach
[01:04:54] <jbk> though it is better than nothing
[01:04:57] <masta> sun should automaticly take care of the path like how zfs autmaticly takes care of mounting... cuz we users too stupid
[01:05:00] <jbk> of course on the flipside
[01:05:07] <jbk> HP-UX uses /etc/PATH and /etc/MANPATH i believe
[01:05:27] <jbk> and i've seen HP-UX boxes whose $MANPATH and $PATH literally take up 3/4 of a 80x24 window
[01:05:39] <g4lt-mordant> jbk uhm, those are basiucally shellscripts that set $PATH and $MANPATH respectively
[01:05:39] <jbk> because of all the stuff in /opt
[01:05:58] * wesolows has had ~/.pathrc since 1994 at least
[01:06:02] <jbk> actually i think it's not a script as much as just 'this is the default value'
[01:06:12] <g4lt-mordant> hp-sux and the "binaries in /etc" :(
[01:06:17] <gcj> jbk, what are the issues?
[01:06:24] <wesolows> g4lt-mordant: we, um, have that too
[01:06:55] <jbk> well either with large number of things in /etc/profile.d potential conflicts, excessive number of files slows down login, or messed up scripts in there break everything
[01:06:58] <masta> yea, /etc is not a registry... not pure config data. is that good or bad?
[01:06:59] <jbk> not unsurmountable
[01:07:02] <g4lt-mordant> wesolows, yes, and I never liked it.  I know they're there for hysteerical raisins, but HP seemed to take a perverse pride in loading /etc with binaries
[01:07:05] <richlowe> wesolows: symlinks thereto, mostly.
[01:07:16] <wesolows> yes, yes.  Still ugly.
[01:07:21] <jbk> well HP-UX is somewhat perverse itself :)
[01:07:22] <richlowe> wesolows: though if eschrock intends to properly clean that up, he's up for lots of fun with wbem.
[01:07:37] <jbk> though at least it still feels like unix
[01:07:41] <alanc> isn't wbem dead yet?
[01:07:44] <jbk> unlike some variants *cough*aix*cough*
[01:07:45] <richlowe> Sadly, no
[01:07:53] <wesolows> He hasn't expressed any intent to do that, but he works us all under the table and could probably do it faster than tell me about it
[01:08:06] <wesolows> alanc: such stuff as dreams are made of
[01:08:51] <jbk> heh doesn't smc use wbem?
[01:09:04] <wesolows> probably
[01:09:08] <wesolows> your point? :-)
[01:09:09] <jbk> i remember trying it, and i thought smoke was gonna start coming off the hd
[01:09:30] <jbk> though i think that could also have been due to jvm stupidity
[01:09:36] <jbk> *cough*single byte reads*cough*
[01:09:44] <wesolows> the stupidity is the use of a jvm
[01:09:58] <wesolows> for something that administers the system, whose interfaces are all C
[01:10:20] <richlowe> if wbem stuck to things that were interfaces, I'd hate it far less.
[01:10:35] <wesolows> and if it used SMA instead of agents
[01:10:45] <wesolows> just say no to usr/*/cmd/agents
[01:11:17] <jbk> well for gui stuff, i'd argue that recent versions of java are a lot easier to write useful guis (though vala looks very interesting)
[01:11:43] <jbk> so with jni, i could see that as being somewhat reasonable
[01:12:11] <jbk> but older jvms were braindead on unix
[01:12:23] <jbk> i actually had to download the source to believe it
[01:12:48] <jbk> but sure enough
[01:13:05] <jbk> read(fd, buf, 1);
[01:13:08] <jbk> for all file i/o
[01:13:47] <masta> wtf
[01:14:02] <masta> so jvm is all streams?
[01:14:14] <jbk> ?
[01:14:35] <masta> perhaps that is to support eh wides range of environments
[01:14:38] <jbk> this was the jni implementation for any file reader/writer classes
[01:14:41] <jbk> on older jvms
[01:14:43] <jbk> they did fix that
[01:15:05] *** SplasPood has quit IRC
[01:15:08] <jbk> the File class wouldn't do that (which I used to get around that)
[01:15:18] *** lisppaste3 has quit IRC
[01:15:50] *** SplasPood has joined #OpenSolaris
[01:16:37] <jbk> i had a utility that parsed showrev output, against a patchdiag.xref file, and would produce a excel spreadsheet that would show each patch, and if they were downrev (and highlight it)
[01:16:56] <jbk> and wrote my own buffered i/o classes based on the File class
[01:16:59] <jbk> so it would actually perform well
[01:17:37] *** Lusitanian has joined #opensolaris
[01:17:54] <masta> I'm skeptical, unless your filesystem was unbuffered (dirrect io)
[01:18:05] <jbk> about what?
[01:18:12] <masta> unless your buffering means something diff like a cache
[01:18:20] *** stevel has quit IRC
[01:18:26] <jbk> no
[01:18:32] *** yippi has quit IRC
[01:18:33] <jbk> in java you had 'buffered' io which was buffered in java
[01:18:48] <jbk> but the older jvms still issues a syscall for every byte of i/o done
[01:19:18] <jbk> however, the java File class was basically a thin wrapper around read(2)/write(2)
[01:19:34] <masta> disgusting
[01:19:38] <jbk> which allowed me to actually read/write data in larger blocks
[01:19:44] <jbk> like i said, they did fix that (thank god)
[01:19:49] <masta> but I can see why to support window
[01:20:08] <jbk> but someone should have been smacked for letting that go in in the first place
[01:21:11] *** lisppaste3 has joined #opensolaris
[01:31:01] <jbk> i really don't see anything wrong with java or jvms -- good code and bad code can be written in any language
[01:31:25] *** jpdrawneek has joined #opensolaris
[01:34:16] <LeftWing> jbk: Indeed.
[01:34:18] *** slowhog has left #opensolaris
[01:38:52] <bda> http://mediatemple.org/webhosting/nitro/ # Dedicated Physical Virtual Server
[01:38:55] <bda> ...what?
[01:40:07] <trygvis> sounds like they're only running a single, virtual machine
[01:40:13] <alanc> from the people who brought you 100% Genuine Imitation Leather?
[01:40:14] <trygvis> (to ease migration)
[01:40:28] <bda> trygvis: Yeah, just... a weird gimmick.
[01:40:47] <trygvis> yeah
[01:43:05] <alanc> I don't know if I should find it sad or hilarious that the official S10U4 release docs on docs.sun.com still point to my personal webpage on freedesktop.org as the docs for the Xserver DTrace provider
[01:43:18] <jbk> haha
[01:43:19] <bda> ha.
[01:44:37] <alanc> a doc writer came to talk to me once about putting them in the dtrace guide on docs.sun.com, but never followed through...
[01:46:29] *** NikolaVeber has quit IRC
[01:46:31] <alanc> and I'll just chalk it up as bizarre that you get to the S10 8/07 what's new by selecting the "Solaris 10 1/06 Release" and then looking for it's what's new guide
[01:52:05] *** Emmedics4 has quit IRC
[01:52:06] *** mmmurf has joined #opensolaris
[01:52:16] <mmmurf> hello, i'm trying to compile asterisk on opensolaris and i'm a bit lost... this is probably an obvious issue, but would someone mind taking a look at the compiler error:  http://pastie.caboo.se/92900
[01:52:48] <jbk> what is crvs?
[01:53:18] <mmmurf> jbk: no clue
[01:53:37] <jbk> i think that's the issue :)
[01:53:46] <mmmurf> :)
[01:53:47] <richlowe> my guess would be "the arguments to ar"
[01:53:53] <mmmurf> ahh
[01:54:05] <richlowe> I'd bet there's an $AR that's expanding to nothing.
[01:54:12] <mmmurf> what package is that from, do you know?
[01:54:22] <jbk> what release are you compiling on?
[01:54:23] <richlowe> SUNWbtool
[01:54:35] <richlowe> it'll be in /usr/ccs/bin (or /usr/bin, on recent builds)
[01:54:36] <jbk> well rather
[01:54:37] <mmmurf> 5.11 snv_67 i86pc i386 i86pc
[01:54:38] <jbk> yeah
[01:54:53] <jbk> might need to add that to your path and rerun ./configure or such
[01:55:27] <mmmurf> jbk: it's in /usr/ccs/bin on my system... i'll try that... thanks!
[01:56:48] <mmmurf> here's an offhand question...  in linux I can type grep -rn pattern * to search the current directory recursively for matching files... the grep installed on opensolaris doesn't like the r option... do you know if there is a different grep that might be installed?
[01:57:12] <jbk> hmm is there? /usr/gnu/bin/grep
[01:57:21] <alanc> or /usr/sfw/bin/ggrep on older builds
[01:57:34] <jbk> ahh yeah
[01:57:40] <alanc> since most linux'es just use ggrep
[01:57:41] *** tg has quit IRC
[01:58:30] <mmmurf> ahah! you are right... ggrep it is..
[01:59:51] <mmmurf> thanks very much jbk and alanc
[02:00:16] <CIA-16> willard: PSARC/2007/277 tsalarm library, 6553178 The tsalarm library is needed for Turgo.
[02:00:17] <CIA-16> willard: PSARC/2007/277 tsalarm library, 6553178 The tsalarm library is needed for Turgo.
[02:00:18] <CIA-16> raghus: 6543604 NIU driver(nxge) support needed for Monza(CP3260)
[02:01:08] <alanc> of course, I'll still stick to using find to do recursive operations (except rm -r), since it works with all commands, not just those that added -r flags
[02:03:58] <wesolows> another day, another wad from gdamore adding support for a NIC
[02:04:17] <mmmurf> alanc:  hmm... how would you use that approach to find all the files recursively that contain the string earthquake, for example? (if you don't mind)
[02:05:34] <alanc> my usual habit is: find . -type f | xargs grep earthquake
[02:06:02] <wesolows> compare -exec
[02:06:03] <alanc> though you can also do: find . -type f -exec grep earthquake \{\} \+
[02:06:13] <wesolows> which is more reliable
[02:06:20] <wesolows> though perhaps slower
[02:06:47] *** migi has quit IRC
[02:07:10] <alanc> the -exec version works better if you have files with weird chars in their names, like spaces or carriage returns
[02:07:23] <wesolows> or too many files for xargs
[02:07:33] <alanc> but I haven't gotten my fingers retrained since I learned the xargs way first
[02:09:40] <mmmurf> alanc:  ahh thanks!
[02:09:47] *** migi has joined #opensolaris
[02:09:47] <gdamore> wesolows: are you complaining?
[02:10:12] <wesolows> gdamore: no!
[02:10:39] <wesolows> your productivity is such that the endless stream of putbacks has become unremarkable
[02:10:53] <wesolows> they do no one harm, and may help people
[02:11:27] <gdamore> i prefer to do more smaller putbacks, then to do "monster" putbacks of a wad of a million lines at a time.
[02:11:44] <wesolows> agreed; if the work lends itself to that, it's better
[02:11:53] <gdamore> (although to be fair, the last one was a couple thousand lines)
[02:12:28] <alanc> there might be some people harmed if he ever gets the ce nemoization putback done, but only if he can find the original authors of the ce code
[02:12:31] <gdamore> actually i have about three more RTIs coming *RSN*
[02:12:42] <gdamore> i know where the original author works.
[02:12:46] <gdamore> and it isn't Sun
[02:12:57] <wesolows> harming the ce author is a public service
[02:12:58] <gdamore> but I've shelved ce nemoification.
[02:13:22] <wesolows> would a case of beer or a bottle of good scotch get that off the back burner?
[02:13:28] <gdamore> NSN is supposedly doing it.  they objected to my removal of 11,000 lines of #ifdef'd out code ^W^W^Wswill
[02:13:45] <wesolows> but removing that code is an important part of the cleanup
[02:13:46] *** nrubsig has joined #opensolaris
[02:13:46] *** ChanServ sets mode: +o nrubsig
[02:14:06] <gdamore> that's what I said.  Joyce Yu politely disagreed with me.
[02:14:09] <wesolows> #ifdef thing_we_never_define ... #endif
[02:14:25] <LeftWing> #ifdef _WIN32
[02:14:44] <wesolows> LeftWing: might as well be, although at least there are build environments in which that might *conceivably* be defined.
[02:14:54] <gdamore> there is crap code that was put in there for Venus (SCA 4000) early testing, that will *never* ever, be useful again, and has never ever shipped to a customer.  but it is still there.
[02:15:01] <wesolows> This is more like #ifdef name_of_engineer_who_wrote_this_inscrutable_thing
[02:15:17] <gdamore> my favorite: #ifdef DITCHED_CODE
[02:15:19] <LeftWing> Vanity ifdefs? ;P
[02:15:26] <wesolows> yeah that's a good one
[02:15:27] <nrubsig> Does anyone know what's that horrible that noone wants to sponsor http://bugs.opensolaris.org/view_bug.do?bug_id=6561901 ?
[02:15:32] <wesolows> that's why we have SCCS
[02:16:10] <alanc> nrubsig: OpenSolaris is sadly lacking in sponsors for many requested bugs
[02:16:11] <wesolows> nrubsig: well, let's put it this way: if I were going to find a 169th hour in which to sponsor stuff, I wouldn't choose that one.
[02:16:13] * nrubsig detects bad english... lets rephrase the questions...
[02:16:35] <nrubsig> wesolows: why ?
[02:16:56] <gdamore> precompilation for shell scripts seems... pointless....
[02:17:01] <richlowe> gdamore: if I were to suggest that the *last* people I want fixing ce are the people responsible for its current incarnation, would that help you?
[02:17:11] <wesolows> (a) the project seems unnecessary, (b) it seems dangerous, (c) it seems like a big old ARC fight which is a waste of my time, (d) there are less controversial and (IMO) more valuable projects out there to sponsor
[02:17:12] <nrubsig> gdamore: erm
[02:17:28] <gdamore> richlowe: not really.  its a political mess, and I am not really a political animal.
[02:18:29] <richlowe> I don't care for politics, I care about the technology.
[02:18:40] <alanc> http://www.opensolaris.org/os/bug_reports/request_sponsor/ is sad - Murayama can't even get anyone to putback his ethernet drivers
[02:18:53] <nrubsig> gdamore: that's ONE if the usages of shcomp. Other usages include l10n message extraction using option -D, the abilty to "lint" shell scripts for dangerous constructs, faster startup time, the abilty to sign scripts and it should form the basis for one of the ksh88--->ksh93 migration tools
[02:18:56] <gdamore> his drivers aren't yet suitable for ON.
[02:18:57] <richlowe> alanc: I can think of several reasons why in their current form, I wouldn't either.
[02:19:03] <richlowe> alanc: but he'll have nicdrv problems, as gdamore used to.
[02:19:06] <gdamore> even my *own* AFE driver isn't quite ready yet.
[02:19:06] <wesolows> nrubsig: All of that, however, just points out how flawed the sponsor process is.  Since you've led a major (apparently successful) project, you should be about one more putback away from being able to do this yourself and putback directly without a sponsor
[02:19:10] <richlowe> "The tests fail, but we won't tell you why.  have fun"
[02:19:11] *** kaiwai has joined #opensolaris
[02:19:26] <kaiwai> eta on SXCE B71?
[02:19:32] <gdamore> i've started helping murayama... so it may get better soon
[02:19:37] <alanc> guess gdamore doesn't need a sponsor for his rge putback on that list anymore
[02:19:42] <richlowe> kaiwai: I doubt they have an opinion on it.
[02:19:50] <gdamore> actually, it was already putback, IIRC>
[02:19:58] <wesolows> nrubsig: Because my personal opinion of that project doesn't matter a good goddamn; if you're willing to do the work and can get it past the ARC and your code and design reviewers, you should be able to putback.
[02:20:32] <richlowe> wesolows: He should be, but he isn't.
[02:20:35] <gdamore> getting shell precompilation past ARC might be ... um... non trivial.
[02:20:48] <richlowe> unless he delays his next putback for, I dunno, enough time for those wanting to make that possible to finish.
[02:20:50] <wesolows> richlowe: Yes, I'm griping.
[02:20:51] <nrubsig> wesolows: re: (a): How can we extract l10n strings from scripts, re (b) why "dangerours" ? There is already javaexec and aoutexec etc. which all do the same stuff
[02:21:12] <richlowe> wesolows: yes, I was wondering if continuing to point it out would cause people to, I dunno, care.
[02:21:13] <gdamore> more code in the kernel == bad.
[02:21:32] <gdamore> unless there is a really compelling argument for it (like hardware support).   I'm not sure there is one in this case.
[02:21:33] <comay> gdamore: besides mxfe & afe, do you have any other drivers under consideration?
[02:21:46] <gdamore> define "consideration"
[02:21:57] <wesolows> nrubsig: Unnecessary because we already have #!, dangerous because adding more code that executes in kernel space is always a risk.
[02:22:11] <wesolows> for the record, javaexec should never have happened
[02:22:29] * gdamore points to nose.
[02:22:31] <comay> gdamore: supports hardware that isn't completely obsolete
[02:22:40] <richlowe> but now opensolaris is about marketing and all sorts of other stupidity, actually making it useful for development is a lost cause.
[02:22:54] <gdamore> i'm planning on porting dmfe to x86.  and rtls to sparc is about to be rti'd.
[02:23:07] <gdamore> or are those completely obsolete? :-)
[02:23:34] <jbk> well unfortunately, the code is still thrown over the wall to a large degree
[02:23:35] <wesolows> I think anything that supports 100Mbit and is still in the field is probably not completely obsolete
[02:23:37] *** mega has joined #opensolaris
[02:23:40] <comay> no, i'm not complaining
[02:23:46] <comay> just *wondering*
[02:24:01] <wesolows> 10Mbit-only, I wouldn't bother putting back today.
[02:24:12] <gdamore> mxfe i did because I had the hardware and wanted support.  i wouldn't spend time on it otherwise (you can't buy them anymore)
[02:24:19] <gdamore> afe is a totally different story.
[02:24:22] <comay> aww, i had a 3mb 3com driver i really wanted to putback
[02:24:29] <jbk> ce would be nice
[02:24:32] <richlowe> I have at least one, and perhaps has many as 4 cards that afe would support.
[02:24:35] <wesolows> comay: 3c503?  ISA?  wow.
[02:24:37] <richlowe> so yeah, afe is a good thing.
[02:24:42] <jbk> since i believe that's still your only option from sun for quad-nic cards
[02:24:43] <gdamore> I actually looked at the effort to GLDv3 pcelx and elx.  pcelx mostly because I want to kill pcgld.
[02:25:09] <comay> just joking wesolows.  last 3mb driver i looked at was for a vax
[02:25:18] <richlowe> jbk: sun have the atlas qgc, I think a quad e1000g, and qfe.
[02:25:28] <wesolows> comay: Yeah, I figured.  Hard to believe anyone still has such hardware in working order.
[02:25:29] <gdamore> most cardbus 100Mbps cards are supported by afe, as well as the highly common Linksys LNE100TX.  (which is probably the highest volume 100Mbps product of all time)
[02:25:31] <jbk> do they still sell qfe?
[02:25:39] <jbk> i mean if you buy a new server
[02:25:48] <jbk> at least up until may
[02:25:57] <jbk> if you bought a new sparc, they were putting in quad ce cards
[02:26:08] <gdamore> i don't think so.  but i looked at porting qfe->x86.  i might do it.  i might not.  undecided at the moment.
[02:26:11] <jbk> perhaps the niagra stuff is different
[02:26:12] <wesolows> I hope you can't buy qfe any more; it should be all atlas now.
[02:26:32] <gdamore> dmfe -> x86 is a little more interesting because of the huge number of DLink 530TX+ boards out there.
[02:26:43] <gdamore> and davicom is still shipping DM9102's.
[02:27:32] *** tg has joined #opensolaris
[02:27:47] <gdamore> i've been approached about helping tehuti out.  need to get approval from Sun mgmt though...  doing some 10G 3rd party drivers would be interesting.  NSN's stuff isn't much fun to work on.
[02:28:05] * nrubsig scratches his head
[02:28:55] <gdamore> actually the ne2000 (realtek 8029) which is *totally* obsolete is slightly itneresting, because qemu and parallels emulate them.
[02:29:21] <gdamore> murayama has a driver for them, though.
[02:29:32] <jbk> i should probably put my name down for the developer summit here soon.. i've got the time off, but need to figure out travel plans
[02:29:49] <nrubsig> gdamore: wouldn't it be easier to add a e1000g emu to qemu ?
[02:30:14] <gdamore> maybe.  realtek would be even easier.
[02:30:23] <nrubsig> gdamore: why ?
[02:30:31] <gdamore> but i have commit privileges to solaris, not to qemu.
[02:30:49] <gdamore> realtek hardware is incredibly simple.  the e1000g hardware is a bit more complex.
[02:31:07] <mmmurf> hello, anybody know the default location of cc on opensolaris 10 ?
[02:31:20] <nrubsig> mmmurf: /opt/SUNWspro/bin/cc
[02:31:22] <kaiwai> richlowe: pardon? is there an eta for SXCE B71?
[02:31:27] <alanc> if you mean Solaris 10, there is no cc by default
[02:31:33] <kaiwai> ETA = Estimated Time of Arrival
[02:31:36] <richlowe> mmmurf: solaris 10 is not open, SUNWspro is unbundled on both 10 and Nevada
[02:31:41] <e^ipi> jbk: "Top Dude" from Sun micro will be there to open it
[02:31:43] <e^ipi> heh
[02:31:47] <richlowe> when you install it, it will be where you installed it. :)
[02:31:50] <alanc> you have to download the Sun Studio compilers and add it
[02:31:51] <mmmurf> richlowe: ahh
[02:32:09] <comay> gcc is available though in both s10 and nevada, mmmurf
[02:32:23] <mmmurf> comay: ahh ok
[02:32:42] <gdamore> e^ipi: any further progress on your link problems with libc_i18n.a?
[02:32:53] <nrubsig> comay: and gcc on Solaris is a p**n in the a** and I wouldn't throw it after people even if marketing would threaten me with wild dragons to do so...
[02:32:56] <gdamore> i was thinking that may be you need to try compiling with PIC.
[02:32:56] <jbk> well it's more a matter of timing -- i'm still paying for a house as well as rent (as my house still hasn't sold), so that puts a bit of a crimp on finances -- if that wasn't an issue, flight & hotel would already be booked
[02:33:16] <e^ipi> hadn't thought of that actually
[02:33:17] <jbk> but i tend to be conservative financially (i.e. i almost never carry a cc balance)
[02:33:20] <alanc> kaiwai: no one here knows how the SXCE releases get to the download center or how to find out their status
[02:33:46] <e^ipi> I'm actually mostly just getting annoyed with my ( lack of real ) build system
[02:33:59] <comay> nrubsig: i don't know whether komodo dragons prefer gcc to suyn studio but i know almost any gcc is better than no compiler at all
[02:34:06] <e^ipi> and I think my next task will be to migrate the code in to usr/src/lib/libc
[02:34:07] <gdamore> you should try integrating your stuff into a real ON tree.
[02:34:07] <nrubsig> alanc: is there no way to hijack a bunch of managers, drag them to guantanamo and then get the answers out of their heads ?
[02:34:10] <wesolows> the current working theory is avain carrier and a mad pigeonkeeper
[02:34:15] <wesolows> *avian
[02:34:18] <gdamore> that would be easier.
[02:34:19] <richlowe> wesolows: close.
[02:34:26] <e^ipi> it would be much easier, heh
[02:34:49] <alanc> nrubsig: sorry, no budget for torture this quarter, and I wouldn't even know which managers to torture for it
[02:35:05] <gdamore> start with Jeff Jackson? :-)
[02:35:10] <comay> wesolows is correct - and when the kobodo dragons are hungy, the pigeons end up being food and the sxde builds are delayed
[02:35:11] <nrubsig> comay: right (but note that komodo dragons won't even care for any computer)
[02:35:12] <kaiwai> e^ipi: is it almost complete? (the language support for libc)
[02:35:17] <wesolows> nrubsig: I assume you've filed bugs against gcc for the problems you've seen?
[02:35:43] <alanc> stevel mentioned something about added delays due to the migration of the Sun download center to new software
[02:35:46] <nrubsig> alanc: not even for a little session of electroshocks
[02:35:59] <e^ipi> kaiwai: it builds, and all the relevant symbols should be defined
[02:36:16] <e^ipi> that said, it needs to live in libc before I can start breaking machines with ti
[02:36:25] <kaiwai> neat, so end of the closed binaries soon :)
[02:36:28] <jbk> though i guess i should see how expensive transportation from sjc->sc is
[02:36:34] <richlowe> of the closed build dependencies.
[02:36:40] <e^ipi> kaiwai: not entirely
[02:36:47] <richlowe> assuming jbk isn't slacking :)
[02:37:05] <nrubsig> wesolows: My complaint is more about a general nature of gcc, e.g. Sun Studio produces faster code and has better options to squish the last 10% out of the code. And C99 support works.
[02:37:06] <jbk> richlowe: i'm just stuck in process land
[02:37:12] <wesolows> jbk: should be no more than $2
[02:37:13] <richlowe> jbk: which bit?
[02:37:31] <jbk> you can grab my code and compile it and drop it in and it works fine
[02:37:43] <richlowe> have you filed a bug, and requested a sponsor for it yet?
[02:37:54] <jbk> wesolows: to santa cruz? that's pretty cheap
[02:37:55] <wesolows> jbk: sjc has a free bus to the santa clara caltrain station and the light rail stations; I usually get the bus (59? 61?  something like that) from the caltrain station, which drops me 1/2 mile or so from SCA
[02:38:04] <wesolows> jbk: oh, sorry, SC != SCA
[02:38:13] <wesolows> that's like $15 or so I think
[02:38:18] <jbk> richlowe: yes, there's a bug now for libdisasm
[02:38:19] *** vmlemon has quit IRC
[02:38:32] <alanc> probably enough people flying in to SJC for the OS devsummit that you could organize ride sharing
[02:38:37] <jbk> but waiting to hear back on if/how i should handle the other existing bugs (w/ closed version) i've found
[02:38:44] <richlowe> file them, fix them.
[02:38:48] <jbk> well
[02:38:49] <richlowe> bugs are bugs, fixes are good.
[02:38:54] <jbk> i can do 1 entry for all of them
[02:39:03] <jbk> or like 3 for the different 'types' i've found
[02:39:11] <richlowe> file a bug per (logical) bug, fix them all in that wad. :)
[02:39:12] <jbk> not sure which would be appropriate
[02:39:27] <jbk> well they are all subcases of 'dis produces incorrect output'
[02:39:52] <richlowe> alanc: I don't see any indication of who intends to be there.
[02:40:24] <wesolows> jbk: yes but individual bugs makes it easier to eval them, and it pads your bug count
[02:40:35] <wesolows> it's all about looking really really productive
[02:40:46] <jbk> hehe.. i'm not a sun employee :)
[02:40:51] <richlowe> the logical was intended to imply grouping of related bugs.
[02:40:51] <jbk> this is all done outside of work
[02:40:56] <wesolows> foo.c line 398 is wrong, foo.c line 399 is wrong, etc.
[02:41:03] <wesolows> well, still
[02:41:08] <jbk> so bug counts don't really concern me
[02:41:10] <wesolows> would you like to be? :-)
[02:41:15] <richlowe> one master "dis has incorrect output" seems broken.
[02:41:20] <jbk> (it would be nice if i could work on it at work)
[02:41:28] <jbk> wesolows: i seem to have bad luck in that regard
[02:41:39] <richlowe> whereas "dis always uses %fp register names", and such seem usefully distinct.
[02:41:40] <jbk> twice now people have tried to hire me on to different positions
[02:41:47] <wesolows> richlowe is correct
[02:41:48] <sbahra> pong?
[02:41:55] <richlowe> wesolows: never.
[02:41:56] <jbk> one case the manager was just silly
[02:42:02] <wesolows> but he's being drily technical
[02:42:02] <jbk> the other case, there was no open reqs
[02:42:15] <richlowe> wesolows: right now, the choice is that and asshole.
[02:42:18] <richlowe> I'm being conservative :)
[02:42:50] <wesolows> richlowe: fair enough
[02:43:04] <jbk> hehe
[02:43:39] <jbk> i am unsure about this current job though -- my friend says i'm a crack addict
[02:43:59] *** jpdrawneek has left #opensolaris
[02:44:16] <wesolows> well, it could be worse; at least crack is cheap
[02:44:17] <jbk> i went from being overworked, forced to work on vacations, etc. to literally having hours with nothing to do but stare are my monitor and do rather trivial stuff
[02:44:27] <wesolows> try being addicted to heroin and coke
[02:45:16] *** jafari has joined #opensolaris
[02:45:59] <jbk> well i guess arguing with development groups over audit controls & compliance issues is somewhat advanced, but that's still a very minor amount of time
[02:46:08] *** alanc is now known as alanc-away
[02:46:13] *** tg has quit IRC
[02:46:14] <nrubsig> wesolows: well, I'll try it when I'm 80 and have at least $1000000 in stocks
[02:47:00] <wesolows> who wants to wait?  everything's more fun when you're young and you still have a liver!
[02:47:23] <jafari> good night everyone, question in solaris 9/10 how do ou can for new luns/hdd on the OS?
[02:48:01] <masta> just present the lun
[02:48:05] <jbk> i think a verb is missing
[02:48:13] <jbk> or for == format?
[02:48:20] <nrubsig> jafari: ask jmcp
[02:48:48] <jafari> ok
[02:48:52] <jafari> hows jmcp
[02:49:01] <nrubsig> !seen jmcp
[02:49:13] <jafari> i mean whos jmcp
[02:49:20] <nrubsig> Drone: wake up!
[02:49:28] <masta> newfs? zfs create foo/volume-name
[02:50:01] *** bengtf has quit IRC
[02:50:03] <jafari> say if you pop in a new hdd how can you detect it for formatting
[02:50:17] <jbk> devfsadm should detect the new disk
[02:50:31] <jbk> if using zfs, it should just depend on how you want to use it in/with a pool
[02:50:40] <jbk> zfs add or zfs create disk..
[02:50:41] <jafari> just type devfsadm with no arguments?
[02:51:01] <jbk> if using ufs or something else; you'll need to run format, select it, and partition (slice) it up
[02:51:04] <jbk> correct
[02:51:11] <jbk> it should rescan stuff and see if anything new is there
[02:51:24] <jafari> ok
[02:51:36] <jafari> thanks
[02:51:56] <masta> anybody ever see their nsswitch.conf get filled with this:
[02:52:01] <masta> [ Aug 23 16:40:07 Executing stop method (null) ]
[02:52:13] <wesolows> nsswitch.conf?!  no, never
[02:52:15] <richlowe> that's part of an SMF logfile.
[02:52:18] <masta> today mine was filled with that when I went from wifi to ether
[02:52:31] <richlowe> my *hope* would be that someone poked that string into the logfile property of that service.
[02:52:36] <richlowe> my fear, is that software did it for you.
[02:52:39] <masta> I think smf somehow overflowed into my nsswitch.conf
[02:52:41] <richlowe> (my bigger fear, would be that something ate your FS)
[02:52:48] <jbk> brb
[02:52:49] <masta> scarry
[02:52:56] <masta> right
[02:53:03] <jafari> after running devfsadm , what do i do after run the command format?
[02:53:14] <masta> I'm scared, lsof'n my nsswitch
[02:53:37] <masta> jafari: newfs, or zpool
[02:54:27] <masta> richlowe: that string is one of many just like it dating back weeks
[02:54:50] <jafari> masta, ok but what is the command format for? do i need to use it or after usinf devfsadm i can just start using newfs or zpool?
[02:54:57] <richlowe> svcprop '*' | grep 'logfile.*nsswitch'
[02:55:11] <masta> I did have a smf event just prior, as I was migrating to another network
[02:56:03] <masta> richlowe: ty
[02:56:13] <richlowe> if it there's output, something bad happened.
[02:56:58] <kaiwai> jmcp: you there?
[02:57:01] <masta> no output
[02:57:06] <richlowe> huh.
[02:57:59] <masta> svcprop outputs, the grep doesn't match the string
[02:58:22] <richlowe> wonder how you're getting that gunk in nsswitch.conf then.
[02:58:55] <masta> a classic spew I guess... I'm probably going to have to reproduce with a traced or something to report it
[02:59:04] <nrubsig> richlowe: maybe someone had a rw filedescriptor open to nsswitch.conf and then a memory corruption mixed the descriptors ?
[02:59:15] <masta> exactly
[02:59:38] <masta> I have lsof going on it, will try to reproduce the condition again
[03:00:01] <sbahra> yo masta
[03:00:15] <masta> went from my iwi0 to iprb0
[03:00:17] <CIA-16> gd78059: PSARC 2007/291 mxfe ethernet driver, 6562372 add mxfe driver
[03:00:37] <masta> hey sammy
[03:00:41] <sbahra> ;]
[03:00:47] <jafari> question if you accidently remove /etc/system and the os is unable to reboot how can you recover the system?
[03:01:09] <masta> you can boot the cdrom, and mount root
[03:01:14] <richlowe> boot with -a, and give it an alternative.
[03:02:30] <jbk> or specify /dev/null instead
[03:03:29] <jafari> what is the option -a? somwone told me to boot the system in to interactive mode and the /etc/system file recreates itself, is that true?
[03:04:17] <jafari> is that what -a does?
[03:04:37] <gdamore> CIA finally caught up. :-)
[03:04:50] <richlowe> gdamore: it does it on the hour.
[03:05:01] <gdamore> oh, makes sense.
[03:05:25] *** mega has quit IRC
[03:05:38] *** sbahra has quit IRC
[03:07:31] <jafari> n/a
[03:07:32] <jafari> ?
[03:08:25] *** tg has joined #opensolaris
[03:08:40] <jbk> hmm.. my test box seems to be down :(
[03:09:54] <masta> jafari: boot -as, specify dev/null for system file
[03:10:11] <masta> answer the questions then fix your system file
[03:10:20] <jbk> hmm..
[03:11:11] <gdamore> i just submitted the RTI make rtls supported on SPARC. :-)
[03:11:23] *** fedorared has joined #opensolaris
[03:13:32] <richlowe> gdamore: shouldn't you be doing all this in 78/79?
[03:13:47] *** bengtf_ has joined #opensolaris
[03:13:49] <richlowe> gdamore: so little chance of you getting getting yelled at because of DE, if you do it now.
[03:13:49] *** bengtf_ is now known as bengtf
[03:14:00] <gdamore> *exactly*
[03:14:11] *** hohum has quit IRC
[03:14:16] <wesolows> probability that my gigantic SMF wad will go back in build 73: 40%
[03:14:31] <gdamore> since 73 closes on Tuesday night.
[03:14:36] <wesolows> yes
[03:14:41] <wesolows> otherwise it would b 5%
[03:15:04] <masta> smf, uh....
[03:15:07] <gdamore> gigantic wads scare me. :-)
[03:15:18] <masta> especially in the nsswitch.conf
[03:15:55] <wesolows> well, I file an ARC case and they tell me I have to fix some other RFE before I can putback, then I go implement this thing and I find another half dozen bugs and I have to fix those too so I can test; lather, rinse, repeat and you get a giant wad
[03:15:58] <kaiwai> hmm, intel is doing a fair amount of contributions
[03:16:27] <wesolows> then Bustos does a code review and finds a bunch more bugs, ... sigh.
[03:16:49] <gdamore> i try to putback smalle wads earlier... i wish more people would fix directly in Nevada, rather than fixing in "some-random-project-gate".
[03:17:00] <wesolows> no project gate here
[03:17:03] <wesolows> just my personal workspace
[03:17:09] <gdamore> ah ok.
[03:17:47] <gdamore> there's a crapload of things in e.g. cleaview or crossbow, that probably could have been putback separately, but they sit in those gates for now.
[03:18:11] <gdamore> it seems like some people like the idea of a "big project milestone" punctuated by a putback.
[03:18:51] <wesolows> in many cases it's impossible to phase things because they'll be incomplete without the other stuff
[03:18:59] <wesolows> it varies by project
[03:19:10] <wesolows> I try not to second guess other project teams too much on this
[03:19:54] <gdamore> i see things like "fix random bug in driver x" in a project gate, and its only there because the project team found it.  they need the fix, but that doesn't mean that the fix isn't universally relevant.
[03:20:10] <wesolows> what is annoying is that so many giant projects get a whole build for themselves, break a whole bunch of shit in that build, and get to stay in because the C-team is afraid of pissing off management who made a big deal of setting aside that build for the project
[03:20:33] <wesolows> there were several of these toward the end of S10
[03:20:36] <gdamore> yes that is really annoying.
[03:20:37] <richlowe> sod that.
[03:20:52] <gdamore> bug again, i bet alot of those cases could go back separately.
[03:21:29] <gdamore> for example, i'm reading power managment fixes  to add DDI_SUSPEND/RESUME support to a bunch of drivers.  Why not put them back as they are reviewed, tested, etc.  No need to hold them all for one giant putback at the end.
[03:21:55] <wesolows> it's a bit of a tradeoff there though
[03:22:04] <gdamore> There are other cases as well... e.g. it would be nice to have support in GLDv3 for NICs that can't do VLANs, but we won't get that until it comes in with Clearview. :-(
[03:22:07] <wesolows> for example if you have different code reviewers for each one, they might find different bugs
[03:22:13] <wesolows> and you end up not getting them all the same
[03:22:25] <e^ipi> theoretically, suspend/resume should just work on sparc, correct?
[03:22:26] <wesolows> also, larger wads help amortise the cost of code review and RTIs
[03:22:32] <gdamore> e^ipi: yes.
[03:22:41] <gdamore> wesolows; strongly disagree.
[03:22:56] <gdamore> I'd rather review 10 small changes separately than one modno review that is 10x as big.
[03:23:12] <wesolows> as would I
[03:23:27] <gdamore> the cost of RTIs I agree with.  but the benefit of getting changes in early, and reducing merge nightmares, is probalby well worth it.
[03:23:37] <wesolows> but if you're making the same change in 100 places, I'd rather see it all at once so I can make sure you did them all the same and that you didn't miss any
[03:24:02] <gdamore> if its in 100 places, it will be hard for me to make sure that you didn't miss any. :-)
[03:24:02] <jbk> hmm... there is one issue with kmdb + libdisasm (which is really more of a 'how do I do it correctly' than anything)
[03:24:07] *** bondolo has quit IRC
[03:24:10] <richlowe> what is it?
[03:24:12] <wesolows> jbk: which is?
[03:24:32] <richlowe> recall that you're working on a change a whole lot of us want to see, there's no shortage of people willing to try and help :)
[03:25:06] <jbk> well i utilize vsnprintf to implement 'bprintf' basically simple wrapper function to printf into whatever buffer was supplied to hold the disassembled output
[03:25:13] <gdamore> I shudder to think how many versions of the mac.c module are around... one version in Brussels, another in Cleaver, another in Crossbow, the official Nevada, and probably a half-dozen or so individual ones with engineers test code.
[03:25:16] <kaiwai> anyone got KDE 3.5.7 binaries?
[03:25:21] <jbk> in usr/src/lib/libdisasm/Makefile.com
[03:25:27] <gdamore> s/Cleaver/Clearview/
[03:25:39] <jbk> they (for kmdb) rename snprintf to mdb_snprintf
[03:26:03] <jbk> however, to do something similar for vsnprintf, i really just need to s/vsnprintf/mdb_iob_vprintf/
[03:26:38] *** mega has joined #opensolaris
[03:26:56] <jbk> so do I try to somehow fanagle STAND_RENAMED_FUNCS (in Makefile.com) to do it, or do I just add my own -D vsnprintf=mdb_iob_vsnprintf to the CPPFLAGS_standalone line
[03:27:09] <e^ipi> kaiwai: probably, but in either case, in case you're curious the lion's share of KDE4 deps. build with studio.
[03:27:56] <kaiwai> e^ipi: oh, I'm probably best to wait for kde 4 then
[03:28:35] <e^ipi> the KDE org is very receptive to patches, so a couple guys are upstreaming stuff to make it work
[03:29:26] *** nachox has joined #opensolaris
[03:29:34] <jbk> or, do i modify kmdb to make mdb_vsnprintf as part of the module api?
[03:36:45] <coraline> Nacho cheese.
[03:37:17] <e^ipi> ... carrying on ...
[03:38:44] <jafari> where is the file located where you set ip/netmask in solaris?
[03:39:00] <jafari> does solaris 10 comes with a JDK?
[03:39:39] <e^ipi> it's sun... what do you think?
[03:40:14] * coraline lols
[03:40:39] <jbk> it comes with _2_ :)
[03:41:02] <jbk> wesolows: any opinion?
[03:43:04] <nachox> jbk, you love seeing wesolows rant dont you?
[03:44:32] <Drone> jmcp is currently online in #opensolaris and last spoke on Fri 31 Aug 2007 01:06 GMT, saying 'jmcp recaffeinates'.
[03:44:45] <e^ipi> umm... thanks Drone
[03:44:50] <LeftWing> hahaha
[03:44:50] <jafari> how do i set it up
[03:44:53] <jbk> well he probably has a better idea as to which approach is more likely to be shot down and which isn't :)
[03:45:09] <jafari> do i need to declare a variable JAVA_HOME?
[03:45:35] <nrubsig> e^ipi: mhhh... only one hour lag today for Drone ... a new... record... =:-)
[03:45:39] <jafari> im want to install/test weblogic and it make require JDK
[03:48:30] <Tempt> Morning everyone.
[03:48:33] *** cypromis has quit IRC
[03:49:04] <kaiwai> good morning tempt
[03:50:56] <jafari> morning
[03:54:40] <jbk> well i have a diff against a pretty current copy of ON that puts everything into place if anyone wants to test it :)
[03:55:26] <e^ipi> hey Tempt
[03:56:01] <Tempt> How's the coffee mission going?
[03:56:08] <Tempt> (I'm waiting for my machine to heat up)
[03:56:39] * LeftWing imagines Coffee Missionaries.
[03:56:56] <Tempt> LeftWing: Settling into the zone?
[03:57:03] <Tempt> LeftWing: (the revdns has been done now)
[03:57:34] <e^ipi> Tempt: not bad... do you think I could PID this little machine?
[03:57:44] <LeftWing> Tempt: I'm effectively settled, but distracted with impending Uni assignments.
[03:57:51] <e^ipi> I only ever see PID's for silvia & gaggio machines
[03:58:09] *** mmmurf has quit IRC
[03:58:26] <Tempt> Probably not worth installing a PID.
[03:59:08] <Tempt> I don't think temperature stability is the limiting factor with that machine. People PID the Silvia and Gaggia machines because the biggest quality difference between a commercial and a Silvia is temperature stability.
[03:59:25] <Tempt> No reason why you couldn't do it though!
[04:01:35] *** coraline has quit IRC
[04:01:49] <nrubsig> Where is mail.opensolaris.org located ?
[04:02:09] *** coraline has joined #OpenSolaris
[04:02:34] <wesolows> San Francisco, I think
[04:02:40] <e^ipi> you & the rest of the coffeegeeks will be the death of me... I already pulled my portafilter to bits & the other day I couldn't get my grinder to go fine enough so there goes the warranty on that thing too
[04:03:01] <nrubsig> wesolows: Thanks! :-)
[04:03:13] * nrubsig still sulks about shcomp
[04:03:25] <wesolows> jbk: Ask on mdb-discuss; I'm not too familiar with that code but I expect mws will answer your question.
[04:03:44] <Tempt> e^ipi: Which grinder?
[04:04:31] <e^ipi> saeco 166
[04:05:00] <e^ipi> err, solis
[04:05:41] <e^ipi> works great at the coarser grinds, i've been using it mostly for press
[04:06:09] <Tempt> Hmmm
[04:06:18] <Tempt> Mazzer time.
[04:06:25] <Tempt> :)
[04:06:26] <e^ipi> yeah, that's what I'm thinking
[04:06:33] <e^ipi> next time I've got the budget for it
[04:06:34] <Tempt> Iberital, to be honest
[04:06:53] <Tempt> No point in spending much $$$ for a Mazzer at home, the Iberital does a suitable job.
[04:08:09] <e^ipi> interesting
[04:09:19] *** Lusitanian has quit IRC
[04:10:19] *** laca has quit IRC
[04:10:50] <e^ipi> hadn't heard of them before, actually
[04:10:59] <cmang> Hmm.  Was wondering if someone could help with this question.  I have a system drive previous in a netra x1 in an ultra 10.  It seems the hard drive device locations are different, as I can't seem to run fsck on the drive: fsck: could not stat /dev/rdsk/c0t0d0s0: No such file or directory
[04:11:23] <cmang> question is, how can I fsck the drive and have it booted properly :)
[04:11:36] <jbk> sam: devfsadm, then run format and see what's there
[04:12:07] <cmang> jason, do I have to remount it as readwrite?
[04:12:27] <cmang> it's the only drive in the system.
[04:12:58] <jbk> ahh hmm.. that's interesting
[04:13:07] <jbk> might try a boot -a and have it rebuilt the path_to_inst file
[04:13:31] <jbk> err rebuild
[04:14:25] <cmang> that's boot -a from the OF prompt?
[04:14:30] <jbk> yeah
[04:15:06] <jbk> there's probably other ways, but that'd be the easiest
[04:15:32] <jbk> what i suspect happened is /etc/path_to_inst keeps track of hardware path->instance #
[04:15:46] <cmang> I had switched the drive between the two machines before (from the ultra to the netra), apparently I had a problem that was "fixed" as I now have /dev/rdsk.old and /dev/dsk.old.  ah, the importance of taking notes :)
[04:15:47] <jbk> so c0 is still referring to the hardware path on the netra
[04:16:00] <cmang> I see
[04:16:26] <jbk> since it's a different box, rebuilding path_to_inst is probably better
[04:16:27] <e^ipi> Tempt: presumably someone, somewhere sells this grinder?
[04:16:57] <jbk> though dunno if -a implies -r (would make sense to)
[04:18:07] <Tempt> Well, I bought one, so I presume there are more around!
[04:19:32] *** Nishaway has quit IRC
[04:19:48] <Tempt> e^ipi: have a morning coffee @ http://www.purplecow.org/?p=48
[04:19:51] *** Red_Cloud has joined #opensolaris
[04:20:41] <e^ipi> that's a nice cup of crema... usually there's espresso below it
[04:20:46] <cmang> jbk: nice.  I tried "boot -a" from OF, it says "Fast Data Access MMU Miss."  does that mean I can go ahead and just "boot," or is this some bad problem :)
[04:21:19] <jbk> grr.. I hate that error
[04:21:24] <masta> anybody have fluxbox working on 5.11?
[04:21:34] <jbk> try:
[04:21:39] <jbk> setenv auto-boot? false
[04:21:40] <jbk> reset-all
[04:21:42] *** nachox has quit IRC
[04:21:45] <jbk> (it'll go through post, etc)
[04:21:48] <jbk> then boot -a
[04:21:59] <e^ipi> masta: yes, it compiles fine
[04:22:06] <masta> e^ipi: ty
[04:22:19] <Tempt> e^ipi: I'm drinking those very shots right this minute.
[04:22:22] <Tempt> Deee-licious
[04:22:41] <e^ipi> what sort of beans do you run through your machines, anyways?
[04:22:50] <Tempt> depends.
[04:22:55] <Tempt> At the moment, Atomica Dark
[04:23:05] <Tempt> (blend)
[04:23:40] <Tempt> I move around a little, there are two local roasters who have their shit together and will get me beans the day of roast.
[04:24:13] <jafari> any Merrill Lynch employees here?
[04:24:23] <Tempt> I tend to do single origin beans more for press or vacuum pot than for espresso.
[04:24:32] *** Red_Cloud has quit IRC
[04:25:00] <cmang> jbk: nice.  it's asking me questions.  crap, what I name the kernel, heh
[04:25:20] *** lagasek has joined #opensolaris
[04:25:27] <cmang> either unix or kernel, I suppose
[04:28:02] <jbk> it should have a default
[04:28:38] <jbk> generally accept the default, but if it asks for the location of path_to_inst, enter /dev/null (or yes if it asks to rebuild it; can't remember what the prompt actually is)
[04:29:31] <masta> dang it, fluxbos failed to build..
[04:29:36] * masta hunkers down
[04:31:01] <masta> e^ipi: was that last comment in referrence to gcc of studio?
[04:31:18] <masta> s/of/or/g
[04:31:19] <e^ipi> i don't actually remember which I built it with
[04:31:34] <masta> umkay
[04:31:52] <e^ipi> i usually go with studio and if it fails badly enough that I don't feel like fixing it, i fall back to gcc
[04:32:03] <masta> nod
[04:32:06] *** lagasek has left #opensolaris
[04:33:06] <jafari> what command do ou guys use to show current memory size one a system
[04:33:09] <jafari> ?
[04:33:22] <Tempt> As in how much RAM? prtconf | grep Mem
[04:33:27] <e^ipi> Tempt: i'd love to blend stuff myself, but as I live an apartment with no patio, roasting presents a whole new set of problems
[04:33:36] <e^ipi> ones which my landlord may not approve of
[04:33:39] <Tempt> e^ipi: Same deal here
[04:33:46] <Tempt> I did a popcorn popper roast once
[04:33:54] <Tempt> It was fine but the smoke and mess was a bit much.
[04:34:03] <jafari> thanks
[04:34:30] <Tempt> You can already smell coffee by the time you get outside my apartment door as it is.
[04:34:39] <e^ipi> heh
[04:34:49] <e^ipi> I have grounds covering pretty much everything I own
[04:35:10] <masta> grounds make good compost catalist
[04:35:18] <masta> the grubs love that
[04:35:59] <jafari> what causes a service to go in to maintenance mode?
[04:36:12] <jbk> depends
[04:36:16] <jbk> what does svcs -x show?
[04:36:23] <jafari> seems like there is something wrong with svc:/system/webconsole:console (java web console)
[04:36:38] <jafari> Reason: Start method exited with $SMF_EXIT_ERR_FATAL.
[04:36:58] <jbk> grrrrrr
[04:37:01] <cmang> jbk: interesting.. booted it up, and it's still saying "No such file or directory."
[04:37:09] <e^ipi> masta: yes, well... if my landlord would have a problem with me roasting coffee in my apartment, he'd probably have a problem with me composting in it too
[04:37:14] <jbk> what does format show?
[04:37:28] <cmang> no disks found
[04:37:29] <jbk> i thought from the os.org website, you could post to any discussion list
[04:37:32] <cmang> when it's booted off one. :)
[04:37:35] <jbk> run devfsadm -C
[04:37:38] <jbk> then try again
[04:37:41] <e^ipi> and the grounds are accidental
[04:37:47] <jbk> apparently you cannot
[04:37:48] <cmang> read only filesystem
[04:37:53] <jbk> hmm
[04:37:57] <jbk> what does 'mount' show
[04:38:06] <jbk> actually
[04:38:11] <jbk> just try mount -o remount,rw /
[04:38:30] <jbk> might have to actually boot off cdrom :)
[04:38:38] *** Shiv__ has joined #opensolaris
[04:38:39] <cmang> it's beautiful... "mount: /dev/dsk/c0t0d0s0 or /, no such file or directory"
[04:38:45] <cmang> cdrom, ouch
[04:39:14] <jbk> what about just devfsadm ?
[04:39:15] <jafari> it wont start sudo svcadm enable system/webconsole
[04:39:27] <Tempt> Composting in an apartment.
[04:39:28] <jafari> $ svcs -a|grep -i webconsole
[04:39:28] <jafari> maintenance Aug_30 svc:/system/webconsole:console
[04:39:31] <Tempt> That's just full of win.
[04:39:40] <jbk> also, what are you running? 10, nevad?
[04:40:13] <cmang> jbk: devfsadm: mkdir failed for /dev 0x1ed: Read-only filesystem.  this is an old build of nevada 28.  I  know I need to upgrade, just want to recover data primarily
[04:40:50] <cmang> suppose I could just wait longer for the netra serial cable to get here, heh.
[04:41:09] <Tempt> netra serial cable?
[04:41:15] <Tempt> you mean just another serial cable.
[04:41:20] <cmang> an rj45 serial cable is what I mean
[04:41:29] <cmang> but yeah
[04:41:46] <Tempt> Same as cisco, so if you can pinch someone's cisco console cable ...
[04:41:49] <jafari> any expert here with weblogic?
[04:41:52] <jbk> why would you need to use the serial console?
[04:42:00] <richlowe> why would you want to use something else?
[04:42:01] <jbk> you can obviously right now get to the obp
[04:42:59] <jafari> does the weblogic server 10 run via the web when its installed or i need some form of web server also like SunONE Webserver or Apache ? is it the same concept as apache/tpmcat?
[04:43:20] <jafari> tomcat
[04:43:31] *** sbahra has joined #opensolaris
[04:43:34] <cmang> jbk: well, I would need the serial console to setup the netra for my network, with this drive/installation.  The reason it's in the ultra right now is so I can use it in the meantime without the serial cable (as the ultra has a keyboard and monitor).  But I'm running in to this issue with the disk location, heh
[04:44:13] <jbk> no cdrom on the ultra 10?
[04:44:25] <cmang> nah.  Suppose I could dig one out if I get desparate.
[04:44:37] <jbk> oh hmm...
[04:44:40] <nrubsig> Shiv__: ping!
[04:44:45] <jbk> what does mount show for / ?
[04:44:57] <Shiv__> nrubsig: Hi
[04:45:04] <nrubsig> Shiv__: Do you know whether Moinak can sponsor patches for OS/Net ?
[04:45:11] <richlowe> nrubsig: sec.
[04:45:18] <nrubsig> richlowe: ?!
[04:45:21] <Triskelios> jafari: not sure what weblogic server is, but the solaris webconsole service mentioned *is* tomcat
[04:45:24] *** mega has quit IRC
[04:45:28] <richlowe> nrubsig: http://opensolaris.org/os/community/on/crt/advocates/
[04:45:33] <richlowe> nrubsig: not that it's necessarily up-to-date
[04:45:51] <cmang> jbk: / on /pci@1f,0/pci@1,1/ide@3/disk@0,0:a read/write/setuid/devices/dev=2200008 on Wed Dec 31 18:00:00 1969
[04:45:52] <Shiv__> I do not know. I do not remember him sponsoring.
[04:46:16] <Shiv__> As seen by me on the mailing list.
[04:46:31] <jbk> hmm.. ok try mount -o remount,rw /pci@1f,0/pci@1,1/ide@3/disk@0,0:a /
[04:46:38] <richlowe> nrubsig: that doesn't necessarily mean he can't though, just that he hasn't.
[04:46:44] <jbk> though it looks like it's already read-write
[04:46:49] <nrubsig> grumpf
[04:46:50] <Shiv__> Yes
[04:46:52] <jbk> so devfsadm shouldn't be failing
[04:46:56] <cmang> something funny is definitely going on!  heh
[04:47:12] <nrubsig> OK, XML demo is online, too.
[04:47:13] <cmang> it boots in to maintenance mode complaining of an unclean /
[04:47:25] <jbk> does touch /test fail?
[04:47:30] <nrubsig> http://svn.genunix.org/repos/on/branches/ksh93/gisburn/scripts/xmldocumenttree1.ksh
[04:47:34] <jbk> actually
[04:47:55] <jbk> you could just fsck -F ufs /pci@1f,0/pci@1,1/ide@3/disk@0,0:a
[04:48:18] *** jcsmith has quit IRC
[04:48:19] <cmang> do I need to escape characters like @ at the prompt? :)
[04:48:24] <jafari> Triskelios weblogic is an Java application server like how tomcat works?
[04:48:25] <jbk> err i guess there should probably be a /devices in front of that
[04:48:26] <cmang> when typing in to bash
[04:48:30] <cmang> I mean, sh
[04:48:33] <jbk> shouldn't need to
[04:48:33] <cmang> *couh*
[04:48:35] <cmang> okay.
[04:48:38] <richlowe> depends on the terminal setup.
[04:48:46] <richlowe> but if you press @ and it does something unexpected, fix it with stty
[04:48:57] <jbk> well it's not hp-ux, or 1978 :)
[04:49:20] <jafari> http://www.bea.com/framework.jsp?CNT=index.htm&FP=/content/products/weblogic/server/
[04:49:21] <richlowe> I wouldn't be shocked if there was a situation you could easily get into where @ and # were still editing characters.
[04:49:24] <nrubsig> When does B73 close ?
[04:49:41] <richlowe> nrubsig: 11PM PDT tuesday
[04:49:47] <richlowe> monday, rather.
[04:49:48] <jbk> richlowe: that's still the default on hp-ux
[04:49:55] <Triskelios> jafari: yeah looks like it
[04:50:02] <jbk> they just change it in /etc/profile, etc.
[04:50:03] <richlowe> nrubsig: http://www.opensolaris.org/os/community/on/onnv_schedule.txt
[04:50:06] <richlowe> nrubsig: bookmark it already :)
[04:50:29] <nrubsig> richlowe: Well, I need a sponsor somehow (demo code patch) ...
[04:50:32] <richlowe> no, I was right, tuesday.
[04:51:05] <jafari> Triskelios looks like it will run on its own?
[04:51:45] <nrubsig> harrrrrrrrr
[04:51:52] * nrubsig sees wesolows on http://opensolaris.org/os/community/on/crt/advocates/
[04:52:02] <richlowe> nrubsig: monday is a US holiday, I think.
[04:52:13] <cmang> jbk: yes, touch /test says "touch: /test cannot create" .. mount or fsck with the same device shown running "mount" says "No such file or directory" for the device.
[04:52:23] * nrubsig rotates his head and looks at wesolows and starts purring...
[04:52:37] <nrubsig> wrrrrrrrrrrwrrrrrrrrwrrrrrrwrrrrrr
[04:52:40] <richlowe> I'd be very surprised if wesolows was still sober.
[04:52:45] <cmang> jbk: additionally, I noticed that /devices/pci@1f,0/etcetra exists.  But rather than disk@0,0:a, it's dad@0,0:a .. (I had originally tried running tools with that, but that also says No such file or directory)
[04:52:53] <nrubsig> richlowe: taser will solve that problem.
[04:53:39] <masta> does fujitsu fab sparc now, or does sun have a fab in usa?
[04:53:48] <kaiwai> sun doesn't have a fab
[04:53:51] <kaiwai> its fabless
[04:53:59] <kaiwai> not the be confused with fabulous
[04:54:02] <masta> what about design?
[04:54:24] <nrubsig> masta: Sun makes it's own designs and Texas Instruments builds the childs
[04:54:38] <richlowe> except OPL, which is Sun/Fujitsu, and using Fujitsu's CPUs
[04:54:39] <masta> ty.. you solved an argument in ##freebsd
[04:54:42] <nrubsig> masta: and Fujitsu makes it's own chipds and has it's own fab
[04:54:55] <nrubsig> richlowe: OPL is Fujitsu deign and fab, too
[04:55:05] <kaiwai> ah, iI remember the freebsd room, good memories
[04:55:12] <kaiwai> quite a nice design too
[04:55:15] <masta> oh, so the T1 and T2 are in fact fugitsu?
[04:55:22] <kaiwai> nope, Sun's
[04:55:37] * masta is confused
[04:55:38] <richlowe> No, the OPL machines using SPARC64-VI's are fujitsus.
[04:55:40] <Triskelios> jafari: sure. I have had good experiences using glassfish (sun java ee server, nicer than tomcat)
[04:56:13] <masta> so ... T1 and T2 are Fugitsu designed and fab?
[04:56:16] <masta> or no?
[04:56:23] <richlowe> No.
[04:56:26] <masta> ty
[04:56:42] <nrubsig> masta: OPL and Niagara T1/T2 are different designs
[04:56:48] <jafari> im not talking about tomcat i wanted to know if weblogic is the setup as tomcat where you need a web server for it to wokr?
[04:57:10] <nrubsig> masta: OPL is high-end large SMP-like machines, giant machines which wheight enougth to crush an elephant
[04:57:24] <kaiwai> real man machines
[04:57:32] <nrubsig> master_o1_master: Niagara T1/T2 are small machines
[04:57:44] <nrubsig> erm
[04:57:57] <nrubsig> s/master_o1_master/masta/
[04:58:06] <kaiwai> @nrubsig: IIRC Rock will be SMP capable
[04:58:08] * nrubsig commits patch 800
[04:58:15] <masta> ok.. so there is a distinction for who fabs the niagra and OPL?
[04:58:25] <nrubsig> kaiwai: yeah, but Rock it at least a year away from now
[04:58:33] <nrubsig> masta: yes
[04:58:59] <masta> ty
[04:59:00] <kaiwai> @nrubsig: true; I guess you were referring to T1/T2 processors today rathe than what could be
[04:59:54] <nrubsig> masta: niagara T1/T2 is Sun design, produced by TI and sold by Sun and Fujitsu ; OPL is Fujitsu design, produced by Fujitsu and sold by Sun and Fujitsu
[05:06:52] *** sponix has joined #opensolaris
[05:09:11] <cmang> I'm screwed.
[05:09:36] <nrubsig> cmang: no
[05:09:40] <kaiwai> cmang: if I was in your room you would have/been already; so you're not screwed (yet)
[05:09:41] <nrubsig> cmang: You're doomed.
[05:09:51] <nrubsig> cmang: We're ALL doomed.
[05:09:56] <nrubsig> cmang: We will all DIE.
[05:10:02] <nrubsig> cmang: (some day)
[05:10:29] <kaiwai> when? I wanna know when :)
[05:10:40] <cmang> I suppose it might be more appropriate to say, "I wish I were screwed, it would make this problem easier to bear."
[05:10:51] <kaiwai> I'd acccelerate it but there is a risk I'll make a mistake and turn into a vegetable (or more than I am now)
[05:11:30] <bda> On a long enough timeline...
[05:11:31] <Shiv__> In the find command, is there a way to exclude a dir being listed by -print
[05:11:36] <Shiv__> find under GNU userland has a -path option
[05:12:22] <kaiwai> what is one looking for?
[05:12:35] <Shiv__> find . -path ./CVS -prune -o -name configure.in -print
[05:14:15] <Shiv__> It is part of a shell script used in a linux app. I want It to be modified so that it works on both SX and Linux.
[05:14:20] <nrubsig> Shiv__: try $ find ... | fgrep -v "/CVS" #
[05:15:20] <kaiwai> hmm B72 is up
[05:15:55] <Shiv__> nrubsig: fgrep !?
[05:16:55] <kaiwai> hmm
[05:17:19] <nrubsig> Shiv__: "fgrep" means "fast grep", e.g. no regular expressions, just plain charatcer string
[05:17:20] <kaiwai> I think he means ggrep
[05:17:27] <kaiwai> ah
[05:17:28] <nrubsig> kaiwai: no
[05:18:09] <nrubsig> kaiwai: "fgrep" is like POSIX's "grep -F"
[05:18:20] <kaiwai> ah
[05:18:25] <nrubsig>  /usr/bin/fgrep
[05:18:25] <kaiwai> an ugly short cut :(
[05:18:34] <nrubsig> erm
[05:18:46] <LeftWing> How is it ugly?
[05:18:49] <nrubsig> kaiwai: does Solaris support -F for /usr/bin/grep ?
[05:18:56] <kaiwai> almost as bad as the litany of crap sprawled through /usr/include and /usr/lib
[05:19:21] <Shiv__> nrubsig: fgrep searches inside a file
[05:19:30] <Shiv__> nrubsig: I believe the command I listed looks for configure.in in all the paths other than CVS.
[05:19:47] <kaiwai> @nrubsig: shouldn't the -F be inside the standard grep command? if so, use grep -F instead of fgrep
[05:19:53] <nrubsig> kaiwai: problem is that /usr/bin/grep is not even near anything defined by POSIX
[05:20:04] <nrubsig> Shiv__: fgrep reads from file or stdin
[05:20:42] <nrubsig> Shiv__: e.g. $ find . | fgrep -v "/CVS" # will print all files except those which start with "CVS"
[05:20:55] <nrubsig> s/files/directory elements/
[05:21:47] <Shiv__> nrubsig: OK.
[05:23:13] <Shiv__> When people say Solaris is posix compliant, but we see that some of the userland are not, what does this mean?
[05:23:31] <Shiv__> /usr/xpg4/bin/find and /usr/bin/find seem to behave the same
[05:23:32] *** BatonT has joined #opensolaris
[05:23:50] <Shiv__> As per man pages
[05:25:21] <nrubsig> Shiv__: except one tiny difference
[05:26:45] <Shiv__> nrubsig: The -print expression?
[05:29:57] * nrubsig checks
[05:30:06] <Shiv__> Is solaris posix compliant only for the system interfaces ?
[05:30:25] <Shiv__> Incase of shell & utilities, there seem to exist deviations.
[05:30:59] <nrubsig> Shiv__: see find(1), section "-name", search for string "xgp" which explains the difference
[05:32:10] <nrubsig> Shiv__: Solaris userland utilities are only POSIX conformant if you use PATH=/usr/xpg4/bin:/usr/bin # as PATH to use the /usr/xpg4/bin/ binaries if their /usr/bin/ counterparts are not POSIX conformant
[05:32:44] <nrubsig> Shiv__: and for SuS3 conformance you need PATH=/usr/xpg6/bin:/usr/xpg4/bin:/usr/bin
[05:34:04] <Shiv__> Am aware of that. Wanted to ask if the support was complete OR
[05:34:29] *** derchris has quit IRC
[05:34:30] <Shiv__> if there are some commands in /usr/bin that are not compliant and without a counterpart in xpg<n>/bin
[05:34:37] <nrubsig> Shiv__: AFAIK support or more or less complete.
[05:34:42] <Shiv__> Is there any test suite for checking posix compliance? Say for both system interfaces and shell utilities?
[05:34:44] <nrubsig> Shiv__: no.
[05:34:45] *** derchris has joined #opensolaris
[05:35:02] <nrubsig> Shiv__: yes... Sun has the POSIX+Sus3 test suites.
[05:35:17] <Shiv__> nrubsig: Accessible to the community members?
[05:35:31] <nrubsig> Shiv__: no clue... please ask in opensolaris-code@ ...
[05:35:40] <Shiv__> nrubsig: ok
[05:35:53] <Shiv__> nrubsig: thanks
[05:38:51] *** sponix has quit IRC
[05:39:42] <Tempt> boyd: PING
[05:40:54] <Shiv__> If any of the community members here are likely to be in India during Dec first week,
[05:41:00] <Shiv__> please have a look at http://www.genunix.org/distributions/belenix_site/?q=node/54
[05:42:09] <Shiv__> This is a 2day OpenSolaris related program planned for http://foss.in
[05:42:33] <Shiv__> by the user group here
[05:43:09] <Shiv__> foss.in is one of the largest community driven conferences that happens in this part of the country.
[05:48:12] <Shiv__> The program listed in the belenix website once approved will appear as a CfP (call for participation)
[05:49:43] *** karrotx has joined #opensolaris
[05:49:46] <Shiv__> And any of the community members may submit sessions on the lines of the program.
[05:51:38] <Shiv__> Incase any of you have any comments, feel free to send it across to me. Today is the last date to submit the final proposal to http://foss.in
[05:53:08] <Shiv__> Comments if any may be mailed to shivakumar.gn at gmail dot com
[05:54:52] <kaiwai> hmm, oss india
[05:55:21] <kaiwai> I get this idea of a guru in a meditative state coming up with code :P
[05:56:34] *** linux_user400354 has joined #opensolaris
[05:57:59] <Shiv__> !?
[06:01:28] <Shiv__> kaiwai: Some of the movies do a bad job of showing what this part of the country is about :)
[06:01:41] *** linux_user400354 has left #opensolaris
[06:02:20] *** linux_user400354 has joined #opensolaris
[06:02:22] <kaiwai> Shiv__: true, but it would be a cool site "coding for an innner peace" :)
[06:02:27] <kaiwai> *sight
[06:03:10] <jbk> i still need to take a vacation to someplace beyond north america at some point
[06:03:16] *** linux_user400354 has left #opensolaris
[06:03:17] <jbk> damn stingy corporate vacation policies
[06:04:57] <Shiv__> kaiwai: "Coding for an inner peace" : that would be going far beyond the freedoms espoused by FSF :P
[06:06:05] <Shiv__> kaiwai: The existing ones itself many find unacceptable :)
[06:11:00] <Shiv__> Back to my original post. If any of you are here. Please try to factor this in in your travel plans. Let us know at ug-bosug at opensolaris dot org OR respond to the CfP that will be announced in the coming days.....
[06:11:05] <jbk> so, here's a bit of a loaded question
[06:11:41] <jbk> i know amds still draw less power typically, but i believe the power savings support is better in nevada for intel
[06:11:57] <jbk> so anyone have any idea how the two end up at the end?
[06:14:45] <kaiwai> jbk: depends on the ACPI implementation though; my processor has speedstep support but isn't supported by Solaris because _PSS symbols aren't exported
[06:16:00] *** Shiv__ has quit IRC
[06:31:20] *** andyshack has joined #opensolaris
[06:31:35] *** kaiwai has quit IRC
[06:36:05] *** dlg has quit IRC
[06:38:32] <richlowe> nrubsig: no, the posix tests aren't accessible.
[06:40:27] *** sbahra has quit IRC
[06:40:50] *** dlg has joined #opensolaris
[06:52:03] *** masta has quit IRC
[06:55:41] <nrubsig> From http://en.wikipedia.org/wiki/F-22#Recent_developments : "... While attempting its first overseas deployment to the Kadena Air Base in Okinawa, Japan, on 11 February 2007, a group of six Raptors flying from Hickam AFB experienced multiple computer crashes coincident with their crossing of the 180th meridian of longitude (the International Date Line).  ..."
[06:55:51] <nrubsig> erm
[06:56:00] <nrubsig> How much does a F-22 cost these days ?
[06:56:05] <nrubsig> $100 million ?
[06:56:07] <nrubsig> $200 million ?
[06:57:13] *** tg has quit IRC
[06:57:19] *** tg has joined #opensolaris
[07:00:30] <jbk> is the driver different between sata & regular ide?
[07:00:43] <jbk> i.e. can i tell by looking at the hw path if it's in emulation mode?
[07:05:35] *** libkeiser has joined #opensolaris
[07:13:19] *** lisppaste3 has quit IRC
[07:13:31] *** comay has quit IRC
[07:24:06] <ofu> nv72 is out, someone please change the topic
[07:28:33] *** helpmehome has joined #opensolaris
[07:28:54] *** oninoshiko has joined #opensolaris
[07:29:54] *** jmcp has quit IRC
[07:30:17] *** jmcp has joined #opensolaris
[07:31:44] *** bengtf has quit IRC
[07:42:22] *** oninoshiko has quit IRC
[07:45:58] *** Wez has joined #opensolaris
[07:51:08] *** Wez has quit IRC
[07:52:26] *** lisppaste3 has joined #opensolaris
[07:53:37] *** bengtf_ has joined #opensolaris
[07:53:38] *** slowhog has joined #opensolaris
[07:53:39] *** bengtf_ is now known as bengtf
[07:53:48] *** slowhog has left #opensolaris
[08:08:38] *** Wez has joined #opensolaris
[08:09:25] *** Wez has quit IRC
[08:09:44] *** Wez has joined #opensolaris
[08:12:52] *** jamesd_ has joined #opensolaris
[08:14:17] *** Triskelios has quit IRC
[08:17:17] *** jamesd has quit IRC
[08:38:11] <WickedWi1ky> morning all
[08:55:25] *** jfndi has joined #opensolaris
[08:59:03] *** Fish- has joined #opensolaris
[08:59:31] *** trendoid has joined #opensolaris
[08:59:44] *** Fish- is now known as Fish
[09:00:15] *** mikefut has joined #opensolaris
[09:00:18] <Fish> hello
[09:02:32] <trendoid> Hello all. I'm going to install opensolaris today. As I'm a little short on disks I'd like to manage some data before the install. Is there a distro (live CDO I can boot that will let me manage a disk with ZFS and mount and ext3 file system? Benelix perhaps...
[09:03:14] <bda> Not sure if Belenix/Schillix have ext3 support.
[09:03:23] <CIA-16> dm199272: 6249663 memory leak in libnsl
[09:05:08] <WickedWi1ky> if I may believe google, belenix does
[09:06:27] <bda> Cool.
[09:06:38] <WickedWi1ky> Cool for them who use it, I guess :P
[09:07:50] <bda> Availability of tools is never a badness. :)
[09:08:03] <WickedWi1ky> that's true
[09:08:42] <WickedWi1ky> Well, I decided to cut my losses and buy a new ipod today. There is no way I am gonna get that thing returned
[09:08:49] <trendoid> I think I'll give Belenix a shot. :) This will save me some time.
[09:13:51] <WickedWi1ky> good luck trendoid :)
[09:20:19] <trochej> Ukhm.qemu is painfully slow
[09:20:30] <trochej> Still, works
[09:43:41] <moazamraja> WickedWi1ky: so...uh, you already bought an ipod toda?
[09:43:48] <moazamraja> today, even...
[09:44:07] *** migi has quit IRC
[09:44:27] <bda> WickedWi1ky: Rumors are they're going to announce new iPods soon. :\
[09:44:59] <bda> (Sept 5th)
[09:45:35] <moazamraja> yah, so the rumors say
[09:45:46] <moazamraja> a tad odd to be buyin an ipod only days before a possible rumored intro
[09:47:09] *** mikefut_ has joined #opensolaris
[09:47:09] *** mikefut has quit IRC
[09:47:53] *** mikefut_ has quit IRC
[09:51:08] <asyd> \_o<
[09:52:05] *** Doomshammer has joined #opensolaris
[09:52:08] <Doomshammer> hi
[09:53:55] <cmang> asyd: nice bird
[09:54:12] <Doomshammer> I've got a little problem which I have no clue why this happens... I just installed a fresh version of 5.11 snv_70 (Solaris Express). After installation it runs GRUB and when I select the Solaris kernel to load, it fails with this error: krtld: failed to open '/platform/i86pc/kernel/amd64/unix'
[09:54:16] <Doomshammer> any ideas?
[09:54:45] <Doomshammer> I checked with failsave, that /platform/i86pc/kernel/amd64/unix is existing
[09:55:08] <Doomshammer> in bootenv.rc the correct bootpath (at least from my point of view) is set
[10:00:11] *** dmarker has quit IRC
[10:00:23] *** dmarker has joined #opensolaris
[10:16:06] *** Mdx4 has joined #opensolaris
[10:18:48] *** Mdx4 has quit IRC
[10:23:50] *** ysss__ has quit IRC
[10:29:47] *** NikolaVeber has joined #opensolaris
[10:31:15] *** FrostCS has quit IRC
[10:31:30] *** frostcs_ has joined #opensolaris
[10:33:06] *** nostoi has joined #opensolaris
[10:45:37] *** nrubsig has quit IRC
[10:49:41] *** nivox has joined #opensolaris
[11:06:50] *** FrostCS has joined #opensolaris
[11:07:15] *** frostcs_ has quit IRC
[11:10:04] *** Gropi has joined #opensolaris
[11:17:08] *** pablomh_ has joined #OpenSolaris
[11:19:49] *** trendoid has quit IRC
[11:20:33] *** nrubsig has joined #opensolaris
[11:20:34] *** ChanServ sets mode: +o nrubsig
[11:21:01] <nrubsig> Stupid question: Who actually had the godforsaken idea for the /usr/gnu/ stuff ?
[11:21:25] *** medar_ has joined #opensolaris
[11:22:46] <Doomshammer> anyone an idea why Solaris doesn't boot (as described above)
[11:23:32] *** medar has quit IRC
[11:23:49] <trygvis> is the device in the grub menu correct?
[11:24:10] *** dthe-decier has joined #opensolaris
[11:24:10] *** the-decider has quit IRC
[11:24:10] <trygvis> does bootadm show the same device?
[11:24:11] *** dthe-decier has quit IRC
[11:25:55] <Doomshammer> i'm not very used to grub.. how can i display the device?
[11:26:49] <trygvis> press "e" on the OS line in the first screen
[11:27:36] <Doomshammer> k.. there is no device... just $kernel /platform/i86pc/kernel/$ISADIR/unix
[11:27:47] <Doomshammer> kernel$ sorry
[11:28:15] <Doomshammer> followed by module$ /platform/i86pc/$ISADIR/boot_archive
[11:31:14] <trygvis> strange, I just realized that my system is similar to your
[11:31:29] <Doomshammer> *g*
[11:31:50] <trygvis> except that it boot :)
[11:31:55] <Doomshammer> hehe
[11:32:29] <Doomshammer> wondering if it has s.th. to do with, that it's a VMWare virtl. machine
[11:32:52] <Doomshammer> but I have a Solaris 10 VMware as well, which boot/works w/o problems
[11:33:01] <oxygene> nrubsig: godforsaken? rather this than that newer idea of polluting /usr/bin
[11:35:14] *** sioraiocht has quit IRC
[11:35:35] *** sioraiocht has joined #opensolaris
[11:35:44] *** pablomh has quit IRC
[11:39:29] <trygvis> Doomshammer: check the bootpath with eeprom
[11:39:36] <trygvis> I have bootpath=/pci@0,0/pci-ide@7,1/ide@0/cmdk@0,0:a
[11:39:42] <nrubsig> oxygene: the problem is that some of that junk invaded /usr/bin
[11:40:46] *** nostoi has quit IRC
[11:41:05] <oxygene> nrubsig: that sucks indeed
[11:41:27] <nrubsig> oxygene: and read http://mail.opensolaris.org/pipermail/request-sponsor/2007-September/002748.html for more trouble ahead.
[11:41:28] <Doomshammer> trygvis: already did so.. bootenv.rc says: /pci@0,0/pci1000,30@10/sd@0,0:a
[11:41:49] <andyshack> will the future hold an option for me to be able to boot from zfs instead of a ufs partition on one of the drives in the pool ? or can that be done already ?
[11:42:18] <Doomshammer> andyshack: works already in the community release of solaris express
[11:42:33] <Doomshammer> http://www.opensolaris.org/os/community/zfs/boot/zfsboot-manual/
[11:42:36] <Doomshammer> for x86 only
[11:42:37] <g4lt-mordant> "ask again, future hazy" is what my magic 8-ball says
[11:42:42] <andyshack> cool thanks, ill look over that :)
[11:42:42] <oxygene> nrubsig: hmm.. sounds like "how about we just drop our stuff and hack compatibility into the gnu tools, then shut down the whole operation?"
[11:43:04] <g4lt-mordant> Doomshammer, okay, your magic 8-ball must be drugs
[11:43:35] <Doomshammer> g4lt-mordant: hmm?
[11:51:12] *** logic has joined #opensolaris
[11:53:11] *** jfndi has quit IRC
[11:57:56] *** Triskelios has joined #opensolaris
[11:58:52] <Triskelios> in case anyone cares, I got Banshee (the music player) working. it should be a bit nicer than rhythmbox...
[12:00:18] *** nivox has quit IRC
[12:03:01] *** nivox has joined #opensolaris
[12:05:44] *** logic_ has quit IRC
[12:09:58] <oxygene> Triskelios: I was about to try that eventually.. any big hacks necessary?
[12:11:42] <Triskelios> oxygene: not really, the major change was the usual dbus-sharp patch and some DllImport changes,  I had to make the avahi package build the mono bindings
[12:12:14] *** nrubsig has quit IRC
[12:12:16] <Triskelios> about to commit a fix for gettext stuff (it's still trying to use libintl)
[12:12:52] <oxygene> yes, found that, too - but I don't know mono enough for a portable solution (beyond an autotools macro that inserts the DllImport calls)
[12:13:00] <Doomshammer> ok, got the "krtld: failed to open '/platform/i86pc/kernel/amd64/unix'" problem fixed
[12:13:12] *** bengtf has quit IRC
[12:13:28] <oxygene> otoh, I know autotools, so adding that test should be easy :)
[12:14:03] <Triskelios> oxygene: yeah, that seems like a reasonable solution really
[12:14:14] *** nrubsig has joined #opensolaris
[12:14:15] *** ChanServ sets mode: +o nrubsig
[12:14:16] <oxygene> they do something like that for the database access already
[12:14:24] <Doomshammer> but now, when I start the kernel the screen wents black and the "caps lock" and "scroll lock" LEDs on my keyboard start glowing and the system seems to hang- any idea?
[12:14:47] * nrubsig curses flash and GNU sleep and wishes both die a horrible death at the hands of the grim ripper
[12:15:57] <Doomshammer> ha.. now the boot process goes on ;-) just took 2 minutes ;-)
[12:16:15] <Doomshammer> is that usual?
[12:16:21] <nrubsig> no
[12:16:34] <Triskelios> oxygene: SFEbanshee is the package if you want to try it
[12:16:46] <Doomshammer> ok, that settles me ;-)
[12:16:55] <nrubsig> Doomshammer: on my old laptop running Solaris within VMware B61 boots in less than 20secs
[12:17:28] <Doomshammer> nrubsig: yeah.. my Solaris 10 VMware boots in the same time... the b70 seems to act different
[12:17:36] <nrubsig> groan
[12:17:53] <Doomshammer> a) the ramdisk was broken so that I had to rebuild it in failsafe mode, and now it boots very slow
[12:18:22] <nrubsig> Doomshammer: is it possible that it's just re-importing the SMF database ?
[12:18:56] <Doomshammer> nop... after it was hanging for 2 minutes, it finally startet the kernel and then ist imported the smf
[12:19:30] <Doomshammer> well.. it's the first boot after installation - i'll give it a second chance when the system is ready
[12:21:31] <Triskelios> huh, banshee seems to lag less than rhythmbox
[12:27:35] *** Vanuatoo has quit IRC
[12:28:20] *** nrubsig has quit IRC
[12:34:26] *** bengtf_ has joined #opensolaris
[12:34:28] *** bengtf_ is now known as bengtf
[12:35:50] *** danny_j has quit IRC
[12:36:34] <oxygene> Triskelios: how is its memory use?
[12:36:58] <oxygene> (given that it uses gtk and pango, it must be horrible, exponentiated by the fact that mono gots its hands in it, too - so horrible^2)
[12:38:28] <Triskelios> RSS is on the high side, 88M
[12:39:16] <oxygene> bah
[12:39:16] <Triskelios> I'm also running tomboy and beagle so the mono overhead is a little bit less of a problem (but it's still about 100M!)
[12:40:06] <oxygene> was there a world wide lobotomy program for coders?
[12:41:07] <Triskelios> rhythmbox is 54M for comparism, sigh
[12:44:50] <Triskelios> yay, crashed gstreamer
[12:44:56] *** Vanuatoo has joined #opensolaris
[12:47:57] <kjetilho> I think many (most?) of these programs keep metainformation for all your music in memory
[12:48:08] <kjetilho> obviously that will use a fair bit of RAM ...
[12:48:09] *** mikefut has joined #opensolaris
[12:48:34] <Doomshammer> to bad, there is no port of foobar2000 for *NIX
[12:48:47] <kjetilho> Rhythmbox even used (uses?) an XML-"database"
[12:48:56] <kjetilho> the bloat makes my head hurt
[12:49:20] <oxygene> kjetilho: meta information.. 300 bytes for the tags, 4 bytes for length, 300 bytes for the filename.. ~600 bytes per entry.. 20000 entries ~ 12mb...
[12:49:21] <kjetilho> for every added album or song, it had to rewrite megabytes
[12:49:53] <kjetilho> oxygene: a bit optimistic, I think.  but it will vary with how "diligent" you are in your tagging
[12:50:02] <kjetilho> oh, no.
[12:50:08] <kjetilho> sorry, I was thinking albums, not songs.
[12:50:10] <LeftWing> kjetilho: iTunes uses an XML backend, too.
[12:50:27] <Doomshammer> well.. iTunes sucks like hell
[12:50:28] <oxygene> and so it must be right? *snicker*
[12:50:36] <Triskelios> Doomshammer: foobar2k looks completely unappealing
[12:50:45] <LeftWing> oxygene: I offer no judgement. ;P
[12:50:57] <oxygene> LeftWing: :)
[12:50:59] <LeftWing> I just know because I had occasion to parse it recently.
[12:51:02] <oxygene> Triskelios: it works - unlike many other players
[12:51:23] <Triskelios> oxygene: mpg123 "works"
[12:51:28] <kjetilho> I use Amarok.  it's sort of a hog, too, but at least it has a *lot* of functionality
[12:51:37] <oxygene> Triskelios: not when playing flacs with inline cue sheets
[12:51:58] * LeftWing uses iTunes on Mac OS for track management and XMMS on Solaris for most of his playing.
[12:52:03] <Doomshammer> Triskelios: I don't care about how the player looks. It's lightweight, it has all features I need, it's fast, uses about 24megs of ram with >10.000 songs in various playlists, there is a plugin to control/upload your iPod.. what do I need more? Fancy album pictures?
[12:52:43] <Triskelios> Doomshammer: an interface that doesn't drive me crazy
[12:52:46] <kjetilho> I thought album covers was silly until I tried it :)
[12:53:03] <Triskelios> amarok is probably best-of-breed right now, but I'm trying to stick with jds-friendly solutions
[12:53:23] <oxygene> Triskelios: what's amarok's memory footprint? (just curious)
[12:53:41] <Doomshammer> doesn't amarok depend on KDE?
[12:53:55] <Doomshammer> or better... on QT
[12:54:02] <kjetilho> oxygene: mine is 32 MiB RSS just now, 10k tracks.  200 MiB VSZ
[12:54:05] <Triskelios> oxygene: similar to rhythmbox, although it uses sqlite like banshee...
[12:54:13] <Triskelios> Doomshammer: amarok uses KDE and Qt, yes
[12:54:24] <oxygene> Doomshammer: Qt isn't an insane memory hog like gtk/pango
[12:54:33] <Doomshammer> well.. then the footprint of the player is probably low, but the QT overhead is massive
[12:54:39] <Triskelios> oxygene: yeah, the gtk situation is rather sad
[12:55:21] <Doomshammer> but i think with $player it's the same as with $os, $editor, $mua, $mta, $browser, to be continued...
[12:55:31] *** pablomh_ is now known as pablomh
[12:55:52] <LeftWing> I know the QT overhead in my MTA is quite high.
[12:56:00] <Triskelios> LeftWing: lol
[12:56:11] <LeftWing> I'll get my coat.
[12:56:43] <Doomshammer> LeftWing: i meant, that $player is a thing that causes flamewars, like $os, $editor....
[12:57:12] <LeftWing> Doomshammer: Well thats several orders of magnitude less amusing.
[12:57:13] <Triskelios> Doomshammer: sure, lots of different tastes
[12:57:48] <Doomshammer> indeed
[13:06:31] *** deather_ has joined #opensolaris
[13:20:04] <Doomshammer> I'll get a customer system next days.. a Dell 1950.. anyone experiences if Solaris 10 does work on this type of machine w/o any manual driver installation or stuff?
[13:20:49] *** JSRJ has joined #opensolaris
[13:23:01] *** deather has quit IRC
[13:49:10] *** hile__ has quit IRC
[14:03:10] *** jcsmith has joined #opensolaris
[14:05:26] *** medar_ has quit IRC
[14:05:43] *** medar has joined #opensolaris
[14:06:36] *** Pietro_S has joined #opensolaris
[14:12:05] *** master_of_master has joined #opensolaris
[14:28:24] *** master_o1_master has quit IRC
[14:30:25] <madhatter> re
[14:31:22] *** JSRJ has quit IRC
[14:32:24] <madhatter> Can somebody give me a hint on what I am doing wrong? I downloaded the rwt driver as I have a D-link DWL-510 in my box, but when I 'ifconfig rwt0 plumb' I get the message that there is no such device...
[14:38:46] <trygvis> does prtconf show that the driver is attached/loaded?
[14:38:51] <trygvis> does modinfo list it+
[14:39:10] <madhatter> I will look in a second. I just removed all other nics
[14:40:17] <madhatter> That no LED is blinkling on the card makes me wondering
[14:40:44] <kjetilho> not even the link indicator?
[14:41:14] <madhatter> No, but I don't know if it ever did
[14:42:44] *** vmlemon has joined #opensolaris
[14:43:28] <madhatter> Hm, in prtconf I don't see anything that looks like a nic. Any guess what I should search for?
[14:44:26] <trygvis> rwt
[14:44:36] <trygvis> try ifconfig -a plumb, it should try to load all drivers
[14:44:50] <trygvis> if not I guess you should make sure that the driver is actually there
[14:45:14] <Berny> morning
[14:45:54] *** kFuQ has quit IRC
[14:45:57] <Berny> anyone got a hint, why nwam won't pick up a dhcp ip after successfully associating to a wpa encrypted ap?
[14:46:51] <madhatter> trygvis: Did so, but modinfo | grep rwt does nothing
[14:48:14] <Berny> madhatter: did the install/pkgadd say something that the driver didn't attach or smething?
[14:48:42] <madhatter> Berny: No, but it asked me if I wanted to install it twice
[14:48:43] <kjetilho> sounds like a dead card (or card slot) to me if there's not LED at all.
[14:49:05] <madhatter> Berny: Can I reinstall it without any problems to make sure?
[14:49:27] <madhatter> kjetilho: I can switch slots. I removed a cable nic a minute ago
[14:49:41] *** kFuQ has joined #opensolaris
[14:50:53] <madhatter> I will try that.
[14:50:58] <madhatter> Don't go away ;)
[14:53:44] <madhatter> Hm, card does nothing
[14:54:19] <madhatter> But iirc it showed up in the list of hardware when booting the install-check cd a few days ago
[14:54:26] *** kFuQ has quit IRC
[14:55:59] *** kFuQ has joined #opensolaris
[14:56:00] <madhatter> Has somebody a tp-link with an atheros chip running under solaris? I have one somewhere
[14:59:48] <kohju> No. I have thinkpad t60 with atheros 11n wireless chips, too.
[15:00:27] <madhatter> kohju: I am just asking because that card is not listed on the hardware list on the ath driver's site
[15:02:51] <kohju> madhatter: here?  http://www.opensolaris.org/os/community/laptop/wireless/ath/
[15:04:10] <madhatter> kohju: Yes
[15:04:57] <madhatter> kohju: That's why I chose to get a d-link 510 because that was listed on the rwt driver's page
[15:06:10] *** kaiwai has joined #opensolaris
[15:06:19] <kaiwai> hmm
[15:06:20] <kohju> madhatter: I'd checked,too, just lastweek. I found an atheros 11n chip supported over NDIS wrapper on the linux, recently. but solaris's one is not supported.
[15:06:21] <kaiwai> B72 is released btw
[15:09:15] <madhatter> kohju: So, now I open prtconf, but I only have a pci168c,1051. That's not from that list
[15:10:00] <jmcp> kaiwai: you wanted to know whether I was around earlier today
[15:10:06] <madhatter> But that damn led on that nic is not blinking either
[15:10:28] <madhatter> Do i have to activate wifi nic support in the bios or what's happening here? ;)
[15:14:29] <jbk> morning
[15:14:41] *** tombhadAC has quit IRC
[15:16:21] <madhatter> ifconfig ath0 plumb did not say anything. So does that mean it works?
[15:16:53] <trygvis> try ifconfig ath0 and see
[15:18:02] <madhatter> trygvis: Hey, it shows up
[15:19:16] <madhatter> Now only wificonfig... that does not seem to accept anything. Always prints the help.
[15:22:48] *** estibi has joined #opensolaris
[15:23:02] <madhatter> wificonfig -i ath0 scan works fine, but when I want to createprofile it prints the help. Do I have to add some parameter?
[15:23:10] <madhatter> wificonfig -i ath0 createprofile essid=mywifi encryption=WEP wepkey1=12345
[15:23:24] <madhatter> That's from the wificonfig website
[15:27:02] <madhatter> Okay, that's wrong. There is no -i for that command
[15:30:34] *** jfndi has joined #opensolaris
[15:32:28] * jmcp finds the sleep button
[15:37:42] *** jmcp has quit IRC
[15:38:03] *** jmcp has joined #opensolaris
[15:43:53] <WickedWi1ky> re's
[15:50:14] *** sioraiocht has quit IRC
[15:51:16] *** Pietro_S has quit IRC
[15:53:02] *** bengtf has quit IRC
[15:55:55] *** andyshack2 has joined #opensolaris
[15:55:58] *** andyshack2 has left #opensolaris
[16:12:46] *** jfndi has quit IRC
[16:14:27] *** bengtf_ has joined #opensolaris
[16:14:29] *** bengtf_ is now known as bengtf
[16:21:19] *** _hsilva has joined #opensolaris
[16:24:33] *** SunTzuTech_AFK is now known as SunTzuTech
[16:25:40] *** tiger11 has joined #opensolaris
[16:27:41] *** Plaidrab has joined #opensolaris
[16:27:42] <tiger11> got some questions about iscsi targets. Any takers?
[16:28:42] *** jmcp has quit IRC
[16:29:06] *** jmcp has joined #opensolaris
[16:32:14] <kaiwai> tiger11: whats wrong?
[16:32:30] <FunkyELF> I was in here before and some people were saying that ZFS on a 32bit machine stinks.  Why is that?  Does it use more than 2Gb of memory?  Its not doing 64bit math or anything is it?
[16:33:28] <Stric> virtual memory range iirc
[16:33:29] <tiger11> I am getting different values for the USED property on the host and clients machines.....
[16:34:02] <kaiwai> FunkyELF: I guess it depends on what you use it for; it does use a lot of memory for speed purposes, and the focus I'd say for Sun is mostly in the direction of 64bit
[16:34:17] <flyingparchment> it uses a lot of virtual address space
[16:34:28] <flyingparchment> (not real memory, although it can use a lot of that too)
[16:34:44] <Tempt> Doesn't everyone have 64 bit machines now anyway?
[16:34:52] <tiger11> The target lists 198M used, the initiator lists 108k
[16:34:54] *** hsilva has quit IRC
[16:35:02] *** mrdeviant has joined #opensolaris
[16:35:09] <kaiwai> tiger11: checked the mailing lists for answers?
[16:35:44] <kaiwai> Tempt: alot of el-cheapo laptops use 32bit processors
[16:36:43] <tiger11> new to OpenSolaris, was told to jump in here. mailing lists off the main site?
[16:36:59] <kaiwai> http://www.opensolaris.org/jive/index.jspa?categoryID=1
[16:37:19] <kaiwai> its the weekend now, so alot of the sun people might not be here
[16:37:56] <tiger11> many thanks. will check there for a bit. may be back.
[16:38:08] <kaiwai> specifically, http://www.opensolaris.org/jive/category.jspa?categoryID=45
[16:38:19] <kaiwai> that is the 'storage' which will cover iscsi
[16:40:13] *** vmlemon has quit IRC
[16:41:29] *** vmlemon has joined #opensolaris
[16:44:30] *** Nishaway has joined #opensolaris
[16:46:29] *** Plaidrab has quit IRC
[16:54:05] <andyshack> can someone point me in the right direction re : serial terminal port that appears to be rj45 and the alom port ? ive been reading up about alom a bit. im just wanting to tell the machine to boot :) its a v240. do i need to use some rj45 to serial cable or is that done somehow over the network ?
[16:54:34] <Stric> is it a new or old v240?
[16:54:41] <andyshack> oh, i understand that the alom port is via the network.
[16:55:08] <Stric> old v240's had the network thingie disabled by default, so you had to go in through the serial-rj45 and enable it
[16:55:19] <andyshack> its not brand new. im actually not sure how old it is although it doesnt appear to have had much use.
[16:55:28] <Stric> newer v240's have network alom enabled by default, with dhcp
[16:55:44] <andyshack> stric is the serial rj45 litterally just a serial rj45 ?
[16:56:27] <andyshack> yeah i had a play with the alom via dhcp although couldnt get it up. i jsut assumed i was doing soemthing wrong.
[16:56:31] <Stric> I had to use an adapter to hook it up to a cyclades
[16:58:16] <Stric> http://www.sun.com/products-n-solutions/hardware/docs/html/819-2445-11/configure.html#51919
[16:58:35] <andyshack> thanks, i was just looking up the cyclades.
[16:59:06] <Stric> although that table is wrong.. pin 5 on rj45 isn't rxd but gnd
[16:59:15] <andyshack> yeah thats the doc i looked over, although it doesnt say 240, just 245 re dhcp
[16:59:26] <Tempt> andyshack: Same as a cisco.
[16:59:38] <Stric> newer ALOM versions are dhcp by default, older are not
[16:59:42] <andyshack> temp cheers. i got a v240!
[16:59:57] <andyshack> although its not going to get me far if i cant boot it :)
[17:00:56] <andyshack> ok, looks like ive got half an hour of finger burning to do tomorrow.
[17:01:21] <Stric> doesn't the cyclades usually come with one adapter?
[17:02:15] <Stric> or if you could quickly steal an adapter from some other machine/equip.. just to enable the network thing
[17:02:58] * delewis hates Flash
[17:03:08] <andyshack> i dont own a cyclades and my serial cross over is not the 9pin
[17:03:11] <delewis> the only time Firefox seems to crash lately is because of Flash.
[17:04:01] <Stric> andyshack: oh.. seems like I misunderstood you when you said you were looking up the cyclades..
[17:04:27] <d-s-d> hi
[17:04:33] <andyshack> oh yer. not Hooking up, Looking up as ive never seen one.
[17:04:40] <andyshack> its also a greek island
[17:05:22] <d-s-d> Is there a detailed list of hardware that is supported by sol10?
[17:05:59] <andyshack> d-s-d there is the sun hcl, google "sun hcl"
[17:06:41] *** onlinebacon has joined #opensolaris
[17:06:44] <onlinebacon> hey
[17:08:07] *** Sup3rkiddo has joined #opensolaris
[17:10:33] <d-s-d> There seems to be no driver for the Ali M5271 USB-Controller... is it possible that another driver supports this controller?
[17:11:04] <quasi> andyshack: why google when you can go bigadmin.com and select the hcl link
[17:11:45] <Sup3rkiddo> hello world
[17:11:46] *** e^ipi has quit IRC
[17:15:29] <andyshack> all i know of sun and usb is to make sure its d-(insert the generic protocol as im too tired to recall) and it will work. i think. maybe someone else might be able to help on that. is it not working or are you just checking first ?
[17:16:29] <Sup3rkiddo> i can still contribute code, compile the latest builds..etc without any problems on an amd athlon xp 2000+, right?..is there anyone with still a lower system config as mine?
[17:16:58] <andyshack> i had dramas with needing a better solution to having many servers and only 1 sun kbd. in the end i just got a larger extension for the serial cable to type GO
[17:17:25] <andyshack> oh, that was re usb.
[17:18:17] <oxygene> Sup3rkiddo: compiling all of ON will take half a day on that configuration, but it works.. where do you want to contribute?
[17:18:18] <g4lt-mordant> Sup3rkiddo, does a 2.4GHz laptop count as worse?
[17:18:21] <andyshack> my favourite dos box is an amd 1600
[17:20:17] <Sup3rkiddo> oxygene: oh no problem then...my openmoko builds take a full day and my favourite distro is gentoo...:P, i am just looking around the opensolaris site, but as a student the kernel fascinates me..
[17:20:30] * Sup3rkiddo is running nevada 57 now
[17:20:34] <g4lt-mordant> oxygene, TBH compilation is more bottleenecked by available memory than processor sped
[17:20:52] <Sup3rkiddo> oh and ih ave 768 - 32 mehs
[17:20:59] <Sup3rkiddo> *megs of RAM
[17:21:42] <g4lt-mordant> Sup3rkiddo, that's kinda ghettor for SXDE.  in fact, it's the bare minimum to install
[17:21:58] <oxygene> sxde *shrug*
[17:22:45] <onlinebacon> does anyone know where a good linux container howto is?
[17:22:47] <Sup3rkiddo> g4lt-mordant: yeah sxde complained something and i had to reboot to something else from GRUB...so you reckon my box is good as brick?
[17:22:48] <g4lt-mordant> oxygene, what else would you use to "contribute code, compile the latest builds..etc "
[17:23:15] <g4lt-mordant> Sup3rkiddo, not really, just going to be a bitch to compile on untile you up the memory
[17:23:44] <g4lt-mordant> onlinebacon, look at the brandz docs
[17:23:45] <oxygene> g4lt-mordant: sxce + studio
[17:23:53] <g4lt-mordant> oxygene, no shit?
[17:24:13] <oxygene> g4lt-mordant: ie. I ignore the installer fuckup and the marketing buzz
[17:24:18] <g4lt-mordant> SXCE + SUNWspro is what?
[17:24:25] <Sup3rkiddo> g4lt-mordant: compiling takes up that much memory..i always thought your processor plays a major rule in the speed than memort atleast...
[17:24:36] <g4lt-mordant> Sup3rkiddo, not IME
[17:24:38] <oxygene> g4lt-mordant: a sane configuration
[17:24:58] <oxygene> g4lt-mordant: sxde is that sane configuration with too many marketers running around screaming while anally raping the installer
[17:25:16] <Sup3rkiddo> oxygene: so i am on SXCE now...and downloading the latest SunStudio...so i am game after i set it up?
[17:25:46] <g4lt-mordant> please tell me you're downloading 11
[17:26:27] <Sup3rkiddo> g4lt-mordant: yup i am :D
[17:26:29] <oxygene> Sup3rkiddo: generally, yes - but nv57 might be too old (look for flag days on opensolaris.org)
[17:26:46] *** linux_user400354 has joined #opensolaris
[17:26:47] <d-s-d> andyshack: I'm just checking first.
[17:27:14] <d-s-d> And haven't yet installed solaris 10.
[17:27:44] <Sup3rkiddo> oxygene: oh...dang, i see the latest is 70, dont have bandwidth to download a dvd..i guess i will settle for a cd and possibly upgrade the current installation
[17:28:15] <oxygene> Sup3rkiddo: it _might_ work.. but don't be too surprised if the build fails somewhere
[17:28:49] <Sup3rkiddo> oxygene: thanks for the headsup...
[17:29:19] <d-s-d> The card seems to be OpenHCI compatible...
[17:34:34] *** derchris has quit IRC
[17:34:49] *** derchris has joined #opensolaris
[17:40:55] <d-s-d> ... so, i guess it's supported on sol10.
[17:49:45] *** mikefut has quit IRC
[17:52:32] *** onlinebacon has quit IRC
[17:53:04] *** lucky_luck has joined #opensolaris
[17:54:51] *** vmlemon has quit IRC
[17:55:29] *** tiger11 has left #opensolaris
[17:55:30] *** mrdeviant has quit IRC
[17:55:32] *** vmlemon has joined #opensolaris
[17:55:44] *** Nishaway has quit IRC
[17:56:42] *** Nishaway has joined #opensolaris
[17:56:52] *** Sup3rkiddo has left #opensolaris
[18:00:52] *** Nishaway has quit IRC
[18:01:03] *** lucky_luck has left #opensolaris
[18:01:57] *** lucky_luck has joined #opensolaris
[18:03:09] 
[18:03:16] <gcj> hi all, i installed CSW less, but it seems to corrupt my files, replacing the letter "y" with ^9. any ideas?
[18:03:31] <gcj> lucky_luck, doesn't come with libc?
[18:03:52] <gcj> do you have the install CD/DVD?
[18:03:53] <lucky_luck> not mine
[18:04:12] <lucky_luck> yes, thanks, I search on cd
[18:04:49] <lucky_luck> but... where can I download packages? in sun.freeware are few
[18:05:04] <gcj> not SUNWGlib?
[18:05:42] <g4lt-mordant> prolly SFWglibc actually
[18:06:09] <oxygene> glibc for solaris? what do you want _that_ for?
[18:06:24] *** onlinebacon has joined #opensolaris
[18:06:33] <lucky_luck> sorry, I try to install mysql 5.x on sparc
[18:06:54] <lucky_luck> but when i try to execute de daemon, the system don't found libc
[18:07:25] <gcj> what's the actual error message?
[18:07:32] <gcj> SUNWcsl
[18:07:45] <gcj> provides /usr/lib/libc.so and .so.1
[18:08:16] *** vmlemon has quit IRC
[18:08:23] <lucky_luck> sorry, this computer is at work
[18:08:40] <lucky_luck> tomorrow I take the exact error
[18:09:17] <gcj> i installed CSW less, but it seems to corrupt my files, replacing the letter "y" with ^9 among others. any ideas?
[18:09:22] *** vmlemon has joined #opensolaris
[18:09:27] <jamesd_> lucky_luck, its time to call in a consultant or reinstall...  nothing will run with a broken libc ...
[18:09:54] <gcj> lucky_luck, i think it might be compiled for a different version of libc that you don't have on your system, try a different package
[18:10:29] <lucky_luck> well, until this morning, my /var/spool/pkg was empty
[18:11:25] <lucky_luck> I have downloaded from sun.freeware and compiled... openssl, bison, and mysql
[18:11:41] <Auralis>  /var/spool/pkg is supposed to be empty
[18:11:48] <lucky_luck> jajajaaj ok
[18:11:51] <lucky_luck> good news
[18:12:28] * g4lt-mordant bets that lucky_luck was trying to install packages without -d
[18:12:54] <lucky_luck> nooo... i try with this flag xD
[18:13:15] *** jafari has quit IRC
[18:13:32] * g4lt-mordant wins!
[18:13:41] <lucky_luck> but I haven't wget and have had to install configure ssh to update files remotly
[18:13:54] <lucky_luck> not lynks
[18:14:37] <g4lt-mordant> that's why you got the /var/spool/pkg error.  -d /path/to/file.pkg installs a package from path/to/file.pkg
[18:15:23] *** piwi has joined #opensolaris
[18:15:33] <lucky_luck> I'll do that with 3 o 4 packages... until mysql can't start
[18:16:17] <lucky_luck> i did
[18:16:22] <lucky_luck> :)
[18:19:00] <lucky_luck> well, thanks for your suggestions
[18:19:12] <lucky_luck> and sorry for my innoperance
[18:19:24] <lucky_luck> I'll try search on cd.... bye
[18:22:39] *** lucky_luck has quit IRC
[18:25:11] <madhatter> re
[18:25:35] <madhatter> So, wificonfig worked. dhcp worked and I can reach the router.
[18:25:44] <madhatter> But not the world outside yet ;)
[18:25:57] <madhatter> btw is there nothing like ping on solaris?
[18:26:35] <Auralis>  /usr/sbin/ping
[18:26:51] <madhatter> Auralis: Oh, that's secure :)
[18:28:46] *** jmcp has quit IRC
[18:29:53] <madhatter> Seems there is no DNS
[18:30:00] <madhatter> I can ping google by ip
[18:30:11] *** jmcp has joined #opensolaris
[18:31:02] *** bengtf has quit IRC
[18:31:07] <trygvis> seems like your dhcp server didn't give you any dns hosts
[18:32:07] *** alanc_away has quit IRC
[18:32:13] <madhatter> trygvis: I don't think that he can do something like that
[18:32:21] *** Fladimir has joined #openSolaris
[18:32:39] <trygvis> mine do :)
[18:32:50] <madhatter> trygvis: Never needed that on OSX and in FreeBSD I had a lookup conf
[18:33:02] <madhatter> trygvis: So what's your essid then?
[18:33:16] <trygvis> essid?
[18:35:13] <madhatter> The name of your wifi network, so I will use that one then ;)
[18:35:34] <trygvis> :P
[18:35:45] <trygvis> I think you can just add them to /etc/resolv.conf
[18:36:29] *** vmlemon has quit IRC
[18:37:14] *** vmlemon has joined #opensolaris
[18:37:44] *** nightswim has quit IRC
[18:37:48] *** nightswim has joined #opensolaris
[18:38:39] <Fladimir>  I downloaded all *.iso of Solaris and started to install the first segment 1 using VMware it should gives me the GRUB loader to choose. But i see only command line GRUB>  and before that it says Minimal Bash-like line edditing is supported
[18:39:12] <Fladimir> would you give an advice how i can start the installation part or atleast to see the GRUP loader
[18:39:35] <jamesd_> Fladimir, make sure you have the iso created properly...  and make sure you have a recent version of vmware,  solaris 10 works fine in vmware, and it should handle booting automaticly..
[18:40:11] *** ichigo has joined #opensolaris
[18:41:28] 
[18:41:38] <Fladimir> .iso
[18:42:04] <madhatter> trygvis: Found some howto on the net. Seems I have to edit /etc/nsswitch.conf also. But now it works. Yay!
[18:42:31] *** gm152 has joined #opensolaris
[18:42:32] <jamesd_> check out the instructions on how to build them, you basicly  combine  abcde  into a single iso...  and vmware can work off that image.
[18:42:49] <Fladimir> well i thought so :) thanks
[18:46:29] <g4lt-mordant> lies, all lies.  you get a dialog "by your powers combined I am captain planet" then the theme song plays out of your speakers
[18:48:55] *** g4lt-mordant is now known as g4lt-paperclip
[18:49:07] <WickedWi1ky> ..
[18:49:15] <WickedWi1ky> so you're that office thingy always bugging me
[18:50:29] <g4lt-paperclip> WickedWi1ky, g4lt-mordant sysdef, sounds like you're trying to get rid of adware
[18:50:30] <g4lt-paperclip> sysdef lol
[18:50:30] <g4lt-paperclip> * g4lt-mordant sets mode -e g4lt-mordant
[18:50:30] <g4lt-paperclip> * You are now known as g4lt-paperclip
[18:50:30] <g4lt-paperclip> g4lt-paperclip can I help?
[18:51:01] <gcj> wow i didn't know the paperclip would talk to itself
[18:51:06] <gcj> is that a new feature?
[18:51:13] <WickedWi1ky> yea, comes with 2007
[18:51:16] <WickedWi1ky> AI
[18:51:25] <gcj> damn i have to go out and buy that right now
[18:51:35] <WickedWi1ky> it morphs into a dog when you wanna use the find function
[18:51:48] <gcj> does it pee all over your tree?
[18:52:05] <gcj> or just the root?
[18:52:08] <WickedWi1ky> that's only in the Ultimate edition
[18:52:32] <WickedWi1ky> which costs 379 euros, I saw in the store today
[18:53:35] *** bengtf_ has joined #opensolaris
[18:53:37] *** bengtf_ is now known as bengtf
[18:53:50] <WickedWi1ky> not that I could buy it if I wanted to... since my bank didnt send me my new credit card and debit card while the expire date of the current card ended yesterday
[18:54:15] <gcj> is this a good place to ask sysv programming questions about signal/sigset? or should i try #c?
[18:54:18] <WickedWi1ky> "We're sorry, looks like something went wrong in our administration"
[18:54:57] <WickedWi1ky> well, dunno, I cant help you.. I'm already jumping in the air when I can fopen() a file without a segvault
[18:56:17] <gcj> ok
[18:57:06] *** piwi has quit IRC
[19:02:35] *** jmcp has quit IRC
[19:03:18] *** jmcp has joined #opensolaris
[19:11:30] *** Fladimir has quit IRC
[19:15:17] *** g4lt-paperclip is now known as g4lt-vigor
[19:16:16] <g4lt-vigor> WickedWi1ky, it sounds like you want to go into edit mode, can I help?
[19:16:50] <WickedWi1ky> hahaha
[19:17:07] <WickedWi1ky> yes, I am looking for my Snatch DVD
[19:17:17] <WickedWi1ky> where did I put it oh mighty paper clip?
[19:17:29] <g4lt-vigor> that's only available with the reegistered version
[19:19:11] <WickedWi1ky> well damn
[19:19:29] <WickedWi1ky> vigor btw eh?
[19:19:39] * WickedWi1ky has a Cradle of Filth shirt saying "Vigor Mortis"
[19:20:46] <jamesd_> ah is that from the aj_z0 collection?
[19:21:43] * g4lt-vigor idly wonders if anyone ported vigor to solaris.  should be pretty trivial now with JDS
[19:23:38] <madhatter> re
[19:29:49] *** kFuQ has quit IRC
[19:32:19] *** mega has joined #opensolaris
[19:35:34] *** Pietro_S has joined #opensolaris
[19:43:58] *** g4lt-vigor is now known as g4lt-sb100
[19:50:04] *** mega has quit IRC
[19:50:29] *** onlinebacon has quit IRC
[19:51:08] *** mega has joined #opensolaris
[19:58:39] *** nachox has joined #opensolaris
[19:58:54] <madhatter> One more question: is there a way to 'rm -rf'? I get asked for every single file although I added the -f
[20:01:12] <Auralis>  /usr/bin/rm  your rm is aliased to rm -i
[20:02:10] <nachox> shouldnt -f override -i?
[20:02:15] <madhatter> Auralis: How does that happen? I am using my own .cshrc
[20:02:21] <Auralis> no
[20:03:30] <madhatter> I am. The Path and Prompt changed since I copied it into my $home
[20:03:57] <WickedWi1ky> the no wad for the "shouldnt -f override -i"
[20:04:00] <WickedWi1ky> I presume
[20:04:25] <Auralis> correct
[20:04:36] <madhatter> Oic
[20:04:48] <Auralis> unalias the alias in your .cshrc to get rid of it
[20:05:47] *** ashner has joined #opensolaris
[20:06:08] <madhatter> Auralis: Thanks a lot. That .cshrc is damn old and it never happened on my other systems. Solaris seems to be the only system that does not ignore the -i when -f is added.
[20:08:47] <WickedWi1ky> that's not entirely true
[20:09:35] *** JSRJ has joined #opensolaris
[20:10:25] <madhatter> WickedWi1ky: Why?
[20:13:19] *** oninoshiko has joined #opensolaris
[20:13:48] <WickedWi1ky> Cause RHEL3.4 and RHEL4.x have the same behaviour
[20:15:11] <WickedWi1ky> why did i spell my nake with a 1 even
[20:15:16] *** WickedWi1ky is now known as WickedWicky
[20:15:17] <WickedWicky> there
[20:16:12] <oninoshiko> i hate it when i misspell my nake
[20:16:56] <WickedWicky> it's better than misspelling rm -rf ./ and actually typing rm -rf /
[20:17:21] <oninoshiko> quite true
[20:17:39] <madhatter> WickedWi1ky: Okay, let me change my sentence: 'Solaris seems to be the only system I installed that .cshrc on that does not ignore the -i when -f is added.' :D
[20:18:25] <g4lt-sb100> WickedWicky, that won't work
[20:18:32] <WickedWicky> no?
[20:18:37] <g4lt-sb100> try it
[20:18:51] <Auralis> why should -f override -i in the first palce anyway?
[20:18:57] <WickedWicky> well damn
[20:18:59] <WickedWicky> rm: Cannot remove any directory in the path of the current working directory
[20:19:00] <Auralis> the manpage makes no such suggestion
[20:19:15] <WickedWicky> so you cant rm -rf / then
[20:19:22] <WickedWicky> that's quite a nice protection
[20:19:32] <nachox> Auralis, hmm, because -f is "force"?
[20:19:49] <Auralis> its force for overriding write proection of files
[20:20:30] <Auralis> and -i explicit states its overriding -f
[20:21:46] <g4lt-sb100> http://cvs.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/rm/rm.c#165
[20:22:39] <nachox> i was right, in linux -f overrides -i
[20:23:00] <WickedWicky> what distro?
[20:23:05] <trochej> Of cource it does.
[20:23:18] <nachox> any
[20:23:19] <trochej> At least RH, Fedora, Ubuntu and DEbian
[20:23:22] <trochej> And I bet any other
[20:23:29] <WickedWicky> I am pretty sure RHEL4 doesnt
[20:23:32] <trochej> Since they use the sabe base tools
[20:23:35] <WickedWicky> since I always have to unalias rm when I get annoyed
[20:23:44] <WickedWicky> (as root)
[20:24:01] <nachox> http://www.koders.com/c/fidA429F12EC2F128F6E8D853AF503886AD6A59FEDD.aspx check line 154
[20:24:45] *** Murmuria has joined #opensolaris
[20:24:52] *** Murmuria has left #opensolaris
[20:25:14] <WickedWicky> hm, interesting
[20:26:44] <WickedWicky> ya
[20:26:45] <WickedWicky> you're right
[20:26:56] <WickedWicky> I jsut checked it on a RHEL server at work
[20:27:02] <WickedWicky> I must be on crack during work
[20:27:21] <nachox> Auralis, xpg4's rm does ignore -i if you also add  -f
[20:28:15] <nachox> or rather it depends on the order of the arguments
[20:30:09] <WickedWicky> Note: The following compile may take a long time.
[20:30:09] *** Kush- has joined #opensolaris
[20:30:09] <WickedWicky> If it fails, re-run configure with --with-low-memory
[20:30:17] <WickedWicky> 7 minutes later...
[20:30:25] <WickedWicky> still waiting :P
[20:31:22] <oninoshiko> atleast it hasn't failed
[20:31:27] <WickedWicky> yet
[20:31:41] <WickedWicky> I think 64MB of RAM qualifies as low-memory
[20:32:00] <oninoshiko> *I* dont, but most everyone else does
[20:32:17] <WickedWicky> then again, everybody qualifies SS20 as a brick, I dont :P
[20:32:55] * Auralis still has a ss10 in use
[20:32:59] <oninoshiko> THATS THE SPIRIT
[20:33:46] <jamesd_> WickedWicky, a properly configured ss20 given the right tasks can really rock...   quad  150mhz cpus and 512MB of ram running solaris 8 or earlier would be one hell of a machine for  firewall or mail server  etc....
[20:33:47] *** mega has quit IRC
[20:34:50] <Pietro_S> WickedWicky: These days everyone would tell you that 64MB for gfx card is really too low ;-)
[20:34:53] <WickedWicky> jamesd_: yea, I am trying to build it up to those specs. But it seems the majority qualifies whatever can't run Solaris 10/Opensolaris as a brick
[20:34:56] <WickedWicky> which is quite silly IMO
[20:35:31] <WickedWicky> Pietro_S: znam
[20:35:38] <oninoshiko> given reasonably written code, it would make a fine web server (what moron though preforking was a good idea?)
[20:35:38] <jamesd_> WickedWicky, not me.. but if you are paying ebay prices for  ss20 parts you would be better off buying a u2 even it its a  2x200.. to run  solaris 8
[20:36:54] <WickedWicky> to be honest with you, I am running Solaris 9 plus the Gnome desktop that comes with it. And I am quite happy with it. Disabling the useless stuff like inetd and appserver of course
[20:37:16] <jamesd_> if you have access to a local hardware reclyler or a company that tosses old boxes fairly often... the ss20 can be configrued quite nicely for cheap
[20:37:25] *** cypromis has joined #opensolaris
[20:37:43] <WickedWicky> there is a hardware company here that sells refurbished (spelling) Sun parts
[20:38:04] * jamesd_ still likes  openlook on solaris  much better than gnome on old boxes.
[20:38:19] <Auralis> anything is better then gnome
[20:38:30] <jamesd_> true...
[20:38:46] <nachox> e16/17
[20:38:55] <WickedWicky> s.nl/index.php?&svc=search&lang=en&manlike=SUN
[20:39:01] <oninoshiko> GO E17!
[20:39:11] <WickedWicky> crap
[20:39:14] <WickedWicky> http://www.rls.nl/index.php?&svc=search&lang=en&manlike=SUN
[20:39:15] <WickedWicky> that
[20:39:27] <WickedWicky> ah yea, E17, I heard it runs neat.. and fluxbox
[20:39:27] <oninoshiko> too bad this version broke my theme
[20:40:27] <jamesd_> i like kde on modern hardware.. no reason not to use the 1ghz+ of cpu plus  1GB of ram to have a few bells and whistles the other wise the cpu and ram just mostly go to waste.
[20:40:43] <WickedWicky> I'm looking for bigger harddisk for the SS20 and CPUs.. the internal disks are SCSI-2 right?
[20:41:56] <jamesd_> they are sca, and limited to  10MB/s if i remember...  7200 rpm drives are better,, the faster create too much heat...  i would limit them to 9 or 18GB max.. for that reason.
[20:42:08] <oninoshiko> how about bells and whistles without being inefficient... or maybe getting only as much machine as you need for what you really do?
[20:43:24] <jamesd_> oninoshiko, i got my main desktop for free...  2.6 ghz opteron  2GB of ram and  my main solaris desktop free as well... 1.0Ghz  blade 1500...  why waste the cpu.. i'm not going to pay to downgrade my hardware.
[20:43:27] <WickedWicky> yea.. like people buying a dual core for word processing
[20:43:51] <oninoshiko> devote the processing to something practical
[20:44:00] <oninoshiko> like cureing cancer
[20:44:01] <WickedWicky> lke what? Seti?
[20:44:26] <nachox> WickedWicky, windows guys need it, they need 2 second proc for their antivirus and anti spyware
[20:44:34] <WickedWicky> hahaha
[20:44:42] <jamesd_> oninoshiko, i do  other stuff, the fact that my desktop has bells and whistles that mostly just sit idle even so, or i move an extra 100MB to swap doesn't effect me...
[20:44:50] <kjetilho> go green, don't waste energy
[20:44:59] *** JSRJ_ has joined #opensolaris
[20:45:08] <oninoshiko> SETI has made scientific contributions
[20:45:31] <jamesd_> kjetilho, my desktop also powers  windoze for my kids on there sunrays... is that green enough?  one desktop instead of 4?
[20:45:32] <WickedWicky> yeah it failed!
[20:45:35] <WickedWicky> \o/
[20:45:54] <oninoshiko> aww poor WickedWicky
[20:46:01] <WickedWicky> nah
[20:46:10] <kjetilho> jamesd_: Sunrays are nice, I set up a set for my triplet nephews
[20:46:24] <WickedWicky> I have to stay home anyway, I have the pager from work
[20:46:26] <kjetilho> very convenient that they can just pull the power plug or whatever and it still keeps state
[20:46:48] <jamesd_> yeap... "dad i need help!" okay bring your card...
[20:47:48] <nachox> "dad, i lost my card"
[20:47:59] <jamesd_> find it.
[20:48:25] *** mikefut has joined #opensolaris
[20:48:30] <nachox> sold it to a dealer for some we..?
[20:48:32] <kjetilho> it's no disaster to use the Sunray without a card, though.  it's just a little less convenient
[20:49:06] * Auralis still wishes for  sunray with good 3d performance
[20:49:38] <delewis> good texture performance, anyway.
[20:49:54] <delewis> programmed IO performance (CAD programs, etc.) should be adequate on a Sun Ray.
[20:50:11] <Auralis> opnegl blows chuncks on them
[20:50:46] *** Kitty has joined #opensolaris
[20:51:42] <delewis> texture performance is also the same reason video performance is bad.
[20:51:42] <delewis> Auralis: are you using textures?
[20:51:47] <Auralis> no
[20:52:37] <delewis> Sun Ray 1s only have Mach graphics, anyway.
[20:53:08] <delewis> I wonder if Sun is doing some sort of acceleration on the Sun Ray 2's and 1g's, which have Radeon 7000s.
[20:53:13] <Auralis> pushing geometrys with a few millions polys sucks ass with software render, and that is all you have on a ray
[20:53:55] *** yeo_coky has joined #opensolaris
[20:55:02] <yeo_coky> hello
[20:55:17] <yeo_coky> need some help please
[20:55:28] *** BatonT has quit IRC
[20:55:30] *** PerterB has quit IRC
[20:55:31] <Auralis> Please state the nature of the Solaris emergency.
[20:55:54] *** BatonT has joined #opensolaris
[20:55:54] *** PerterB has joined #opensolaris
[20:56:04] <delewis> Auralis: I submitted a 'hardware open documentation' request for MAJC and the XVR-1000... be interesting to see where that goes.
[20:56:12] <Auralis> yeah
[20:56:32] <yeo_coky> hello , i have ultrasparc III 64bit , is there any linux distro compatible on it
[20:56:34] <yeo_coky> ?
[20:56:37] <delewis> they're considering the XVR-500 at the moment, so maybe the XVR-1000 isn't too much of a stretch.
[20:56:42] <delewis> yeo_coky: wrong channel.
[20:56:52] <delewis> and why you would run Linux on UltraSPARC-III is questionable.
[20:56:53] <Auralis> the 1000 should be much simpler then the 500
[20:57:00] *** onlinebacon has joined #opensolaris
[20:57:13] <delewis> Auralis: I should think so, given Sun "should" own all of the IP involved.
[20:57:18] <coffman> yeo_coky: we cant support something cruel like that
[20:57:19] <Auralis> yeah
[20:57:26] <delewis> and even then, we should definitely be able to get MAJC technical documentation.
[20:57:27] *** rmorse has joined #opensolaris
[20:57:37] <yeo_coky> i'd prefer used linux or bsd than solaris as the platfoorm to keep my program running
[20:57:52] <delewis> yeo_coky: then throw the UltraSPARC gear out.
[20:57:59] <coffman> yeo_coky: then you should not come in a opensolaris channel ffs
[20:58:02] <g4lt-sb100> then join some #openlinux or something
[20:58:03] <delewis> you run Solaris on such hardware, not Linux.
[20:58:04] <nachox> yeo_coky, try openbsd then or buy x86 hardware
[20:58:11] <Auralis> sparc linux blows chunks, its badly brocken in to many places
[20:58:20] <yeo_coky> ic ic , thanks folks
[20:58:21] *** richlowe has left #opensolaris
[20:58:37] <delewis> and besides, if you're developing applications, you should seriously consider what Solaris has to offer.
[20:58:48] <g4lt-sb100> delewis, no, have them send the useless sparc gear to ME ;P
[20:58:48] <nachox> the t1 support in linux seems to be much better than earlier sparcs though
[20:58:55] *** andyshack has quit IRC
[20:59:16] <nachox> at least now there is a supported linux version
[20:59:23] <delewis> why you'd even think about running Linux on a T1 (where application performance is very important) is beyond me, given you've got zero observability on Linux.
[20:59:47] *** timeless has quit IRC
[20:59:52] <jamesd_> if you have 1GB of ram in the box, it will rock using solaris 10 or later, and have features that linux only dreams of.
[20:59:57] <delewis> good luck figuring out which thread on the system is causing all of the disk IO and thus, causing itself to sleep.
[21:00:01] <nachox> delewis, i guess people are used to a linux environment
[21:00:04] <g4lt-sb100> delewis, channelling brendang again? ;P
[21:00:21] <delewis> nachox: no, they're used to not having tools. That's bad. And that should change.
[21:00:32] <delewis> The tools are there on the right OS. Use them.
[21:01:37] <nachox> delewis, no, usually if the only difference between two tachnologies is the available tools in one and the other, transitions are quick, look somewhere else
[21:01:56] <delewis> nachox: there are no tools in Linux.
[21:02:09] <delewis> you've got vmstat, iostat, and oprofile. That's about it.
[21:02:14] <jamesd_> yeo_coky,  http://video.google.com/videoplay?docid=-8002801113289007228   check out that video... it will change the way you think about debugging problems....
[21:02:21] <delewis> *cough* the state of Solaris in 1995 *cough*
[21:02:39] <yeo_coky> solaris is closed to BSD , but i never found any suitable bsd/linux tht could run good on ultrasparch III/64
[21:02:52] <delewis> well before Solaris 8 anyway, which introduced microstate accounting.
[21:02:57] *** yeo_coky has left #opensolaris
[21:03:12] <jamesd_> yeo_coky, huh?  freebsd has  dtrace and ZFS ...    not too closed in my book
[21:03:24] <nachox> there is sar too and strace and some other stuff but that is besides the point, all the better reason to switch quickly but that has not happened, so there has to be another appealing reason to people prefering linux
[21:03:25] <delewis> jamesd_: DTrace is currently a problem.
[21:04:26] <delewis> nachox: most people that use Linux aren't in environments that depend on an optimal applications. Most people that use Linux are using it on the desktop, and thus spoiled with resources, rather than in a server room with a lack of resources.
[21:04:29] <jamesd_> delewis, they just need to  flex a little or create a work around... they are only complaining about linking to include files that have a cddl  header they will find a work around.
[21:04:36] <delewis> Hence why most Gnome applications are bloated pieces of shit.
[21:05:04] <delewis> and those people have never been faced with performance issues.
[21:05:17] <delewis> and therefore, they don't need such tools as DTrace (at least they think they don't)
[21:06:21] <jamesd_> its okay if it take 10 minutes to start  gnome  " i only restart it once a month, when i upgrade my kernel"  ...haha
[21:06:31] *** JSRJ has quit IRC
[21:06:55] <delewis> jamesd_: every applet seems to assume that there are plenty of resources for it to acquire whenever it wants to.
[21:07:18] <jamesd_> 3173 jamesd    20   0 1019m 368m  19m R  6.6 18.3 176:38.94 firefox-bin
[21:07:19] <delewis> mixer-applet2 does hundreds of ioctl's per second just refreshing the mixer settings.
[21:07:22] <jamesd_> ^^^ you think?
[21:08:02] <delewis> don't get me started on acroread, even, which exectutes getrusage syscalls hundreds of times a second to get its resource usage when *it* is consuming resources.
[21:08:31] <delewis> and that's the state of things in the Linux world, where it's assumed you're on a desktop with plenty of resources and just you.
[21:08:45] <jamesd_> delewis, did you watch the  google video, on what brought a  e4500 to its knees?  a stock market ticker, was doing  refreshes once a  microsecond.
[21:08:51] <delewis> and until those developers developing those applications change from that perspective, they're going to keep churning out applications that do silly things.
[21:09:03] <delewis> jamesd_: yes..
[21:09:29] <nachox> delewis, and i agree with you there, dont get me wrong. i just think people use what they are used to using even if they are sacrifising some performance
[21:10:19] <delewis> nachox: but they don't see it as sacrificing performance. They're on desktops (thus single-user interactive systems) spoiled with resources. Mixer-applet2 can refresh the mixer-settings 200-300 times/second, and that's fine for that workload.
[21:10:40] *** JSRJ_ has quit IRC
[21:10:47] <nachox> delewis, no, they see it as increasing their inmediate productivity
[21:10:52] <delewis> on the other hand, when you've got a Sun Ray Server with 150 users running 150 mixer-applet2s your syscall count is now 150*250 (avg of 200-300 times/second)
[21:11:13] <delewis> and that's just for a *single* applet.
[21:12:00] <delewis> nachox: no, it's not. mixer-applet2 doesn't need to refresh the mixer settings 200-300 times a second.
[21:12:10] <delewis> it's just poorly written applications. period.
[21:13:21] <delewis> the person developing mixer-applet2 doesn't even think about the consequences of what's he done in doing that, just because he doesn't notice it on his one-user desktop.
[21:13:25] <nachox> delewis, not what i meant, it increases their productivity because they know where to find the tools they learnt to use, while solaris and linux are similar things just dont work the same way for good or bad
[21:14:25] <delewis> nachox: I'm just saying that unless developers get off of Linux, because there's a lack of observability tools, poor applications will continue to be churned out by the dozens.
[21:14:36] <delewis> and that's bad for us, because we generally use Solaris in environments that are cramped by resources.
[21:15:41] <nachox> delewis, oh, come on, poor applications happen everywhere, look SMC, check solaris' installer
[21:15:54] <nachox> delewis, hell, check nightly
[21:16:15] <delewis> nachox: but we can see the problems with those applications. Linux folk cannot.
[21:17:00] <delewis> and the only reason we're finding such huge performance issues with Gnome is Solaris.
[21:17:54] <oninoshiko> actually i was already aware of huge performance issues with gnome. the only difference is that rather then tracking it, i just didnt use it.
[21:19:49] <nachox> we've only had the right observability tool (dtrace) for a few years and many developers are not used to them because there are other appealing reasons to use something else
[21:20:27] <nachox> for example, i can still see a lot of people asking for valgrind in solaris
[21:20:36] <jbk> uugh
[21:20:45] <delewis> hopefully, that'll change, too. You can answer a lot of simple questions that you were unable to answer before with a DTrace one-liner.
[21:21:02] <delewis> nachox: those people haven't been exposed to dbx and RTC.
[21:21:14] <delewis> if they had, they wouldn't be asking for Valgrind.
[21:21:15] <nachox> sun studio was not free till sun released opensolaris, no dbx without it
[21:21:36] <delewis> nachox: that's no longer the case, therefore, it makes little sense to be interested in Valgrind at this point.
[21:22:41] *** tek-ops has joined #opensolaris
[21:22:52] <delewis> and there's libumem.
[21:23:01] <nachox> delewis, no, the problem is that now it seems that solaris is not appealing enough for them to switch, they are used to their current tools
[21:23:30] <nachox> auto(crap) for example
[21:25:27] <jamesd_> i think they are just trying to get there enterprise customers off of  forte workshop ver 3.0 ....  yes there are people still using it, and they work for the same company i do.
[21:25:30] <delewis> in that case, maybe someone will have the balls to bring DTrace to Linux, which will most likely be Google or someone else.
[21:26:10] <delewis> I really don't care if they 'switch' to Solaris or not. I just want them to use observability tools to produce better applications.
[21:27:49] <delewis> and the users of Linux that are using their applications can't see what's going on -- it's the endless cycle of "i-can't-understand-my-system"
[21:28:00] <nachox> my inmediate concern is about IT in the next 3 years, people are not used to solaris and that's what i've been asked for the last few interviews
[21:28:40] <nachox> the new professionals will suck... bad
[21:28:52] <delewis> sure, they will.
[21:29:08] <jamesd_> i heard that recruiters consider only linux experience, no experience.
[21:29:18] <delewis> and they'll find out how bad they suck when they're seeing some application (they don't know which one) causing a disk (or disks) to be 99% busy and they can't locate it with their traditional tools and experience.
[21:29:36] <nachox> jamesd_, that's what happened in my case at EDS
[21:30:29] <delewis> jamesd_: good policy, IMO.
[21:31:02] <nachox> only that cant be good for solaris
[21:31:40] <jamesd_> delewis, yes, if all you know is one way to solve a problem...  its the best way... and you miss out a lot thinking that way.
[21:31:43] <nachox> a whole generation of it pros that do not have any solaris experience cant be good for business
[21:32:08] <flyingparchment> i think we should discuss solaris vs linux again, that's an awesome discussion
[21:32:17] <flyingparchment> oh wait, you already are!  great
[21:32:23] <delewis> and that's why those of us that do have Solaris experience will be able to find and retain jobs. :-)
[21:32:28] <nachox> hahaha
[21:33:10] <jamesd_> delewis, and they are going to be quite nice paying jobs.. because we are rare.
[21:33:11] <nachox> only for a while, till companies start replacing solaris with they their employees know
[21:33:26] <delewis> nachox: and then they'll run into the same issues I mentioned earlier.
[21:33:44] <delewis> 'we have a performance issue? What do we do? reboot the system. No way we can isolate the pid that's causing it.'
[21:33:59] <delewis> or 'strace every pid! Wait, noooooo!'
[21:34:05] <nachox> hehe
[21:34:09] <jamesd_> nachox, yeah.. not likely,  my company tried moving apps to linux  and then had to move them back 3 months later when  linux couldn't handle the apps.
[21:34:21] <nachox> that assumes there is no work being done in linux which is not true
[21:34:51] <delewis> nachox: not it just means that companies that are using Linux are using it in very specific places, and often not for core application services, like database servers.
[21:35:00] <vmlemon> I've been idle mostly, but couldn't someone like Google (with their current alliance) coerce Sun into dual-licensing DTrace?
[21:35:04] <jamesd_> nachox, yes we have a dozen  linux boxes that are doing fine... but we also have 130 solaris boxes that aren't moving anywhere...
[21:35:09] <delewis> more likely round-robined web servers, which can deal with failure.
[21:35:26] <jbk> or those that are are subjecting themselves to needless pain :)
[21:35:29] <delewis> vmlemon: more likely Google will ignore the non-existent legal issues.
[21:35:34] <jbk> vmlemon: why would they need to?
[21:35:35] <delewis> or keep the port strictly in-house.
[21:35:46] <jamesd_> vmlemon,  watch this  especially the end ,  http://video.google.com/videoplay?docid=-8002801113289007228
[21:35:54] <nachox> jamesd_, not in the short run but it will happen once linux evolves a bit more, look where linux was 5 years ago
[21:35:58] <delewis> and I think Bryan did an excellent job of informing Google there are no legal issues with porting DTrace to Linux.
[21:35:59] <vmlemon> I wouldn't be too surprised if they didn't have a Google version of DTrace
[21:36:06] <oninoshiko> DTrace is not going into Linux because of licensing issues (real or imagined) unless SUN releases it under GPLv2. get over it.
[21:36:09] <delewis> vmlemon: they don't, as of 2 weeks ago.
[21:36:14] <vmlemon> OK
[21:36:24] <delewis> oninoshiko: there are no legal issues.
[21:36:40] <jbk> i don't think there's a dispute about legal issues
[21:36:44] <vmlemon> This is going to be fun, video playback over NX, but here goes...
[21:36:45] <delewis> there can be no legal issues because the GPLv2 has never been tested in court. That's that. And nothing more out of me in this license discussion.
[21:36:52] <nachox> delewis, the key will be ISV support and they have google, ibm and oracle behind
[21:36:59] <jamesd_> nachox, look where solaris is now... and guess where its heading in 5 years....   the apps are not going to move to solaris, the machines can't handle it.. we run  6-10 instances of oracle on our servers....  linux cant handle it.
[21:37:00] <LeftWing> vmlemon: I watched it on my Sun Ray.
[21:37:01] <oninoshiko> delewis: and you are licensed to practice law in which state?
[21:37:04] <jbk> just those that would 'bless' it in the linux community won't integrate it
[21:37:30] <vmlemon> LeftWing: Gives me some reassurance, then ;)
[21:37:39] <delewis> oninoshiko: I don't need to be licensed to interpret law and apply it to myself. I'm a citizen that lives in the states and thus make it a point to know law and interpret it to the best of my ability as a citizen.
[21:37:46] *** tek-ops has quit IRC
[21:37:51] <delewis> I'm tired of this 'you're not a lawyer, you can't say anything about licensing'
[21:38:17] <WickedWicky> our law actually says every citizen is suposed to know the laws :P
[21:38:18] <delewis> that's bullshit.
[21:38:41] <LeftWing> WickedWicky: Perhaps more strictly that ignorance of the law is not a defence?
[21:38:56] <WickedWicky> yea,
[21:39:11] <WickedWicky> like, you cant defend yourself with saying "I didnt know it was against the law"
[21:39:14] <WickedWicky> things like that
[21:39:45] <nachox> jamesd_, and how many solaris professionals will there be in 5 years if the trend continues? i'm happy there are some things happening in china and japan about solaris in the university
[21:39:57] <oninoshiko> then, with all due respect, I think I will listen to what the lawyers say. Listening to some yahoo in IRC does not extactly qualify as due diligence.
[21:40:01] <jamesd_> WickedWicky, sure you can... just have to convince a jury of your peers (or 12 loosers that are not smart enough to avoid jury duty) that you didn't break the law
[21:40:02] <delewis> case in point -- if you don't want to hear license speak from those that aren't lawyers, then you shouldn't trust any free software program you use and its license, because most likely the copyright holder didn't 'consult' with a lawyer before that program was licensed under $WHATEVER_LICENSE_YOU_PREFER
[21:40:17] <WickedWicky> we dont have juries here :D
[21:40:20] <vmlemon> Watching the video now
[21:41:02] <LeftWing> vmlemon: I wouldn't concentrate on the licensing or the porting issues, though -- the bulk of the video itself is much more entertaining.
[21:41:05] <delewis> oninoshiko: you should listen to several noteworthy people who have stated there are no legal issues between DTrace integration in Linux.
[21:41:08] <vmlemon> OK
[21:41:54] <nachox> there is always a way to test, integrate something and wait for people to sue you  :P
[21:42:09] <delewis> I'd welcome that. At least we could see the GPL tested in court.
[21:42:19] <delewis> (which has yet to happen)
[21:42:34] <jamesd_> and if they really wanted to do it with least legal issues...  port the dtrace userland to solaris not a problem,  re-write  dtrace kernel parts in 2 pieces,  1 part cddl  which is legal  and 1 part gpl using  kprobes  and uprobes
[21:43:20] <vmlemon> Couldn't you make an Intelligent Installer (R), similar to classic UNIX installers that links precompiled kernel modules together, to try and work around the license issues?
[21:43:21] <nachox> jamesd_, ohh, come on, there is a NIH problem there :P
[21:43:59] *** Jondice has quit IRC
[21:44:01] <vmlemon> So, it would link the ZFS and DTrace stuff with the rest of the system, after linking other components together
[21:44:09] <vmlemon> If it makes sense
[21:44:11] <delewis> vmlemon: that's possible, but the question then becomes who would distribute those precompiled kernel modules.
[21:44:22] <vmlemon> Good Point
[21:44:27] <delewis> the GPL is broad that distributed the precompiled kernel modules on media might be seen as a violation of the GPL.
[21:44:34] <delewis> distributing, rather.
[21:44:45] <oninoshiko> and are kernel modules derivative works.
[21:44:56] <delewis> whatever 'derivative work' is.
[21:45:13] <vmlemon> I wouldn't call it a derivative work, if it doesn't contain code from the Kernel, but that's my view
[21:46:00] <delewis> vmlemon: the definition of 'derivative work' is also not clearly defined within the texts of the GPL (at least no towards any application that people use it for, nowadays) -- another ambiguity, and all the more reason I want the GPL tested in court to evaluate these ambiguities.
[21:46:11] <nachox> not so long ago i think some linux developers prevented a guy from distributing in his distribution the precompiled nvidia drivers
[21:46:16] <oninoshiko> every attorney I've head speak on the subject.
[21:46:27] <quasi> http://undeadly.org/cgi?action=article&sid=20070901041657
[21:46:32] <vmlemon> Alternatively, you could put glue into the kernel (probably Modified BSD licensed), and compile the modules separately on Solaris
[21:46:32] *** linux_user400354 has quit IRC
[21:46:42] <vmlemon> so that it would run them through an ABI layer
[21:46:46] *** linux_user400354 has joined #opensolaris
[21:47:14] <delewis> oninoshiko: does the definition of 'derivative work' in the GPL contain the word 'linking'? (it does not as the entire text of the GPL does not contain one instance of the word 'linking')
[21:47:17] <oninoshiko> "derivative work" does not need to be defined within the confines of the GPL, it is used within the context of copyright law (which there is a definition)
[21:47:28] <delewis> if it does not, then it pertains to no application that it's commonly used for, nowadays.
[21:47:46] *** rmorse has quit IRC
[21:47:46] <delewis> oninoshiko: does the definition of 'derivative work' in copyright law contain the word 'linking'
[21:47:50] <delewis> ?
[21:47:53] <vmlemon> Aah, what about the GPLv3, though?
[21:47:58] <delewis> if not, then that's still an ambiguity.
[21:48:07] <delewis> and something to be evaluated by the court system.
[21:48:14] <delewis> vmlemon: plenty of instances.
[21:48:21] <delewis> (most of which are in the right manner, though)
[21:48:53] <vmlemon> Why do they do such stupid things in the name of "Freedom"?
[21:49:10] <delewis> because they aren't lawyers.
[21:49:42] <delewis> and the lawyer that was consulted for to author the GPLv2 (Eben Moglen) has said on many occasions his interpretation of the GPL and how it disagrees with the current intrepretation.
[21:50:58] <delewis> perhaps we need to pick up a habit of the ACLU and create a 'test case' for the GPL. :-)
[21:51:03] <delewis> *cough* DTrace *cough*
[21:51:11] <vmlemon> If they didn't design such ambiguity into it
[21:51:12] <vmlemon> ...
[21:51:45] <nachox> delewis, didnt you know? linux devels just found a way, they just strip the license and paste gpl there instead ;)
[21:52:23] <vmlemon> They stole the code? ;)
[21:52:40] <g4lt-sb100> vmlemon, deadly.org has the whole story
[21:52:46] <nachox> no, they liberated it! :P
[21:52:56] <vmlemon> Wasn't there a DTrace knockoff being built for Linux?
[21:53:12] <g4lt-sb100> that would be systemtap, and jamesd was part of it
[21:53:23] <delewis> nachox: even if that hadn't already been fixed, I doubt the application of 'derivative work' in the GPL would need to be tested.
[21:53:42] <delewis> that's a separate copyright issue, I would think.
[21:54:00] <delewis> porting DTrace to Linux would be an ideal test-case.
[21:54:29] *** Jondice has joined #opensolaris
[21:54:54] <delewis> provided you could actually get somebody to sue over it, which hasn't happened yet (despite a number of GPL 'violations')
[21:55:03] <nachox> hmm, no, it wouldnt, it would be too hard to integrate
[21:55:06] <delewis> in the MPlayer project, alone, we've had a number of violations.
[21:55:20] <delewis> yet, no one sues, and I've always wondered why there are violations to begin with.
[21:55:37] <delewis> surely these companies that are violating are getting some sort of reassurance from their legal consultation that there is no issue.
[21:55:38] *** ajocrudo has joined #opensolaris
[21:55:45] <vmlemon> I'd asked in another channel, if by using proprietary modules with the kernel (as I do daily for VMware Workstation), am I violating some top secret, unwritten rule?
[21:56:03] <delewis> vmlemon: no, because you aren't distributing.
[21:56:04] <vmlemon> (Probably, thou' shall not mix open and closed code)
[21:56:09] <delewis> the GPL only pertains to you, if you're distributing.
[21:56:18] <nachox> delewis, BUT i think the test case is already available, check what is happening with vmware
[21:56:19] <delewis> in fact, you only 'accept' the text of the GPL once you distribute.
[21:56:32] <delewis> VMware is distributing, because of ESX.
[21:56:37] *** ajocrudo has quit IRC
[21:56:46] <delewis> nachox: yes, I had forgotten about that. Be interesting to see how that plays out.
[21:57:02] <delewis> VMware will definitely go to court if it means they can't sell ESX "legally" anymore.
[21:57:33] <vmlemon> Isn't it their "bread and butter"/premium product as so to speak?
[21:57:37] <delewis> just 'using' GPL code; however, does not mean that it applies to you, which means Google could do an in-house DTrace port and not have any legal issues whatesoever.
[21:57:40] <delewis> vmlemon: yes.
[21:57:53] <nachox> depends, is going to court more expensive than porting it to opensolaris/bsd?
[21:57:58] <delewis> of course, for VMware to go to court, they have to be sued first :-)
[21:58:08] <bda> delewis: Pretty sure Cantrill says just that in his DTrace Google TechTalk.
[21:58:20] <vmlemon> The fun thing is, they just went through an IPO/yet another ownership change
[21:58:23] <delewis> and I'm sure VMware has already gotten some reassurance from its legal consultants that there is no issue with the GPL and ESX.
[21:58:34] <delewis> which gives me credence that the bulk of the legal community thinks the GPL is a joke.
[21:58:44] <delewis> (as there are so many violations)
[21:58:45] <g4lt-sb100> !seen useful license
[21:59:05] <nachox> delewis, there is never a problem till the first court order says so
[21:59:06] <vmlemon> I use Linux heavilly, but I question the intentions of the FSF and their GPL almost daily
[21:59:18] * g4lt-sb100 smacks Drone: WAKE UP
[21:59:44] <delewis> nachox: right, but it's the job of a legal consultant to (a) look at the current law for violations and (b) look at the current law to see how a judge deciding on something might apply it to you
[21:59:45] <g4lt-sb100> lazy bot!
[21:59:50] <WickedWicky> they should make Milka chocolate opensource
[22:00:08] <vmlemon> Open Sauce Ketchup, anyone? ;)
[22:00:08] <delewis> in this case the majority of violators that have legal consultants see no issues with (b)
[22:00:14] <nachox> they should free coca-cola under the gpl :)
[22:00:16] <g4lt-sb100> WickedWicky, it is.  you can make your own trivially
[22:00:33] <WickedWicky> well I meant more like, make the package openable without cutting your fingers
[22:00:44] <Drone> I've never seen useful license talk in #opensolaris.
[22:00:54] <g4lt-sb100> funally
[22:00:56] <nachox> lol
[22:00:59] <delewis> hah. :-)
[22:01:00] <vmlemon> Just use the MIT License, and problem solved ;)
[22:01:00] <WickedWicky> I'm sure not helping with my Milka talk eh
[22:02:01] <WickedWicky> Milka sure tastes better than Linux and Solaris though
[22:02:30] <vmlemon> If only it had asprintf() ;)
[22:02:35] <WickedWicky> haha
[22:02:37] <vmlemon> But that's another matter
[22:03:17] <vmlemon> Hah, he's going on about the "trash applet"
[22:03:42] <WickedWicky> I wish Fido would get a clue or two about the cool offers they make me while I pointed out 50.000 times already I am not in Canada anymore
[22:06:19] <g4lt-sb100> well, given that linux is a detergent, yeah
[22:06:31] <vmlemon> So is vim and AJAX ;)
[22:06:35] <WickedWicky> hey
[22:06:47] <WickedWicky> my fav. soccer team is named Ajax
[22:06:49] <WickedWicky> watch it
[22:06:50] <vmlemon> Nothing wrong with vim, though
[22:07:15] <g4lt-sb100> other than its wussified vi, not at all ;P
[22:07:32] <WickedWicky> you're a paper clip...
[22:07:51] <WickedWicky> or a sound blaster at the moment
[22:07:54] *** linux_user400354 has quit IRC
[22:08:14] <g4lt-sb100> no, sb100 in htis case is quite relevant
[22:08:22] <g4lt-sb100> SunOS andrea-doria 5.11 snv_55b sun4u sparc SUNW,Sun-Blade-100
[22:08:41] <delewis> hah. andrea-doria.
[22:08:41] <WickedWicky> ah! I thought the Kenwood speakers
[22:08:42] <WickedWicky> :P
[22:08:48] <delewis> I would ask why 'andrea-doria'
[22:08:49] <nachox> hmm, did anyone hever had problems with the screws in a dell laptop?
[22:09:09] <WickedWicky> No, I say "screw you" quite convinient at Dell's stuff
[22:09:17] <nachox> the head of one is brokem and i'd like to remove it
[22:09:30] <g4lt-sb100> delewis, all my computers anre named after spectacular shipwrecks
[22:09:39] <delewis> nice.
[22:10:03] <vmlemon> Japanese prefectures, here
[22:10:19] <delewis> star names in the constellation orion, here.
[22:10:20] <vmlemon> at least for Virtual Machines
[22:10:21] *** cypromis_ has joined #opensolaris
[22:10:22] <WickedWicky> I've got Lilith, Eva, Elisabeth and Pebbles
[22:10:40] <WickedWicky> Pebbles being the rare duck in the gang
[22:10:43] <delewis> betellatrix, betelgeuse, saiph, hatsya, etc.
[22:11:05] <WickedWicky> I think I am gonna call my U5 "Medusa"
[22:11:09] * g4lt-sb100 has a practically infinite store of names
[22:11:27] <WickedWicky> I'll keep it with goddesses/daemons/vampires
[22:11:31] <delewis> so do I. I just go over to the next constellation.
[22:11:36] <WickedWicky> should be enough names for the coming 340 years
[22:11:39] <delewis> although, my domain name is orion.const.
[22:12:03] <nachox> i'm pissed :( i needed to remove the screw to change the disk drive
[22:12:33] <WickedWicky> drill it out
[22:12:53] <nachox> i dont think i have a suitable drill
[22:13:51] <WickedWicky> me and drills.. the memories.. *sigh*
[22:14:00] <oninoshiko> they make some special drill bit thingie that is suppose to grab the screw or something... then again i trust the people on TV as far as i can throw them
[22:14:07] <WickedWicky> one time I tried to make an 80cc out of a 50cc Puch
[22:14:24] <WickedWicky> needless to say it worked but the cylinder crashed into the block
[22:15:19] <nachox> and this screw is small, way too small, i'm affraid to break the case of the laptop
[22:15:49] <WickedWicky> I'm afraid you're sorted out of luck
[22:16:08] <oninoshiko> screwed, so to speak
[22:16:18] <WickedWicky> oh look it! Snakes on a plane is on tv today
[22:16:37] <nachox> lol
[22:16:38] <WickedWicky> that'll be a nice time killer while compiling and patching stuff
[22:17:19] *** Gekkko[PDA] has quit IRC
[22:17:51] <WickedWicky> I heard the movie is worse than I can possibly imagine, so I have some expectations
[22:18:51] <oninoshiko> you'll be disappoined, i have faith in your imagaination
[22:19:03] <WickedWicky> it's really that bad?
[22:20:03] *** jmcp has quit IRC
[22:20:35] *** jmcp has joined #opensolaris
[22:21:01] <oninoshiko> it is what it is.... "bad" is a matter of expectations... i mean Evil Dead 3 wasn't exactly a great movie, but it's a cult classic.
[22:21:23] <WickedWicky> so true
[22:22:02] <WickedWicky> Dusk till dawn same thing, IMO
[22:22:12] <WickedWicky> I loved it though
[22:22:27] <oninoshiko> but if your expecting it to be worse then you expect, and it doesn't meet those expectations, then you might be disappointed because its better then you thought it would be
[22:22:35] <WickedWicky> hahaha
[22:22:51] <WickedWicky> you're right. I'll try to watch the movie as blank as possible
[22:23:05] <oninoshiko> one of these days I'm gonna get arrested for making someone's head explode
[22:23:53] <oninoshiko> hrm... i have a bad QLA4010
[22:24:20] <oninoshiko> stupid "last thing i could possibly check
[22:26:03] <oninoshiko> ok, im hungry... i wonder of that chilli/beer eating/drinking thing is today
[22:26:47] *** cypromis has quit IRC
[22:27:11] <oninoshiko> not till the 22nd :(
[22:27:30] <oninoshiko> oh well... chilli and beer for the day after my birthday
[22:28:45] <vmlemon> Strange question, but can DTrace be used as a reverse engineering tool?
[22:29:37] <coffman> i bet it is forbiden now in germany :\
[22:30:12] <quasi> vmlemon: partially
[22:30:28] *** mikefut has quit IRC
[22:31:06] *** nostoi has joined #opensolaris
[22:31:11] *** perlmonk has joined #opensolaris
[22:35:19] *** cypromis_ has quit IRC
[22:36:04] *** nachox has quit IRC
[22:38:37] *** oninoshiko has quit IRC
[22:39:37] *** Lusitanian has joined #opensolaris
[22:42:32] *** Plaidrab has joined #opensolaris
[22:44:31] *** Gropi has quit IRC
[22:46:02] *** charlesnw has joined #opensolaris
[22:47:44] *** jmcp has quit IRC
[22:48:04] *** jmcp has joined #opensolaris
[22:56:50] <vmlemon> Any ideas as to whether or not it's possible to sniff USB on Solaris? Google isn't helping me, at the moment :|
[22:58:58] <quasi> vmlemon: no idea - perhaps you can find some hints in http://blogs.sun.com/colin/entry/usb_documents_for_solaris_users
[22:59:04] <vmlemon> Thanks
[22:59:48] *** Mdx4 has joined #opensolaris
[23:01:19] <quasi> http://www.sun.com/io_technologies/usb/USB-Faq.html (if it wasn't on that page)
[23:01:31] *** JSRJ has joined #opensolaris
[23:01:45] <vmlemon> I'm just looking at that page at the moment
[23:02:00] *** TFKyle has joined #opensolaris
[23:04:16] *** tokyoeye has joined #opensolaris
[23:08:07] *** Kush- has quit IRC
[23:15:00] *** jmcp has quit IRC
[23:15:26] *** jmcp has joined #opensolaris
[23:17:54] <WickedWicky> okay so.. this movie is pretty bad.. :s
[23:22:59] <holcomb> which one
[23:24:35] <vmlemon> Interesting view about DTrace on Linux
[23:24:59] <WickedWicky> Snakes on a plane
[23:25:06] <holcomb> haha
[23:25:30] <WickedWicky> I thought I'd watch some tv during patching/compiling
[23:30:07] *** henriknj has joined #opensolaris
[23:30:35] <vmlemon> Nice to see that people are working DTrace and Perl integration, albeit at a glacial pace (probably due to the complexity involved)
[23:30:41] <vmlemon> *working on
[23:32:50] *** nrubsig has joined #opensolaris
[23:32:51] *** ChanServ sets mode: +o nrubsig
[23:33:28] * nrubsig wonders whether wesolows is around
[23:34:10] *** rasputnik has joined #opensolaris
[23:35:11] *** rasputnik has quit IRC
[23:37:46] *** Plaidrab has quit IRC
[23:39:31] *** mega has joined #opensolaris
[23:41:45] <nrubsig> gdamore: ping!
[23:51:09] <nrubsig> Anyone awake and willing to check the bug triage queue for new bugs ?
[23:51:36] *** FunkyELF has quit IRC
[23:53:33] * nrubsig thinks about writing an ARC case "Depreciate weekends, engineers shall work 24h/7d" ...
[23:54:45] *** gdamore has quit IRC
[23:59:03] *** Mdx4 has quit IRC





top