[00:00:06] <ocr> tomww: when i executed /sbin/bootadm update-archive -R /a i didn|t get an error
[00:00:18] <codestr0m> NP had problem.. what do you mean?
[00:00:20] <e^ipi> it is travelling salesman
[00:00:34] <ocr> (and echo $? == 0)
[00:00:54] <e^ipi> http://en.wikipedia.org/wiki/NP-hard
[00:01:06] <tomww> ocr: if you want to be safe, you could touch a file and then rebuild.    touch /a/rpool/etc/patch_to_inst
[00:01:43] <tomww> ocr: bootadm detects the changes timestamp and will rebuild the archive then (should take some time then)
[00:01:48] <wonko2> tomww: i was able to get the script working flawlessly in 'env -i', but SMF still doesn't want to work
[00:02:03] <codestr0m> ok. so the next package manager should be named om (oracle machine ) right?
[00:02:09] <ocr> tomww: and exec /sbin/bootadm update-archive -R /a again after touching the file?
[00:02:33] <tomww> wonko2: the script runs under the user-id mlmonkey without errors?
[00:02:35] <e^ipi> codestr0m: no, because it likely will still be run on a turing machine;
[00:02:43] <e^ipi> also, Oracle would probably sue you for it
[00:02:50] <tomww> ocr: yes, run the same command after the touching
[00:02:51] <codestr0m> hehehehe
[00:02:55] *** ehtom has joined #opensolaris
[00:03:06] <ocr> snap, timed out from elom :>
[00:03:09] <ocr> sec
[00:03:34] <codestr0m> well. I'm not sure how many people would mistake oracle machine which builds/manages packages for a database, but still.. you'd get sued for sure :P
[00:04:46] <ehtom> whats oracle machine?
[00:05:06] <ocr> tomww: and init 6 after reboot?
[00:05:19] <ocr> err after rebuild ;)
[00:05:23] <e^ipi> http://en.wikipedia.org/wiki/Oracle_machine
[00:05:26] <e^ipi> google.
[00:05:26] <oxygene> basically, most package managers are like cvs: you have independent version graphs for each file/package, and when you update the entire set (tree / system), you might not get a consistent state
[00:05:29] <e^ipi> it is wonderful
[00:05:39] <tomww> ocr: yes. if it still complains with the ufs error, then the ramdisk should not be faulty
[00:05:40] <wonko2> tomww: yes
[00:06:03] <ocr> tomww: awesome, will reboot once its completed updating :))
[00:06:10] <ocr> tomww: you rock! ;)
[00:06:33] <codestr0m> oxygene: , but you're talking about things which might happen and can be avoided
[00:06:44] <codestr0m> RH has an exhaustive process of automated QA
[00:07:11] <codestr0m> I think it mostly comes down to tools and policy
[00:07:14] <tomww> ocr: hey, that worked?
[00:07:20] <tomww> ocr: congrats!
[00:07:26] <codestr0m> give the right people the right tools and it'll come out well
[00:07:28] <ehtom> that seems somewhat useless
[00:07:32] <ehtom> since you need an oracle
[00:07:36] <ehtom> for an oracle machine...
[00:07:53] <oxygene> codestr0m: the right tool would not allow inconsistent states without lots of flags to override sanity checking
[00:07:56] <ocr> i celebrated early :>
[00:08:05] <oxygene> codestr0m: as is, most tools have no sanity checking to override in the first place
[00:08:13] <e^ipi> ehtom: it is a construct in complexity theory
[00:08:33] <ocr> panic[cpu0]/thread=ffffffffffbc5cca0: cannot mount root path /ramdisk:a
[00:09:39] <ehtom> i see
[00:10:27] <ocr> and <off> genunix:rootconf+122 () <off> genunix:vfs_mountroot+65 () \ genunix:main+dd () \ unix:_locore_start+80 ()
[00:10:37] <codestr0m> anyway.. I'd just be happy if the damn packages are in the tree and they don't segfault :P
[00:10:38] <ocr> where <off> is some memory offset
[00:10:59] <ocr> it also states: NOTICE: mount: not a UFS magic number (0x0)
[00:11:09] <codestr0m> I only need about 1000 packages.. 20% of which don't have mainstream usage :P
[00:12:29] <e^ipi> codestr0m: if you can solve a package dependency graph in polynomial time, you should probably call the turing institute
[00:12:43] <e^ipi> they'd probably like to get that problem sorted out
[00:13:11] <codestr0m> e^ipi: you're completely exaggerating the problem..
[00:13:17] <e^ipi> i'm not actually
[00:13:22] <codestr0m> yes you are
[00:13:29] <oxygene> .o0O am not
[00:13:41] <codestr0m> even worst case you can recursive build the dep tree
[00:14:08] <codestr0m> and the number of results is limited at max to those in the tree (yes that's making some assumptions)
[00:14:12] <oxygene> codestr0m: as long as it's a tree...
[00:14:50] <e^ipi> dependency solving is mathematically 3-SAT
[00:14:55] <e^ipi> 3-SAT is NP-complete
[00:15:18] <e^ipi> add one more package to your tree, you get an exponential increase in search time
[00:15:28] <codestr0m> oxygene: by tree I mean the structure.. anyway. .I've messed with calculating dependencies efficiencies in sqlite
[00:15:40] <TomJ> can a zone be a NIS server?
[00:15:56] <oxygene> codestr0m: tree or graph?
[00:16:24] <codestr0m> oxygene: technically a graph
[00:17:51] <e^ipi> anyways, the whole point of this exercise is that I don't trust network package repos
[00:18:17] <codestr0m> and my point is you can't blame the tools.. it's policy and people who are behind it
[00:18:18] <codestr0m> :)
[00:18:27] <e^ipi> i don't blame the tools
[00:18:31] <e^ipi> i blame the theory behind the tools
[00:18:43] <e^ipi> which is much worse than if apt were just broken
[00:18:58] <e^ipi> apt is broken before it even came in to existence
[00:19:00] *** Fish has quit IRC
[00:19:58] *** the-decider has quit IRC
[00:20:04] *** storycrafter has left #opensolaris
[00:21:20] <oxygene> one of the reasons why my source packaging system never got anything to do with dependencies except for storing them in the resulting binary package (derived from analyzing executables, etc)
[00:22:23] <codestr0m> oxygene: wouldn't that be a logical separation of concern?
[00:23:33] <oxygene> codestr0m: hm?
[00:25:43] <codestr0m> well. there's different "concerns" when building a package.. resolving a dependency isn't directly related to building a source for example.. so sometimes it may not be the most efficient way to store that dep with the instructions of how to build the source. if that makes sense
[00:27:13] <oxygene> codestr0m: that dependency isn't stores with the build recipe. dependencies are derived from the output of the build (eg. ldd output). if you have packages missing when starting the build, the build either fails, or lacks features, depending on whether the source can adapt
[00:28:40] <codestr0m> oxygene: in my mind. the deps are created at build time.. not as a result of missing ldd output.. I mean. what happens if you have  a static binary. that methodology breaks down pretty quick
[00:29:02] <oxygene> codestr0m: a static binary doesn't rely on another package - the stuff is linked in
[00:29:45] <codestr0m> oxygene: not true.. it still depends on basic things
[00:29:58] <codestr0m> (does a static binary depend on ld at all?)
[00:30:02] <ocr> anybody here have b97 booting with xVM properly?
[00:30:31] <codestr0m> anyway. I'm back to building my env so I can rant more about this on my blog
[00:30:40] <oxygene> codestr0m: truly static binaries are rather hard to create on solaris 10 and newer
[00:37:28] <ocr> e^ipi: probably more PSAT than 3-SAT
[00:38:29] <TomJ> I've just implemented NIS, and it's working ok. but I've noticed that a non-root user can now do: ypcat passwd and get the crypted passwords, whereas they could not read /etc/shadow. is this normal?
[00:39:42] <e^ipi> ocr: well, nSAT where n > 3 anyways
[00:40:05] <TomJ> oh, nvm
[00:40:13] <e^ipi> either way, it's an NP-complete problem
[00:40:40] <ocr> aye
[00:41:51] * e^ipi idly wonders if the ubuntu forums ever have discussions about complexity theory
[00:45:24] <ocr> either way theres a million dollars awaiting whomever proves you wrong
[00:46:22] <e^ipi> presumably in addition to the turing award and fields medal
[00:46:44] <ocr> probably a nobels prize aswell
[00:46:51] <e^ipi> no nobel prize in math
[00:46:56] <ocr> no?
[00:46:59] <codestr0m> you guys are being so dramatic :P
[00:46:59] <e^ipi> nope
[00:47:37] <e^ipi> Nobel had a thing against mathematicians
[00:47:40] <ocr> who cares about medicine - they should have a nobels prize for maths ;)
[00:48:00] <ocr> he was a physicist, small mans complex
[00:48:15] <e^ipi> this is assuming the NSA or similar doesn't make you disappear before you publish
[00:50:09] <codestr0m> you know I'm understanding this packaging problem a lot better.. instead of breaking things into manageable chunks. the blob the whole thing under svn+ssh://anon at svn dot opensolaris.org/svn/jds/spec-files/branches/gnome-2-18/
[00:50:10] <ocr> conspiracy theorist
[00:50:25] <codestr0m> (as an example)
[00:51:41] *** rand7 is now known as Barack_Wabama
[00:51:54] <ocr> wambama
[00:53:42] <e^ipi> codestr0m: do you know where a developer delivers things that are integrated in to Solaris ?
[00:53:49] <e^ipi> it's called the WOS... wad-of-stuff
[00:54:26] <codestr0m> are you kidding?
[00:54:34] <e^ipi> nope
[00:54:50] <codestr0m> s/stuff/shit/g
[00:55:27] <codestr0m> ok. whatever. I won't rant here. I'll just piss off those who I may end up needing help from
[00:56:05] <e^ipi> *shrug* rant all you like, i don't get pissed off when people disagree with me
[00:56:06] <codestr0m> (this is more painful than trying to reproduce a jboss build)
[00:56:35] *** h3sp4wn has joined #opensolaris
[00:57:09] <e^ipi> it's fundamentally a problem of desired goals; reproducability and guaranteed inter-stability vs. convienience
[00:57:28] <e^ipi> that's at the heart of the matter and it's an opinion of values
[00:57:59] <e^ipi> both have value, but the opinion bit of it comes via ranking
[00:58:15] <codestr0m> a bunch volunteers can maintain the portage tree which has thousands of packages in relatively organized and highly reproducible building state
[00:58:51] <e^ipi> yeah, if you're going to make an example of a robust package tree, perhaps try something other than portage
[00:59:15] <e^ipi> which is constantly broken, it's just a matter of if this week's breakage affects you
[00:59:16] <codestr0m> such as?
[00:59:55] <codestr0m> heh.. most things will build.. that's rarely the problem I had. it's the obscure bug you had to track down which followed
[01:00:16] <codestr0m> things much worse than jumping mouse pointer :P
[01:00:20] <e^ipi> are you arguing your point or mine now?
[01:01:02] <e^ipi> ;)
[01:01:02] <codestr0m> I'm not even at obscure bug phase.. I'm at just wanting it to build.. anyway. I'm verbosity off on this..
[01:01:10] *** alanc is now known as alanc-away
[01:01:16] *** luc^ has quit IRC
[01:01:25] <e^ipi> JDS's build system isn't so bad actually
[01:01:31] <codestr0m> pkgtool download SUNWgnome-terminal.spec
[01:01:34] <e^ipi> you use CBE and there's a top-level make thingy
[01:01:47] <codestr0m> I don't want to build the whole mess of shit
[01:01:53] <codestr0m> one package any any required deps
[01:02:11] <e^ipi> at least, the last time i used it it was like that
[01:02:36] <e^ipi> "grab wad. install CBE. ./make-jds (?). done"
[01:02:38] <codestr0m> I pulled the sources directly from upstream, but got
[01:02:43] <codestr0m> terminal.c:3608: error: `G_GNUC_FUNCTION' undeclared (first use in this function
[01:03:10] <codestr0m> which lead me to believe there's some patching downstream
[01:03:22] <e^ipi> almost certainly
[01:03:31] <codestr0m> so instead of duplicating effort I thought I'd hunt down the original source and give that a spin
[01:03:36] <e^ipi> not least of which because almost everything is built with sun studio
[01:03:38] <e^ipi> rather than gcc
[01:05:04] <ocr> codestr0m: be happy you're using jboss and not websphit
[01:05:22] <codestr0m> ocr: I rm -rf jboss* a long time ago
[01:05:29] <e^ipi> i've got glassfish *shrug*
[01:05:39] <e^ipi> mostly just because it's SMF integrated
[01:05:52] <e^ipi> and it was there
[01:05:54] <ocr> e^ipi: then you're not entitled to whine!
[01:06:29] <codestr0m> ocr: hehe.. you can't even get websphit to run on os2008.. doesn't it still require jdk5?
[01:06:35] <e^ipi> ocr: i wasn't whining
[01:06:59] <ocr> ibm jvm 5 ak jdk5 aka not really java5-compatible
[01:07:26] <codestr0m> I think it should require icedtea6 :)
[01:08:17] <e^ipi> it's IBM, so the developer tools should cost half your salary
[01:08:54] <codestr0m> yeah, but it's worth half your salary :P
[01:09:02] <ocr> and be 2 years more retarded than the open source versions
[01:09:10] <e^ipi> codestr0m: XLC isn't really
[01:09:23] <e^ipi> it's a great compiler suite for PC
[01:09:25] <e^ipi> *PPC
[01:09:37] <e^ipi> but worth 9 grand a seat? ... i dunno, probably not
[01:09:50] <ocr> at least sun got the point
[01:09:55] <e^ipi> you'd think they'd want as many people using the best tools possible to dev. for their platform, but again... IBM...
[01:10:10] <e^ipi> ocr: yeah, studio used to not be free either, that was retarded.
[01:10:14] <ocr> sadly ibm has better salespeople than opensource
[01:10:21] <ocr> so i still need to use it
[01:10:48] *** yarihm has quit IRC
[01:11:21] <e^ipi> eh, gcc sucks anyhow
[01:17:43] *** ahe has quit IRC
[01:28:11] *** ninjaslim has quit IRC
[01:31:08] <wonko2> tomww++
[01:31:17] <wonko2> e^ipi: i knew i liked you for a reason
[01:31:18] <wonko2> ;)
[01:34:58] <tomww> wonko2: it was a pleasure :-)
[01:35:24] <wonko2> :)
[01:36:43] <e^ipi> i dunno what I did, but cool
[01:37:32] <wonko2> 19:13:07 <e^ipi> eh, gcc sucks anyhow
[01:38:00] * wonko2 has been fighting to get OSS built with MIPSpro, Sun Studio, DEC C, etc for more years than he cares to admit to
[01:39:51] <e^ipi> oh, yeah
[01:43:41] *** estibi has joined #opensolaris
[01:44:06] <e^ipi> hey sweet, that pay hiccup got resolved
[01:44:08] * e^ipi gets drunk
[01:45:12] *** coffman_ has joined #opensolaris
[01:45:22] <tomww> good night all!
[01:45:30] <coffman_> off
[01:45:55] <coffman_> tomww: i come on you leave!
[01:47:47] <tomww> coffman_: already did my goal today (2 problems solved)
[01:47:57] <coffman_> heh
[01:48:28] <coffman_> i got some major trouble with live upgrade on a sxce with zfs root and zones
[01:48:39] <coffman_> :/
[01:49:09] <tomww> which was the old release and which the new one?
[01:49:23] <coffman_> b96 to b97
[01:51:12] <tomww> I think liveupgrade was upgraded first? (Installers/liveupgrade20)
[01:51:33] <coffman_> never got a clean lu working with zones till now...
[01:51:41] <coffman_> tomww: installed the new packages
[01:51:42] <tomww> I have to upgrade from 88 to say 98 or later shortly
[01:52:06] <tomww> zones-root-fs on zfs
[01:52:22] <coffman_> yes
[01:52:24] <tomww> last time i moved with zoneadm move to ufs, upgraded and moved back :)
[01:52:55] <coffman_> wah
[01:52:57] <coffman_> crap
[01:53:07] <tomww> this time I'll try it with one zone on ufs, copied to zfs and on staying on zfs before+after
[01:53:36] <tomww> what was the problem with the zfs + zones?
[01:53:45] <coffman_> well, i got sparse zone only, so ill do set them to configured in the index file normaly
[01:54:05] *** bahumbug has joined #opensolaris
[01:54:24] <tomww> shouldn't that be "installed" ?
[01:55:12] <coffman_> if i set it to installed, lu will go after the filesystem and will fail
[01:55:29] <coffman_> but now that hack does not work any more :(
[01:57:12] <coffman_> http://pastebin.ca/1202187
[01:59:06] <estibi> there is 'update zones on attach' feature so you can easly detach zones, do LU, and attach zones :)
[02:00:30] <tomww> coffman_: hmmm. maybe an error from using this hack before.
[02:00:54] <tomww> you could backup and then try zoneadm detach and then attach with allowing to upgrade the zone content
[02:01:08] <tomww> that is on the old relase, to have zones in sync with the global zone
[02:02:52] <tomww> but I don't know how far the code for the automatic update on attach has eveolved.
[02:03:33] <estibi> tomww: it upgrades zones from S10 to the latest SXCE, works great
[02:04:19] *** postwait has joined #opensolaris
[02:05:04] <tomww> yes, I've used it befor on Lab machines because telling customers to use liveupgrade requires to test this before telling them that this is a good thing (lieupgrade)
[02:05:11] <tomww> oh,,,, liveupgrade :-)
[02:05:33] <tomww> I'm too tired. n8!
[02:07:42] *** FurnaceBoy has joined #opensolaris
[02:10:59] <wonko2> ls
[02:11:01] <wonko2> dammit
[02:16:11] <niq> wonko2: next time, your root password, right :-)
[02:16:52] <wonko2> heh
[02:16:53] <wonko2> ok
[02:16:54] <wonko2> ;)
[02:16:58] <estibi> :)
[02:20:50] *** ehtom has quit IRC
[02:25:20] <ocr> is there a decent opensol/sxce forum anywhere?
[02:28:18] <coffman_> ocr: http://www.opensolaris.org/jive/index.jspa?categoryID=1
[02:30:16] *** LordIllpalazo has quit IRC
[02:32:59] <coffman_> lu is running...
[02:33:07] <ocr> thanks
[02:35:58] <ocr> that forum was annoying
[02:36:07] <coffman_> which one?
[02:36:10] <ocr> too many restrictions on registering :)
[02:36:41] <ocr> i couldnt use god for password - thats retarded :)
[02:38:45] *** Auriel_ has joined #opensolaris
[02:45:13] *** niq has quit IRC
[02:45:37] *** estibi has left #opensolaris
[02:50:27] *** not-me-guv has joined #opensolaris
[02:51:56] *** ninjaslim has joined #opensolaris
[02:54:14] *** Auriel has quit IRC
[03:00:53] *** bahumbug has quit IRC
[03:02:19] *** AxeZ has quit IRC
[03:08:44] *** Auriel_ is now known as Auriel[A]
[03:44:51] *** postwait has quit IRC
[03:49:07] *** balbirs has joined #opensolaris
[03:54:22] *** RElling1 has joined #opensolaris
[04:02:36] *** RElling has quit IRC
[04:25:29] *** myrkraverk has quit IRC
[04:31:21] *** stux|away is now known as stux
[04:34:28] *** balbirs has quit IRC
[04:36:36] *** not-me-guv has quit IRC
[04:43:39] *** medar has quit IRC
[04:46:07] *** Kernel86 has quit IRC
[04:46:48] *** spiki has quit IRC
[04:53:14] *** criso has joined #opensolaris
[04:53:49] *** gnut has joined #opensolaris
[04:54:02] <gnut> hi all
[05:01:42] *** coffman__ has joined #opensolaris
[05:02:08] *** Odin- has joined #opensolaris
[05:03:30] *** storycrafter has joined #opensolaris
[05:04:14] *** coffman_ has quit IRC
[05:04:51] *** storycrafter has left #opensolaris
[05:05:25] *** freakazoid0223 has quit IRC
[05:08:09] *** kohju is now known as KOHJU
[05:10:36] *** freakazoid0223 has joined #opensolaris
[05:16:08] *** rtor has quit IRC
[05:33:01] *** zenbalrog has quit IRC
[05:41:23] *** zenbalrog has joined #opensolaris
[05:45:38] *** Trede has joined #opensolaris
[05:45:45] *** Trede has joined #opensolaris
[05:48:46] *** fr4g has joined #opensolaris
[05:54:10] *** Odin- has quit IRC
[05:58:50] *** FurnaceBoy has quit IRC
[06:22:28] *** rtor has joined #opensolaris
[06:37:53] *** anilg has joined #opensolaris
[06:42:08] *** Gekz_ has joined #OpenSolaris
[06:43:00] *** Gekz has quit IRC
[06:45:48] *** balbirs has joined #opensolaris
[06:46:07] *** Gekz_ is now known as Gekz
[06:56:07] *** c00p has quit IRC
[06:58:59] *** xRaich[o]2x has left #opensolaris
[07:02:56] *** storycrafter has joined #opensolaris
[07:03:49] *** qmrw has joined #opensolaris
[07:04:00] *** storycrafter has quit IRC
[07:11:40] <ipfw> Ram: Kingston or CORSAIR ?
[07:41:31] *** rgiltrap has joined #opensolaris
[07:42:02] <rgiltrap> Hi All, Just upgraded to 97
[07:42:18] <rgiltrap> But now I need to reconfigure X so I can set up my dual screens
[07:42:29] <rgiltrap> Haven;t played with x in about 3 years
[07:42:43] <rgiltrap> So would appreciate some pointers on where to start
[07:43:16] <e^ipi> man xorg.conf
[07:43:40] <e^ipi> ipfw: the cheaper the better
[07:48:04] <rgiltrap> Where is the right xorg.conf found?
[07:48:51] <e^ipi> read the man page
[07:49:29] <rgiltrap> I did and it confused my poor old brain
[07:50:34] <e^ipi> lovely, the information is there anyways
[07:52:23] *** eljak has joined #opensolaris
[07:54:26] <e^ipi> there is also google.
[07:54:52] <rgiltrap> There's also the friendly and helpful folk on irc!
[07:55:04] <eljak> hi, how can i mount an opensolaris2008.5 partition from solaris ex? i am trying the following command: mount -oufstype=sunx86 /dev/dsk/c2t0d0s0 /mnt/ , and it gives me : mount: /dev/dsk/c2t0d0s0 is not this fstype
[07:55:44] <e^ipi> rgiltrap: that is for after you've exhausted your other resources
[07:56:55] <rgiltrap> If you're not going to help stop talking. I have man'd and googled without much luck so I thought I would ask some helpful people here
[07:57:11] <erast> hi
[07:57:24] <erast> anyone has idea what might break if we compile ON with gcc?
[07:57:48] <erast> i know from the past that - there were some issues with binary compat, is it fixed now?
[07:57:53] <e^ipi> since i know for a fact xorg dual monitors is well documented on peoples' blogs, i will refer you back to google.
[07:57:59] <gnut> rgiltrap: there is no xorg.conf by default. you have /etc/X11/.xorg.conf
[07:58:12] <e^ipi> erast: it'd be slow as crap, other than that keith took great pains to making it work
[07:58:40] <erast> e^ipi: i bet it will be faster if you will use -O3
[07:58:55] <erast> sunstudio is not good at inlining..
[07:59:10] <gnut> is there a good backup solution for zfs systems besides send/receive?
[07:59:18] <gnut> one that's free?
[08:00:11] <e^ipi> erast: no, gcc's pretty bad at code gen, even with -OMGFAST
[08:00:41] *** gnut has quit IRC
[08:01:23] <erast> e^ipi: i did a lot of comparision - sunstudio 10 vs. gcc, and gcc was generally 10% faster
[08:01:43] <erast> i.e. not gcc-3, gcc-4.2
[08:02:04] <rgiltrap> SunStudio can be fast.. I used it to provide the validation code http://www.mersenne.org
[08:02:30] <rgiltrap> Gave almost 20% over gcc
[08:02:59] <rgiltrap> SunSTudio 12 that is
[08:03:19] <erast> well, then my test was a bit different
[08:03:38] <e^ipi> if you can find a place where gcc is faster, it's a bug, and you should report it to Sun as such
[08:03:39] <erast> anyway, i'm mostly concerned about binary compat. issues
[08:04:11] <e^ipi> anything with C++ in it is studio-only;
[08:04:20] <e^ipi> gcc breaks compat way too often
[08:04:35] <erast> e^ipi: yes.. that is my concern too
[08:05:14] <erast> we are working on ubuntu official port for nexenta and trying to figure out how we could build nexenta on nexenta without closed binaries involved
[08:05:29] <erast> ideas appreciated
[08:05:40] <e^ipi> it would involve rewriting large parts of ON
[08:05:53] <e^ipi> well, not really
[08:05:58] <e^ipi> but a chunk of it
[08:06:01] *** fr4g has quit IRC
[08:06:03] <erast> i know
[08:06:06] <erast> but what chunk
[08:06:27] <erast> well, we not going to drop closed-bins entirely though
[08:06:41] <erast> the will be separated into APT restricted
[08:07:04] <erast> so, the main will be clean open source then
[08:07:48] <e^ipi> so not ON then?
[08:08:40] <erast> ON... just re-packaged a bit
[08:09:08] <e^ipi> you understand that you can't build libc without closed bins, yes?
[08:10:18] <erast> i'm hoping that we will work it out and ubuntu folks will make an exception for us till the issue is fixed
[08:10:41] <e^ipi> ubuntu folks?
[08:10:43] <erast> but i need to figure out totally minimal set of closed bins required, so the effort will be focused
[08:10:45] <e^ipi> for nexenta?
[08:11:04] <erast> you don't know that nexenta is ubuntu derivative?
[08:11:17] *** Auralis_ has joined #opensolaris
[08:11:33] <e^ipi> what do the ubuntu people have to do with it though?
[08:12:43] <DerJoern> ubuntu? nothing. nexenta uses their userland
[08:13:44] <erast> e^ipi: there is a lot of excitement in ubuntu campus wrt. nexenta and its potential adoption as official debian/ubuntu port
[08:13:48] *** Auriel[A] is now known as Auriel
[08:13:59] <erast> but we need to work out stupid closed binaries..
[08:14:40] <e^ipi> keep the sysv userland.
[08:15:23] <erast> e^ipi: its a matter of preference these days
[08:15:35] <e^ipi> well, and standards
[08:15:39] <e^ipi> and compatibility
[08:16:13] <e^ipi> and don't forget that the gnu tools are pretty broken on solaris
[08:16:15] <erast> yep, we trying to preserve ABI
[08:16:22] <e^ipi> non xattr's, no extended system attributes, no acl's
[08:16:32] <erast> i know that
[08:16:44] <erast> but we have SUN_PERSONALITY libc switch
[08:16:45] <e^ipi> there's a reason the sysv userland exists, and it's not simply 'preference'
[08:17:37] <e^ipi> in addition, a lot of the GNU tools don't even pass POSIX
[08:17:44] <e^ipi> and forget about SUS
[08:17:45] <erast> so, we are totally fine with sysv
[08:18:48] <e^ipi> i'm just saying, be careful with using GNU stuff as default, because you may end up causing problems for your user
[08:18:54] <erast> by default, sysv is disabled in nexenta
[08:19:09] <e^ipi> i know that, and so you can lock people out of files and they won't even know why
[08:19:24] <erast> e^ipi: we fixed all critical issues with gnu userland
[08:19:45] <e^ipi> you gave them NFSv4 acl support?
[08:19:52] <erast> no.. this is feature
[08:20:00] <e^ipi> sounds like a bug
[08:20:02] <erast> needs to be implemented
[08:20:19] <e^ipi> or you can just use the proper userland
[08:20:19] <erast> well, missing ACL support is not a bug
[08:20:28] <e^ipi> yes it is
[08:20:28] <erast> setacl still there
[08:20:40] <e^ipi> setfacl is POSIX draft acl's
[08:20:48] <e^ipi> i'm talking about nfsv4/ZFS acl's
[08:20:55] <e^ipi> which are accessed via chmod(1) and ls(1)
[08:21:06] <erast> really, i thought it extended with zfs/nfs acls
[08:21:26] <e^ipi> nope
[08:21:27] <e^ipi> $ /usr/gnu/bin/ls -V
[08:21:27] <e^ipi> /usr/gnu/bin/ls: invalid option -- V
[08:21:44] <erast> right it doesn't
[08:22:10] <erast> well, in nexenta you can do this: /usr/sun/bin/ls -V
[08:22:29] <erast> so, technically it is provided
[08:22:53] <e^ipi> yeah, but why hide the feature complete tools in a subpath, why not have the gnu tools in a subpath instead like nevada does it?
[08:24:16] <erast> exactly ... feature
[08:24:17] * e^ipi fails to see why one would want the broken gnu tools as default
[08:24:39] <e^ipi> on solaris, ls missing the -V switch is a bug
[08:24:44] <erast> i think over time ls/chmod will be extended with this missing functionality
[08:24:58] <erast> and i bet, initial implementation comes from ubuntu/nexenta crowd
[08:25:22] <erast> e^ipi: opensolaris != solaris
[08:25:24] <e^ipi> or you can have it now, by using the sysv tools :)
[08:26:03] <erast> again, i have it if I want it: echo export SUN_PERSONALITY=1 > ~/.bash_profile
[08:26:36] <erast> but then we will break ubuntu userland compatibility
[08:26:44] <e^ipi> well, it's GNU after all
[08:26:54] *** Auralis has quit IRC
[08:26:55] <e^ipi> it's userland compatibility breaks itself pretty frequently
[08:27:00] <e^ipi> so no loss there really
[08:27:02] <erast> so, the right way would be to add functionality to gnu ls/chmod
[08:27:19] <erast> e^ipi: who told you that?
[08:27:20] <e^ipi> that's the long way around
[08:27:28] <e^ipi> erast: experience with linux, mostly.
[08:27:48] <erast> well, userland and ABI pretty stable last 3 years
[08:28:28] <e^ipi> i have a copy of mosaic browser 0.9 for solaris 2.5 ( 1993 ) that runs just fine on nevada
[08:29:04] <e^ipi> 3 years of not breaking too much just screams unstable to me
[08:29:26] <erast> well, than you will be probably excited by Windows - because binaries compiled 10 years ago still works
[08:30:07] <e^ipi> they don't actually, but windows isn't UNIX so i don't really pay too much attention to it
[08:30:49] <erast> not all, but a lot of binaries will work
[08:33:17] <Adamant> e^ipi: here's a question - why do you care about running a web browser made in 1993 on the modern Internet?
[08:33:52] <e^ipi> Adamant: i don't really, it's just an example
[08:34:05] <Adamant> how many Unix apps do people actually use in 2008 that haven't been updated with new binaries in the last 10 years?
[08:34:30] <Adamant> aside from custom one-offs build for a company or things like that
[08:34:37] <Adamant> *built
[08:34:39] <ipfw> In the Military, 80% ;)
[08:34:43] <e^ipi> Adamant: probably quite a bit, and don't discount one-off builds for a company
[08:34:58] <e^ipi> there's a lot of critical stuff that is in-house in a lot of places
[08:35:10] <e^ipi> anyhow erast, you were asking about what has to stay and what can go in closed_bins
[08:35:30] <e^ipi> most of the drivers you can probably do away with; the xpg userland stuff can probably be replaced
[08:35:34] <erast> e^ipi: just minimal stuff
[08:35:36] *** ipfw has quit IRC
[08:35:42] <e^ipi> libc_i18n.a has to stay or be rewritten
[08:35:54] <e^ipi> there's some stuff about encrypted drivers in there
[08:36:11] <e^ipi> if you modify the kernel such that it always returns success, you can do without Sun's signing key
[08:36:22] <erast> any work been don on libc_i18n.a rewrite so far?
[08:36:23] <e^ipi> it's in the crypto stuff, go looking and you'll find it
[08:36:33] <e^ipi> erast: yes, and then the work was flushed because I went about it the wrong way
[08:36:52] *** Auriel has quit IRC
[08:37:00] <e^ipi> i updated the page with a new list of symbols and an hg changeset that should apply against a fresh checkout of onnv-gate
[08:37:26] <e^ipi> it just removes the checks in the makefiles, and patches some private-private stuff out of the mapfile
[08:37:41] <e^ipi> everything that fails after that is public symbols and they need to be rewritten
[08:38:00] <e^ipi> how i'm going to go about it now that I know better is just get POSIX/C working... worry about other locales later
[08:38:02] <erast> do you have a file shared with missing symbols somewhere?
[08:38:17] <e^ipi> yaeh, it's on the emancipation site
[08:38:23] <erast> ok
[08:38:37] <erast> thanks a lot e^ipi
[08:38:38] <e^ipi> http://opensolaris.org/os/project/emancipation/tasks/libc_i18n/i18n_symbols/
[08:39:04] <e^ipi> things like _toupper can just be synonyms
[08:39:08] *** eljak has quit IRC
[08:39:32] <erast> can we substitute some with gettext?
[08:39:41] <e^ipi> ( eg: wchar_t _toupper(wchar_t a) { return toupper(a); } )
[08:41:35] <oxygene> erast: you could look at netbsd's citrus project, which does i18n
[08:41:46] <e^ipi> yeah, that's a dead end
[08:42:02] <e^ipi> i wasted a bunch of time there too, it was good times
[08:42:02] <oxygene> it is?
[08:42:43] <e^ipi> like i said, i tossed a bunch of stuff because i'm firmly convinced that the best way to get that stuff done is to worry about POSIX/C first, and add functionality later
[08:44:15] <erast> why Sun can't open source this stuff?
[08:44:57] <erast> anyway... i don't really care
[08:45:17] <e^ipi> because they don't own it
[08:45:47] <e^ipi> and i assume that the contract they've got it under doesn't allow them to open it up
[08:48:40] *** stux has quit IRC
[08:50:56] *** stux has joined #opensolaris
[08:52:29] *** tsoome has quit IRC
[08:57:17] *** jfndi_ has joined #opensolaris
[09:06:40] *** twisti has quit IRC
[09:11:46] *** alibb has joined #opensolaris
[09:13:32] *** balbirs has left #opensolaris
[09:14:03] *** Auriel has joined #opensolaris
[09:17:00] *** Cookie` has quit IRC
[09:17:36] *** Openfree has quit IRC
[09:32:40] <codestr0m> e^ipi: the reason it's closed source normally comes down to a simple equation.. if it costs more than it's financially worth then they don't open source it
[09:32:44] <codestr0m> that's what I've been told
[09:33:20] <codestr0m> erast: I'm reading up and I'm quite interested what you work out in terms of dropping those encumbered bins
[09:33:25] <erast> should we donate to Sun, so they open source it for us? :-)
[09:33:39] <codestr0m> also if you want java check out the latest icedtea6 ebuild from gnu_andy
[09:33:50] <codestr0m> erast: ask Simon
[09:33:55] <erast> codestr0m: its all just plans, plans...
[09:34:11] *** twisti has joined #opensolaris
[09:34:25] <codestr0m> and you're talking about something they'd measure in millions and we could ship the specs to china and get it rewritten for much less :P
[09:35:05] <e^ipi> codestr0m: trust me on this one.
[09:35:28] <e^ipi> it is not as simple as it may appear from the outside.
[09:36:45] <codestr0m> erast: for what it's worth.. I'm pretty agnostic about which side of the GNU hate fence and for a month I've been testing the difference between sysv and gnu userland.. I'd 100% say put gnu in a subpath and make a guide that tells people how to create alias and make an interactive terminal which feels more intuitive
[09:37:14] *** ninjaslim has quit IRC
[09:38:25] <codestr0m> e^ipi: yes I realize that, but what I was trying to say is the cost of new development is probably less than the cost for sun to open source.. it's really pick between the two very time consuming projects :P
[09:38:59] <erast> codestr0m: ideologically, nexenta userland == ubuntu userland, and we should at least try to preserve our upstream compatibility
[09:39:55] <erast> Debian community is huge.. a lot of scripts written, a lot of tools, etc - thats our legacy
[09:41:00] <codestr0m> erast: there's two sides to this... 1) interactive (eg gnome-terminal or something) 2) scripts for the scripts if Ubuntu/deb is doing anything strange.. it's just wrong.. from the interactive pov. it's just retraining people. that simple.. I don't care about ideologies.. this is simple about doing it right.. because the loss of sysv is greater than the gain for people who can't adjust
[09:41:41] <codestr0m> erast: the deb community scripts probably only depends on sh right?
[09:42:42] <e^ipi> oh, they probably make all sorts of GNU assumptions
[09:42:49] <erast> codestr0m: sorry i don't agree with both of your points
[09:43:03] <codestr0m> e^ipi: if you are going to be flippant be quiet
[09:43:17] <codestr0m> erast: I wouldn't have agreed with both of my points before either
[09:43:34] <e^ipi> codestr0m: who's being flippant
[09:43:46] <e^ipi> i'm pointing out that the scripts probably make all sorts of assumptions on a gnu userland
[09:44:04] <erast> debian is about open-source
[09:44:08] <e^ipi> shell scripts aren't typically self-contained
[09:44:15] <Tempt> yawn. Debian is about politics, for the most part.
[09:44:16] <erast> what that means: 1) no closed binaries
[09:44:32] *** alibb has quit IRC
[09:44:36] <codestr0m> erast: yeah I got that. no need to explain that
[09:44:39] <e^ipi> Tempt: yeah, pretty much
[09:45:09] <erast> anyway, I think we are good as long as we allow SVR4 packages to be cleanly installed in Nexenta :-)
[09:45:20] <Tempt> and they work.
[09:45:36] <erast> yes, and they actually will work
[09:45:50] <codestr0m> erast: btw.. is the nexenta full build process open source?
[09:45:58] <Tempt> And if only people could refrain from the /bin/sh=/bin/bash assumption, the world would be happier.
[09:46:05] <codestr0m> I mean the scripts used to bootstrap and everything?
[09:46:26] <erast> codestr0m: as much as we can do that
[09:46:42] <e^ipi> codestr0m: you cannot build ON without a good lot of closed-source software
[09:46:43] <erast> anyone can grub the package, fix it and upload it back
[09:46:53] <Tempt> Heh, given the choice between purity of open source and having decent video drivers, most people take the simple choice.
[09:46:57] <erast> ON build is sucks...
[09:47:12] <erast> that's why i started this longish thread here
[09:47:31] <e^ipi> Tempt: in fairness if intel started selling discrete graphics cards i'd buy 2 dozen of them
[09:47:52] <codestr0m> erast: well. I know you're using stuff which isn't public to bootstrap.. and that you can't release it because of the storage company backing you
[09:47:54] <Tempt> Nvidia too fast for you? Need to slow things down a little?
[09:48:04] <e^ipi> precisely because the driver is open source, so i can throw whatever weird OS i can conjure up and expect it to work with some degree of coaxing
[09:48:21] <erast> codestr0m: these patches will be released soon
[09:48:39] <erast> without them, we can't be accepted as official ubuntu port
[09:48:56] <Tempt> See, I don't get enough hours in my day to worry about video driver source. At least you get paid to code away on such things.
[09:48:57] <codestr0m> erast: I heard that 6 months ago. .it's what really turned me off to nexenta.. I'd have contributed, but my impression is the project was controlled by some company
[09:49:18] *** Fish has joined #opensolaris
[09:49:27] <codestr0m> I needed a more recent kernel and w/o a bunch of duplicate effort was stuck
[09:49:31] <erast> but not anymore
[09:49:39] * Tempt drifts off again
[09:57:20] <oxygene> hmm.. installing solaris from nfsv2 - go or no go? ;)
[09:58:24] <e^ipi> what, is the server linux or something? doesn't speak nfsv3 properly ?
[09:59:43] <oxygene> e^ipi: http://www.novatech.net/jgaa/warnfsd.htm
[10:01:10] <e^ipi> SUA ?
[10:01:18] <e^ipi> or SUS on older systems
[10:01:35] <kaleb> my opensolaris is complaining about shared irq, and my mouse does not work. In linux I usually just restart the /etc/init.d/uhci thingy. Is there a similar way to restart usb services in opensolaris?
[10:01:50] <_mary_kate_> e^ipi: i think you mean SFU on older systems ;)
[10:01:57] <e^ipi> it's possible
[10:02:03] <e^ipi> too many TLA's
[10:02:36] <kaleb> TLA's?
[10:02:50] <e^ipi> three letter acronyms
[10:02:58] <kaleb> I guess you are not adressing me :)
[10:03:02] <_spike> what do I use if I want to connect to a vnc type ordeal running on my solaris machine, but I want it to create a -new- non-physical x-session for me...
[10:03:15] <_spike> i.e. if someone is using the physical machine it doesn't interfere with them.
[10:03:23] <_mary_kate_> _spike: use vnc?
[10:03:37] <kaleb> Is there a way to restart a driver in opensolaris?
[10:03:42] <_spike> well i was looking at like x11vnc but it controls the current physical session
[10:03:53] <_mary_kate_> _spike: use the normal Xvnc
[10:04:06] <_spike> what vnc software allows one to create virtual x sessions, i.e. 5 people could connect and have there own x sessions, detach, and then return later and everyting would be the same.
[10:04:09] <_spike> Okay perfect, thanks.
[10:04:15] <_mary_kate_> _spike: what you want is exactly what VNC does
[10:04:22] <_spike> perfect.
[10:04:25] <_mary_kate_> if it can control an existing session, that's additional functionality, not how it usually works
[10:04:32] *** evocallaghan has joined #opensolaris
[10:04:50] *** evocallaghan has left #opensolaris
[10:04:53] <_spike> And then i gotta get virtual terminal or whatever it's called installed so that I can run windows on one of the x sessiosn as well, i need to be able to run this program for school under windows
[10:05:10] <e^ipi> _spike: virtualbox
[10:05:16] <e^ipi> _spike: or xvm ( xen )
[10:05:20] <_spike> yeah virtualbox that's what i meant :)
[10:05:59] <_spike> as soon as I get Xvnc online i'm gonna work on virtualbox
[10:07:31] <e^ipi> there are also packages for wine which you might have some success with
[10:07:59] <e^ipi> last time i needed to run windows only software for a class was this digital logic design thing written by some third rate canadian university
[10:08:05] <e^ipi> i threw it in wine
[10:08:25] <_spike> i have some stuiped math software
[10:08:35] <e^ipi> maple?
[10:08:43] <_spike> Xvnc isn't built in right? I can't seem to find it in either sfw nor blastwave
[10:08:58] <_spike> i dunno what it is yet, haven't opened it, but i wanna get it running on my solaris machine so that i can use it from my mac laptop via vnc
[10:09:10] <_spike> i don't htink its maple though
[10:09:15] <_spike> it's like something calc or something like that
[10:10:03] <e^ipi> virtualbox runs on the mac, you know?
[10:10:20] <e^ipi> PSARC 2007/545
[10:10:44] <_spike> yeah but not my -old- powerbook g4...
[10:10:45] <_spike> heh
[10:10:46] <e^ipi> refers to bug 6572087
[10:10:48] <_spike> it's only got 512mb ram
[10:10:58] <e^ipi> State: 10-Fix Delivered:Verified (Fix available in build)
[10:11:15] <e^ipi> Fixed In    snv_76
[10:11:23] <e^ipi> looks like xvnc should be there
[10:11:44] <e^ipi> for that matter, here's the bug; it's got instructions v
[10:11:51] <e^ipi> erm, that should've been a control-v
[10:11:53] <e^ipi> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6572087
[10:11:55] <e^ipi> there you go
[10:13:22] <_spike> how come i don't have the pkg command?
[10:13:28] *** anilg has quit IRC
[10:13:37] <_spike> yeah i found it is there, svcs enable xvnc-inetd
[10:13:38] <_spike> is what i did
[10:13:43] <e^ipi> cat /etc/release
[10:13:53] <_spike> its svn97
[10:14:00] <e^ipi> lovely, that tells me nothing.
[10:14:02] <e^ipi> cat /etc/release
[10:14:04] <_spike> assembled 27 august 2008, thats my b-day
[10:14:13] <_spike> Solaris Express Community Edition snv_97 X86
[10:14:24] <_spike>   Assembled 27 August 2008
[10:14:27] <e^ipi> that's why you don't have pkg.
[10:14:36] <e^ipi> SXCE doesn't have IPS, only indiana does
[10:15:03] <_spike> oh okay, hope i don't need it then
[10:15:03] <_spike> heh
[10:15:15] <e^ipi> i've been fine without it all this time
[10:15:16] <_spike> i'm still pretty new to solaris in general, i'm a bsd guy
[10:15:33] <_spike> thats just to add xvnc anyhow and i already seem to have it.
[10:16:08] <e^ipi> sxce ships with everything and the kitchen sink, and anything not in it isn't in the pkg(5) repos yet anyways
[10:16:22] <_spike> nic
[10:16:23] <_spike> nice*
[10:16:32] <e^ipi> since the method of adding things to the pkg repo is to go through SXCE
[10:16:38] <_spike> that's one thing i liked about bsd origionally (and still do to an extent) it comes pretty bare bones.
[10:16:45] <_spike> at least obsd does (thats what i'm most familiar with)
[10:16:55] *** Openfree has joined #opensolaris
[10:17:05] <e^ipi> yeah, you can strip things out of the WOS if you know what you're doing
[10:17:27] <_spike> i just put the fileservers behind an obsd firewall
[10:17:40] <e^ipi> yeah, whatever works for you
[10:17:57] <e^ipi> oBSD's a good OS
[10:18:15] <e^ipi> if i weren't using solaris i'd probably be an obsd guy
[10:18:18] <_spike> i'm finding that opensolaris is better all the time
[10:18:53] <_spike> yeah obsd's nice in its own ways. I find it -very- simple to use.
[10:20:21] <_spike> which do you think is easier all in all to get up an running, xen or virtualbox?
[10:20:28] <sickness> yeah, obsd is so clean and simple
[10:21:51] <sickness> http://www.opensolaris.org/jive/message.jspa?messageID=272281#272281
[10:22:43] <e^ipi> probably virtualbox
[10:22:44] <rgiltrap> _spike: You're not Spike from Weelington are you?
[10:22:47] <e^ipi> it's dead simple
[10:23:06] <_spike> thx
[10:23:12] <_spike> and no i'm not rgil
[10:23:15] *** _spike is now known as spike-
[10:23:21] <rgiltrap> k
[10:23:32] <spike-> i also can't see my nick from this bnc...so i didn't realize it was _spike, sorry bout that
[10:23:35] <spike-> i hate it when people do that
[10:23:48] <rgiltrap> VirtualBox is real easy
[10:23:54] <rgiltrap> If I can do it anyone can
[10:24:18] <spike-> cool as soon as i finish figuring out xvnc i'll get that online.
[10:24:32] <rgiltrap> E^ipi: I
[10:24:36] <spike-> ld.so.1: Xvnc: fatal: librfb.so.0: open failed: No such file or directory
[10:24:37] <spike-> ld.so.1: Xvnc: fatal: relocation error: file /usr/X11/bin/Xvnc: symbol __1cDrfbYSSecurityFactoryStandardRvncAuthPasswdFile_: referenced symbol not found
[10:24:38] <spike-> oh great
[10:24:41] <rgiltrap> Still struggling with x.org
[10:24:48] <spike-> that's what i get when i goto http://server:5900
[10:24:58] <spike-> and when i try to connect with tightvnc viewer it says invalid protocall
[10:25:36] <rgiltrap> Would someone please tell me where I find the x conf file these days???
[10:26:03] <rgiltrap> I followed the man page and still can;t find the bugger
[10:27:00] <e^ipi> rgiltrap: there probably won't be one there by default. x autodetects
[10:27:15] <rgiltrap> how do I get it to reauto detect then
[10:27:27] <rgiltrap> I changed hardware after install
[10:27:35] <rgiltrap> I don;t think it;'s picked up the drivers
[10:27:38] <e^ipi> it detects each time it launches
[10:28:48] <rgiltrap> hmmmm.. it;s picking up my 1280*1024 on vga by default then and missing my DVI
[10:29:30] <e^ipi> nvidia?
[10:29:34] <rgiltrap> now my 1680*1050 gets a copy on the other screen
[10:29:36] <rgiltrap> ADM
[10:29:38] <rgiltrap> AMD
[10:29:42] <rgiltrap> Radeon
[10:29:49] <e^ipi> oh
[10:29:51] <e^ipi> *shrug*
[10:30:16] <rgiltrap> The release info says it's now support
[10:30:17] <rgiltrap> ed
[10:30:45] <e^ipi> i don't doubt that it is
[10:30:50] <e^ipi> i just have never bought ATI
[10:30:58] <e^ipi> so the particulars of getting it working i have no idea
[10:31:31] <rgiltrap> I had nvidia but the ATI Radeon doesn't work on my wife's Vista PC so we had to swap
[10:31:43] <spike-> does tightvnc do the same thing as Xvnc? Cause xvnc isn't working for me and i have no idea why...i don't often use X
[10:32:09] <_mary_kate_> spike-: Xvnc comes from VNC.  there are several varieties of VNC: RealVNC, TightVNC, UltraVNC...
[10:32:12] <_mary_kate_> they all provide an Xvnc
[10:32:24] <spike-> oh okay perfect
[10:32:31] <spike-> i'll go with tightvnc then
[10:32:56] <rgiltrap> I've tried the randr commands but it appears that the driver isn't being seen by X so I wanted to check what X does have from it;s auto config
[10:34:13] *** qmrw has quit IRC
[10:34:36] *** sailorvrz has joined #opensolaris
[10:34:39] <spike-> i've used it in the past, just a long time ago
[10:35:16] *** noyb has joined #opensolaris
[10:36:52] *** jfndi_ has quit IRC
[10:42:01] *** Auriel is now known as Auriel[A]
[10:43:53] *** rgiltrap has quit IRC
[10:46:18] *** anilg has joined #opensolaris
[10:50:40] *** cypromis has quit IRC
[11:06:09] *** derchris has quit IRC
[11:06:15] *** derchris has joined #opensolaris
[11:08:17] *** pjfloyd has joined #opensolaris
[11:08:28] *** c00p has joined #opensolaris
[11:09:26] *** plavcik has joined #opensolaris
[11:21:04] <plavcik> hi, I like to install OpenSolaris to use ZFS, could I install basic system for console use only via network?
[11:22:37] <e^ipi> no, the installer doesn't support that currently
[11:22:50] <e^ipi> but you can remove the unnecessary packages after install
[11:23:08] <e^ipi> and more to the point, disk is cheap; having them kicking around doesn't hurt
[11:23:42] <plavcik> sorry, I not get the last comment
[11:24:07] <plavcik> you mean, that X installed and not used is no issue about space?
[11:24:36] <plavcik> I see it as possible security issue
[11:25:49] <e^ipi> sitting on the disk, it's not an issue
[11:25:53] *** RElling has joined #opensolaris
[11:26:10] <plavcik> the iso of OpenSolaris 2008.05 is then the right option?
[11:27:15] <e^ipi> sure
[11:27:39] <e^ipi> I use the SXCE dvd
[11:27:47] <e^ipi> that's just how I roll
[11:27:48] <plavcik> I'm use to install my OpenBSD systems via serial console 9600 bauds, so it will be a big change to do a graphical installation to have storage installed
[11:28:02] <plavcik> ok, I will look at SXCE dvd
[11:28:10] *** mubex has quit IRC
[11:30:46] *** criso has quit IRC
[11:32:20] <plavcik> is raid-z a safe option on 4 disks if one of them will die?
[11:33:29] <e^ipi> yes
[11:33:37] <e^ipi> at about 6 you want to think about raidz-2
[11:33:50] <e^ipi> since MTBF of the pool will pass the MTBF of a single disk
[11:34:06] <e^ipi> however, keep in mind that your root pool can only be a single disk or a mirror
[11:34:11] <e^ipi> raidz or stripes are not supported
[11:34:38] *** RElling1 has quit IRC
[11:36:30] <plavcik> root can be in slice of one disk and then everythin else in raidz?
[11:36:47] <plavcik> root can be in slice of one disk and then everything else of this disk part of raidz?
[11:37:24] <e^ipi> yes
[11:38:04] <plavcik> thank you for answers and raidz-2 comment
[11:38:10] <e^ipi> no problem
[11:39:34] <noyb> zfs needs the whole disk to utilize the disk cache, no?  if you slice the disk, zfs will not utilize the disk cache.  please correct me if I'm wrong, thanks.
[11:45:05] <e^ipi> well, zfs can manage the disk cache better if you do, yeah
[11:50:18] *** not-me-guv has joined #opensolaris
[11:54:22] <noyb> so zfs will use the cache, but it won't be as good as when using the whole disk?  I had thought zfs would not use the cache at all unless it had the whole disk in the pool.  I may have a wrong memory clanking around up there.
[11:56:20] <noyb> found this: http://mail.opensolaris.org/pipermail/indiana-discuss/2007-June/001229.html
[11:57:17] <noyb> "Using the Slice notation leaves using disk's internal write-cache for ZFS off." it says
[12:00:27] <e^ipi> presumably the disk logic might do it's own thing
[12:00:41] <e^ipi> either way, giving zfs the whole disk allows zfs to manage the cache itself
[12:00:48] <e^ipi> and it's probably a lot smarter than disk logic
[12:00:59] <noyb> agreed
[12:02:06] <noyb> it "knows" about what it's going to write, and can order things before said writes, so as to optimize them for example.
[12:03:38] <Stric> the thing is.. if the write cache is on, and the filesystem is unaware of this.. when the fs thinks data is on the platters, it might not be..
[12:03:56] <Stric> so if you get a power outage then, your fs might be corrupted
[12:04:12] <trygvis> is it possible to have roles without password when using RBAC?
[12:04:15] <Stric> zfs enables write cache, writes a bunch of data and then turns it off to flush the transaction
[12:04:33] *** luc^ has joined #opensolaris
[12:04:49] <Stric> because it's safe (corruption wise) to have it enabled during that write.. but if you have other fs's as well, it might not be safe - thus leaving it off.
[12:05:53] <Stric> so when zfs is all alone (whole-disk), it can enable the disk write cache safely (for higher performance during that write).. but when using slices, it can not. but if you want to enable it yourself, use 'format -e'
[12:06:52] <Stric> warranty void etc
[12:08:34] 
[12:14:36] *** scimmia has joined #opensolaris
[12:14:57] *** scimmia has quit IRC
[12:19:06] *** mega has joined #opensolaris
[12:19:45] *** tsoome has joined #opensolaris
[12:29:14] *** Auralis_ has quit IRC
[12:32:52] *** scimmia has joined #opensolaris
[12:33:01] *** cypromis has joined #opensolaris
[12:33:02] *** scimmia has quit IRC
[12:34:56] *** Cookie` has joined #opensolaris
[12:37:52] *** Auralis has joined #opensolaris
[12:44:07] *** mega has quit IRC
[12:58:04] <lanc> plavcik: have a look at pulsaros, it is doing things like this: https://developer.berlios.de/projects/pulsaros
[12:58:54] *** alibb has joined #opensolaris
[12:59:01] *** bahumbug has joined #opensolaris
[13:01:13] *** dom_ has joined #opensolaris
[13:01:59] <plavcik> lanc: thx
[13:03:30] *** pjfloyd has left #opensolaris
[13:07:55] *** not-me-guv has quit IRC
[13:08:23] <JWheeler> can anyone tell me what the special requirements are for cloning zones? every combination of zoneadm -z zoneFOO clone zone1 has resulted in the new zone being copied rather than cloned
[13:08:36] *** cypromis has quit IRC
[13:09:04] *** pjfloyd has joined #opensolaris
[13:09:31] *** Vagrant has joined #opensolaris
[13:09:47] <Vagrant> does anyone of you make update on laptop 2008.05?
[13:10:07] <Vagrant> some icons are missing and fonts are fscked up after update :<
[13:10:45] <tomww> JWheeler: the zones-filesystems reside all on a zfs filesystem with each zone having it's personal very own filesystem?
[13:10:50] <tomww> otherwise cloning would not work
[13:11:01] *** eject has joined #opensolaris
[13:11:07] <eject> Hi all
[13:11:29] <pjfloyd> does anyone know the fdisk partition id used for ZFS partitions (on x86 of course)?
[13:11:34] <JWheeler> yes they are all on a zfs file system, or at least they should be
[13:11:50] *** dom has quit IRC
[13:12:32] <JWheeler> I have /Z as a zpool, and /Z/zones as a fs under that, that isn't mounted, and finally /Z/zones/zonename here
[13:12:53] <JWheeler> I'm using zfs root, so I guess there isn't any way it can't be on zfs actually :)
[13:13:25] *** Trede has quit IRC
[13:14:11] *** mubex has joined #opensolaris
[13:14:42] <eject> Can anybody help me with Xorg on Solaris 10 and HP Compaq 6720s. I tired a lot of variants but no luck. I can make only XSun working with 640x480 (Intel Vesa driver)
[13:15:16] <eject> I can guess that I need use something like 915resolution
[13:16:16] <Vagrant> JWheeler: sxce or 2008.05
[13:16:17] <Vagrant> ?
[13:16:38] <JWheeler> sxce 95
[13:17:05] <JWheeler> I already have one clone, maybe the problem is simply that the clone is booted and running too?
[13:17:39] <Vagrant> im wondering what should i install as a desktop machine on my laptop
[13:17:41] <Vagrant> any ideas?
[13:18:02] <Vagrant> sxce/2008.05 or what ;[
[13:18:32] <JWheeler> Vagrant, I'm more a fan of sxce myself, and that's what I use on both my servers and desktops
[13:19:04] <JWheeler> pjfloyd, do you need it for partitions specifically, or will a whole disk do?
[13:21:38] *** medar has joined #opensolaris
[13:22:09] <tomww> when ZFS uses the whole disk then there is a EFI label instead of a dossish fdisk partition IIRC
[13:22:25] <JWheeler> that's my understanding too tomww
[13:22:40] <tomww> JWheeler: every zones-root-fs noted in /etc/zones/<all-zones-configs> should be a separate zfs mountpoint
[13:22:44] *** noyb has quit IRC
[13:23:18] <tomww> if yes, then is may detect the fact, that every filesystem could be cloned (but I never used this personally)
[13:23:19] <JWheeler> the weird thing is, even if I don't clone it, it still isn't making a zfs dataset... so something is wrong here
[13:24:16] <tomww> JWheeler: I'm wonderung why you haven't mounted the /Z/zones
[13:24:24] <tomww> *ing
[13:24:32] *** dnm has joined #opensolaris
[13:24:37] <JWheeler> it's really just an organisational fs, to be honest
[13:25:07] <JWheeler> /Z is a huge zpool, and I wanted to have all my zones under their own tree
[13:25:16] <tomww> hm. does this confuse the logic? if /Z is the top filesystem in this hierarchy, this sets the mountpoint.
[13:25:30] <tomww> better let the mountpoint be inherited down to the zones-fs
[13:25:39] <JWheeler> I could mount /Z/zones, but there wouldn't actually be anything in it, because the zones themselves are in a zfs below them
[13:25:47] <tomww> so zfs get mountpoint /Z/zones
[13:26:08] <tomww> shows "inherited from..."  and the zones itself "inherited from..."
[13:26:12] <JWheeler> as a test, I tried mounting it elsewhere, but it didn't seem to make any difference
[13:26:41] <pjfloyd> JWheeler: currently I mix UFS and ZFS, but I'm planning on going to all ZFS, perhaps raidz
[13:27:34] <pjfloyd> I have two ZFS fdisk partitions, but I can't remember how I crreated them and why they have a partition id of 38 decimal
[13:27:38] <pjfloyd> 28 I mean
[13:27:42] <JWheeler> hrm, interesting result:
[13:28:10] <JWheeler> I think you might be right about it needing to inherit a zfspath
[13:28:20] <JWheeler> this time it worked (not a clone, just a plain test)
[13:28:48] <pjfloyd> fdisk.h in /usr/include has nothing for 28, so perhaps it wasn't with Solaris fdisk that I created the partitions
[13:29:36] <JWheeler> maybe
[13:29:43] <JWheeler> either that, or 28 is EFI??
[13:30:34] <pjfloyd> don't think so, the disks use old fashioned fdisk partitions
[13:30:46] *** eject has quit IRC
[13:31:46] <JWheeler> ah, I think I know what's going on now
[13:32:03] <JWheeler> and this is a conceptual thing that I keep running into, and it's damned confusing with zfs
[13:32:47] <JWheeler> the rpool is different to /Z. So I have 2 zpools on this system - rpool for OS, and Z (raidz) for all storage
[13:33:28] <JWheeler> while I do have Z/zones/zone{1,2,3} they are mounted as /zones/zone{1,2,3}.
[13:33:58] <JWheeler> now the directory /zones, despite being part of the /Z namespace, is _ACTUALLY_ a directory from rpool, which is mounted as /
[13:34:56] *** pjfloyd has left #opensolaris
[13:35:16] <JWheeler> zfs clone isn't going to be able to work across pools, and I expect the logic is simply looking at who owns /zones, in order to create zone4, and since the inheritance of the folder structure isn't configured within the rpool, it's bailing out
[13:36:05] <sailorvrz> any chance of getting sxce as cd isos? my ultra doesn't have a dvd-rom
[13:36:22] <McBofh> sailorvrz: no
[13:36:30] <sailorvrz> k thanks
[13:36:36] <McBofh> sorry :|
[13:37:03] <sailorvrz> nwm
[13:37:30] <sailorvrz> it's probably a stupid idea
[13:37:43] *** Rarok has joined #opensolaris
[13:37:43] <tomww> sailorvrz: the singel-cdrom-iso are still build, but dunno if they are downloadable... (-rw-rw-r--   1 root     root     554762240 Sep  8 12:45 solaris_1.iso
[13:38:02] <McBofh> nah, not a stupid idea, just one whose time has past
[13:40:04] *** master_o1_master is now known as master_of_master
[13:40:10] *** jv5181 has quit IRC
[13:40:31] <sailorvrz> I like the Ultrs 10 :) the keyboard and the mouse just rock
[13:40:37] <sailorvrz> *ultra
[13:41:13] <sailorvrz> I wanted to use it as a little home httpd with lighttpd
[13:41:59] <coffman__> wah
[13:42:10] <coffman__> you like the u10
[13:42:11] <coffman__> wtf
[13:42:21] <coffman__> reattching zones...
[13:42:31] <sailorvrz> Yep, rock solid hardware :)
[13:42:34] <coffman__> Remove 116 of 1070 packages
[13:42:51] <coffman__> sailorvrz: not rly, u10 and u5 are the most hated sparc machines
[13:43:07] <coffman__> ide without dma is all i say
[13:43:17] *** coffman__ is now known as coffman
[13:43:19] *** yarihm has joined #opensolaris
[13:43:21] <sailorvrz> As I don't have any other sparc machines... I like the one I have
[13:43:27] <sailorvrz> true, ide sucks
[13:43:27] <coffman> :P
[13:43:47] <sailorvrz> but scsi is too expensive
[13:43:48] <JWheeler> what command would one use to work out which FS the pwd, or a given directory is in?
[13:44:03] <JWheeler> zfs really starts to blur that line, and I'm getting damn confused
[13:44:09] <JWheeler> du -d is about the only tool I have so far
[13:44:30] <McBofh> JWheeler: pwdx no help for you?
[13:45:13] <JWheeler> I've not heard of that one, but no it's not really
[13:45:40] <JWheeler> the confusing thing is that I'm using zfs root, but I have 2 zpools, and it's hard to work out what exists where
[13:46:24] <JWheeler> for example, my Z zpool has a fs mounted in /zones/zonename, and even though /zones doesn't actually exist (in either rpool) it still happily mounts it
[13:46:24] <McBofh> JWheeler: I don't really understand your question, or why "du -d" appears to "solve" it
[13:46:41] <JWheeler> so where is /zones coming from? I was expecting an error
[13:46:54] <McBofh> zfs get mountpoint Z
[13:46:58] <McBofh> zfs get mountpoint Z/zones
[13:47:08] <McBofh> df /zones/zonename
[13:47:12] <JWheeler> oh du -d doesn't really solve it, but it helps to understand when I'm crossing fs boundries... in some circumstances
[13:48:16] <JWheeler> the mountpoint of Z is /Z, Z/zones is set to none, Z/zones/zone1 is set to /zones/zone1
[13:48:55] * McBofh waves hands
[13:49:01] <JWheeler> which would normally mean that /zones isn't on the Z zpool at all, however even when I remove it from / (ie rpool), it still mounts, which means that zones is being created by zfs mount... somewhere!
[13:49:08] <McBofh> ZFS does a whole *heap* of funky stuff under the hood
[13:49:26] <JWheeler> yeah... so I'm finding
[13:49:56] <McBofh> fr'instance, I noticed recently that if you boot -m milestone=none, and have a zfs dataset for your dump device, running "dumpadm -u" will *fail* because the dataset isn't mounted at that point
[13:50:09] <McBofh> But run a zpool list or zfs list before you run dumpadm -u ... all good!
[13:50:38] <JWheeler> I mean, I guess it's great that it's just "taking care of this" for me... except that it appears that the whole logic of zoneadm install is built on working out if the zonepath is on zfs (it is either way) however I guess it's also checking to see if it's crossing zpools
[13:50:44] <JWheeler> and I'm not even sure of that one myself
[13:51:05] <McBofh> probably
[13:51:13] <JWheeler> woah... now that's weird, with extra weird sauce
[13:51:14] <McBofh> and you do know you can utsl for zoneadm
[13:51:20] <JWheeler> and weird sprinkles!
[13:51:34] <JWheeler> I can what?
[13:51:40] <McBofh> Use The Source, Luke
[13:51:41] <McBofh> :)
[13:51:46] <JWheeler> ROFL
[13:52:02] <McBofh> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/zoneadm/
[13:52:14] <JWheeler> <<< the source is weak, in this one
[13:52:26] <McBofh> there's a zfs.c file in that directory, so clearly there's some handling going on
[13:52:40] <JWheeler> I did try truss quickly, and worked out that it was looking at /etc/mnttab
[13:54:06] <McBofh> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/zoneadm/zfs.c#92
[13:54:22] <McBofh> yes, it calls getmntent()
[13:54:31] * JWheeler blinks
[13:54:41] <McBofh> and Does Stuff(tm)
[13:54:55] <McBofh> so is it a problem you've got, or do you just want to understand what's going on?
[13:55:46] <JWheeler> well, I guess I've worked past it, and discovered some holes in my understanding; so suddenly my problem has forked children
[13:55:53] <McBofh> heh
[13:56:21] <JWheeler> on the one hand, mounting the parent directory (fs) has solved the problem of making my clone
[13:57:01] <JWheeler> and my primative code reading skills tell me that from the source that you've pointed me at, zoneadm requires that the parent directory be a _mounted_ zfs...
[13:57:18] <JWheeler> which I kinda worked out by trial and error... I don't really understand why this should be a requirement
[13:58:10] <JWheeler> and on the other - I've discovered that zfs mount seems to "make" missing parents, that are needed for nested mounts to succeed... which isn't strictly related at all, but now I'm even more perplexed
[13:58:17] *** alibb has quit IRC
[13:58:20] <McBofh> if it's not mounted, how would you know what directories are underneath it?
[13:58:38] <JWheeler> child fs?
[13:58:40] <McBofh> yes, ZFS will do some default mount things
[13:59:06] <McBofh> if I've got /a, /a/b and /a/b/c, but /a/b is not mounted, how do I get from /a to /a/b/c ?
[13:59:15] *** yarihm has quit IRC
[13:59:22] <McBofh> assuming that there's no underlying dir structure created in /a/
[14:00:00] <JWheeler> oh, I understand why it's handy that /a/b be "created" for me
[14:00:25] <McBofh> so zoneadm is playing it safe
[14:00:26] *** communicator has joined #opensolaris
[14:00:38] <JWheeler> well we're talking about two different things, aren't we?
[14:00:56] <McBofh> possilby
[14:01:02] *** gottadoit has joined #opensolaris
[14:01:03] * McBofh falls asleep @ kbd
[14:01:05] <JWheeler> zoneadm creates it's own zfs under the parent, so mounted or not, it'll succeed
[14:01:13] <McBofh> I don't think so
[14:01:32] <JWheeler> there is the possibity that the namespace will clash, but you won't know that until the parent dataset is mounted
[14:02:57] <McBofh> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/zoneadm/zfs.c#708
[14:03:01] <McBofh> and on that note, gnite
[14:03:13] <JWheeler> thanks for your input McBofh :)
[14:03:17] <JWheeler> I should do the same
[14:04:37] *** scimmia has joined #opensolaris
[14:04:45] *** scimmia has quit IRC
[14:06:40] *** stukag has joined #opensolaris
[14:28:01] *** lewq has quit IRC
[14:31:39] *** bahumbug has quit IRC
[14:38:05] *** mega has joined #opensolaris
[14:38:31] *** ksyl_ has joined #opensolaris
[14:38:33] <ksyl_> hello
[14:38:42] <sickness> hello
[14:39:48] <ksyl_> Someone now what is the simple way for adding a kernel module in OpenSolaris?
[14:39:57] <ksyl_> I found some tutorial on the net
[14:40:18] <ksyl_> but I m need to rebuild all the kernel in this howto..
[14:40:34] <ksyl_> does it exist simple way to do that ?
[14:41:23] <sickness> download just one of the drivers of masayuki murayama, and look at the script inside the package :)
[14:41:58] *** eugene has joined #opensolaris
[14:41:59] <ksyl_> the driver have sources included in the package ?
[14:42:11] <ksyl_> it is not a binary package ?
[14:42:15] <sickness> yeah
[14:42:20] <ksyl_> ok
[14:42:25] <ksyl_> thanks for the help
[14:42:25] <sickness> they have sources and instructions to compile and install
[14:42:27] <sickness> yw
[14:46:15] *** scimmia has joined #opensolaris
[14:46:23] *** scimmia has left #opensolaris
[14:46:25] *** eugene has left #opensolaris
[14:49:28] *** mega has quit IRC
[14:51:14] <ksyl_> thanks sickness I download the tarball and its exactly what I need :)
[14:52:29] <sickness> yw :)
[15:06:15] * codestr0m has broken down and decided to do things the "SUNW" way and use /usr/local/src/SUNWspro/bin/cc
[15:09:47] *** Fish has quit IRC
[15:14:32] *** jgracin has joined #opensolaris
[15:15:54] <codestr0m> anyone able to advise jds-cbe-1.6.1 with SS12..  http://rafb.net/p/Nk80Wv24.html  I'm getting /export/home/asura/packages/BUILD/libtool-1.5.22/missing[52]: aclocal-1.9a: not found [No such file or directory]
[15:16:47] <codestr0m> I've no idea why it's appending the a at the end
[15:18:08] *** LordIllpalazo has joined #opensolaris
[15:23:28] *** jv5181 has joined #opensolaris
[15:24:19] *** Ivan_Shih has joined #opensolaris
[15:24:20] *** zenbalrog has quit IRC
[15:27:31] *** Vagrant has quit IRC
[15:43:56] <glance> ksyl_: the kernel modules i have compiled from source has bin quite easy...
[15:48:18] *** kimc has quit IRC
[15:53:52] <TomJ> is it possibly to restrict ssh to login with ssh key only, while allowing password-based logins by ssh for non-root users?
[15:54:53] *** alibb has joined #opensolaris
[15:59:33] *** anilg has quit IRC
[16:02:41] *** xRaich[o]2x has joined #opensolaris
[16:04:18] <xRaich[o]2x> hi folks
[16:10:27] *** spiki has joined #opensolaris
[16:11:35] <codestr0m> xRaich[o]2x: hej
[16:11:50] <xRaich[o]2x> hey codestr0m
[16:12:27] <codestr0m> TomJ: that didn't make any sense.. root shouldn't be able to directly login remotely
[16:12:43] <TomJ> the answer is PermitRootLogin without-password
[16:12:54] <TomJ> and for the moment, I want remote root login by ssh, if they have a predefined key
[16:12:57] *** edgy has joined #opensolaris
[16:13:45] <codestr0m> xRaich[o]2x: when you were messing with tcsh did you give periodic a try? (tcsh periodically run command as defined by tperiod)
[16:14:00] <xRaich[o]2x> yep
[16:14:11] <codestr0m> I can't get it to work
[16:14:32] <codestr0m> I want every minute for the history to save and merge
[16:14:37] <codestr0m> doing it manually works
[16:15:06] <xRaich[o]2x> i think i works like this: after counting down the time you defined you need to issue another command to make periodic execute.
[16:15:28] <xRaich[o]2x> sorry, just woke up, so i'm not fully there ^^
[16:15:35] <TomJ> in bash I do that with PROMPT_COMMAND="history -a"  #  PROMPT_COMMAND being a command that is executed just prior to a new prompt being printed.  maybe tcsh has an equivalent
[16:16:02] <edgy> Hi, eeprom displays a value for bootpath but man eeprom doesn't mention this variable and instead there is something called boot-device, are they the same? why the confusion?
[16:16:08] <codestr0m> TomJ: base has append history so you don't even have to do that
[16:17:02] <TomJ> setting append history usually means that it only appends on shell exit.  doing history -a after each command is better, it means you dont lose any history if your shell terminates or becomes inaccessible
[16:17:03] <codestr0m> , but it does give me a good idea. thanks
[16:17:08] *** PicCard has joined #opensolaris
[16:17:39] <TomJ> in Bash, I do both:  shopt -s histappend  and the PROMPT_COMMAND.
[16:19:14] <edgy> TomJ: so in solaris the history would be gone if I won't save it by history -a before quitting?
[16:20:03] <codestr0m> edgy: which shell are you using?
[16:20:11] <TomJ> edgy: by default, Bash (and most other shells I think) will do a history replace. that means on exit they truncate the .bash_history and replace it with the contents of its own history buffer
[16:20:13] <TomJ> that is very bad
[16:20:18] <TomJ> it means if the shell terminates early, nothing is saved
[16:20:27] <edgy> codestr0m: bash
[16:20:33] <TomJ> it also means that if you are logged int wice to the same account, the history for the last shell to exist will overwrite the previous one
[16:20:35] <codestr0m> well. for me it's a major pita since I'm using multiple tabs
[16:20:43] <codestr0m> so I close 10 tabs and last one wins
[16:20:45] <codestr0m> that's no good
[16:20:45] <TomJ> edgy: read http://wooledge.org/mywiki/BashFAQ/088
[16:20:47] *** cypromis has joined #opensolaris
[16:21:07] <codestr0m> the other problem is when I close the tab. it doesn't exit. it just like disappears
[16:21:24] <codestr0m> I was going to change the gnome-terminal source, but that lead me down this path.. which is just a mess
[16:21:40] <TomJ> the two solutions are to set history append, and to run history -a after every command.    the former ensures that on close, history files just append their content not overwrite, so multiple tabs don't overwrite.  the second saves history after every command, meaning that premature shell death/disconnection will not prevent history writing
[16:21:46] <TomJ> gnome terminal sucks
[16:22:05] <codestr0m> ok. ok.. we don't need to go down this road.. :P
[16:22:13] <TomJ> it does though
[16:22:14] <edgy> TomJ: forget about multi-tabs or if I am logging twice. Let's say I logged once, typed some commands and rebooted, would I find the history?
[16:22:17] <TomJ> this is not a gnome vs kde argument
[16:22:20] <TomJ> gnome terminal just blows
[16:22:24] <codestr0m> TomJ: thanks
[16:22:46] <TomJ> edgy: if you follow the instructions in that link, yeah
[16:22:48] <codestr0m> edgy: I don't know if reboot calls exit
[16:22:56] <TomJ> PROMPT_COMMAND="history -a" will take care of the reboot case
[16:23:28] <edgy> TomJ: but in bash in linux I don't need this history -a, no?
[16:23:38] <codestr0m> edgy: yes you do
[16:23:42] <TomJ> edgy: you do
[16:23:46] <TomJ> I dont think any distro implements that yet
[16:23:48] <TomJ> my OpenSUSE didnt
[16:23:52] <TomJ> didnt even do histappend
[16:24:14] *** Ivan_Shih has quit IRC
[16:24:17] <TomJ> the other thing is that the default history file size of 500 is woefully inadequate
[16:24:22] <TomJ> so my link above also teells you to increase that
[16:24:30] <TomJ> I have history files of 15,000+ lines now
[16:24:51] <edgy> codestr0m, TomJ: I am using ubuntu now and I never do history -a but when I reboot I find my history
[16:24:55] <TomJ> hell I've written 2000 history lines in one box in 2 days
[16:25:13] <TomJ> edgy: shrug, maybe they have put it or something similar in then. who knows
[16:25:22] <codestr0m> TomJ: fyi. I just checked. tcsh calls it precmd
[16:25:32] <TomJ> cool
[16:26:43] <edgy> TomJ: I am new to solaris but I used linux (redhat, mandrake/mandriva/, ubuntu) since ages and never need to do any thing manual to change the history, but when I tried solaris I discovered the history is gone and never knew there is an option I need to set till I saw your discussion now ;)
[16:27:16] <houst0n-> tcsh>*
[16:29:12] <edgy> TomJ: also do you have any idea why the append option is not set by default in bash?
[16:30:17] *** Stun__ has joined #opensolaris
[16:30:23] <codestr0m> edgy: because unix admins in general are so personal about this stuff that it's safer to set nothing and let everyone screw it up on their own
[16:30:50] <edgy> codestr0m: good reason ;)
[16:31:05] <TomJ> yeah something like that
[16:31:11] <TomJ> i cant think of any sane reason why histappend shouldnt be default
[16:31:16] <TomJ> it drove me mad the default behaviour
[16:31:35] <codestr0m> TomJ: yeah. you think this, but what if someone has a security concern for not wanting it saved
[16:31:47] <codestr0m> first time they don't check. they'll yell at someone and have it removed
[16:31:51] <TomJ> well then he woudlnt want history at all , would he :)
[16:32:07] <codestr0m> yeah, but put that in the default and someone will complain
[16:32:53] <edgy> so in my new session if I want to exit without saving the history of that session, can I?
[16:33:06] *** alibb has quit IRC
[16:33:11] <TomJ> edgy: with shopt -s histappend and history -a in PROMPT_COMMAND, you can do whatever you like
[16:33:25] <TomJ> you will never lose more than the last line of history
[16:33:36] <TomJ> and you only lose the last line if it didnt complete execution
[16:33:57] <edgy> TomJ: no, I want to lose my history of the session, shopt -s histappend won't do it, would it?
[16:34:01] <TomJ> that's because PROMPT_COMMAND is run after the command, and before the next one is processed.  so if hte last command exits or never completes, its history is not written
[16:34:12] <TomJ> you want to LOSE your history?
[16:34:25] <edgy> TomJ: yes, for the current session only
[16:34:34] *** jgracin has quit IRC
[16:34:51] <TomJ> then you dont want history -a being run, and you would do HISTFILE=/dev/null or unset HISTFILE in that session
[16:35:16] <edgy> TomJ: ah! ok
[16:35:43] <TomJ> if you set histappend then history is appended at the end of a session, not overwritten. so you can stil choose not to have the history for a session by killing the shell improperly, or changing HISTFILE as above
[16:35:53] <Auralis> i never ever used a shell history from file
[16:36:02] <TomJ> if you do the history -a PROMPT_COMMAND thing also, there is no way to stop history for the current session being written
[16:36:18] <TomJ> or at least, not historically - you can unset PROMPT_COMMAND HISTFILE  # to stop it from this point forward
[16:36:28] <TomJ> Auralis: you dont use a shell history file?
[16:36:36] <Auralis> nope, not at all
[16:36:39] <TomJ> wow, I cant imagine that
[16:36:48] <TomJ> Control-R (search backwards in history) must be two of my most used keys
[16:36:53] <TomJ> I never retype anything of any length
[16:37:21] <codestr0m> TomJ: since we're way off topic.. what do you think of auto-color ;)
[16:37:29] <TomJ> especially when you're doing stuff like:  for zone in $(zoneadm list | grep -v global) ; do zlogin $zone 'svcadm restart blah' ; done
[16:37:37] <Auralis> i type faster then i can find stuff in a long history
[16:37:40] <TomJ> I could function that I suppose but I tend to just history search and repeat on the command line
[16:37:50] <edgy> TomJ: I tried PROMPT_COMMAND="hi " in solaris and it claims command not found, this is bash btw
[16:38:02] <TomJ> edgy: err yeah, obviously?  what is 'hi' meant to be?
[16:38:05] <jamesd_> Auralis, then you must be a damm good typist, and have a perfect memory or never do the same job twice...
[16:38:14] *** sola has joined #opensolaris
[16:38:29] <edgy> TomJ: just a value for a variable, can't I set any value arbitrary
[16:38:32] <TomJ> codestr0m: auto-color for ls?  I like it, but I also can't lose ls -V -v  so I have alias ll='gls -al' and normal ls is uncoloured
[16:38:39] <TomJ> edgy: no, it's a COMMAND.. hence the name
[16:39:02] <codestr0m> TomJ: that's why I switched over to tcsh.. has built-in color and I get a proper ls :)
[16:39:05] <Auralis> jamesd_: hehe, no, i just the session history, but not a history file
[16:39:08] <TomJ> codestr0m: I thought about writing a function called ls which would called gls unless ls -V or -v was requested, but havent got around to it yet
[16:39:16] <sola> help please, i installed SXDE and trying to start dtrace , but when run from terminal command not found err coming, how to enable or install
[16:39:34] <Auralis>  /usr/sbin/dtrace
[16:39:49] <TomJ> codestr0m: hehe, I'd rather find a function or whatever solution and use bash or ksh93. isn't tcsh a little.. crusty?
[16:39:52] *** SpyKee has joined #opensolaris
[16:39:54] *** spiki has quit IRC
[16:40:01] <codestr0m> TomJ: not at all
[16:40:11] <codestr0m> the completion and interactive capabilities are nice
[16:40:17] <edgy> TomJ: how can I set a value to variables in solaris? in linux bash I would just do PROMPT_COMMAND='whatever'; export PROMPT_COMMAND
[16:40:29] <im_alone> i use bash for interactive consoles, tcsh for background scripts, ...
[16:40:40] <Auralis> zsh ftw! :)
[16:40:48] <codestr0m> zsh has bugs :)
[16:40:58] <Auralis> hint: every shell has
[16:41:09] <edgy> codestr0m: but it has the best autocomplete feature ;)
[16:41:09] <TomJ> can it parameter expansion?  stuff like:   for html_file in *.html ; do  html_to_pdf "$html_file" > "${html_file%.*}".pdf ; done ?   i.e. run  html_to_pdf blah.html > blah.pdf
[16:41:13] <codestr0m> xRaich[o]2x: did you say zsh had *critical* bugs ?
[16:41:29] <TomJ> edgy: exactly the same.. it's the same shell
[16:41:32] <xRaich[o]2x> codestr0m: i fixed that one a few days ago
[16:41:40] <codestr0m> did you?
[16:41:44] <edgy> TomJ: in solaris it complains command not found
[16:41:46] <codestr0m> did it make it into 97?
[16:41:46] <xRaich[o]2x> i found a horrible memory leak
[16:41:56] <TomJ> edgy: then you're doing something wrong
[16:41:57] <xRaich[o]2x> don't think so.
[16:41:58] <sola> Auralis, thank you..will set the path to /usr/sbin
[16:42:04] <codestr0m> have a patch?
[16:42:05] <edgy> TomJ: let me recheck ...
[16:42:08] <xRaich[o]2x> sure
[16:42:12] <codestr0m> and are you back to using zsh?
[16:42:19] <im_alone> my brain is harm and hurt
[16:42:24] <TomJ> edgy: remember PROMPT_COMMAND is asking you for a command to run before every prompt. a command!  so PROMPT_COMMAND='hi' is gogint to fail because there is no such commadn as hi!
[16:42:36] <xRaich[o]2x> right now, i'm using zsh yes. tried ksh for some time. which is also a great shell
[16:42:44] <xRaich[o]2x> k93 that is
[16:42:46] <codestr0m> I couldn't bare ksh
[16:42:53] <TomJ> ksh93 is superb for programming, the best of all of them I think. not sure how good it is interactively though
[16:43:02] <im_alone> why not at unifying shell?
[16:43:16] <xRaich[o]2x> TomJ: the ksh93 integration on opensolaris is pretty nice
[16:43:18] <codestr0m> TomJ: for interactive I couldn't deal with k93
[16:43:20] <jamesd_> i even use history to watch pr0n^wmovies  history | grep mplayer | grep -v history | head -n 1
[16:43:21] <jamesd_>    58  find . -type f -size +10000 -mtime -17 -atime -3   -not -name '*[ZzJRjUuPpam][RIPpDdIi][PpFfGgk]'  -exec mplayer -idx  -vo gl2  "{}" \;
[16:43:26] *** storycrafter has joined #opensolaris
[16:43:41] <xRaich[o]2x> im_alone: different users want different shells
[16:43:51] <xRaich[o]2x> same goes for desktops etc
[16:44:14] <codestr0m> xRaich[o]2x: is your patch going to make it upstream? (I assume so)
[16:44:55] <TomJ> xRaich[o]2x: yeah it looks it. I'm on Solaris 10 and I rolled mty own but I'm looking to try and backport the snv integration to sol10 at some point
[16:44:58] <xRaich[o]2x> codestr0m: pretty sure about that.
[16:45:21] <xRaich[o]2x> codestr0m: if you want the patch and completion for ips check my blog raichoo.blogspot.com ;)
[16:45:21] <codestr0m> xRaich[o]2x: if you post any patch you have to fix it I'll give zsh a shot
[16:45:30] <codestr0m> k
[16:45:32] <edgy> TomJ: yes I got it thanks a lot
[16:45:47] * codestr0m reading german
[16:45:53] <xRaich[o]2x> wrote a script for it yesterday
[16:46:20] <xRaich[o]2x> codestr0m: just pick copy the patch. it's a twoliner you can even do that by hand ^^
[16:46:33] <TomJ> the never-lose-a-history line has really revolutionised my admining.  I can recall a command I wrote 9 moonths ago at 3.47am during an outage which I forgot to make a note of and suddenly I need again urgently.
[16:46:48] <edgy> svcs displays enabled service in the default milestone but if I change to multi-user e.g svcs would even display disabled services, any idea what does this means?
[16:46:52] <codestr0m> xRaich[o]2x: I see that..  do you have the link to the zsh source + patches for os2008* ?
[16:47:37] <xRaich[o]2x> that patch is against the newest version of zsh, just download that
[16:48:10] *** jv5181 has quit IRC
[16:48:16] <xRaich[o]2x> 4.3.6 that is
[16:49:27] <codestr0m> is it just me or is the completion for zsh more ugly than tcsh ?
[16:49:44] *** sola has quit IRC
[16:49:53] <xRaich[o]2x> Auralis: want ips completion for zsh? ;)
[16:50:14] <Auralis> xRaich[o]2x: not using ips yet, but that would be handy i bet
[16:50:17] *** mui has joined #opensolaris
[16:50:19] *** Rarok has quit IRC
[16:50:33] <xRaich[o]2x> Auralis: http://docs.google.com/View?docID=dfdnr8dn_1f89hktdc&revision=_latest
[16:50:49] <xRaich[o]2x> wrote that yesterday might still have some bugs but well. patches welcome
[16:51:08] <xRaich[o]2x> writing completion for zsh is pretty horrible ^^
[16:51:18] <edgy> codestr0m: for me it looks nicer because it doesn't scroll you terminal
[16:51:24] <mui> beep
[16:51:30] <codestr0m> (horrible)
[16:51:36] <mui> anybody have any experience using areca raid controllers on solaris/opensolaris?
[16:51:50] <xRaich[o]2x> codestr0m: it's pretty powerful though
[16:52:05] <xRaich[o]2x> codestr0m: and most of it is already there
[16:52:15] <codestr0m> xRaich[o]2x: if you ^r  to search back in history and then ^right arrow.. do you get ";5C"
[16:52:43] <edgy> codestr0m: with zsh you don't even need to complete the command manually as in bash, it cycles automatically in all of them, isn't that fancy?
[16:52:48] <codestr0m> tcsh and zsh do it, but bash doesn't.. because the other two are doing some vim escape thing
[16:53:13] *** jv5181 has joined #opensolaris
[16:53:17] * xRaich[o]2x is using zsh with vi
[16:53:20] <xRaich[o]2x> vimode
[16:53:46] <codestr0m> edgy: what I mean is you find the patch. .and want to just jump to the end of the word.. and it prints that out instead of jumping to the end of the word
[16:54:07] <xRaich[o]2x> i'd still prefer ksh93 with completion an some minor changes for prompts
[16:54:31] <TomJ> xRaich[o]2x: you ever find any ksh93 completion?
[16:54:55] <xRaich[o]2x> TomJ: not really. searched the whole net. nothing found :/
[16:55:50] <xRaich[o]2x> which is sad. because ksh is a pretty fast shell. zsh is... well slow
[16:56:24] <edgy> codestr0m: sorry I didn't understand you. in zsh you cannot jump to the end of the word?
[16:56:41] <Auralis> edgy: sure you can
[16:57:09] <edgy> Auralis: yes I know but codestr0m is complaining about something I don't know what
[16:57:11] <codestr0m> edgy: ^r first "bck-i-search: _" and then try to ctrl+right arrow
[16:57:23] <xRaich[o]2x> you can do pretty much everything in zsh.... some might say you could code a kde replacement in it :P
[16:58:35] <edgy> codestr0m: sorry but I don't understand this bck-i-search stuff :(
[16:58:41] <edgy> codestr0m: any doc?
[16:59:11] <xRaich[o]2x> edgy: search for zshguide on the net
[16:59:32] *** smtms has quit IRC
[16:59:33] <xRaich[o]2x> you will get a 300+ pages book about the shell :P
[16:59:34] <codestr0m> edgy: press ctrl + r and you should see
[17:01:06] <xRaich[o]2x> mmh 5pm.... time for breakfast
[17:02:25] <coffman> hmm
[17:02:32] <coffman> 17.01
[17:03:09] <coffman> still time for breakfast
[17:03:28] <xRaich[o]2x> no! wie only server breakfast till 5pm
[17:03:31] *** KOHJU is now known as kohju
[17:03:43] <xRaich[o]2x> my gosh..... what was that!
[17:03:50] <xRaich[o]2x> let's try again:
[17:04:08] <xRaich[o]2x> no! we only serve breakfast till 5pm
[17:04:21] <xRaich[o]2x> i should drink less coffee -_-
[17:04:29] <coffman> hmm coffee
[17:09:25] *** sporq has quit IRC
[17:09:28] *** Gnu_Raiz has joined #opensolaris
[17:10:43] *** storycrafter has quit IRC
[17:11:07] *** sporq has joined #opensolaris
[17:11:52] <edgy> codestr0m: aha! that's nice
[17:12:28] <edgy> codestr0m: the first time I used right-ctrl and this is why I didn't see any thing ;)
[17:12:38] *** sporq has quit IRC
[17:13:27] <codestr0m> edgy: ? ctrl + right arrow should move forward one word
[17:15:03] *** LordIllpalazo has quit IRC
[17:15:41] <codestr0m> xRaich[o]2x: do you have zpool autocomplete?
[17:16:08] <TomJ> zfs fs complete would be much more useful
[17:16:10] <xRaich[o]2x> zsh already has. but i think is b0rked
[17:16:26] *** pjfloyd has joined #opensolaris
[17:16:46] <codestr0m> TomJ: I have a list of things I'm going to write completes for. that's on it
[17:17:07] *** seanmcg has joined #opensolaris
[17:17:14] <xRaich[o]2x> codestr0m: zsh on solaris comes with zfs und zpool completion
[17:17:24] <xRaich[o]2x> pfexec also
[17:17:58] <coffman> xRaich[o]2x: i have no coffee, but i have club mate!
[17:18:05] <codestr0m> xRaich[o]2x: umm. mine must be missing.. cause I type zfs tab-tab and I just start cycling through directories
[17:18:31] <xRaich[o]2x> codestr0m: run autoload -U compinit
[17:18:39] <xRaich[o]2x> and then compinit
[17:18:53] <xRaich[o]2x> voila compsys is loaded and completion should work
[17:19:12] <pjfloyd> any idea which borked GNU header defines G_HAVE_ISO_VARARGS? glib or something I think
[17:19:21] <xRaich[o]2x> you can cache compinit with -d cachefile that should speed up zsh startup
[17:20:07] *** LordIllpalazo has joined #opensolaris
[17:20:16] <codestr0m> how do you go back if you tab tab past the option you want?
[17:20:57] <TomJ> hopefully shift tab?
[17:21:16] <xRaich[o]2x> still searching for it should be possible but i don't know how
[17:21:21] <xRaich[o]2x> ping me if you find out ^^
[17:21:49] <codestr0m> #zsh has a lot of people
[17:22:06] <houst0n-> ugh zsh, does it still have a built in ftp client?
[17:22:07] <xRaich[o]2x> a lot of people that say very little
[17:22:19] <xRaich[o]2x> houst0n-: only if you load it
[17:22:28] <houst0n-> bloatware if you ask me
[17:22:41] <houst0n-> but no one did, so i'll remain silent
[17:22:45] <xRaich[o]2x> it's pretty bloated yes
[17:22:45] <houst0n-> ;)
[17:23:34] <xRaich[o]2x> i still prefer it over bash since the completion system is better.
[17:24:00] <houst0n-> I started life with freebsd so stuck with tcsh
[17:24:22] <houst0n-> I really like the way my history browsing works
[17:24:27] <houst0n-> I'm sure you can do the same on basj
[17:24:40] <xRaich[o]2x> i started out with bash so c-shell syntax feels weird to me ^^
[17:25:07] <houst0n-> Basically my history browsing is like, s<up> browses all commands starting with s etc
[17:25:20] <houst0n-> very handy if i need to recheckout something from svn or w/e, i just hit s<up><enter>
[17:25:51] <xRaich[o]2x> history browsing in tcsh is awesome i agree
[17:25:55] <TomJ> i much prefer Control-R myself, then you can search any part of the line
[17:26:28] <TomJ> I think you can get that kind of up arrow completion in bash also, though, if you like it
[17:26:34] <codestr0m> TomJ: yeah I prefer the complete as you type as well
[17:26:39] <houst0n-> Well, at the end of the day, our personal pref in shells is always going to be our own as it's the main way we've interacted with our systems for.. forever
[17:26:45] <jamesd_> i knew C before  shell, yet i still dont stick to tcsh.
[17:26:54] <houst0n-> On a personal note though - bash is disgusting
[17:26:55] <houst0n-> ;)
[17:26:58] <houst0n-> It's like emacs
[17:27:03] <houst0n-> </flamebait>
[17:27:59] <TomJ> bash is a great general purpose shell. its coding is better than most, nearly as good as ksh93, but it is nicer interactive than ksh93.   the only one I'm unsure of zsh, it seems to offer nearly all the coding of bash but with better completion. and the bloat, of course
[17:28:12] *** Stun__ is now known as Axz
[17:28:35] <TomJ> anyway, all of these shells are very good. whatever you use, just make sure you learn all it can do
[17:29:27] <houst0n-> What sort of prompts does everyone use? I'm playing around trying diff ones atm
[17:29:41] <houst0n-> set prompt = "%{\033[0;1;34m%}(%{\033[0;1;36m%}%m%{\033[0m%}:%{\033[0;1;36m%}%c3%{\033[0;1;34m%}) %{\033[0;1;36m%}%n%{\033[0;1;34m%} >%{\033[0m%} "
[17:29:46] <houst0n-> is my current, bit... hard to read
[17:29:49] <xRaich[o]2x> just a basic one with a few colors
[17:30:03] <xRaich[o]2x> houst0n-: what shell?
[17:30:06] <houst0n-> tcsh
[17:30:12] <xRaich[o]2x> ah ok forgot sorry
[17:30:16] <Auralis> prompt='%n@%m:%~ > '   zsh
[17:30:36] <TomJ> I just use "\u@\h \w:$ "  and then I set a PROMPT_COMMAND that echoes xterm codes so that the same display appears in xterm tabs/window title
[17:30:56] <houst0n-> I've never bothered setting that up, it is a nice idea though
[17:31:06] <Auralis> i have that as well
[17:31:08] <houst0n-> What do you have set as prompt_command?
[17:31:19] <edgy> codestr0m: ctrl + right arrow doesn't jump one word in zsh and bash, where does it do this?
[17:32:02] <TomJ> PROMPT_COMMAND='history -a ; echo -ne "\033]0;${LOGNAME}@${HOSTNAME} : ${PWD}\007"'
[17:32:33] <codestr0m> edgy: bindkey "^[[1;5C" forward-word
[17:33:09] *** Odin- has joined #opensolaris
[17:33:26] <edgy> codestr0m: man this is getting complicated! I don't know how bind keys!
[17:33:41] <xRaich[o]2x> edgy: zsh is complicated ^^
[17:33:54] <codestr0m> edgy: that's for tcsh
[17:34:08] <codestr0m> so you'll need the bind keys for bash or whatever you use
[17:34:16] <edgy> codestr0m: and on tcsh it works?
[17:34:32] <codestr0m> press ctrl+v and then the keystroke you want to get the unescape encoding
[17:34:58] <edgy> codestr0m: in tcsh ctrl+r shows nothing
[17:34:59] <codestr0m> edgy: if you are in tcsh and just paste what I wrote.. (using gnome-terminal) you'll get ctrl+right arrow to jump one word forward
[17:35:25] <codestr0m> edgy: right.. that would be bindkey "^r" i-search-back
[17:35:39] <codestr0m> I prefer that over the bindkey "^r" vi-search-back
[17:36:03] <xRaich[o]2x> TomJ: are they still actively developing bash?
[17:36:16] <houst0n-> What's the ^[[1: for left arrow?
[17:36:53] <pjfloyd> any ideas where I should report a bug in gstreamer delivered in SXCE 97?
[17:37:07] <edgy> codestr0m: I changed to csh and tried your bindkey command, then ^r still reveals nothing
[17:37:17] <codestr0m> csh != tcsh
[17:37:18] <houst0n-> pjfloyd: bugs.opensolaris.org ?
[17:37:37] <pjfloyd> I'll give it a go
[17:38:50] <edgy> codestr0m: in tcsh tab even doesn't work to complete the command, how can one live with this shell? ;)
[17:39:10] <houst0n-> edgy: Errr....
[17:39:15] <codestr0m> edgy: 15K   /export/home/asura/.tcshrc
[17:39:16] <houst0n-> wat?
[17:39:32] <codestr0m> 55K   /export/home/asura/.tcsh_complete
[17:39:39] <houst0n-> codestr0m: that's ... massive...
[17:39:44] <houst0n-> could I get a look at it?
[17:39:46] <houst0n-> ;)
[17:40:01] <codestr0m> houst0n-: yeah. let me finish some of my completes to work for opensolaris
[17:40:10] <houst0n-> cool =)
[17:40:14] *** communicator has quit IRC
[17:40:35] <xRaich[o]2x> codestr0m: did you enhance some of my completions?
[17:40:44] <xRaich[o]2x> i'd like to take a look at it :)
[17:41:24] <edgy> codestr0m: I don't have .tcsh_complete what's that? I am feeling now like an alien visiting earth the first time
[17:41:58] <houst0n-> edgy: Well, I don't have that either and I've been using tcsh for some 8 years....
[17:42:45] <edgy> houst0n-: so they are implementing their own completion or what?
[17:42:53] * houst0n- shrugs
[17:43:34] <edgy> ok I am willing to forget this shell stuff. it's confusing me more in this stage of my learning curve ;)
[17:43:44] <xRaich[o]2x> they just have a separate file for the completion stuff
[17:44:11] <edgy> I have a simple question, eeprom displays a value for bootpath but man eeprom doesn't mention this variable and instead there is something called boot-device, are they the same? why the confusion?
[17:47:03] *** jfndi_ has joined #opensolaris
[17:47:47] <edgy> let me answer myself with a guess. boot-device is the device used for booting
[17:47:53] <Auralis> yes
[17:47:55] <edgy> bootpath is the partition in that device
[17:48:16] <edgy> but it's not in man!
[17:49:07] <Auralis> beats me, don't have bootpath on sparc
[17:49:36] *** desai has joined #opensolaris
[17:51:02] <codestr0m> edgy  houst0n- I'm just doing source ~/.tcsh_complete at the bottom of tcshrc to keep the completions a bit more clean and separate
[17:51:17] *** sponix has joined #opensolaris
[17:51:44] <sponix> when is the next OpenSolaris release due out, the one to take the place of os200805 ?
[17:51:52] <codestr0m> my tcshrc is just a bunch of stuff I've nabbed from the net and various other bits I've tweaked to my own design
[17:52:17] <codestr0m> sponix: http://www.genunix.org/ normally respins with each update
[17:52:51] <edgy> Auralis: so in sparc when you type eeprom only you don't see bootpath listed?
[17:53:20] <Auralis> edgy: right, no bootpath listing at all
[17:53:53] <edgy> Auralis: strange! I am using x86 now and it's there,
[17:54:29] <edgy> Auralis: so how do you set the partition to boot from in sparc?
[17:54:46] <Auralis> edgy: with boot-device
[17:55:00] <Auralis> boot-device=/pci@8,600000/SUNW,qlc@4/fp@0,0/disk@w210000203742561e,0:a
[17:55:21] *** Barack_Wabama is now known as rand7
[17:55:26] <edgy> Auralis: Umm! this is the value at my bootpath ;)
[17:55:55] <edgy> so it's a bit confusing
[17:56:23] <sponix> codestr0m:  you ran any of those pre-releases ?
[17:56:27] <desai> hi. i started with an os-200805 install inside of virtualbox. I've done an upgrade (following the manual upgrade process). After the upgrade, the b97 kernel fails to boot, but the b86 kernel still works fine. I think there is an error message printed, but vbox takes me back to the grub menu instantaneously, so i can't see it. I have noticed that b86 boots 64bit but b97 boots 32bit. Can anyone point me in the right direction?
[17:56:54] <xRaich[o]2x> desai: did you follow the upgrade guidelines?
[17:56:57] <houst0n-> boot-device=/pci@1f,0/pci@1,1/ide@3/disk@0,0:a disk:a disk net <-- bit easier to read
[17:57:06] <codestr0m> sponix: yeah. I'm currently on the one which was from august and yesterday tested the sept release on the 2nd dell laptop
[17:57:18] <desai> xRaich[o]2x: yeah, i did the manual beadm create and manual activation
[17:57:54] *** pjfloyd has left #opensolaris
[17:57:55] <houst0n-> desai: Did you upgrade virtualbox between these?
[17:58:13] <desai> nope,  i just did the install last night on vbox 2.0.2
[17:58:44] <houst0n-> Weird that the kern arch changed, if it was installed on 2.0.2 is should have used 64bit kern all the time :/
[17:58:54] <sponix> codestr0m:  these releases offer more updated programs as well, or just the os core updates ?
[17:58:56] <houst0n-> well, by default
[17:59:07] <codestr0m> sponix: to the best of my knowledge it's just a respin
[17:59:25] <desai> houst0n-: yeah, that is what i was thinking
[17:59:27] <sponix> codestr0m:  I was thinking the os200805 is getting very dated, and that they should be releasing another major release soon, but haven't seen anything posted as to when
[17:59:41] <codestr0m> sponix: just install it and update
[17:59:44] <codestr0m> what more do you want?
[17:59:54] <houst0n-> There are current-ish iso's on genunix.org
[18:00:02] <Auralis> sponix: the 2008.11 prerelease is on the opensoalris download page
[18:00:13] <sponix> houst0n-:  codestr0m already pointed me to those
[18:00:24] <sponix> Auralis:  really, I missed that :(
[18:01:14] *** rand7 has quit IRC
[18:01:49] <sponix> Auralis:  I see, they are posting to genunix.org, and codestr0m already steered me in that direction
[18:02:00] <sponix> codestr0m:  thanks btw
[18:02:03] * Auralis nods
[18:02:22] * desai didn't know about the 2008.11 prerelease; I will give that a spin. thanks for the info
[18:02:30] *** master_of_master has quit IRC
[18:02:32] *** psychonate has joined #opensolaris
[18:03:14] *** lanc has left #opensolaris
[18:05:11] *** yarihm has joined #opensolaris
[18:05:26] *** master_of_master has joined #opensolaris
[18:05:45] *** desai has left #opensolaris
[18:06:30] <sponix> Man I love torrents, http was getting 35K/s, torrent is getting well over 100K/s ;)
[18:07:26] <edgy> Auralis: svcs displays only the enabled services for you or includes disabled too?
[18:07:56] <Auralis> enabled only -a shows all installed including disabled ones
[18:19:19] *** anilg has joined #opensolaris
[18:19:49] <sailorvrz> is there a way to install sxce on a ultra 10 through network?
[18:22:20] <Auralis> sure, jumpstart
[18:22:25] *** Adamant_ has joined #opensolaris
[18:24:01] *** Adamant has quit IRC
[18:24:19] <jamesd_> sailorvrz, my prefered way is too...  burn   cd1  or dvd, and boot,  on another solaris box... mount loop back mount the dvd iso,  and share via nfs.. and use it to do the actual install...  a fast x86 box can do an install that way in less 20 minutes.
[18:26:07] <sailorvrz> jamesd_: yeah my problem is actually that I have opensolaris running on my laptop (x86) and need to install sxce on my ultra 10 as I don't really want to run linux :)
[18:26:39] <sailorvrz> and I don't have a spare dvd drive...
[18:27:04] <TomJ> google solaris jumpstart JET
[18:27:11] <TomJ> install that and it makes jumpstart setup a breeze
[18:27:38] *** alibb has joined #opensolaris
[18:27:38] <sailorvrz> thanks
[18:27:48] <TomJ> you can then boot the ultra 10 from network and either then do a manual install at keyboard, or go the whole way and have unattended install via jumpstart scripts
[18:28:15] <sailorvrz> yeah just found a nice howto on google :D
[18:31:46] *** ShadowHntr has joined #opensolaris
[18:33:52] *** jgracin has joined #opensolaris
[18:36:04] <edgy> Auralis: can you please change your milestone to multi-user and see whether svcs now displays all the disabled services? for me it does
[18:36:22] <Auralis> i'm in multiuser
[18:43:17] <edgy> Auralis: multi-user or mult-user-server?
[18:44:58] <edgy> Auralis: I just did svcadm milestone milestone/multi-user and svcs displays disabled services !!
[18:46:21] *** ksyl_ has left #opensolaris
[18:47:29] *** rand7 has joined #opensolaris
[18:50:26] *** edgy_ has joined #opensolaris
[18:51:59] <Auralis> multi-user-server, i -THINK- if you have not reached the defaukt run level/milesotne, svcs displayed the disabled services from the not reached run levels as well as default. checked with svcs -xv if somethign had failed?
[18:52:40] *** smtms has joined #opensolaris
[18:56:27] *** Asako has joined #opensolaris
[18:56:38] <edgy_> Auralis: now you are in milestone/multi-user-server, right? what do you expect would happen if you type svcadm milestone milestone/mult-user-server?
[18:56:47] <edgy_> s/mult/multi
[18:56:59] <Auralis> nothing should happen
[18:57:19] *** crivensbot has joined #opensolaris
[18:57:34] *** chicken has joined #opensolaris
[18:57:49] <edgy_> Auralis: ok please try it
[18:58:08] <chicken> crivensbot: math 6+6
[18:58:09] <crivensbot> 12
[18:58:28] *** Auralis has quit IRC
[18:58:40] <edgy_> crivensbot: 10+11
[18:58:45] <chicken> crivensbot: math sin(67.9)
[18:58:53] <edgy_> crivensbot: math 10+11
[18:58:53] <crivensbot> 21
[18:59:20] <chicken> crivensbot: date
[18:59:20] <crivensbot> 2008-09-14, 18:59:38h/CEST
[19:00:05] <chicken> crivensbot: uuid
[19:00:06] <crivensbot> 0a60111096d8e1882798e58872abd0cfc02e65d0e85f31f26ad7a98fe7b867ea
[19:00:32] <chicken> crivensbot: stocks
[19:00:32] <crivensbot> "JAVA" 9.39 "9/12/2008" "4:00pm" +0.33 9.04 9.55 8.9699 20465136
[19:01:25] <chicken> does anyone have suggestions for more bot commands ?
[19:01:49] <gottadoit> reboot ?
[19:01:58] <chicken> uhm
[19:02:27] <chicken> crivensbot: reboot key=d98a0a4f498afe164edd1316f61778b6
[19:02:29] *** crivensbot has quit IRC
[19:02:29] <gottadoit> or did you only want sensible suggestions ?
[19:02:53] <Asako> you running supybot?
[19:03:57] *** crivensbot has joined #opensolaris
[19:04:02] *** chicken is now known as nrubsig
[19:04:27] <nrubsig> Asako: guess again.
[19:04:44] <gottadoit> how about url
[19:04:48] <nrubsig> Asako: try 'help'
[19:05:20] <gottadoit> crivensbot: help
[19:05:21] <crivensbot> Hello, this is shircbot. Subcommands are 'say hello', 'math <math-expr>', 'stocks', 'uuid' and 'date'.
[19:05:58] <nrubsig> crivensbot: reboot key=d988e58872abd0cfc02e65d0e85b6
[19:06:02] *** crivensbot has quit IRC
[19:06:09] *** Auralis has joined #opensolaris
[19:06:35] <Auralis> that was entirely unexpected
[19:07:06] *** crivensbot has joined #opensolaris
[19:07:08] <nrubsig> Auralis: ?!
[19:07:32] <nrubsig> Auralis: what happened ?
[19:07:54] <Auralis> nrubsig: going from milestone multi-user-server to multi-user-server droped me right into mainance mode with failed login service
[19:08:08] <nrubsig> ah
[19:09:14] *** edgy has quit IRC
[19:09:19] <nrubsig> crivensbot: reboot key=e58872abd0cfc02e65d02e65d0e85b6
[19:09:23] *** crivensbot has quit IRC
[19:09:57] <sickness> hi nrubsig :)
[19:10:15] *** crivensbot has joined #opensolaris
[19:10:17] <nrubsig> sickness:
[19:10:24] <nrubsig> sickness: Hi!
[19:10:29] <sickness> hi :)
[19:10:30] <nrubsig> crivensbot: help
[19:10:30] <crivensbot> Hello, this is shircbot, written in ksh93 (Version M 93t 2008-07-25). Subcommands are 'say hello', 'math <math-expr>', 'stocks', 'uuid' and 'date'.
[19:10:52] <sickness> nice, a bot written in ksh? =)
[19:11:11] <edgy_> Auralis: I know and sorry I didn't warn you ;)
[19:11:14] <nrubsig> crivensbot: math tan(1)
[19:11:14] <crivensbot> 1.55740772465490223050697480745836
[19:12:11] <nrubsig> crivensbot: math hypot(9,1)
[19:12:11] <crivensbot> 9.05538513813741662657380816698407
[19:14:13] <nrubsig> gottadoit: yes, sensible suggestions, please
[19:14:14] *** jfndi_ has quit IRC
[19:14:17] <Asako> anybody know much about the dtrace ip provider?
[19:14:25] <nrubsig> gottadoit: and what is the "URL" command ?
[19:15:37] *** ahe has joined #opensolaris
[19:16:44] *** ipfw has joined #opensolaris
[19:17:13] *** crivensbot is now known as roguebot
[19:17:19] <gottadoit> simple suggestion to return an address of a webserver where we can view/search history of the channel from when the bot has been around
[19:17:24] <nrubsig> uh
[19:17:28] *** AxeZ has joined #opensolaris
[19:17:36] <nrubsig> how did the bot rename itself ?
[19:17:46] <sickness> gottadoit: hey there was a site which did just that some time ago...
[19:18:25] *** roguebot has quit IRC
[19:21:42] *** crivensbot has joined #opensolaris
[19:21:52] <nrubsig> mhhh
[19:22:33] *** crivensbot has quit IRC
[19:23:19] *** jtmuzix has joined #opensolaris
[19:23:20] *** crivensbot has joined #opensolaris
[19:23:29] <nrubsig> crivensbot: exitbot 8df69a19926648dab7cd92adcb485454d0372844
[19:23:29] *** crivensbot has quit IRC
[19:24:35] <jtmuzix> how come solaris 5/8 won't boot 64 bit mode with vmware fusion when trying to install??  it has 64 bit support and nexenta boots 64 bit..
[19:24:42] *** crivensbot has joined #opensolaris
[19:25:54] <edgy_> Auralis: so when you are there, did you try svcs?
[19:28:05] <Auralis> no, and i'm not going to repeat that experiment
[19:28:31] <codestr0m> xRaich[o]2x: which variable did you use originally to detect if the current user was root from tcsh?
[19:29:33] <xRaich[o]2x> codestr0m:  $uid
[19:29:56] <codestr0m> hmm. that didn't return 0 for root
[19:31:05] <nrubsig> crivensbot: exitbot 8df69a19926648dab7cd92adcb485454d0372844
[19:31:05] *** crivensbot has quit IRC
[19:31:20] *** crivensbot has joined #opensolaris
[19:32:35] *** ChanServ sets mode: +o nrubsig
[19:34:56] *** sponix has quit IRC
[19:36:30] <Asako> Cannot find Gtk+2! If you are using binary packages based system, check that you
[19:36:30] <Asako> have the corresponding -dev/devel packages installed.
[19:36:34] <Asako> anybody know how to fix that?
[19:36:39] *** crivensbot has quit IRC
[19:36:51] *** crivensbot has joined #opensolaris
[19:37:16] <Asako> SUNWgnome-base-libs is installed
[19:37:19] *** crivensbot has quit IRC
[19:37:41] *** ahe has quit IRC
[19:37:42] *** crivensbot has joined #opensolaris
[19:38:25] <Asako> or anybody compiled audacious on open solaris?
[19:39:20] <oxygene> Asako: maybe there's a SUNWgnome-base-libs-devel, too?
[19:40:05] <Asako> didn't know they had devel packages
[19:46:23] <Asako> Package xrender was not found in the pkg-config search path.
[19:46:26] <Asako> guess that's my problem
[19:47:06] *** alibb has quit IRC
[19:47:12] <codestr0m> Asako: not to intrude, but if you don't have the package installed locally you can add -r pkg search -r foo and it'll also look on remote authorities
[19:47:20] <Asako> yeah, did that
[19:47:28] <Asako> no package matches xrender
[19:48:04] *** crivensbot has quit IRC
[19:48:49] *** nrubsig has quit IRC
[19:54:20] *** houst0n- has quit IRC
[19:54:34] *** qmrw has joined #opensolaris
[19:56:43] <edgy_> ls
[19:56:50] <trygvis> pwd
[19:58:04] <anilg> cat /dev/random
[19:58:17] <ShadowHntr> i made the mistake once
[19:58:23] <ShadowHntr> of cat /dev/random > /dev/audio
[19:58:24] <ShadowHntr> ;)
[19:58:37] <edgy_> Auralis: yes, I understand you are not going to repeat it ;)
[19:58:38] <ShadowHntr> i think i heard God. ;)
[19:59:20] *** jbasse has joined #opensolaris
[19:59:53] *** cypromis has quit IRC
[20:00:14] *** jtmuzix has quit IRC
[20:00:49] *** nitrile has quit IRC
[20:01:20] <edgy_> Auralis: can you now understand why this is happening?
[20:01:32] *** edgy_ is now known as edgy
[20:02:36] *** paul has quit IRC
[20:13:07] *** erast has quit IRC
[20:14:31] <sailorvrz> is nano not in ips?
[20:15:45] <Asako> did you search for it?
[20:16:53] <Asako> doesn't look like it's there, should be easy to compile
[20:17:57] *** anilg has left #opensolaris
[20:19:32] <sailorvrz> Asako, yeah, even google doesn't deliver anything
[20:20:22] *** LordIllpalazo has quit IRC
[20:20:30] <Asako> google doesn't have much about open solaris
[20:21:15] <sailorvrz> yeah it's a pity
[20:21:39] <Asako> not very popular
[20:24:26] *** LordIllpalazo has joined #opensolaris
[20:24:29] <sailorvrz> it will become popular... it's an awesome os
[20:25:35] <Asako> yeah, I like it
[20:25:39] <ShadowHntr> Asako: try opensolaris w/o spaces
[20:25:54] <Asako> yeah, that might work
[20:26:05] <Asako> I've switched my servers to nexenta
[20:33:29] *** qmrw has quit IRC
[20:39:06] *** paul___ has joined #opensolaris
[20:39:16] *** jtmuzix has joined #opensolaris
[20:39:50] *** LordIllpalazo has quit IRC
[20:40:33] *** Adamant_ has quit IRC
[20:40:56] *** paul___ is now known as paul
[20:42:16] *** LordIllpalazo has joined #opensolaris
[20:56:37] *** alibb has joined #opensolaris
[21:10:53] *** jbasse has quit IRC
[21:11:23] *** jbasse has joined #opensolaris
[21:11:38] *** LordIllpalazo has quit IRC
[21:14:04] *** xinkeT has quit IRC
[21:15:33] *** alibb has quit IRC
[21:16:40] *** hircus has quit IRC
[21:18:46] *** sailorvrz_ has joined #opensolaris
[21:19:27] *** GhostDWORD` has joined #opensolaris
[21:19:48] *** jlc has joined #opensolaris
[21:27:02] *** nitrile has joined #opensolaris
[21:28:21] *** erast has joined #opensolaris
[21:28:37] *** mikl has joined #opensolaris
[21:29:48] *** jv5181 has quit IRC
[21:30:36] *** fr4g has joined #opensolaris
[21:32:17] *** MeP3aBeu has joined #opensolaris
[21:32:32] *** gottadoit has quit IRC
[21:32:50] *** MeP3aBeu has quit IRC
[21:32:52] *** ninjaslim has joined #opensolaris
[21:35:51] *** sailorvrz has quit IRC
[21:37:18] <jlc> Does Solaris express have the ability to LU anymore?  E.g. going from Build 94 to 97.  I have not used solaris for awhile so not sure if that is possible with it or not.
[21:37:37] <trygvis> sure, I use it all the time
[21:37:48] <jlc> cool
[21:38:27] <jlc> 2008.05 just isnt fun trying to get mp3/dvd and all that working so its a bit annoying to me.
[21:38:39] <jlc> I just like the old opensolaris
[21:38:48] * trygvis too
[21:38:49] <Asako> I have svn97
[21:39:03] <Asako> you need the codec for mp3
[21:39:17] *** syamajala2 has joined #opensolaris
[21:39:34] <Asako> or install pkg-get and get xmms
[21:39:36] <h3sp4wn> You can get the spec-file from the gnome-2-18 branch for SUNWgnome-media-mp3
[21:39:50] <Asako> binary works too
[21:40:11] <h3sp4wn> But I don't think it sounds that good - the libmad plugin is not that hard to build with SFE either
[21:40:26] <Asako> ah, ok
[21:40:47] <jlc> I used to use opensolaris and sfe all the time, but for some reason got away from it
[21:40:53] <Asako> https://shop.fluendo.com/product_info.php?products_id=40&osCsid=uoptt6ho47fd88p4hhg5v9eil5
[21:41:09] <h3sp4wn> Yeah but that is badly linked to libgcc
[21:41:13] *** master_of_master has quit IRC
[21:41:21] <h3sp4wn> without RPATH
[21:41:37] <Asako> I didn't even have to mess with that
[21:42:07] *** yarihm has quit IRC
[21:42:08] <jlc> 2008.05 just seems to be a pita to get the stuff like mp3/dvd codecs to work
[21:42:11] <h3sp4wn> Perhaps os2008.05 has libgcc in /usr/lib somewhere or they fixed it
[21:42:38] <jlc>  in /gnu something i thought
[21:42:49] *** RElling1 has joined #opensolaris
[21:43:16] <Asako> libgcc_s.so.1 =>	 /usr/sfw/lib/libgcc_s.so.1
[21:43:21] <Asako> they must have rebuilt it
[21:43:56] <Asako> I could almost ditch windows for solaris at home
[21:44:15] *** GhostDWORD` has quit IRC
[21:44:17] <Asako> wonder if ushare will compile
[21:45:13] <h3sp4wn> What do you need upnp media for with Solaris
[21:45:25] <Asako> for my tv shows
[21:45:32] <xRaich[o]2x> codestr0m: my patch for zsh is already in cvs ;)
[21:45:35] <Asako> I use Twonky right now
[21:45:36] *** sailorvrz_ has quit IRC
[21:45:50] <Asako> bah, too much work to reinstall
[21:46:32] <h3sp4wn> xRaich[o]2x: What does it do ? pfzsh would be nice
[21:46:46] <codestr0m> xRaich[o]2x: how did you track it down.. I'm not sure if it was just obscure or what
[21:47:06] <xRaich[o]2x> h3sp4wn: fixes a memory leak
[21:47:10] <xRaich[o]2x> codestr0m: dtrace ;)
[21:47:16] <xRaich[o]2x> and mdb
[21:47:26] <Asako> dtrace rules
[21:47:35] <xRaich[o]2x> it first showed up in pmap output
[21:47:37] <codestr0m> xRaich[o]2x: you should make a blog entry on how you did it
[21:48:02] <xRaich[o]2x> codestr0m: i posted the dtrace skript that i used to isolate the bug in my blogentry with the patch
[21:48:17] <codestr0m> ok. I just saw the two liner
[21:48:25] <xRaich[o]2x> 2 liner?
[21:48:57] <codestr0m> xRaich[o]2x: I thought you said it's two lines?
[21:49:01] <codestr0m> I just glanced at it
[21:49:20] <xRaich[o]2x> the fix was a 2 liner ^^
[21:49:31] <codestr0m> yeah
[21:49:32] <xRaich[o]2x> tracing it down was something else
[21:49:36] <codestr0m> oh sure
[21:50:05] <xRaich[o]2x> i wrote a shitload of scripts in the end it was totaly obvious but if you don't know the code... :P
[21:50:15] <xRaich[o]2x> even the devs didn't know what was happening ^^
[21:50:22] <xRaich[o]2x> the thought it was solaris' fault
[21:50:23] <codestr0m> hehe..
[21:50:50] <Asako> has anybody tried building galeon?
[21:50:57] <xRaich[o]2x> since they didn't see any weird behavoir on linux (which i doubt when you look at the nature of the bug)
[21:51:26] <xRaich[o]2x> but when i posted the bug they magically could reproduce the bug on linux :P
[21:51:40] *** RElling has quit IRC
[21:51:43] <seanmcg> xRaich[o]2x, got a pointer to your blog entry ?
[21:51:46] <codestr0m> xRaich[o]2x: you should take a look at fish some time.. it's got some really awesome features on the list, but the performance sucked when I gave it a shot
[21:52:01] <xRaich[o]2x> seanmcg: it's all in german raichoo.blogspot.com
[21:53:03] <xRaich[o]2x> codestr0m: i found some other way to speed up zsh startup compinit has the option -C which prevents an update of the cachefile
[21:53:29] <xRaich[o]2x> i'm still way to new to the shell to know all the tricks but dang... there are a lot
[21:53:45] <codestr0m> with my little 67K .tcshrc I had to inline the source file for completion. it was causing a problem
[21:53:55] <codestr0m> you said 300 page manual
[21:53:57] <codestr0m> that's evil
[21:54:07] *** gnut has joined #opensolaris
[21:54:08] <Asako> I hate how everything assumes gnu
[21:54:08] <seanmcg> thanks xRaich[o]2x
[21:54:18] <gnut> hello everybody
[21:54:19] <xRaich[o]2x> codestr0m: fish looked nice i compiled in on solaris after you mentioned it. how is the syntax? bourne like?
[21:54:21] <Asako> xmms shows tar errors on skin files
[21:55:04] <gnut> anyone been able to play WMV-9 files?
[21:55:29] <codestr0m> xRaich[o]2x: I've very very minimally played with it.  on gentoo it was setup very sanely by default
[21:55:45] <codestr0m> when I built it on os2008* I realized it would have a learning curve
[21:55:53] <xRaich[o]2x> seanmcg: no problem. i also have an IPS completion script if you care ;)
[21:56:15] <xRaich[o]2x> it's just a day old so i could use some patches :P
[21:56:22] <codestr0m> gnut: you may need to pull from another authority != opensolaris.org
[21:56:24] *** fr4g_ has joined #opensolaris
[21:56:39] *** edgy has left #opensolaris
[21:57:01] <xRaich[o]2x> codestr0m: i mean syntax is very important to me. you know how i whined about tcsh ^^
[21:57:11] *** fr4g has quit IRC
[21:57:13] *** gottadoit has joined #opensolaris
[21:57:20] <codestr0m> hehe.. sure I appreciate that
[21:57:28] <codestr0m> I look at bash completion and zsh and think yuck
[21:57:41] <xRaich[o]2x> the zsh completion is.... well....
[21:57:43] <xRaich[o]2x> yuck
[21:57:44] <xRaich[o]2x> :P
[21:57:46] <Asako> I'm bash for life
[21:58:13] <xRaich[o]2x> very powerful but it's almost a language of it's own
[21:58:33] <Asako> kind of like sed/awk
[21:58:36] <Asako> hehe
[21:58:41] <seanmcg> xRaich[o]2x, i'd use it, though still using sxce and .. um (ba)sh.
[21:58:58] <gnut> codestr0m: know of any authorities ?
[21:59:02] <xRaich[o]2x> seanmcg: ah ok. ^^
[21:59:18] <gnut> I have totem from blastwave...
[21:59:20] <codestr0m> gnut: Blastwave.org                       http://blastwave.network.com:10000/
[21:59:20] <codestr0m> opensolaris.org (preferred)         http://pkg.opensolaris.org:80/
[21:59:20] <codestr0m> sunfreeware.com                     http://pkg.sunfreeware.com:9000/
[22:00:04] <Asako> blastwave uses IPS now?
[22:00:21] <xRaich[o]2x> i still have some kind of love/hate relationship with zsh. in most cases i think "wtf?" but it has a lot of features i like a lot
[22:00:34] <Asako> bash sure makes my job easier
[22:00:43] <mui> zsh completition is way nicer than bash
[22:00:56] <e^ipi> yeah, bash is pretty antiquated and kinna crap
[22:00:58] <gnut> yes... zsh is much way nicer than bash
[22:01:01] <gnut> i use zsh
[22:01:07] <e^ipi> i use ksh93
[22:01:12] <Asako> how about tcsh?
[22:01:16] <mui> :p
[22:01:18] <xRaich[o]2x> mui: dunno never coded completion in bash. but zsh is a pita to code ^^
[22:01:30] <mui> it is not(:
[22:02:04] <xRaich[o]2x> i meant completion. well. I'm not a script guru after all :P
[22:02:35] <e^ipi> evidently you can, through .profile, make ksh93 indistinguishable from bash from the users perspective
[22:02:41] <xRaich[o]2x> Asako: tcsh is great if you like c-shell syntax. i hate it with a passion
[22:02:47] <xRaich[o]2x> not the shell but the syntax
[22:03:19] <xRaich[o]2x> ksh93 is great. but i really miss programmable completion that's why i stick with zsh for now
[22:04:09] *** syamajala2 has quit IRC
[22:04:43] <e^ipi> the programmable completion in bash et al is typically wrong on non-gnu systems
[22:04:52] <e^ipi> no --gnu-long-options
[22:05:05] <e^ipi> except with the ksh93 builtins, which do... oddly
[22:05:27] <e^ipi> error404@gomorra:[ /export/home/error404 ] $ cat --version version         cat (AT&T Research) 2006-05-17
[22:07:44] <Asako> lots of stuff is wrong on non-gnu systems
[22:08:00] <e^ipi> or rather, GNU is wrong in a lot of places
[22:08:10] <e^ipi> there are published standards, which the GNU tools fail
[22:10:10] <Asako> I've always liked BSD better but linux gets all the attention
[22:10:40] <codestr0m> it's because Solaris doesn't have an arrogant bastard dictator stealing the limelight ;)
[22:11:17] <codestr0m> or maybe it was because in the areas people were looking it was advancing
[22:11:18] <Asako> like Jonathon?
[22:11:29] <codestr0m> he's too politically correct
[22:11:39] <Asako> he has to be
[22:11:43] <codestr0m> you'll never see him call the obsd devs masterbating monkeys
[22:11:47] <Asako> lulz
[22:12:14] <Asako> I think opensolaris is playing catch-up
[22:12:24] *** storycrafter has joined #opensolaris
[22:12:40] <codestr0m> Asako: it has potential.. it's a better technology by far in most areas
[22:12:53] <Asako> yeah, I like it
[22:13:09] <Asako> configuring X was really easy
[22:13:09] <e^ipi> playing catch-up how exactly?
[22:13:27] <e^ipi> solaris is deployed in finance in much larger numbers than linux
[22:13:32] <codestr0m> e^ipi: market share
[22:13:33] <Asako> would Sun have released the code if Linux/GPL didn't exist?
[22:14:00] *** steleman has quit IRC
[22:14:01] <codestr0m> Asako: Sun did this because of marketing
[22:14:14] <e^ipi> Asako: probably not, but who cares? would linux had existed if not for BSD being open source? probably not
[22:14:32] *** cypromis has joined #opensolaris
[22:14:34] <Asako> sure
[22:14:40] <Asako> Linus never used BSD
[22:15:10] <Asako> all I know is I wouldn't be where I am today without free software
[22:15:13] <e^ipi> codestr0m: look at it this way; when you help someone out... donate to charity, whatever; do you do it because it makes you feel good or because it's the right thing to do?
[22:15:30] *** steleman has joined #opensolaris
[22:15:35] *** Auriel[A] is now known as Auriel
[22:15:40] *** storycrafter1 has joined #opensolaris
[22:15:43] <codestr0m> e^ipi: in general and this is personal not business. you give not to receive
[22:15:56] <e^ipi> Asako: linux was just in the right place at the right time; there would be no open source if not for BSD, and there would be no Linux if BSDi weren't being sued at precisely the time that linux was establishing itself
[22:16:01] <codestr0m> Sun is a corporation with a responsibility to shareholders
[22:16:10] <codestr0m> it's nothing about giving. it's about money. bottom line
[22:16:23] <e^ipi> codestr0m: does this affect the end result at all?
[22:16:29] <e^ipi> for whatever reason, solaris is open source
[22:16:31] <e^ipi> and that's great
[22:16:34] *** xinkeT has joined #opensolaris
[22:16:51] <codestr0m> e^ipi: who are you trying to argue with.. we're on the same side
[22:17:00] <e^ipi> it's not an argument
[22:17:05] <e^ipi> cypromis: hey
[22:17:16] <cypromis> hi :)
[22:17:31] <e^ipi> cypromis: you're doing some solaris telephony stuff lately, yeah?
[22:17:36] <cypromis> codestr0m: I made the mirrors, details tomorrow
[22:17:42] <cypromis> e^ipi: yes
[22:17:46] <cypromis> msot of my time actualy
[22:17:47] <cypromis> :)
[22:18:07] <codestr0m> cypromis: awesome.. I've been reading various things about the state of building/packaging in os2008*
[22:18:12] <e^ipi> i have 2 Dialogic D/41E-PCI cards that came in a machine i bought from fleabay
[22:18:16] <e^ipi> can you do anything with them?
[22:18:21] <e^ipi> ( if not i'll just ebay them )
[22:18:38] <cypromis> ebay them
[22:18:44] <e^ipi> okiedokie
[22:18:45] <cypromis> :)
[22:18:51] <Asako> does asterix work on solaris?
[22:18:52] <cypromis> ok, bbl
[22:18:53] <cypromis> :)
[22:18:56] <cypromis> thanks for the offer
[22:19:02] <cypromis> asko: yes
[22:19:06] <codestr0m> Asako: yuck.. don't say that around here
[22:19:07] <Asako> sweet
[22:19:07] <e^ipi> np
[22:19:07] <cypromis> if you can cal asterisk something that works
[22:19:08] <cypromis> :)
[22:19:14] <e^ipi> heh
[22:19:15] <Asako> lol
[22:19:15] <cypromis> anyway, bbl
[22:19:15] <cypromis> :)
[22:19:19] *** cypromis has quit IRC
[22:19:21] <e^ipi> Asako: freeswitch
[22:19:28] <gnut> e^ipi: why didn't you name yourself e^ipi/2?
[22:19:46] <gnut> then you could say "e^ipi/2 like this, or e^ipi/2 like that" and it would make sense.
[22:19:47] <codestr0m> hopefully.. between the few people I know that care about telephony in here we can keep the assterisk crowd out
[22:19:50] *** xRaich[o]2x is now known as Guest80085
[22:20:48] <e^ipi> codestr0m: *shrug* i know nothing about telephony; just people that i know that are telephony geeks all seem to slag on asterisk and inform me that freeswitch is much better
[22:21:11] <e^ipi> since i'm not in that field, i defer to those that are
[22:21:25] *** xRaich[o]2x has joined #opensolaris
[22:21:34] <codestr0m> (shrugs) I'd rather just advance the state than complain about others too much
[22:21:50] <codestr0m> occasionally it's fun though ;)
[22:22:12] <coffman> asterisk is a pice of crap
[22:22:21] *** storycrafter has quit IRC
[22:22:41] <Asako> I'll remember that if I ever set up a PBX
[22:22:51] <codestr0m> coffman: have you had to write code for asterisk?
[22:24:18] *** fr4g_ has quit IRC
[22:24:19] <coffman> codestr0m: i had it deployed in my old company
[22:25:55] *** mikl has quit IRC
[22:27:23] *** houst0n- has joined #opensolaris
[22:28:08] <wonko2> does anyone know if the vbox kernel modules are suspend safe?
[22:28:50] *** storycrafter1 has left #opensolaris
[22:30:33] *** hircus has joined #opensolaris
[22:30:34] *** Guest80085 has quit IRC
[22:32:54] <coffman> wonko2: good question, it should or?
[22:34:05] *** edgy has joined #opensolaris
[22:34:14] <wonko2> i wonder where i could find that out.  time to hit goodle
[22:34:17] <wonko2> google even
[22:34:52] <seanmcg>  forums.virtualbox.org maybe ?
[22:34:59] <edgy> Hi, pwdx `pgrep cron` shows /var/spool/cron/atjobs. I don't understand what's the current working directory of  a process means, any help?
[22:37:06] <wonko2> well, from the looks of the source code it does, but that's the trunk, not sure if that's in 2.0.2 or not
[22:39:36] *** techqbert has quit IRC
[22:40:12] *** edgy has quit IRC
[22:40:51] *** gottadoit has quit IRC
[22:41:51] <wonko2> well, looking back through the log entries the last time anything was changed to the attach/detach/suspend/resume code was a minor patch over 2 months ago
[22:42:00] <wonko2> i'll have to assume that's all in 2.0.2 ;)
[22:42:33] *** kimc has joined #opensolaris
[22:46:46] *** hircus has quit IRC
[22:50:11] *** PicCard has quit IRC
[22:52:08] *** ninjaslim has quit IRC
[22:53:50] *** djgregor has quit IRC
[22:59:46] *** madnex has joined #opensolaris
[23:03:57] *** madnex has quit IRC
[23:06:15] *** derchris has quit IRC
[23:06:20] *** derchris has joined #opensolaris
[23:09:25] *** PicCard has joined #opensolaris
[23:09:28] *** Adamant has joined #opensolaris
[23:09:47] *** ShadowHntr has quit IRC
[23:12:30] *** Adamant has quit IRC
[23:13:35] *** Adamant has joined #opensolaris
[23:13:41] *** Asako has quit IRC
[23:14:05] *** madnex has joined #opensolaris
[23:21:07] *** jgracin has quit IRC
[23:21:54] *** madnex has quit IRC
[23:22:37] *** gottadoit has joined #opensolaris
[23:24:55] *** madnex has joined #opensolaris
[23:28:56] *** PicCard has quit IRC
[23:32:23] *** alibb has joined #opensolaris
[23:35:55] *** FurnaceBoy has joined #opensolaris
[23:37:07] *** Gekz has quit IRC
[23:39:05] *** madnex has quit IRC
[23:49:37] *** [1]kimc has joined #opensolaris
[23:57:10] *** AxeZ has quit IRC
[23:57:30] *** bahumbug has joined #opensolaris





top