[00:00:52] *** Kernel86|Laptop has joined #OpenSolaris
[00:01:09] *** fik has joined #opensolaris
[00:01:37] <irc00123> solaris express version post 35 have brandz included ?
[00:03:10] <Error_404> wasn't it included in build 49?
[00:05:34] <richlowe> yes.
[00:06:22] *** axisys has quit IRC
[00:13:57] <Error_404> that's what i thought
[00:14:59] *** irc00123 has quit IRC
[00:15:10] <sahafeez> delewis: when i find that i kill the dl and restart. 90% of the time it goes back to 200kb
[00:26:33] *** koolniczka has quit IRC
[00:28:58] *** gisburn has joined #opensolaris
[00:29:09] <gisburn> steleman: ping!
[00:29:44] <steleman> pong!
[00:29:58] <steleman> wassup
[00:30:09] *** fik has quit IRC
[00:30:16] *** gisburn has quit IRC
[00:31:31] *** gisburn has joined #opensolaris
[00:31:40] <gisburn> steleman: sorry, chatzilla hang
[00:31:47] <gisburn> steleman: --> /msg
[00:35:51] *** jsubl2 has joined #opensolaris
[00:48:01] *** th3tick has joined #opensolaris
[00:51:16] *** krozinov has left #opensolaris
[00:54:25] *** denko has joined #opensolaris
[00:55:34] *** dekop has joined #opensolaris
[00:56:58] <dekop> hi everyone
[00:58:04] <dekop> i have a question related to Intel SCSI controllers
[00:58:22] *** th3tick has left #opensolaris
[00:59:54] <dekop> any help would be really appreciated
[01:02:44] *** spawrq has joined #opensolaris
[01:07:29] *** jamesd has quit IRC
[01:07:37] *** jamesd has joined #opensolaris
[01:07:38] *** ChanServ sets mode: +o jamesd
[01:10:36] <delewis> heh, I love how NetBSD and OpenBSD claim portability, but they only supported SMP on very few of the ports.
[01:10:47] <delewis> s/ed//
[01:11:11] <jamesd> delewis, one processor and 640k should be enough for everyone.
[01:11:22] <delewis> jamesd: :-)
[01:13:10] <Tpenta> richlowe: teh clone doesn't happen until 11pm PST, can't really do much until then unless I generate from a gate snapshot
[01:19:18] *** nwf has quit IRC
[01:19:50] <richlowe> Tpenta: yeah, I saw steve's reply.
[01:20:42] * gisburn looks at sickness
[01:20:55] <gisburn> sickness: ping!
[01:24:45] <sahafeez> delewis SMP on openbsd was not the goal. they just added it like 2 releases ago. and i386 only. sparc is in the works next.
[01:25:10] <sahafeez> and SMP on netbsd works fine on sparc. i had a SS20 with 4 procs running it.
[01:25:59] <delewis> sahafeez: not SPARC64, though
[01:26:16] <delewis> and NetBSD reports that SPARC64 SMP support exists, but is currently broken because of recent "changes"
[01:26:48] <sahafeez> well i run linux on my U80 since drag and drop is broken in JDS on SPARC64 ;)
[01:28:37] *** Fish-- has joined #opensolaris
[01:29:47] <gisburn> Gman: ping!
[01:29:56] <Gman> pong
[01:30:11] <gisburn> Gman: !summon kupfer
[01:30:15] <gisburn> please
[01:30:21] <gisburn> problem
[01:30:22] <gisburn> big
[01:31:02] <jamesd> sahafeez, isn't it better just to drag and drop the machine out the window instead of running linux.
[01:31:41] <sahafeez> runs fine. even got dual head xorg working with 2xElite3d
[01:31:58] <sahafeez> now i will say i was quite shocked at that.
[01:32:14] <Gman> gisburn, ping'd him, but not sure he's around
[01:32:29] <gisburn> Gman: thanks!
[01:32:37] <sahafeez> i have a new 36gb in the box waiting for D&D to be fixed on SNV - tho i think i am going to just install U2 on it tonight
[01:33:58] <sahafeez> from the view of just using the box as a desktop running gnome there is no difference to the user vs solaris with JDS (cept D&D works!)
[01:34:25] <sahafeez> that being said, i want solaris back on the box because i want to play wiht zfs
[01:34:29] <sahafeez> with even
[01:41:17] *** Fish-- has quit IRC
[01:42:33] *** kupfer has joined #opensolaris
[01:42:33] *** ChanServ sets mode: +o kupfer
[01:45:18] *** Fish- has quit IRC
[01:46:33] *** spackes1 has joined #opensolaris
[01:46:38] *** spackes1 has left #opensolaris
[01:51:56] *** paul has quit IRC
[01:54:09] *** sahafeez has quit IRC
[01:55:47] *** sahafeez has joined #opensolaris
[01:56:13] <sahafeez> wow! my ibook just crash. 1st time ever.
[01:57:12] <gdamore> heh.  some poor schmuck is looking for support with my radeonfb on NetBSD (IBM X31).  too bad i resigned from NetBSD. :-)
[01:57:26] <icon> evening all
[01:57:49] <icon> gdamore: as if i had any doubt before, it looks like i will be going to hell soon :/
[01:58:00] <sahafeez> netbsd is dead. just ask netcraft ;)
[01:58:07] <gdamore> icon: why?
[01:58:16] <delewis> hey, they at least have *partial* SPARC64 SMP support
[01:58:17] <icon> gdamore: tcl is looking more and more viable for the ports system
[01:58:19] <delewis> that counts for something :-)
[01:58:20] <gdamore> heh.  netbsd has been dead for a long time.  but I've done some cool work with it
[01:58:27] <gdamore> they do?  since when?
[01:58:28] <delewis> I say partial, because it's broken atm.
[01:58:36] <delewis> gdamore: check the SPARC64 page
[01:58:46] <delewis> apparently it was working sometime ago
[01:58:50] <delewis> but someone decided to "fix it up"
[01:58:53] <delewis> and broke it in the process
[01:59:04] <icon> delewis: 'works on my box' :P
[01:59:21] <gdamore> ah, so phone@ and martin@ are still working on it.  yeah, i knew they were working on it.
[01:59:37] <gdamore> at one point i offered to lend a hand.  but then I got fed up with NetBSD politics/BS and resigned. :-)
[01:59:44] <delewis> hehe
[02:00:05] <delewis> well, I figured I would play around with NetBSD and OpenBSD on my Ultra 2 before I install Solaris on it to put it into production for good :-)
[02:00:12] <gdamore> icon: what "ports" are you referring to?
[02:00:13] <delewis> note that I associated production with Solaris.
[02:00:22] *** elflord has joined #opensolaris
[02:00:40] <icon> gdamore: the opensol ports collection
[02:00:47] <gdamore> NetBSD smp sparc64 is non-existent.  even when they "fix it", it will still suck due to Giant Lock
[02:01:05] <gdamore> opensolaris has a ports collection? what for?  isn't that why we have blastwave?
[02:01:09] <delewis> gdamore: I figured
[02:01:19] <icon> it does. i havent marked the project public yet
[02:01:24] <delewis> Solaris seems to have not only one of the most scalable SMP implementations, but also a fairly portable one
[02:01:33] *** elflord has quit IRC
[02:01:34] <delewis> given it works properly on both x86, x86_64, SPARCv8, and SPARCv9.
[02:01:37] <gdamore> Solaris SMP effing rocks.
[02:01:50] <delewis> and I think "scalable" and "portable" go hand-in-hand when you're talking about SMP.
[02:01:55] <icon> it has different goals than blastwave etc.
[02:01:56] * gdamore is still planning on doing a mips64 port, including SMP
[02:02:01] <delewis> gdamore: nice :-)
[02:02:14] <gdamore> delewis: no, scalable and portable don't necessarily go hand in hand.
[02:02:35] <gdamore> you can easily have a portable system that doesn't scale well.
[02:03:01] <delewis> gdamore: take the case of NetBSD -- poor scalability, and very poor portability, but I could just over-generalizing it, as the NetBSD kernel layout may not be portable, period -- I'm sure you'd be familiar with that.
[02:03:09] <delewis> be*
[02:03:18] <gdamore> icon: so why are you working on a ports collection?  between /usr/sfw, /opt/sfw, and blastwave, do we really need _another_ package system?
[02:03:45] <gdamore> actually, NetBSD is quite portable.  the vast majority of things are shared.
[02:03:55] <delewis> there's only one packaging system.
[02:03:58] <icon> gdamore: its an attempt to unify everything under a opensol project. eventually things will be migrated from the consolodation
[02:04:11] <gdamore> which consolidation?  sfw?
[02:04:17] <delewis> how various repositories decide to interface with that packaging system is well, numerous.
[02:04:28] <icon> the idea is to take the benefits of a source based ports system like darwinports, freebsd ports, etc. and integrate with solaris' packaging tools
[02:04:29] <delewis> gdamore: Blastwave, CCD, etc. all use standard, SVR4 packages.
[02:04:36] <delewis> and a proper ports system should create SVR4 packages after the build
[02:04:38] <gdamore> yes, i know that.
[02:04:42] <icon> delewis: exactly
[02:04:45] <delewis> like the NetBSD ports system does.
[02:04:53] <gdamore> NetBSD ports are a PITA
[02:04:58] <delewis> hehe
[02:04:59] <icon> the idea is to be more responsive than blastwave
[02:05:04] <delewis> I still like make package, though
[02:05:13] <icon> the idea is basically:
[02:05:20] <icon> portadm install vim
[02:05:31] <icon> thats it... everything else i taken care of. fetch, compile, package creation, package install
[02:05:47] <gdamore> i don't think it will necessarily be better/more responsive than blastwave.
[02:05:57] <gdamore> the problem is that _someone_ still has to maintain the system.
[02:05:59] <delewis> Blastwave packages have a lot of cruft
[02:06:04] <icon> its done in a way that you can have multiple repositories for your own custom software, as well as using it as a build system for distributing binary packages
[02:06:07] <delewis> because they aren't only targetting Nevada or Solaris 10
[02:06:13] <delewis> they're also targetting Solaris 8 and 9
[02:06:16] <icon> true
[02:06:17] <libkeiser> and because they have dependency hell
[02:06:19] <delewis> which lack all the GTK libraries
[02:06:24] <icon> in addition
[02:06:28] <gdamore> you can build S11 versions, though.
[02:06:33] <delewis> a ports system could look at what packages already on the exist on the system
[02:06:36] <delewis> and build accordingly.
[02:06:42] <icon> the ports system has direct dependencies on packages, so if JDS is installed, there is no need for alternate gtk packages etc.
[02:07:08] <delewis> icon: yes, I like that.
[02:07:11] <icon> its using a respun version of pkgbuild too
[02:07:13] <gdamore> will the ports system be svr package based?
[02:07:16] <icon> so jds will port directly into the ports system
[02:07:19] <delewis> just give me two things and I'll be satisifed with the ports system
[02:07:19] *** kFuQ has quit IRC
[02:07:19] <icon> gdamore: yup
[02:07:23] <delewis> (1) make it SVR4-based
[02:07:26] <gdamore> okay, then i'm okay with it.
[02:07:31] <delewis> and (2) make it inspect the existing packages upon build
[02:07:36] <icon> delewis: you already got it :)
[02:07:36] <delewis> so libraries, etc. aren't duplicated like in Blastwave.
[02:07:39] <icon> exactly
[02:07:42] <icon> that annoys the crap out of me
[02:07:56] <gdamore> i want binary packages though.  doing source builds to install stuff is a major PITA.
[02:08:03] <delewis> of course, you'd only be able to inspect the official Sun packages
[02:08:06] <gdamore> i mean, who wants to wait around for firefox to build?
[02:08:07] <delewis> which isn't really a problem.
[02:08:15] <icon> gdamore: okay, i can do that
[02:08:24] <delewis> gdamore: the ports system could offer a binary repository of pre-build packages
[02:08:25] <icon> it makes things more complicated for hosting
[02:08:27] <delewis> similar to what FreeBSD does.
[02:08:32] <icon> yeah
[02:08:57] <icon> i agree, firefox builds annoy me.
[02:09:06] <icon> i prefer pulling down a contrib package from ftp.mozilla.org
[02:09:11] <delewis> try building Firefox or Mozilla on SPARC..
[02:09:14] <delewis> talk about masochism.
[02:09:15] <icon> delewis: i have :)
[02:09:17] <delewis> to the n-th degree.
[02:09:27] <gdamore> the problem is, once you offer binary packages, you need to offer some for each version of Solaris.  otherwise you'll have the same problem as blastwave.  (someday Solaris 12 wll introduce e.g. GTK3 or somesuch)
[02:09:28] <icon> i did on my old sparcstation 20 years ago with mozilla 1.5
[02:09:51] <delewis> gdamore: Blastwave has one set of packages for all Solaris versions (at least that was the case)
[02:09:55] <delewis> a ports system could have multiple versions
[02:10:02] <delewis> which is what FreeBSD/OpenBSD/etc. do
[02:10:05] <gdamore> no, they don't.  for some things blastwave has multiple versions.
[02:10:06] <icon> gdamore: unlike blastwave, the ports system is geared towards the current revision with backdate hacks if neccssary
[02:10:19] <delewis> gdamore: that was not always the case
[02:10:22] <gdamore> but most of the package maintainers are too lazy to do builds for anything other than a single version.
[02:10:27] <icon> once you are on a ports system, its very rare that you cannot run the most current version of a port
[02:10:39] <icon> take a look in freebsd ports
[02:10:59] <icon> out of 18,000+ ports, maybe 1% are not supported on a specific revision of the os
[02:11:10] <icon> 1% is probably overshooting it by a lot
[02:11:24] <gdamore> yes, but for that 1% (180 packages in your example) you need to offer multiple versions
[02:11:31] <gdamore> (binary versions)
[02:11:33] <delewis> icon: not to hurt your argument, but out of those 18,000 ports how many of those ports are supported on legacy releases of FreeBSD
[02:11:36] <icon> the ultimate idea is that all non-core software from sun belongs to sun. everything else goes into the ports system.
[02:11:44] <delewis> I vividly recall how quickly support dropped for FreeBSD 4.x with a lot of the ports
[02:11:47] <icon> delewis: most of them, there are a few where they arent supported
[02:11:55] <delewis> "Sorry, this port is not supported on FreeBSD 4.x. Upgrade to 5.x"
[02:12:05] <icon> it really just depends on the port
[02:12:16] <icon> something like gtk3 is pretty platform neutral
[02:12:23] <gdamore> icon: i don't think that will work necessarily.  Sun has to ship 3rd party software, and they have to have stability, revision control, QA, etc.
[02:12:32] <icon> of course
[02:12:37] <delewis> ATLAS, wine, Octave, SCILAB all had 4.x support dropped fairly quickly.
[02:12:53] <delewis> (ATLAS, Octave, and SCILAB are linear algera-related softwares)
[02:13:14] <icon> that doesnt mean they cannot use the ports collection for the build setup. there is tagging/patch support so someone could create a respository that is officially supported
[02:13:28] <gdamore> i just wonder if the problem we are trying to solve with a "ports system" isn't better handled by getting package maintainers in blastwave to do the work to build S10 or nevada specific packages.
[02:13:40] <icon> you loose flexibility
[02:13:45] <icon> heres a good example
[02:13:50] <delewis> I still wouldn't use Blastwave for that reason.
[02:13:52] <icon> in JDS we have rhythmbox
[02:14:03] <icon> i want daap (itunes) support so i can stream my music from my nas
[02:14:17] <icon> right now im screwed because binary packages dont allow for any customization
[02:14:20] <delewis> with Blastwave there's two extremes -- you either get everything and kitchen-sink or you get almost everything (the one vital configure-option you needed) and a lot of shit that you don't need.
[02:14:22] <icon> with a ports system i would just do:
[02:14:32] <icon> portadm install rhythmbox --with-daap
[02:14:35] <delewis> which is why I choose to maintain my own package repository
[02:14:40] <delewis> appropriately named 'DEL'
[02:14:41] <delewis> :-)
[02:14:54] <icon> i still have a bunch of SES packages laying around :)
[02:14:57] <gdamore> this means the port system really needs to be a source based system.
[02:15:09] <icon> it is, but that doesnt mean it cannot support binary packages
[02:15:12] <icon> its also a build system
[02:15:13] <gdamore> as such, i don't think you can _replace_ the binary systems that already exist
[02:15:35] <gdamore> i.e. sun (or whomever) want to make sure they are getting a "fixed" set of bits that have been thru quality checks, etc.
[02:15:47] <icon> which is fine
[02:15:54] <gdamore> if the bits can change due to configure options, compiler choice, etc., then it is hard to guarantee much
[02:15:56] <icon> the ports system supports pulling down binary packages
[02:16:07] <icon> supports multiple compiler sets etc.
[02:16:21] <icon> whoever runs the repository can tweak whatever they need
[02:16:40] <icon> the idea is that there is a primary opensol repository
[02:16:56] <icon> anything else is up to whoever whats to run one. part of that is distributing binary packages
[02:17:17] <icon> so if sun has an 'official' set for solaris 11, they would have a repository of those ports/packages
[02:17:25] <gdamore> that's true for blastwave too, though they don't provide source packages for others.
[02:17:45] <icon> but that really messes with those of us that need to customize things etc.
[02:18:13] <icon> theres no such thing as a generic package that works for every system - you end up with extreme bloat, or no features at all
[02:18:21] <gdamore> can this "ports" system be handled as set of enhancements to pkg-get, though?
[02:18:34] <icon> gdamore: it works similarly, but its a full replacement
[02:18:42] <icon> portadm would be the command of choice
[02:19:10] <gdamore> so then the idea is that nobody uses blastwave anyrmore, i guess.
[02:19:25] <icon> yes and no, it really depends
[02:19:34] <gdamore> i want the fact that you are installing a "non-default" package to be really, really obvious.
[02:19:37] <icon> im really hoping we can unify things, and packages on blastwave be migrated over to the ports system
[02:19:47] <icon> gdamore: it will be ... VERY obvious
[02:19:48] <gdamore> have you talked to dclarke?
[02:19:50] <icon> yup
[02:19:51] <icon> quite a bit
[02:19:56] <icon> he was one of the first ones to approve
[02:20:01] <gdamore> cool.
[02:20:13] <icon> hes a great guy. i dont want to replace blastwave, i just want to unify everything
[02:20:27] <icon> hes offered hardware support etc. too which is great
[02:20:43] <gdamore> i'm just a little afraid of anarchy resulting from 18 gazillion different packages with different configuration options.
[02:20:50] <icon> yeah i am too
[02:21:01] <icon> delewis: do you think that binary packages are going to be very important to users?
[02:21:02] <gdamore> e.g. "package X depends on package Y, and depends on package Y being configured _with_ option Z"
[02:21:27] <icon> gdamore: its not as bad as it sounds
[02:21:38] <gdamore> icon: this is absolutely critical.  otherwise you won't get people to transition from blastwave.
[02:21:43] <icon> okay
[02:21:46] <gdamore> (binary packages i mean)
[02:21:48] <icon> compile farms here we come :)
[02:22:11] <delewis> sounds like a job for my E4500
[02:22:12] <icon> thats good to know - i really want to see this happen. it could really improve things on solaris
[02:22:19] <gdamore> a lot of people don't even _install_ compilers by default...
[02:22:22] <icon> ahh if only. i just have a little opteron here
[02:22:38] <icon> gdamore: i know. this will have to change that
[02:22:51] <icon> unless they want to use binary packages only
[02:22:52] <gdamore> you need to get Studio into the WOS, I think. :-)
[02:23:06] <delewis> gdamore: I doubt that will happen
[02:23:20] <delewis> not with the movement of "getting the install set media as small as possible, while keeping JDS"
[02:23:21] <gdamore> i dunno, its possible.
[02:23:21] *** eboutilier has left #opensolaris
[02:23:32] <gdamore> but a source-only ports system is doomed to failure without it.
[02:23:46] <delewis> there's always /usr/sfw/bin/gcc :-)
[02:23:50] <gdamore> (or some other standard compiler in the system)
[02:23:54] <icon> well
[02:24:11] <icon> the idea would be to provide binary repositories with default settings
[02:24:28] <icon> sans that, you will need compiler packages installed on the system and use the source based approach
[02:24:29] <gdamore> that would be fine.
[02:24:32] <icon> it would be a flag to portadm
[02:24:48] <icon> maybe.. portadm installpkg ?
[02:24:56] <gdamore> i'd prefer to have the default to use the binary, and flag to recompile
[02:25:00] <gdamore> e.g. portadm compile ?
[02:25:13] <gdamore> (portadm build)
[02:25:22] <icon> well, id like to cut down on commands to portadm, since as it is, it will be pretty large
[02:25:53] <icon> pkg repositories will be massive
[02:25:55] <gdamore> imagine 3 commands:  portadm build or portadm download (pick), followed by portadm install
[02:26:14] <icon> well it would be one command, but behind the scenes it would invole multiple
[02:26:22] <icon> ports have around 7 phases or so
[02:26:30] <gisburn> how does svn handle the case when I remove a file and then add it again and then do a commit ?
[02:26:32] <gisburn> what will happen ?
[02:26:43] <gdamore> the world will explode.
[02:27:01] <delewis> the ksh93 repo will be destoryed.
[02:27:06] <delewis> destroyed, rather.
[02:27:10] <gdamore> gisburn: same difference.
[02:27:11] <icon> but the idea is the core command is the same no matter if its source or binary
[02:27:38] <icon> trying to think of a good name to specify for that
[02:27:49] <gdamore> icon: anyone other than a FOSS zealot wants just a binary package, and they want it by default.
[02:28:02] <icon> ouch ;)
[02:28:17] <icon> well ive spent the last 10 years on BSD for the most part, so i just take it as a matter of course
[02:28:20] <gdamore> this is why blastwave is so damned popular.
[02:28:22] *** Cybernd has quit IRC
[02:28:43] <gdamore> otherwise everyone would just download; do ./configure; make install; ad naseum.
[02:28:47] <icon> ugh
[02:28:58] <icon> which i have to do on a few things because there are no packages available
[02:29:05] <icon> i miss fortune ;)
[02:29:26] <icon> thinking out load here, but there would be a command to add/remove repositories etc.
[02:29:43] <icon> ie: portadm addrepo <type> <location>
[02:30:00] <gdamore> are you expecting a source vs. binary choice to be based on the repo?
[02:30:00] <icon> something like... portadm addrepo pkg rsync://some/rsync/host
[02:30:07] <icon> well having different repo types
[02:30:12] <icon> so you can support both on the same system
[02:30:17] <gdamore> i'd rather not to do that.
[02:30:30] <icon> it cant be one or the other, then youre back to having the split we have now
[02:30:33] <icon> freebsd has done this for years
[02:30:39] <gdamore> i'd far rather see a "port" consist of 3 bits: 1) default x86 bits, 2) default sparc bits, 3) source bits
[02:31:05] <gdamore> and a given repo having a package would have all 3 bits for that package
[02:31:26] <icon> gdamore: hmm. thats a lot of load on hosting providers
[02:31:33] <icon> not that it doesnt make sense
[02:31:34] <gdamore> (this could be broken down further, by some ports having different versions of packages for different versions of solaris)
[02:31:48] <gdamore> icon: yes, its a lot of storage load.
[02:31:55] <icon> well for now, everything is targeting opensol, so for specific versions, you would have specific repositories
[02:32:03] <gdamore> the download doesn't have to download all three bits, of course.
[02:32:04] <icon> ie: a solaris 10 repo
[02:32:09] <icon> of course
[02:32:18] *** kupfer has quit IRC
[02:32:21] <icon> the ports system isolates everything too
[02:32:40] <icon> so compiled packages sit somewhere in /var/sadm separate from the port buildfiles/patches etc.
[02:32:58] <icon> the other nice part is the ability to include things like smf scripts etc. in the port itself
[02:33:08] <gisburn> again... when I remove a file in svn and add a file with the same name - will the svn server only see a cange in the file or a remove/add sequence ?
[02:33:09] <gdamore> sure.
[02:33:11] <icon> so we can solarify bits like samba etc. that still rely on the old sysvinit crap]
[02:33:28] <icon> gisburn: between commits? i couldnt tell you. try and find out
[02:33:31] <gdamore> gisburn: i dunno.  try it and see what happens. :-)
[02:34:18] <icon> hrm
[02:34:40] <Error_404> gack, how on earth do i get in contact with someone on os.o
[02:34:55] *** Gadzooks has joined #opensolaris
[02:34:55] <Error_404> i have a name & profile, but no way to email him
[02:34:58] <icon> it might make more sense to just add an entry to a repo list for the binary location
[02:35:01] <gdamore> it would be nice if we could have a way to do "provides" relationships for alternate choices.  e.g. different SMTP MTAs or somesuch.  but that needs something that SVR4 package lacks
[02:35:30] <icon> well, ports uses svr4 in addition to its own database
[02:35:45] <icon> makes things like searches etc. much more efficient and how we group things
[02:35:46] <gdamore> okay, i'll just wait and see, then.
[02:35:55] <icon> no, please - bitch and complain :D
[02:36:10] <icon> the idea is to make it better, not add yet another way to install software
[02:36:20] *** sahafeez_ has joined #opensolaris
[02:36:27] <Error_404> guys?
[02:36:30] <gdamore> i think i've done enough of that.  make it use binary packages by default, and you have me sold
[02:36:32] *** sahafeez_ has quit IRC
[02:36:50] <Error_404> if someone has blogs.sun.com/foo, is it  a fair bet that foo at sun dot com is the correct email?
[02:37:01] <gdamore> NFI.
[02:37:12] <gdamore> usually its first.last at sun dot com
[02:37:33] <icon> gdamore: would a different command name bother you?
[02:37:45] <gdamore> if it is not the default, yes.
[02:38:20] <Tpenta> error_404:  no
[02:38:25] <Tpenta> eg tpenta at sun dot com won't work
[02:38:36] <Tpenta> (btw tp at sun dot com will)
[02:38:44] <icon> okay
[02:38:45] <gdamore> i.e. recompilation is a PITA already, and recompiling may mean you get different untested bits.  So i'd rather have a subcommand or flag that makes it really explicit.
[02:38:55] <icon> gdamore: then how about pkgget support?
[02:39:10] <gdamore> what about it?
[02:39:19] * jamesd wonders if  tp would get forwarded to the head of  the janitorial staff ;-p
[02:39:20] <icon> pkgget could support the port name and/or the svr4 package name
[02:39:37] <gdamore> sure.  that would make it easier than scrapping pkg-get altogether.
[02:39:42] <Error_404> Tpenta: so essentially there's no way of knowing
[02:39:42] <icon> ie: pkgget wget and pkgget SUNWwget
[02:39:46] <icon> yeah
[02:39:49] <icon> same usage basically
[02:39:52] <icon> users are used to that
[02:39:52] <Tpenta> nope that goes to me (and i get all kinds of mail not destined for me, not as much as a friend of mine who has i at sun dot com, he gets all kinds of interesting stuff)
[02:39:55] <gdamore> that's okay.
[02:40:09] <icon> hmm
[02:40:14] * icon is taking notes
[02:40:30] <icon> alright
[02:40:39] <Error_404> oh, here we go... after trawling through the communities, i found the email i was looking for
[02:40:44] <icon> how do you feel about using something like portadm to manage information about a repository etc.
[02:40:54] <icon> or would a config file somewhere be more appropriate
[02:41:00] <gdamore> Tpenta: this is why I don't have garrett@ as an alias.  gdamore@ is much less SPAM prone than gdamore@. :-)
[02:41:08] <Error_404> I apparantly have a bug assigned to me & a sponsor for it which nobody bothered to inform me about
[02:41:23] <gdamore> a command is fine.  probably nicer than editing a config file.
[02:41:27] <icon> done
[02:41:36] <icon> need to check... is bdb part of the base sol install?
[02:41:45] <gdamore> i don't think so.
[02:41:48] <icon> damn.
[02:41:52] <gdamore> there is ndbm.
[02:41:59] <icon> hrmm
[02:42:16] * gdamore wonders if bdb is installed with sendmail....
[02:42:28] * icon checks
[02:42:50] <icon> SUNWbdb is on here
[02:42:55] <icon> this is a fresh box with a full install
[02:43:05] *** fik has joined #opensolaris
[02:43:10] <gdamore> hmmm. not on my S10 box.
[02:43:12] <icon> so far dependencies arent looking too bad
[02:43:27] <icon> SUNWbdb, SUNWwget, SUNWTcl
[02:43:50] <gdamore> yep, there it is on SNV.  It indicates "sun-private".
[02:43:53] <icon> probably going to use rsync for syncing recipies etc.
[02:43:56] <icon> ahh
[02:44:16] <icon> okay, what does a sun-private package entail?
[02:44:23] <gdamore> i have no idea.
[02:44:36] <gdamore> i'm guessing they just don't want 3rd parties depending on it.
[02:44:39] <icon> rsync might be a problem
[02:45:08] <icon> i really want to keep this to existing solaris packages only
[02:45:23] <icon> makes it easy to install a single package (since its just tcl) and off you go
[02:45:33] <gdamore> btw, I don't have SUNWwget on my SNV 50 system
[02:45:45] <icon> theres -r and -u
[02:45:48] <icon> SUNWwgetr
[02:45:51] <icon> SUNWwgetu
[02:46:06] <gdamore> You should write this in java or perl, if you want it to "just work"
[02:46:13] <icon> well here are the problems
[02:46:30] <icon> java tends to be pretty slow with heavy io, and cli is clumsy as hell
[02:46:41] <icon> perl is chaotic on a production system
[02:46:49] *** hile_ has joined #opensolaris
[02:46:53] <icon> you have no idea how much crap is goign to be installed on there, and depending on CPAN packages is a no-no
[02:46:59] <gdamore> you will likely to find tcl to be just as chaotic
[02:47:08] *** hile_ has quit IRC
[02:47:09] <icon> tcl isnt as bad provided that you stay with the system libraries
[02:47:29] <alanc> *mumble, mumble* stupid bdb license won't allow others to link with it...
[02:47:30] <gdamore> not everyone will have tcl, and some users will have different versions
[02:47:34] <icon> python tries to do entirely too much and tries very hard to forget its scripting roots
[02:47:45] <icon> tcl is probably one of the most stable scripting languages
[02:47:48] <gdamore> alanc: wtf is bdb in the default sun install then?
[02:47:49] <icon> i think the last major release was 03?
[02:48:02] <alanc> so that Sun programs can link with it!
[02:48:31] <gdamore> Sun should make a deal to get a license that is more open, or provide its own database instead.
[02:48:38] <gdamore> (fixing ndbm would be nice.)
[02:48:39] <icon> yeah but its bdb
[02:48:45] <Error_404> like javadb
[02:48:48] *** hile_ has joined #opensolaris
[02:48:52] <icon> damned near everything uses bdb (for good or bad)
[02:49:05] <gdamore> icon: are you trying to write a script or a program?
[02:49:20] <Tpenta> alanc: so you've got your case approved, when are we likely to see the probes? ;)
[02:49:23] <gdamore> i think portadm should really be a "program" :-)
[02:49:28] <alanc> Tpenta: nv_53
[02:49:30] <icon> portadm is more glue than anything else
[02:49:44] <icon> solaris already provides everything needed to make it work
[02:49:49] <gdamore> and fwiw, I don't think java is slow if you don't try to start up a GUI
[02:49:50] <Tpenta> cool, along with the jds for hal stuff
[02:50:03] <alanc> waiting for code review to finish now - if that's done by Monday, it will make our code freeze for nv_53
[02:50:07] <gdamore> the GUI classes are terribly slow.  but otherwise it starts up pretty quick
[02:50:10] <icon> gdamore: im a j2ee arch on the outside ;)
[02:50:19] <icon> java in a server env is fantastic
[02:50:29] <icon> but file/disk io isnt the hottest when it comes to some systems
[02:50:40] <Tpenta> cool, and I see that danmcd has fixed the punchin NAT issue, COOL, now we just need it pushed out to the global servers
[02:50:44] <icon> not to mention cli support just isnt there (its not a scripting language)
[02:51:03] <gdamore> but do we really need super fast file/disk i/o?  i mean, this isn't oracle we're trying to replace.
[02:51:12] <gdamore> (and, hopefully it is getting better with each release)
[02:51:20] <icon> not at all, *but* there is a lot of io going on
[02:51:29] <icon> especially with repository resync's, etc.
[02:51:34] <icon> env's get tricky to manage too
[02:51:38] <gdamore> what is the problem with cli support?  you mean lack of getopts(3)?
[02:51:48] <icon> nah, getopt is trivial to replace
[02:51:55] <icon> more about creating your commands to execute
[02:51:56] <gdamore> curses?
[02:51:57] *** sahafeez has quit IRC
[02:52:04] <icon> ProcessBuilder is an improvement
[02:52:18] <gdamore> write a dumb class that does System.exec() or whatever then. :-)
[02:52:27] <icon> but the rub is this... you have to buffer all output to stdio and stderr before getting the exitValue
[02:52:34] <icon> which means memory consumption goes *way* up
[02:53:02] <icon> im finishing up an intellij driver that has to do that because MKS are a bunch of assholes that dont use their own crufty api's
[02:53:33] <gdamore> do you need the output?  if not just redirect to /dev/null.  (maybe sh -c '$command >/ dev/null')?
[02:53:41] <icon> not a good idea
[02:53:47] <icon> if something blows up you need to show the output
[02:54:05] <icon> you could just echo everything to stdio, which is a bit ugly or to a log file
[02:54:14] *** sommerfeld has quit IRC
[02:54:17] <gdamore> sure, a log file would work.
[02:54:26] <icon> java just isnt intended to be used like a scripting language
[02:54:36] <gdamore> its not a scripting language.
[02:54:52] <icon> when 90% of the work is done by invoking other commands, theres no real need to use java
[02:54:57] <gdamore> but i don't really understand the problem.  i think you should be writing a real program, not a script
[02:55:16] <icon> tcl can be written to function as a full blown program
[02:55:22] <icon> darwinports is based off of tcl
[02:55:26] <icon> i never understood why until now
[02:55:51] <gdamore> tcl is very powerful, but it also can't decide whether its a programming language or scripting language.
[02:56:00] <icon> no argument there
[02:56:05] <icon> but its simple and explicit
[02:56:10] <icon> you dont have code creep like you do in perl
[02:56:20] <icon> and it doesnt have grandiose ideas of what it is supposed to do like python
[02:56:28] <gdamore> i've spent waayy to much of my life debugging/sustaining tcl scripts.  i can't stand it.
[02:56:31] <icon> it doesnt forget its scripting roots
[02:56:38] <gdamore> that's true.
[02:56:49] <icon> i dont have much to pick from is the big problem
[02:57:02] <gdamore> most of our GUI platform management software (for Tadpole) is written in Tcl
[02:57:03] <hile_> if one isn't going to use sed/awk/etc, I think I'd rather use python than perl for a scripting language
[02:57:07] <icon> basically its sh, tcl, python, and perl
[02:57:12] <gdamore> I'm still having a hard time understanding the objection to java.
[02:57:12] <hile_> i'm not a tcl experienced person
[02:57:19] <icon> gdamore: too much cli invocation
[02:57:24] <icon> its a waste of what java is good at
[02:57:45] <icon> almost everything the ports system does is done by other programs/scripts
[02:57:47] <gdamore> you mean you spend too much time doing exec() some other program?
[02:57:54] <icon> pretty much
[02:57:58] <gdamore> ksh then.
[02:58:11] <icon> ksh isnt quite there
[02:58:18] <gdamore> what do you need that ksh lacks?
[02:58:19] <icon> rather, its not rich enough
[02:58:34] <icon> no politics :)
[02:58:40] <gdamore> ksh88, not 93.
[02:58:45] <icon> ouch even worse
[02:58:54] <gdamore> ksh88 == no politics.
[02:59:11] <gdamore> but if you're spending ~all your time just sitting in wait() for some other program, ksh should do the job
[02:59:22] <icon> id like it to be more expressive
[02:59:29] <gdamore> how so?
[02:59:38] <icon> i want to treat it like a program which uses exec's only when neccessary
[02:59:50] <gdamore> so you want to write a real program, not a script. :-)
[02:59:51] <icon> things like using bdb or ndbm for supporting the repo index
[02:59:52] <icon> etc.
[03:00:03] <icon> gdamore: i need both :)
[03:00:08] <Tpenta> more expressive? How about:
[03:00:09] <icon> thats why tcl is an ideal choice
[03:00:09] <Tpenta> export PS1="Well Daaaaarlings "
[03:00:18] <icon> ouch ;)
[03:00:38] <icon> if [ -x ${SANITY} ] then; ....
[03:00:54] *** ndroux_ has joined #opensolaris
[03:01:06] <gdamore> it sounds to me like the only objection to java is that exec is somewhat awkward
[03:01:16] <icon> i think thats a very valid point
[03:01:27] <icon> envp gets very difficult too
[03:01:30] <gdamore> i also think you can work around that with a wrapper class
[03:01:36] <icon> its not worth the effort
[03:01:37] <gdamore> ditto
[03:01:50] <icon> why would i go to so much pain just to use java, when another language is better suited?
[03:01:51] <gdamore> but you avoid the package dependency problem
[03:01:59] <icon> i have that no matter what
[03:02:06] <gdamore> i disagree.
[03:02:24] <icon> i need a fs based db for index information
[03:02:31] <icon> unless you want to wait 5 minutes everytime you do a search
[03:02:39] <gdamore> in fact, I think you could build this whole system without doing even one exec(). :-)
[03:02:48] <gdamore> except maybe for "make"
[03:02:51] <icon> ahh just because i can, doesnt mean i should :)
[03:03:18] *** uebayasi has joined #opensolaris
[03:03:22] <icon> id be reinventing the wheel and taking away advantages to using other programs that are already established and stable
[03:03:26] *** kFuQ has joined #opensolaris
[03:03:35] <icon> good example: rsync
[03:03:44] <icon> or using bdb or indexing
[03:03:59] <icon> those two by themselves are sizeable enough to rewrite
[03:04:00] <gdamore> in bdb's case, i'd use a java database.
[03:04:08] <icon> meh
[03:04:10] <gdamore> (i.e. there are free ones)
[03:04:19] <icon> they dont persist particularly well
[03:04:26] <gdamore> i would not try to use bdb in any case.  better to stick with dbm
[03:04:33] <icon> now on most portadm invocations, you have major time wasted for vm startup
[03:04:40] <icon> gdamore: probably right
[03:05:18] <gdamore> i would choose perl over tcl.  it has a richer core library set, and is much more likely to "just be there".
[03:05:25] <icon> well
[03:05:29] <gdamore> especially if you use the fully encoded path to it.
[03:05:32] <icon> what about collisions for perl users?
[03:05:40] <icon> not to mention the language basically encourages code cruft\
[03:05:51] <gdamore> and tcl doesn't!?!
[03:05:52] <icon> ... wait a second, what does $_ mean HERE ?
[03:05:55] <icon> its not as bad
[03:06:14] <icon> command renaming is evil as hell
[03:06:23] * icon is reminded of forth
[03:06:25] <gdamore> no, but you at least _can_ write structured perl code if you are disciplined.  that's pretty much impossible with tcl.
[03:06:34] <gdamore> i can't stand perl, but i like tcl even less.
[03:06:35] <icon> how so?
[03:06:56] <icon> ive been digging into it more lately, and it seems ideal - but ive only maintined lots of perl in the past, not tcl
[03:07:27] <gdamore> i contend that if you think tcl is what you need, you really should be looking at ksh. :-)
[03:07:48] <gdamore> but anyway, i think i need to give this discussion a rest.  its dinner time. :-)
[03:07:54] <icon> ie: set pwd [pwd] versus my $pwd = `pwd`; seems more expressive to me
[03:08:15] <icon> gdamore: ill dig around and see what i can find on ksh
[03:08:40] <gdamore> if you do, consider writing a few simple C programs or something to help you with the gaps.  (e.g. database access)
[03:08:53] <icon> takes away cross platform support :/
[03:08:56] *** ndroux has quit IRC
[03:09:01] <icon> first thought was pure C
[03:09:05] <icon> but sparc throws in a nice wrench
[03:09:08] *** adp has quit IRC
[03:09:43] <gdamore> how so?  deliver it as two binaries.  no problem.
[03:10:14] <gdamore> any way, time to go.
[03:10:16] <gdamore> g'nite.
[03:10:17] <icon> take care
[03:12:21] *** knightblader has joined #opensolaris
[03:12:22] *** fik has quit IRC
[03:13:58] *** ndroux has joined #opensolaris
[03:18:44] <icon> brb
[03:18:45] *** icon has quit IRC
[03:19:53] *** icon has joined #opensolaris
[03:20:30] *** ndroux_ has quit IRC
[03:23:06] *** jsubl2 has quit IRC
[03:31:28] *** wesw has joined #opensolaris
[03:33:00] *** paul has joined #opensolaris
[03:35:24] *** Tridde has joined #opensolaris
[03:57:40] *** darkcmd has joined #opensolaris
[03:58:00] <darkcmd> is there only one opensolaris distribution packaging the gnu utilities as its userland
[03:58:50] <delewis> darkcmd: I believe so, though, there's a movement within Sun to get a /usr/gnu into Solaris
[03:58:51] <jamesd> so far
[03:59:07] <delewis> I imagine once Sun implements it, the other distributions will pick it up
[03:59:39] <darkcmd> other distributions of opensolaris?
[04:00:24] <richlowe> The actual interesting part of that proposal (and stuff Joe has said), is getting stuff the hell out /usr/sfw.
[04:00:39] <delewis> yes
[04:00:41] <richlowe> so (done right) /usr/gnu would actually be fairly minimal.
[04:00:46] <richlowe> non-conflicting stuff would be in /usr/bin
[04:01:15] <delewis> gsed, ggrep, etc.
[04:01:25] <delewis> would be in /usr/gnu as sed and grep, respectively
[04:01:38] <delewis> (which is the way it should've been)
[04:01:41] <darkcmd> "/usr/sfw ?
[04:01:52] <delewis> tossing that stuff in /usr/sfw was well, not well-planned, IMO.
[04:01:58] <delewis> darkcmd: Sun Freeware
[04:02:07] <richlowe> Solaris Free Ware.
[04:02:16] <richlowe> since "Sun Freeware" implies the SMC* packages.
[04:02:20] <richlowe> (isn't this Fun!)
[04:02:51] <delewis> but SMC is the initials of the Sun Freeware maintainer :-)
[04:02:56] <delewis> lovely.
[04:03:06] <richlowe> SMC is the initials of the Sun Freeware maintainer, and the CCD gatekeeper.
[04:03:18] <richlowe> .. but I don't think he has the slightest thing to do with SFW
[04:03:20] <richlowe> :)
[04:03:36] <delewis> though, he could as sfwnv is out in the public now
[04:03:48] <delewis> (which puts into question the purpose of the CCD)
[04:04:01] <richlowe> the stuff in SFW is supported.
[04:04:05] <richlowe> the stuff in the CCD (SFW*) is not.
[04:04:08] <delewis> and the CCD stuff isn't?
[04:04:09] <delewis> ah
[04:04:36] <delewis> well, that pretty much invalidates my reason for using the CCD, though, I haven't used it in a year or so.
[04:04:44] <richlowe> ideally the CCD would be extensive, up-to-date, easy to install...
[04:04:50] <richlowe> and everyone would work together.
[04:04:51] <richlowe> (hah)
[04:06:55] <delewis> yeah. right.
[04:07:20] <delewis> we all know with the emergence of 50,000 package repositories for Solaris (each with their own perspective of what's "right") that has a 0% chance of happening.
[04:08:18] *** gnu2it2 has joined #opensolaris
[04:20:35] *** esproul has quit IRC
[04:20:38] *** esproul has joined #opensolaris
[04:21:16] <delewis> heh, lovely.
[04:21:35] *** eboutilier has joined #opensolaris
[04:21:36] <delewis> the Solaris 9 setup_install_server doesn't look the df output that's given for a ZFS pool.
[04:21:49] <delewis> (there's no leading slash on ZFS pools in df output)
[04:23:08] <jamesd> it think its a remotely mounted filesysme
[04:23:10] <jamesd> m
[04:23:32] <delewis> jamesd: because of the lack of a leading slash
[04:23:37] <jamesd> yeap
[04:23:52] <delewis> stat64("export/jumpstart/sol-9-905hw-ga-sparc-dvd", 0x08047930) Err#2 ENOENT
[04:23:55] *** laca has quit IRC
[04:24:01] <delewis> that's easy to fix, though
[04:24:06] *** hile_ has quit IRC
[04:24:06] <delewis> I'll just copy the script off and edit it
[04:27:36] <Doc> just comment out the following :
[04:27:43] <Doc> > #if [ ! -b "${dfout}" ] ; then
[04:27:44] <Doc> > #    echo "${myname}: \"${dfout}\" is not a local mount of ${IMAGE_PATH}"
[04:27:44] <Doc> > #    echo "           cannot export \"${dfout}\" for install clients"
[04:27:44] <Doc> > #    cleanup_and_exit 1
[04:27:44] <Doc> > #fi
[04:27:51] <delewis> ah, thanks Doc
[04:33:46] *** alanc is now known as alanc-away
[04:35:12] *** sahafeez has joined #opensolaris
[04:35:20] *** OnkelSchorsch has joined #opensolaris
[04:48:43] *** dclarke has joined #opensolaris
[04:50:06] *** lplatypus has joined #opensolaris
[05:14:38] *** darkcmd has quit IRC
[05:22:37] *** darkcmd has joined #opensolaris
[05:22:50] <darkcmd> hi again, just had to reboot my computer
[05:23:15] <dclarke> why ?
[05:23:29] <darkcmd> was trying out a livecd opensolaris
[05:23:49] <dclarke> uh huh .. anyone we would know ?
[05:24:08] <darkcmd> no
[05:27:14] *** darkcmd has quit IRC
[05:31:33] <gisburn> is there an option to stop gcc complaining about a "Prototype mistach" ?
[05:31:48] <richlowe> Yeah, fix the code.
[05:33:03] <dclarke> ha ha
[05:33:17] <dclarke> never were truer words spoken .. er .. typed
[05:33:35] <gisburn> richlowe: please
[05:33:42] *** qbit has joined #opensolaris
[05:33:44] <gisburn> richlowe: stop it, ok ?
[05:34:03] * dclarke runs and hides
[05:34:09] <gisburn> richlowe: I have more urgend things do fix than this one, ok ?
[05:34:20] <dclarke> try to get rid of the -Wall option ?
[05:34:23] <gisburn> is there an -W* option to stop gcc complaining about a "Prototype mistach" ?
[05:34:36] <dclarke> thats what I was looking up
[05:34:42] <dclarke> can you use Studio 11 ?
[05:34:47] <dclarke> no .. probably not
[05:34:56] <gisburn> dclarke: gcc shadow build
[05:35:05] <dclarke> ah
[05:35:20] * dclarke nods ..
[05:35:24] <gisburn> gcc rubs it's godforsaken nose on things which are too subtle to explain.
[05:35:25] <dclarke> looking for the man page
[05:35:29] <gisburn> it's stupid
[05:35:47] <gisburn> and sometimes I curse wesolows for enabling the gcc shadow build this early
[05:35:50] <dclarke> I was recently bootstrapping GCC 3.4.6 on Solaris 2.5.1 for x86
[05:35:51] <gisburn> or whoever did this to me
[05:35:54] <dclarke> its been a real fight
[05:36:10] <gisburn> yeah
[05:36:20] <gisburn> today I am ||-close to cancel the project
[05:36:22] <gisburn> finally
[05:36:58] <gisburn> nothing but pain
[05:37:21] <dclarke> thats one of the reasons I use Studio for _everything_ at Blastwave
[05:37:25] <dclarke> less pain
[05:37:32] <gisburn> nah
[05:37:40] * dclarke except KDE .. but thats another story
[05:37:40] <gisburn> gcc is sometimes usefull
[05:37:42] <richlowe> "This early"
[05:37:49] <richlowe> they worked for close to a *year* to make that work.
[05:38:04] <richlowe> I'll let you describe it as "This early" to wesolows.
[05:38:07] <richlowe> but can I watch? :)
[05:38:17] <gisburn> and sometimes the gcc warnings seem to be designed by people who have their a** glued together
[05:38:21] <dclarke> make popcorn
[05:38:38] <dclarke> gisburn:  you are funny as hell
[05:38:43] <gisburn> ?!
[05:38:46] <dclarke> gisburn:  thanks .. I needed a giggle
[05:38:57] <gisburn> lplatypus: *THANKS*
[05:38:58] <lplatypus> oops wrong one
[05:39:02] <gisburn> ?!
[05:39:05] <gisburn> lplatypus: ?!
[05:39:14] <gisburn> lplatypus: ah
[05:39:21] <gisburn> lplatypus: I want to disable the warning.
[05:39:25] <lplatypus> http://gcc.gnu.org/onlinedocs/gcc-4.1.1/gcc/Warning-Options.html#Warning-Options
[05:39:50] <richlowe> I'll note now, just for kicks, that when it comes to code review, I'm going to call out each and every warning you disabled and ask why you didn't fix it.
[05:40:01] <richlowe> so it'd be best to include comments explaining *why*, not just *what*
[05:40:12] <richlowe> (noting now for kicks, not complaining for kicks)
[05:40:13] <richlowe> cursed commas.
[05:41:03] <lplatypus> the -Wblah options are inverted using -Wno-blah
[05:42:31] <lplatypus> I can't see an option for disabling the mismatched prototype warning... I guess it's bad to have mismatched prototypes anyway
[05:43:00] <gisburn> umpf
[05:43:04] <gisburn> yeah
[05:43:15] <gisburn> the problem is that the AST code is right and gcc wrong
[05:43:24] <gisburn> fixed in gcc 4.x
[05:43:36] <lplatypus> where is the wrong prototype coming from?  is it a gcc builtin?
[05:44:11] <richlowe> gisburn: so explain that in the comment. :)
[05:44:18] <gisburn> it's coming from the ast sources
[05:44:31] <gisburn> richlowe: give me the option to turn it off, please
[05:45:01] <dclarke> what ver of GCC ?
[05:45:06] <dclarke> 4.1.1 ?
[05:45:26] <gisburn> 4.1.blaba
[05:45:33] <gisburn> (AFAIK)
[05:45:41] <dclarke> please type gcc -v
[05:45:49] <richlowe> dclarke: Nevada's
[05:45:55] <richlowe> 3.4.3 (csl-sol210-3_4-20050802)
[05:46:02] <dclarke> old
[05:46:03] <dclarke> okay
[05:46:32] <dclarke> see my comment about about building GCC 3.4.6
[05:46:49] <dclarke> and do yourself a favour and install the GCC 3.4.5 from Blastwave
[05:47:02] <dclarke> at least until we get 3.4.6 hammered out
[05:47:54] <gisburn> umpf
[05:48:02] <gisburn> I am building OS/Net, remeber ?
[05:48:21] <dclarke> yeah
[05:48:28] * dclarke reads about -Wstrict-prototypes
[05:48:31] <dclarke> one sec
[05:49:02] <richlowe> there's nothing more irritating than the way TML/osol.o molests code.
[05:49:07] <richlowe> well, there is, it's the way b.o.o molests everything.
[05:49:08] <richlowe> but...
[05:49:14] <dclarke> oops
[05:49:20] <dclarke> NMI
[05:49:26] <dclarke> gisburn : see http://gcc.gnu.org/onlinedocs/gcc-3.4.6/gcc/Warning-Options.html#Warning-Options
[05:49:30] <dclarke> be right back
[05:49:41] *** Gman is now known as GmanAFK
[05:49:46] * dclarke : NMI service vectors
[05:50:42] <lplatypus> yeah dclarke I looked through the gcc-4.1.1 version of that page and found nothing appropriate... looks like 3.4.6 isn't any different
[05:53:23] <lplatypus> there is -w which means "Inhibit all warning messages" :-)
[05:54:36] *** eboutilier has quit IRC
[05:56:47] * gisburn sets CW_NO_SHADOW and gives up.
[05:57:28] <Tpenta> as long as you dont intend to integrate like that roland
[05:57:37] <gisburn> umpf
[05:58:30] <gisburn> Tpenta: To be honestly: I no longer think this project will succeed.
[05:58:35] *** gisburn has quit IRC
[05:58:50] <richlowe> I guess I'll take the blame for that one.
[05:58:50] <Tpenta> I'm much more optimistic that it will
[05:59:55] <lplatypus> this is the project to have ON compiling with gcc right?
[06:00:03] <richlowe> No, it already does that.
[06:00:31] <lplatypus> what's the project then?
[06:00:55] <richlowe> Roland's ksh93 stuff.
[06:01:46] <lplatypus> oh okay
[06:02:12] <richlowe> which is why I think disabling every warning encountered is Wrong.
[06:02:28] <richlowe> Well, I'd think that either way, but I'd be more open to argument, I guess.
[06:06:34] <lplatypus> I can understand disabling some warnings that can sometimes be triggered just because the compiler isn't clever enough, such as "use of possibly uninitialised variable", but I can't see why a "prototype mismatch" warning should be ignored
[06:08:29] *** MikeE has joined #opensolaris
[06:12:21] *** MikeE has quit IRC
[06:14:53] *** axisys has joined #opensolaris
[06:18:18] <richlowe> I wouldn't ignore either, since even the compiler maybe wrong, it's easy enough to silence.
[06:23:00] <sahafeez> question - can you do a jumpstart that is a like a netboot - i want to make it the normal install questions - i just want to boot from the net
[06:23:48] <richlowe> jumpstart without a sysidcfg or profile, and it'll ask.
[06:24:34] <sahafeez> ok. the setup script for a jumpstart server is on the sol 10-u2 cd? (i have not done this since 2.51)
[06:25:56] <dclarke> question .. has anyone had JDS crash Nautilus when you try to move a icon on the desktop ?
[06:26:03] <dclarke> this looks to be GNOME 2.14
[06:26:53] <sahafeez> hahaha. yes. this is my drag+drop bug i have been bitching about. ping Gman with the info
[06:27:12] <richlowe> dclarke: on sparc?
[06:27:16] *** karrotx has joined #opensolaris
[06:27:21] <dclarke> okay .. it looks to be common to GNOME 2.14 and not JDS
[06:27:26] <dclarke> on Sparc .. yes
[06:27:31] <sahafeez> no, no issues on x86
[06:27:38] <richlowe> yeah, ping Gman anyway, maybe it'll help 'em find it. :)
[06:27:57] <sahafeez> sparc only - prob xsun related
[06:28:02] <dclarke> I think the issue lay with the GNOME codebase
[06:28:19] *** bobbyz has quit IRC
[06:28:21] <Tpenta> sahafeez: do you still have the problem if you use a them other than nimbus?
[06:28:23] <dclarke> because I have GNOME 2.14 here from Blastwave and it has the same problem at the moment
[06:28:34] <sahafeez> nimubs?
[06:28:39] <richlowe> sahafeez: the default theme.
[06:28:41] <dclarke> my snv_49 box does the same thing too
[06:28:48] <Tpenta> oops i meant that for dennis
[06:28:52] <richlowe> I'dd restored much of my gconf stuff by the time I'd noticed, buut I don't know if that included the theme.
[06:28:56] <richlowe> hmmmm
[06:29:01] <sahafeez> well i only see it on an box 48-51 and v52s on sparc in any theme
[06:29:06] <richlowe> Tpenta: hey, do you have the md5sum of the snv51 respin handy?
[06:29:07] <richlowe> (x86)
[06:29:08] <Tpenta> there was a bug in nimbus which would cause it to restart with a null pointer deref
[06:29:18] <Tpenta> isnt it on the site?
[06:29:26] <richlowe> Tpenta: ah, seee.
[06:29:27] <Tpenta> i don't have it handy, no
[06:29:41] <richlowe> Tpenta: yes, it is, but Karyn had said it defininetly wouldn't be up today, not until monday-ish.
[06:29:53] <richlowe> just wanted to double check this wasn't the original 51
[06:30:02] <Tpenta> the orig 51 did not go up
[06:30:15] <richlowe> then they beat their guess  by 3 or 4 days, nice. :)
[06:31:21] <richlowe> Tpenta: for what it's worth, I'm giving thought to the flag day stuff, but there's no really elegent solution.
[06:31:35] *** bobbyz has joined #opensolaris
[06:31:40] <richlowe> you  either teach people that closed v. src matters (not going to work)), or you auto-build closed-bins more often. (kinda cruddy, but better than thee former)
[06:32:05] <Tpenta> about the only thing I can think of is that when we see something like this that affects closed builds, that we do a run off the following nightly
[06:32:40] <richlowe> That's rather manual though.
[06:32:51] <axisys> is there a xls or csv to html converter out there?
[06:32:54] <Tpenta> given that for teh most part you CAN use older closed bins; it's just the odd one like this one that is importanmt
[06:33:05] <richlowe> Tpenta: exactly, it's just flags that matter.
[06:33:06] <axisys> oxygene sent this http://www.45.free.net/~vitus/software/catdoc/ which i just got
[06:33:18] <richlowe> Tpenta: but teaching people to announce them seems unlikely to work, and the process can't *stay* manual...
[06:33:31] <sahafeez> i can jumpstart a sparc box from a x86 box right? only reason i ask is some of the docs i found say they have to be the same
[06:33:34] <axisys> i used xls2csv and now looking for csv to html
[06:33:55] <richlowe> sahafeez: you can, yes.
[06:34:02] <richlowe> sahafeez: it's easier if you use the DVD image.
[06:34:05] <Tpenta> why did you just not load it up into openoffice and export as html?
[06:34:10] <richlowe> if you're using CD's, you'll need a sparc machine to read the slice with the miniroot on it.
[06:34:18] <richlowe> (it's UFS, and thus endian dependent)
[06:34:35] <axisys> Tpenta: the csv file?
[06:34:50] <Tpenta> the xls file
[06:34:57] <sahafeez> hum. shit. i just realized my jumpstart is x86 cdrom only and my box is sun...
[06:35:16] <axisys> i want something really fast.. so that I can read an attachment within mutt
[06:35:35] *** wesw has quit IRC
[06:36:03] <axisys> Tpenta: i like to integrate it w/in mutt
[06:36:35] <axisys> Tpenta: that is why i was looking for a converter.. i can read html using lynx w/ in mutt.. works pretty well
[06:36:48] <lplatypus> richlowe: you can read ufs-sparc with an x86 box running linux :P
[06:38:25] <axisys> richlowe: i finally were able to get the mercurial going.. hg clone ssh://anon at hg dot opensolaris dot org/hg/onnv/onnv-gate
[06:39:17] <richlowe> Yeah, some of it was being down for maint, some of it was just being down.
[06:39:52] <richlowe> Bugger.
[06:40:11] <axisys> now i can just move the whole usr dir into /export/mybuild before step 8 here http://www.blastwave.org/articles/BLS-0050/index.html
[06:40:14] *** dakrone has joined #opensolaris
[06:41:05] <axisys> i guess adding this part before step 8 should make that page uptodate
[06:42:09] *** dakrone has quit IRC
[06:43:41] <richlowe> I don't know why you'd move stuff around, but sure.
[06:44:11] *** dakrone has joined #opensolaris
[06:45:03] <sahafeez> i have an ss5 i could use for jumpstart
[06:48:21] <axisys> richlowe: i had to move usr/src from /export/mybuild/onnv-gate to /export/mybuild  .. opensolaris.sh script was looking for it there
[06:48:55] <Tpenta> the idea is that you modify the opensolaris.sh script to point at the right place
[06:51:25] <axisys> Tpenta: i did change there as http://www.blastwave.org/articles/BLS-0050/index.html suggested.. so i just kept complying w/ it
[06:53:23] <Tpenta> :)
[06:56:45] *** estibi has quit IRC
[07:05:05] *** vmhobbes has quit IRC
[07:08:43] <axisys> anyone use mutt? i get the indicator bar red working fine on sol 8 but not on sol b48
[07:09:18] <axisys> i start mutt like this on both OS COLORFGBG="green;black" TERM=xterm-color mutt
[07:09:29] *** gnu2it2 has quit IRC
[07:10:20] *** Xh4 has left #opensolaris
[07:10:24] *** Xh4 has joined #opensolaris
[07:10:37] <axisys> the indicator suppose to be red according this theme I use http://www.rafb.net/paste/results/souJdr87.html
[07:11:00] *** hyde has joined #opensolaris
[07:11:16] <axisys> aacording to line 13 on http://www.rafb.net/paste/results/souJdr87.html
[07:12:23] <hyde> I can't get netbackup bpcd to work in solaris 10, in svcs -x, it always shows up as in maintenance mode, and it seems the xml config file got corrupted. How to remove it from svcs and recreate it using inetconv?
[07:12:24] <axisys> i ssh to the systems and then run mutt.. if i just connect thru a video/kbd/mouse to sol b48 the color works fine
[07:13:08] <axisys> when i ssh to sol 8 .. i have no color issue
[07:13:15] <axisys> may be some terminfo bug ?!
[07:22:11] <dwc-> probably
[07:29:36] *** sahafeez has quit IRC
[07:39:40] *** hyde has left #opensolaris
[07:52:34] <dclarke> hello
[07:52:38] <dclarke> http://www.blastwave.org/dclarke/GNOME/GNOME_2.14.png
[07:52:42] <Tpenta>             hi dennis
[07:52:55] <dclarke> I think we have a build of GNOME for Solaris users on Solaris 8 or 9
[07:53:10] <Tpenta> :)
[07:53:11] <dclarke> Sol10 can stick with JDS
[07:53:27] <dclarke> but thats looking okay for other Solaris users
[07:53:38] <dclarke> well fixup the catalog and push this out tomorrow
[07:54:03] <dclarke> it comes with FireFox and Gaim and a pile of stuff .. all the goodies
[07:54:21] <dclarke> now I'm tired ..
[07:54:30] <Tpenta> it would be getting late over there
[07:55:38] <dclarke> 2AM ... not too bad for me
[07:55:45] <dclarke> but .. its been a long long day
[07:55:50] <dclarke> nighty
[07:56:05] *** dclarke has quit IRC
[07:56:10] <Tpenta> g'nite
[08:06:29] *** ndroux has quit IRC
[08:10:54] *** uebayasi has quit IRC
[08:30:00] *** vmhobbes has joined #OpenSolaris
[08:41:59] *** slowhog has left #opensolaris
[08:45:09] *** sahafeez has joined #opensolaris
[08:45:19] *** OnkelSchorsch has quit IRC
[08:47:29] *** dakrone has quit IRC
[09:00:34] *** bank has joined #opensolaris
[09:30:58] *** Fish has joined #opensolaris
[09:31:27] <bank> hi Fish
[09:31:58] <Fish> hello
[09:41:29] *** Azureus has joined #opensolaris
[09:43:08] <bank> bye
[09:43:11] *** bank has quit IRC
[10:10:16] *** alanc_away has quit IRC
[10:10:17] *** GmanAFK has quit IRC
[10:11:01] *** alanc_away has joined #opensolaris
[10:11:17] *** GmanAFK has joined #opensolaris
[10:38:14] *** alobbs has quit IRC
[10:57:00] *** robj has joined #opensolaris
[11:06:55] <Doc> *** robj (i=robj@nat/sun/x-ae18d16bf0a5c22d) has joined channel #opensolaris
[11:07:02] <Doc> hmm.. @nat/sun/xxxx - is that new?
[11:08:52] <robj> dunno - but I'm going through a lot of walls to get to freenode - ipsec tunnel from home to sun then back out through sun's firewall
[11:09:28] *** sahafeez has quit IRC
[11:10:03] *** qbit has left #opensolaris
[11:10:10] <quasi> Doc: I've seen others with the same
[11:10:18] *** Doc2 has joined #opensolaris
[11:10:28] <Doc2> hrmm..  and meanwhile i get an IP address.. go figure
[11:10:31] *** Doc2 has quit IRC
[11:11:42] *** Doc2 has joined #opensolaris
[11:11:42] <quasi> mine says n=none@apache/committer/mads - freenode cloaks
[11:11:52] <Doc2> hmm.. interesting
[11:12:00] *** LordKing has joined #opensolaris
[11:12:25] <quasi> I think they give them to supporters of various kinds
[11:13:49] <quasi> something like http://freenode.net/faq.shtml#projectcloak
[11:14:35] <Doc> yah.. just never seen them come up for sun ppl before
[11:15:00] <quasi> I've seen them over the last month or so
[11:36:04] *** elflord has joined #opensolaris
[11:45:16] *** Griff has joined #opensolaris
[11:45:17] *** Griffous has quit IRC
[11:50:21] *** bank has joined #opensolaris
[11:51:06] <bank> hey
[11:51:23] <jmcp> heyhey
[11:51:39] <bank> wow
[11:51:42] <bank> hi
[11:51:45] <bank> long time no see.
[11:51:48] <bank> !
[11:52:03] <jmcp> yeah, had a fairly chaotic existence for the last few weeks
[11:52:35] <bank> :O :O
[11:52:38] <bank> what is that!
[11:52:42] <bank> chaotic existence
[11:52:56] <bank> about project manager?
[11:53:07] * jmcp snorts
[11:53:10] <jmcp> yes and no
[11:53:48] <bank> ah i see
[11:53:53] <bank> dictionary help me.
[11:54:06] *** Ezechial_ has joined #opensolaris
[11:54:08] <bank> what's happen
[11:54:09] <jmcp> there are some which are complete f!@#$!$%!^!%$!#!ng idiots and others who are quite good
[11:54:19] <jmcp> I got asked to help on a performance issue.
[11:54:41] <jmcp> the tech lead was adamant that his solution from 3 years ago was quite fine now and that the monitornig system was the problem
[11:54:56] <jmcp> I re-educated him
[11:58:02] <jmcp> I told the tech lead that my team and the openview team refused to give the ok for him to put these boxes into production
[11:58:09] <jmcp> so the next day the project manager escalated the issue
[11:58:19] <jmcp> I took the gloves off
[11:59:20] <jmcp> and I escalated right back
[11:59:46] <jmcp> and it got to the point yesterday where I cc'd the local Director as well as my management chain
[11:59:59] <jmcp> I *nearly* wrote "what part of 'no you may not do this' do you not understand?"
[12:00:32] <bank> what is "openview " team
[12:00:39] <Doc> i warned you, didnt i? :)
[12:01:09] * twincest yawn
[12:02:28] <bank> by the way ,real life english is hard to understand
[12:03:21] <jmcp> bank: HP OpenView - a monitoring solution
[12:03:32] <jmcp> Doc: yeah, you and J and gregp and damned near everybody
[12:05:28] <bank> ahhh
[12:05:54] <bank> what is different between adament and insist?
[12:06:03] <jmcp> good question
[12:06:12] <jmcp> have you looked at dictionary.com for the meanings?
[12:06:17] <jmcp> it's generally quite good
[12:06:43] <bank> yes ... it translate in thai the same... ( insist .. adament )
[12:06:43] <jmcp> http://dictionary.reference.com/browse/adamant compared with http://dictionary.reference.com/browse/insist
[12:06:50] <jmcp> oh, darn
[12:06:59] <bank> yeun gran
[12:07:26] <bank> ok it may better to throw english to thai dictionary on 21 year old
[12:07:37] <bank> :D
[12:07:39] * jmcp nods
[12:08:00] <jmcp> the English language is a real p.i.t.a. a lot of times
[12:08:02] <jmcp> sorry :(
[12:09:25] <bank> jmcp is dreamming for a better day
[12:09:34] <jmcp> definitely
[12:10:27] <lasseoe> hrm
[12:11:58] <bank> jmcp
[12:12:03] <jmcp> that' sme
[12:13:34] <bank> I ever install php by ..   extract / tar -xvf ./configure / make / make install ....  but I need to reconfigure some parameter ... so I would like to delete it..  is that a properly way to just rm -rf that folder?
[12:13:55] <jmcp> cd  to the directory above the folder, then run "rm -rf foldername"
[12:14:13] <jmcp> general case, your mileage may vary, i am not a lawyer, always test your code before deploying
[12:14:14] <bank> just simple remove that?
[12:14:15] <jmcp> etc etc
[12:14:22] <jmcp> oh hang on
[12:14:28] <jmcp> you want to remove the *installed* php?
[12:14:33] <bank> yes ..
[12:14:46] <bank> that way I install was ( make , make install as describes above)
[12:14:50] <jmcp> ah
[12:14:52] <jmcp> more difficult
[12:14:58] <bank> oh. !
[12:15:07] <bank> what if I already .... rm -rf ?
[12:15:23] <bank> and I going to install newer version.
[12:15:24] <jmcp> you'll need to figure out which files were installed, and remove them
[12:15:37] <bank> oh.
[12:16:04] <bank> I was make my global zone dirty.
[12:17:47] <bank> jmcp: so you are now in normal state?
[12:18:02] <bank> working atmosphere
[12:18:12] <jmcp> kinda
[12:18:19] <jmcp> depends on what you call "normal" :)
[12:19:34] *** alanc_away has quit IRC
[12:20:43] <twincest> apparently raid10 is a "less used" md raid scheme (linux kernel raid)
[12:20:50] <twincest> i wonder what linux people use instead on raid10?
[12:28:33] *** cneira has joined #opensolaris
[12:29:17] <cneira> anyone knows how to get the equation of a line in a staroffice graphic?
[12:36:43] *** bunker has joined #opensolaris
[12:37:58] *** Tpenta changes topic to "Latest builds: SXCR: 51a, ON build: 51, ON nightly: 20061030"
[12:41:07] *** stevel has joined #opensolaris
[12:41:07] *** ChanServ sets mode: +o stevel
[12:41:18] <Tpenta> hello steve, in korea?
[12:41:25] <stevel> yup
[12:41:32] <stevel> just got here
[12:41:39] <Tpenta> i'm just tarring up my copy of yesterdays clone to start building
[12:42:04] <stevel> i'll kick off some builds now too
[12:42:10] <Tpenta> I suspect that we should probably do a full "nightly" given todays flag day
[12:42:44] <Tpenta> i cant wait for punchin.aus to get this, then we can have more than one person punching in from the syd04 wireless net
[12:43:10] * stevel wonders if punchin.aus is closer than punchin.sfbay
[12:43:17] <Tpenta> so what takes you to korea?
[12:43:18] <stevel> eh. i'm going to .sfbay hosts in the end anyway, so it's probably a wash
[12:43:20] <stevel> tech days
[12:43:30] <Tpenta> lucky you
[12:43:35] <jmcp> stevel: having fun over there?
[12:43:46] <stevel> jmcp: so far, but i've only been here all of an hour
[12:43:48] <Tpenta> evening james
[12:43:49] <jmcp> :)
[12:43:52] <jmcp> hi Tpenta
[12:43:59] <jmcp> I see danmcd has been busy
[12:44:35] <stevel> ugh. so much email already
[12:45:12] <stevel> oh this looks fun.
[12:45:23] <stevel> the first 20 emails are all regarding various opensolaris.org servers not responding
[12:45:36] * stevel must have missed a fun day
[12:45:36] <Tpenta> o dear
[12:45:52] <Tpenta> someone also needs to update the on downloads page to point it at sxcr 51a
[12:45:59] <jmcp> can't wait to get my hands on snv_53
[12:46:12] <stevel> uh-oh.
[12:46:14] <stevel> the bridge failed
[12:46:15] <stevel> damn
[12:46:28] <jmcp> a tacoma narrows-style failure?
[12:46:35] <jmcp> or something not quite so bad ?
[12:46:51] <Tpenta> my fave film shown to me in engineering statics
[12:46:52] <stevel> not sure... i may have to rollback and retry one commit to see
[12:47:03] <jmcp> hm
[12:47:24] <jmcp> wow, I've somehow managed to accrete 56gb of packages in my /scratch/pkg directory
[12:47:29] * jmcp investigates
[12:47:54] <Tpenta> stevel - you do know what the reference to the tacoma narrows bridge was about don't you?
[12:48:21] <stevel> yeah
[12:48:31] <stevel> well, i know the disaster - i don't know the details of said disaster
[12:48:40] <Tpenta> gotta watch those resonant frequencies
[12:49:49] <Tpenta> i didnt know that, ... "the first version of the bridge, nicknamed 'galloping gertie", ..."
[12:49:57] <trs81> twincest: I imagine LVM striping across RAID 1 arrays is how most people do it in linux - more flexibility  as to whether each volume is striped or not
[12:51:53] <elflord> hi * ^^
[12:52:34] <elflord> is there any way to edit the menus in jds? i mean... i used gnome for quite a while, but i cannot find the menu editor anywhere o_O
[12:53:23] <sickness> morning all
[12:54:28] <jmcp> elflord: not at present
[12:54:33] <jmcp> elflord: hassle Gman next time he's not afk
[12:55:22] <elflord> oh, okay, thanks ^^
[12:55:35] <elflord> so terminal is my dear friend i assume, right? ^^
[12:55:49] <jmcp> dunno
[12:55:50] <jmcp> probably
[12:55:56] <elflord> okay
[12:58:51] <elflord> thanks ;)
[13:00:29] <bank> I have to compile PHP with support for the mysqli extension. I finding mysql.so and found at /usr/sfw/lib/ser/modules/mysql.so.
[13:01:06] <bank> but I not sure is this enough. I see on other linux distribution it have something like php5-mysql  package to install
[13:03:31] <quasi> well, ignore the lunix dists for now - they have a bad habit of splitting stuff into a million little packages
[13:04:01] <stevel> has anyone built onnv-gate since cindy mcguire's putback?
[13:04:28] <Tpenta> i would imagine danek has
[13:04:40] <oxygene> quasi: compared to blastwave's huge packages with even bigger dependency lists, the linux way is quite sensible
[13:04:43] <stevel> sorry, from the mercurial gate i mean
[13:04:53] <Tpenta> oh
[13:05:03] <stevel> ohhh. it was undone
[13:05:19] <sickness> is the file to be downloaded named sol-nv-b51-x86-dvd-iso-?.zip or sol-nv-b51a-x86-dvd-iso-?.zip
[13:05:50] <sickness> I can't find b51a on the site, only b51 (manually writing this url: http://javashoplm.sun.com/ECom/docs/Welcome.jsp?StoreId=7&PartDetailId=Sol-Express_b51-DVD-x86-SP-G-B&TransactionId=try )
[13:06:02] *** spawrq has quit IRC
[13:06:39] <Tpenta> that's it
[13:07:13] <stevel> ugh. yeah. the bridge definitely failed
[13:07:18] <stevel> woohooo. first bug found in a while
[13:08:01] <sickness> Tpenta: so, no 51a in the file name, but it is 51a?
[13:08:08] <stevel> sickness: yeah
[13:08:19] <Tpenta> yea, if you go to the page it's 51a on the web page
[13:08:33] <sickness> tnx :)
[13:09:12] <sickness> uhm, well, now that I have a better look at it, only sparc is 51a, intel is 51, is this ok? (I'm going to download intel)
[13:09:43] <Tpenta> it was a sparc only problem
[13:09:55] <sickness> oh, tnx! :)
[13:11:45] <quasi> oxygene: there is that
[13:15:35] <bank> erm ..
[13:15:46] <bank> if i have /usr/sfw/lib/ser/modules/mysql.so.is that right to configure like this ./configure --with-apxs2=/usr/apache2/bin/apxs --with-mysql=/usr/sfw
[13:17:10] <oxygene> bank: uh, btw, is that solaris10 or solaris express/opensolaris?
[13:17:27] <oxygene> on solaris 10, you need an updated libxml2 for php5
[13:19:08] <oxygene> bank: looks fine, though I doubt it's looking for that ser/modules/... thing - it's more likely that it looks for /usr/sfw/lib/libmysqlclient.so
[13:21:59] *** Triskelios has quit IRC
[13:24:45] <bank> opensolaris,
[13:24:54] <bank> I opensolaris b50
[13:25:11] <bank> oh because I search for mysql.so
[13:25:31] <bank> yes I already have libxml2. but I just want to install phymyadmin for users
[13:25:46] <bank> that tell me it can't find mysql , mysqli extension ...
[13:25:53] <bank> so I need to reconfigure and install php again.
[13:26:02] <bank> ( I have mysql on another non-global zone)
[13:26:26] <bank> oxygene: Am I do configure parameter correctly?
[13:26:29] <oxygene> it needs this client library
[13:26:38] <oxygene> looks fine
[13:28:42] <oxygene> it's what I find in my package tree, too
[13:29:25] <stevel> damnit. the bridge missed something in the FMA amd64 putback
[13:29:27] <stevel> ugh.
[13:29:34] <stevel> i think i need to rollback all the way to february and regenerate
[13:30:48] <stevel> oh that sucks so hard
[13:30:49] <stevel> :(
[13:31:52] <jmcp> sucks so hard it blows
[13:32:01] <jmcp> that putback was a looooong time ago
[13:32:10] * jmcp sends a stevel a sympathy packet
[13:32:20] <stevel> yeah :(
[13:33:06] <stevel> oh well. i'll tackle it on monday possibly
[13:35:42] *** LordKing has quit IRC
[13:37:32] <jmcp> stevel: can't wait to see you grow that second head
[13:37:39] <stevel> it's grown
[13:37:43] <stevel> and it's got a throbbing headache
[13:39:16] <quasi> have a few drinks, that might dull the pain (and if not, you'll at least have had fun trying :)
[13:40:01] <jmcp> stevel: what's the local drink in Korea?
[13:40:07] <stevel> soju
[13:40:12] <elflord> damnit
[13:40:13] <jmcp> have a few rounds
[13:40:34] <elflord> i ran into some problems with svm... mis-configured the vfstab... is there any way to fix it?
[13:40:44] <jmcp> elflord: docs.sun.com really is your friend
[13:40:51] <elflord> failsafe told me that c1t0d0s0 is under md control...
[13:41:03] *** koolniczka has joined #opensolaris
[13:41:04] <stevel> yeah i will
[13:41:08] <stevel> i gotta go out and find some dinner too
[13:41:11] <stevel> i'm starving
[13:41:11] <elflord> followed by "no solaris os instances found"
[13:42:01] *** lplatypus has quit IRC
[13:42:20] <jmcp> elflord: do you have a failsafe boot option?
[13:42:27] <elflord> yeah
[13:42:30] <elflord> already booted it
[13:42:42] <jmcp> so what's printing that message?
[13:42:58] <elflord> told me that my root slice is under md control
[13:43:12] <elflord> followed by no solaris instances found
[13:43:38] *** Ezechial_ has quit IRC
[13:43:48] <bank> erm ... I rm -rf old php on global zone , then I install newer version with new configuration.
[13:44:04] <elflord> ah okay, got my root slice mounted, well maybe i can fix the vfstab...
[13:44:09] <bank> will each apache+php in each non-global zone update to new version after reboot?
[13:44:49] *** nachox has joined #opensolaris
[13:44:49] <jmcp> elflord: remember to fix both sides
[13:44:59] <elflord> yepp, thx
[13:45:11] <nachox> morning everyone
[13:45:15] <elflord> good morning ^^
[13:45:22] <jmcp> hi nachox
[13:45:32] <elflord> this is my first try with svm... should have given the slices better names ><
[13:45:43] <elflord> okay, i guess it's some case of learning by doing ;)
[13:45:48] <jmcp> elflord: did you print out the docs on how to get svm setup?
[13:45:58] <elflord> yeah
[13:46:45] <jmcp> follow them as if they're Moses' stone tablets
[13:46:54] <jmcp> those docs are *very* good
[13:48:02] <elflord> ^^
[13:48:05] <elflord> okay ^^
[13:48:20] <Doc> thanks. i try
[13:48:27] <jmcp> Doc: ha!
[13:48:34] <jmcp> elflord: it's not only Doc who contributed to them
[13:48:37] <Doc2> me too
[13:48:44] *** Doc2 has quit IRC
[13:48:46] * jmcp thwaps Doc *and* Doc2
[13:49:31] <elflord> jmcp: is it possible to start from scratch with the metadbs?
[13:49:36] *** darkcmd has joined #opensolaris
[13:49:38] <jmcp> elflord: should be
[13:49:43] <jmcp> elflord: metaclear is your friend here
[13:49:48] <stevel> okay i'm off.  gotta go find something to eat, find something to drink, and then find something to bang my head against for a while
[13:49:49] <elflord> oh, okay
[13:49:52] <stevel> 'night folks
[13:49:52] *** cneira_ has joined #opensolaris
[13:49:56] <elflord> i'll try, thanks for the hint :)
[13:49:59] <jmcp> stevel: have fun and sleep well
[13:50:10] *** stevel has quit IRC
[13:50:15] <bank> quasi:  after I configure with --with-mysql=/usr/sfw phpmyadmin still said "Could not load either mysql or mysqli extension"
[13:50:26] <jmcp> elflord: /usr/sbin/metaclear -a     <<< deletes all metadevices and configured hotspare pools
[13:50:33] <elflord> cool okay
[13:50:42] <elflord> this also applies if i have NOT booted from md?
[13:50:57] <jmcp> that's my recollection, yes
[13:51:03] <elflord> okay, thanks
[13:51:27] <jmcp> you might need a "-f" too
[13:52:06] <elflord> k
[13:52:14] *** dekop has quit IRC
[13:52:37] <elflord> another question: yesterday i noticed that smc does not work as expected when using a root password longer than 8 chars... is this normal behaviour?
[13:53:01] <jmcp> elflord: probably
[13:53:05] <jmcp> unfortunately
[13:53:26] <jmcp> SMC is a layered product (spit!) which ... has to meet password requirements on multiple versions of Solaris
[13:53:31] <jmcp> the common denominator there is 8 chars max
[13:53:42] <elflord> whoa okay
[13:54:08] <elflord> i mean... when you have root just as role... this seems to be okay.. but if its a normal user... kinda insecure, aint it?
[13:54:36] <jmcp> I don't follow you
[13:55:48] <elflord> okay err...
[13:56:37] <elflord> i configured the system in a way that root is just a administrative role
[13:57:27] <elflord> so that i have to log in with another user to become root
[13:57:53] <jmcp> ok
[13:58:00] <elflord> and i mean, if root just were a normal user, then 8 chars for pw seems to be a bit "small"... or have i got something wrong?
[13:58:19] *** koolniczka has quit IRC
[13:58:34] <jmcp> elflord: how many chars are you expecting? and what are your expectations about how secure Solaris password stuff is?
[13:58:38] <twincest> elflord: then don't use crypt() passwords
[13:58:45] <twincest> elflord: solaris supports md5 and blowfish as well
[13:58:45] <jmcp> and of course, it's only as secure as the password you supply
[13:59:00] <twincest> oh, SMC bug, who knows
[13:59:09] <twincest> i'd be dubious of smc's security anyway myself
[13:59:15] <elflord> okay ^^
[14:00:37] *** fik has joined #opensolaris
[14:00:54] *** bank has quit IRC
[14:01:40] <nachox> hmm, where i work the policy is 10 chars with numbers and at least 1 upercase letter and at least a symbol
[14:02:07] *** cneira has quit IRC
[14:03:31] <elflord> sry if i ask dumb questions, i'm new to solaris and still learning ^^
[14:03:34] <elflord> but i'm very impressed
[14:04:00] <jmcp> they're not dumb questions
[14:04:24] <nachox> not asking at all is worse
[14:04:43] <Doc> not asking google first is even worse tho...
[14:04:48] <elflord> ^^
[14:04:50] <elflord> okay ;)
[14:05:31] * nachox puts Doc in the top drawer
[14:06:57] <jmcp> Doc: are you back in Sydney?
[14:07:10] <Doc> yes
[14:07:35] <elflord> umm... i guess i did something wrong ><
[14:07:47] <elflord> i put back my old vfstab
[14:07:54] <elflord> rebooted
[14:08:02] * elflord ran a metaclear -a
[14:08:25] <elflord> and then reverted the entry for / back to "normal", without md
[14:08:42] <elflord> now mdconfig and sac are failing during bootup and logins are disabled, argh
[14:09:05] <jmcp> elflord: so what messages do you see from mdconfig?
[14:09:22] *** laca has joined #opensolaris
[14:09:25] <elflord> well, hold on... just booted failsafe...
[14:09:56] *** laca_ has joined #opensolaris
[14:10:49] *** laca_ has quit IRC
[14:11:43] <elflord> so...
[14:12:05] <elflord> at first pmconfig complains about a bad file number while accessing /etc/.cpr_config
[14:12:14] *** bank__ has joined #opensolaris
[14:12:23] <elflord> than cron jumps in with cannot create fifo queue
[14:12:58] <elflord> then svc says that /system/sac:default failed with status 16
[14:13:11] <elflord> and at last it says that mdmonitor failed fatally
[14:13:39] <jmcp> which step did you fail to do correctly while following the metaroot documentation?
[14:14:25] <elflord> the problem was
[14:14:50] <elflord> while a did metainit, i got a typo in while doing metainit for /usr
[14:15:12] <elflord> it just got a wrong number which i couldnt remember than
[14:15:23] <elflord> so bootup failed because it could not mount /usr
[14:15:24] *** TBCOOL has quit IRC
[14:15:39] <elflord> i booted failsafe and copied back my old vfstab
[14:15:45] <elflord> there, just / was md'ed
[14:15:52] <elflord> rebooted
[14:15:59] <elflord> transitioned to single user
[14:16:05] * elflord ran metaclear -a
[14:16:12] <jmcp> elflord: why have you put /usr on a separate slice?
[14:16:42] <elflord> i'm using freebsd for quite a while... was common practice there
[14:17:57] <elflord> so when setup created the partitions for me, i told it to create seperate partitions for /,  /usr and /var
[14:18:08] <jmcp> ok
[14:18:10] <jmcp> fair enough
[14:18:16] <jmcp> it's not necessary on Solaris though
[14:18:17] <darkcmd> Does anyone happen to know if most residential ISPs would be agree if someone ran a mail server for personal use
[14:18:32] <jmcp> darkcmd: depends on how clueful said ISP is
[14:18:41] <jmcp> internode.on.net is mine, they're ok with it
[14:18:46] <nachox> i've heard of some isps that actually block that kind of traffic
[14:18:57] <jmcp> nachox: optusnet is one, telstra bigpond another
[14:19:01] <darkcmd> my ISP, is a crappy ISP, it's AT&T/Yahoo
[14:19:16] <elflord> jmcp: should i try removing the svm configuration files in /etc/lvm?
[14:19:19] <nachox> jmcp: and it actually makes sense if you ask me
[14:19:47] <jmcp> elflord: I actually suggest you do a reinstall from scratch, as long as you haven't done too much customisation
[14:20:01] <elflord> in fact i'd like trying to fix it
[14:20:08] <elflord> just for learning purposes
[14:20:20] <jmcp> nachox: for a lot of cases, yes. my problem is that I'm not joe average user, and I resent being treated like a fuckwit who doesn't understand anything
[14:20:35] <jmcp> elflord: did you do a metaclear -a -f
[14:20:45] <elflord> yeah, i did
[14:21:02] <elflord> but the mddb.cf and md.cf are still in place o_O
[14:21:47] <jmcp> elflord: it's a looooong time since I did this sort of thing
[14:21:54] <elflord> okay
[14:22:04] <elflord> i'll give deleting the files a try..
[14:22:10] <jmcp> and since it's 00:25 here (Australia/NSW) I'm going to have to call it a day and let somebody else help you
[14:22:13] <jmcp> good night all
[14:22:18] <elflord> okay
[14:22:20] <elflord> good night :)
[14:22:25] <elflord> and thanks for your help :)
[14:22:33] <nachox> jmcp: as long as they open it for you if you ask them politely it should be ok, but that could be a problem if you get your address via dhcp but if you count the tachnically able users and compare it to the infected machines that rely mail i'm guessing our technical users will have to get used to not hosting an mta
[14:24:20] *** koolniczka has joined #opensolaris
[14:26:33] <bank__> :(
[14:29:01] <bank__> hello
[14:29:06] <elflord> hey ^^
[14:29:23] <darkcmd> opensolaris has openssh by default right?
[14:29:30] <bank__> I have apache2+php at global zone ... and I place php.ini at /etc/apache2 in non-global zone
[14:29:51] <bank__> but phpinfo() still show Configuration File (php.ini) Path 	/tank/local/php-5.2.0/lib  ( GLOBAL ZONE)
[14:30:30] <bank__> but at global zone I disable apache2 and also there're no php.ini
[14:30:48] <bank__> pls take a look my phpinfo() http://mail.kranular.com/
[14:33:43] <Vanuatoo> Is it hard to create new locale to solaris?
[14:34:12] <Vanuatoo> I'm from georgia and I'm unable to type georgian characters. Where should I start?
[14:34:39] *** fik has quit IRC
[14:49:13] *** uebayasi has joined #opensolaris
[14:57:08] <twincest> that might be hard, libc_i18n is not open
[14:58:27] *** koolniczka has quit IRC
[14:59:17] <bank__> I solve that problem now.
[14:59:18] <bank__> thank you.
[15:07:12] *** doniphon has joined #opensolaris
[15:08:21] *** doniphon has left #opensolaris
[15:10:17] *** esproul_ has joined #opensolaris
[15:12:26] *** Somethingelse has joined #opensolaris
[15:12:43] *** esproul has quit IRC
[15:28:25] *** tfb is now known as [tfb]
[15:28:26] *** [tfb] is now known as tfb
[15:28:47] *** tfb is now known as [tfb]
[15:29:35] *** [tfb] is now known as tfb
[15:29:45] *** tfb has quit IRC
[15:34:34] *** gdamore has left #opensolaris
[15:51:18] *** rydis has quit IRC
[15:52:30] *** deather_ has joined #opensolaris
[15:53:17] <PerterB> hmm, th aussies are giving wales a bit of a kicking
[15:54:18] <Somethingelse> rugby?
[15:55:12] <PerterB> aye
[15:58:06] *** [tfb] has joined #opensolaris
[15:58:17] <jamesd> they are kicking the poor sea creatures,  HOW RUDE!!
[15:58:32] <jamesd> ;-P
[15:59:32] <PerterB> the welsh, on the other hand, we don't care about so much ;)
[16:03:38] <icon> hmm is it just me, or does this seem a little... off?
[16:03:39] <icon> http://www.mwave.com/mwave/amdjewl.hmx?
[16:05:01] <delewis> icon: to say the least
[16:09:40] *** deather has quit IRC
[16:12:35] <bank__> hey
[16:13:18] <icon> morning
[16:13:41] <PerterB> hmm, bus.opensolaris.org appears..... bugged (well, unusably slow)
[16:13:47] <PerterB> *bugs
[16:13:50] <bank__> I don't know why ... I do configure during php installation. and phpinfo also said  php configuration point to /etc/apache2/php.ini  ... I also add php.ini both location ( global and non-global zone) but the configuration didn't effect ...
[16:14:17] <icon> did you restart apache?
[16:14:21] <bank__> also I delete both files (php.ini) phpinfo() still working .
[16:14:23] <bank__> yes
[16:14:28] <bank__> I reboot non-global zone
[16:14:34] <bank__> restart both apache
[16:14:41] <bank__> (non-global and global
[16:14:52] <Doc> did you consider trying #php ?
[16:14:58] <bank__> I edit session.save_path  ... It have no effect to phpinfo(). both location.
[16:15:52] <bank__> ok ...
[16:18:24] *** bank has joined #opensolaris
[16:22:15] <bank> the #php doesn't know about global , non-global zone
[16:22:24] <bank> is that related to my problem?
[16:22:39] <bank> but I try by place php.ini both global , non-globla zone.
[16:22:40] <bank> dunno
[16:27:41] <jamesd> you need to see where it expects to read php.ini from
[16:28:11] <bank> I already set ./confgiure during installation
[16:28:20] <lasseoe> it's a PHP problem, has nothing to do with Solaris
[16:28:55] <bank> as phpinfo() return the Configuration File (php.ini) Path 	/etc/apache2/php.ini
[16:29:06] <bank> then I place php.ini both global zone and non-global zone
[16:29:09] <elflord> bank: it is possible to configure some php options via htaccess files in apache
[16:29:15] <jamesd> bank: get a copy of brendan's dtrace tool kit and find the file monitoring script and see where it is trying to read php.ini from
[16:29:21] <elflord> have a look at the php documentation
[16:30:54] *** wdhath has joined #opensolaris
[16:31:01] <bank> elflord: at global zone?
[16:32:00] *** loke has joined #opensolaris
[16:32:43] <elflord> i'm new to solaris, never done anything with zones... but php only reads its configuration from one source, p.e. /etc/php.ini... so you might put a .htaccess file into the directory of the php scripts in your zones so that you can configure your php options there
[16:36:42] <axisys> opensolaris build took 2 and 1/2 hrs on my ultra 20 http://www.rafb.net/paste/results/d0RmjH57.html
[16:40:26] *** jfndi has joined #opensolaris
[16:41:44] <wdhath> wow, the genunix.org frontpage news is about 7 months old
[16:42:14] <Doc> i started an opensolaris build on my ultra-5 when it was first released - i think it's almost finished!
[16:44:22] <bank> elflord:
[16:45:03] <bank> I switch back and forward between #opensolaris and #php ... solaris tell php wrong php said solaris wrong
[16:45:24] <bank> you see ? I configure config file path at /etc/apche2/php.ini './configure' '--prefix=/tank/local/php-5.2.0' '--with-apxs2=/usr/apache2/bin/apxs' '--with-config-file-path=/etc/apache2/php.ini' '--with-mysql=/usr/sfw'
[16:45:44] <bank> and I place in both global , non-global zone.  then I create .htaccess as you recommend
[16:46:03] <bank> vi /var/apache/htdocs/.htaccess
[16:46:18] <bank> and I place this php_value session.save_path /tmp then restart apache
[16:46:59] <elflord> yeah
[16:47:20] <bank> so I done everything correctly for php?
[16:47:28] <elflord> i think so, yes
[16:47:33] <bank> so .... what wrong with my solaris ..
[16:47:36] <bank> T-T
[16:49:25] <Doc> that's like saying "what's wrong with the road?" because you cant unlock the door to your car
[16:49:59] <bank> exactly
[16:51:45] <axisys> bfu# halt  is the last step .. can I run init 6 instead to boot into my first opensolaris?
[16:51:53] <lasseoe> it's an Apache/PHP issue, NOT solari
[16:51:54] <lasseoe> s
[16:52:01] <bank> OK
[16:52:12] <bank> I will belive.
[16:52:16] *** bank has left #opensolaris
[16:52:23] <axisys> Doc: heh
[17:06:36] <axisys> Doc: i build my w/ debug enabled
[17:06:48] <axisys> otherwise it would probably take half the time
[17:08:33] <Doc> i'm building mine with 128 megs of memory - i think more might have helped
[17:08:39] *** Kernel86_ has joined #OpenSolaris
[17:09:40] *** Kernel86|Laptop has quit IRC
[17:10:09] <axisys> Doc: :-)
[17:10:31] <axisys> Doc: i thought opensolaris needs more memory
[17:10:44] <axisys> Doc: i guess it will still run
[17:11:03] <Doc> it probably does - i'm actually cross-compiling it under windows 98
[17:11:06] <axisys> Doc: so how long has it running
[17:11:15] <axisys> it been*
[17:18:50] <axisys> now that i have snv_48 what is the recommened way to upgrade?
[17:24:05] <delewis> Live Upgrade or a regular one
[17:24:16] <twincest> bfu!
[17:26:04] <axisys> delewis: to say upgrade to next build
[17:26:19] <delewis> ON or Nevada?
[17:26:24] <twincest> axi: live upgrade and regular upgrade and both file to upgrade to the next build
[17:26:44] <twincest> s/file/fine
[17:26:57] <axisys> delewis, twincest I am lost guys.. not sure of the answer
[17:27:10] <twincest> axi: do you know what live upgrade is?
[17:27:39] <delewis> axisys: do you not understand the distinction between ON and Nevada?
[17:27:53] <axisys> no i don't *sigh*
[17:27:58] <axisys> http://www.blastwave.org/articles/BLS-0050/index.html I followed this
[17:28:01] <delewis> ON is a single OpenSolaris bit
[17:28:08] <delewis> Nevada is an entire OpenSolaris distribution
[17:28:23] <twincest> i wish blastwave would come up with better titles than 'bls-0050'
[17:28:32] <twincest> it's not hard to put a descrition of the document in the url :)
[17:28:33] <axisys> so which is i did when i followed this http://www.blastwave.org/articles/BLS-0050/index.html
[17:28:34] <delewis> ON merely includes a userland, libc, kernel, etc.
[17:28:43] <delewis> axisys: I suggest you read the actual OpenSolaris documentation
[17:28:50] <axisys> twincest: i know live upgrade in solaris environment.. not opensolaris
[17:28:50] <delewis> rather than bothering with 3rd party HOWTOs
[17:28:55] <delewis> that leave out important distinctions like this
[17:29:08] <twincest> axisys: solaris express _is_ solaris
[17:29:32] <axisys> twincest: so then i installed opensolaris on top of it
[17:29:39] <twincest> axisys: how did you do that?
[17:29:49] <axisys> using http://www.blastwave.org/articles/BLS-0050/index.html
[17:30:24] <axisys> basically it explains how to install opensolaris.. i followed the first link here http://opensolaris.org/os/community/documentation/doc_index/install/
[17:30:25] <twincest> oh dear
[17:30:33] <twincest> you installed it using BFU
[17:30:43] <twincest> you really should not do that if you don't even know what ON and Nevada are
[17:30:48] <delewis> which means you effectively cannot upgrade to a newer Nevada release
[17:30:49] <axisys> ok .. cuz i did use bfu
[17:30:53] <twincest> anyway, after bfu the only upgrade path allowed is another bfu
[17:30:54] <delewis> your package database is fried.
[17:30:57] <delewis> if you want to BFU
[17:31:00] <delewis> create a separate LU BE
[17:31:03] <twincest> and that only upgrades ON, not the rest of the OS
[17:31:10] <delewis> BFU that, but keep an un-bfu'd BE around
[17:31:17] <delewis> so that you can upgrade to the next Nevada release
[17:31:20] <delewis> (and then re-bfu that if you like)
[17:31:47] <delewis> this also protects you from brickifying your system.
[17:32:04] *** uebayasi has quit IRC
[17:32:22] <axisys> i guess then i sh'd stick w/ Nevada all the way.. its my workstation.. i can rebuilt np
[17:32:43] <delewis> axisys: just leave an extra ~10GB slice laying around for a Live Upgrade BE and you can play with ON all you want
[17:32:44] <axisys> is there a good doc on how to install nevada?
[17:32:53] <delewis> documentation for installing Nevada?
[17:32:56] <delewis> why do you need that?
[17:32:57] <twincest> axisys: insert CD, follow prompts
[17:33:30] <axisys> twincest, delewis; insert solaris xpress cd and install.. thats it?
[17:33:41] <twincest> what else would it need?
[17:33:50] <delewis> Nevada is just the codename for Solaris 11, which is currently in the Solaris Express program.
[17:33:59] <delewis> it *is* Solaris
[17:34:10] <delewis> which is an OpenSolaris distribution
[17:34:16] <delewis> which implies it makes use of several OpenSolaris bits
[17:34:19] <delewis> namely ON, JDS, etc.
[17:34:38] <axisys> in other words just install solaris xpress :-)
[17:34:43] <axisys> ok
[17:34:56] <twincest> yes, if you want to install solaris express, install solaris express
[17:35:29] <delewis> the Nevada "Developer's Reference" explains all of these distinctions quite clearly.
[17:35:37] <delewis> I'm just really understanding where the difficulty is coming from.
[17:35:42] <delewis> not*
[17:36:09] *** alanc-away has quit IRC
[17:36:13] *** GmanAFK has quit IRC
[17:36:14] *** jacotton has quit IRC
[17:36:18] <delewis> http://www.opensolaris.org/os/community/on/devref_toc/
[17:36:29] <delewis> no need for watered-down 3rd party HOWTOs
[17:37:02] *** robj has quit IRC
[17:37:05] <delewis> http://www.opensolaris.org/os/community/on/devref_toc/
[17:37:06] <delewis> er
[17:37:09] <delewis> http://www.opensolaris.org/os/community/on/devref_toc/devref_glossary/
[17:37:12] <delewis> this is especially useful.
[17:37:36] <axisys> ok.. so before i go read that url let me ask one last question
[17:37:48] *** GmanAFK has joined #opensolaris
[17:38:05] <axisys> so to build nevada u install sol xpress and then add opensolaris bits on top?
[17:38:21] <delewis> *no*
[17:38:32] <delewis> Nevada is currently synonymous for Solaris Express
[17:38:39] <delewis> there is no *building* Nevada.
[17:38:55] <twincest> axisys: after you install SX, there is no need to install anything on top
[17:38:57] <delewis> it includes OpenSolaris bits which are open but many closed bits as well
[17:38:58] <axisys> delewis: now i am clear
[17:39:00] <twincest> you already have the "opensolaris bits"
[17:39:01] <delewis> just as CDE and Xsun
[17:39:13] <delewis> s/just/such/
[17:39:29] <axisys> now i follow.. maaann .. uff
[17:44:56] *** Kernel86|Laptop has joined #OpenSolaris
[17:45:21] *** Kernel86_ has quit IRC
[17:47:40] *** Kernel86_ has joined #OpenSolaris
[17:49:06] *** Kernel86_ has quit IRC
[17:49:15] *** jfndi has quit IRC
[17:49:16] *** kleppari has quit IRC
[17:49:30] *** Kernel86_ has joined #OpenSolaris
[17:50:14] *** kleppari has joined #opensolaris
[17:50:31] *** Kernel86|Laptop has quit IRC
[17:50:49] *** wdhath has quit IRC
[17:54:57] *** Fish- has joined #opensolaris
[18:11:15] *** Fish has quit IRC
[18:11:30] *** dakrone has joined #opensolaris
[18:20:53] <bank__> hi all, I solve that by override some configure
[18:21:10] <bank__> thank you and sorry if I disturbing you sometime
[18:21:11] *** bank__ has quit IRC
[18:28:24] *** fik has joined #opensolaris
[18:32:12] *** fik has quit IRC
[18:34:50] *** nachox has quit IRC
[18:36:07] *** alanc has joined #opensolaris
[18:43:20] *** mega has joined #opensolaris
[18:48:38] <axisys> i just enabled apache2 using smf
[18:49:03] <axisys> how do i enable browsing a dir
[18:49:08] *** fik has joined #opensolaris
[18:50:00] <quasi> you ask in #apache ;)
[18:50:59] <axisys> quasi: :-)
[18:53:22] <quasi> you could of course also read the Fine manual - http://httpd.apache.org/docs/2.2/mod/mod_autoindex.html
[18:59:11] <Error_404> as for enabling it, just edit the config file
[18:59:22] <Error_404> smf picks it up & launches apache
[19:00:13] *** Cybernd has joined #opensolaris
[19:01:17] *** dakrone has quit IRC
[19:03:38] * quasi makes note to self of not using default profile for httpd then
[19:05:08] * quasi smells unplanned downtime and other interesting side effects ;)
[19:11:50] *** elflord has quit IRC
[19:14:17] *** mega has quit IRC
[19:14:54] <quasi> hmmm, the default on the s10 I'm looking at does not depend on the conf file
[19:15:03] <dwc-> I hope at the very least it doesn't restart unless it passes the syntax check
[19:16:02] <quasi> well, the problem is that there's several types of restart
[19:16:19] <quasi> +other
[19:17:07] *** Elendal has quit IRC
[19:18:50] *** laca is now known as lacaAFK
[19:18:52] *** coffman has joined #opensolaris
[19:20:55] *** djgregor has joined #opensolaris
[19:36:34] *** robj has joined #opensolaris
[19:36:39] *** jacotton has joined #opensolaris
[19:36:42] *** alanc-away has joined #opensolaris
[19:42:31] *** loke has quit IRC
[19:43:41] *** zarathustra has quit IRC
[19:45:29] <richlowe> movement: linking to monaco for Bug ID's is bad and wrong.
[19:48:10] *** DrewMarin has joined #opensolaris
[19:48:22] <DrewMarin> Hey
[19:48:54] *** ofu has quit IRC
[19:49:16] *** ofu has joined #opensolaris
[19:50:32] <DrewMarin> can anybody tell me if I can install linux apps on open solaris
[19:50:49] <jamesd> check out the brandz community on opensolaris.org
[19:51:14] <quasi> http://opensolaris.org/os/community/brandz/
[19:52:26] <DrewMarin> thx for the url but I found it
[19:53:44] *** kimc has quit IRC
[19:55:05] <axisys> quasi: alias and directory directive did what i needed.. thnx
[19:56:55] <axisys> quasi: i just want my crawler go thru all dirs
[19:57:37] <movement> richlowe: I know. lapse of judgement.
[19:57:53] <axisys> does apache 2 that comes w/ SX uses mpm?
[19:59:16] <axisys> all i see are these core.c, prefork.c, http_core.c and mod_so.c .. so I guess not
[20:00:12] <axisys> i guess if i want to run it on T2000 i should compile w/ mpm
[20:01:49] <axisys> oh wait i take it back .. prefork.c is MPM
[20:03:26] <quasi> axisys: yes, it uses the prefork mpm - which is fine if you want something that isn't threadsafe like php - if you want performance, then either the worker or the event mpm are worth a look
[20:05:06] <axisys> quasi: so in T2000 if i dont have php added i should compile w/o prefork.c correct?
[20:05:30] <quasi> you should either use --with-mpm=worker or event
[20:06:03] <quasi> the default is the same as --with-mpm=prefork
[20:07:09] <axisys> quasi: ok
[20:08:15] <icon> worker does well in smp systems
[20:08:22] <quasi> event doesn't support ssl, so you'd either need kssl or worker if you're doing ssl
[20:08:24] <icon> its a good cross between threaded and prefork support
[20:08:38] <quasi> event is pretty nifty with keepalives
[20:09:02] * icon is used to sites with too much traffic to properly support keepalives
[20:09:06] <icon> ugh that brings back ugly memories
[20:11:06] <quasi> the event mpm fixes most of that problem
[20:12:58] <axisys> i am planning to use only http .. no ssl.. i have only two IPs that will be allowed to access the http server..
[20:13:27] <axisys> can I benefit from using event mpm on ultra 20?
[20:14:42] <quasi> depends on wether keepalive makes any sense for you
[20:15:21] <icon> sure
[20:15:41] <icon> sweet
[20:15:46] <icon> sxcr 51 was released?
[20:17:00] <quasi> see /topic ;)
[20:17:36] <icon> double checking ;)
[20:17:42] <icon> i thought it was delayed until next week
[20:18:57] <axisys> quasi: i guess i should build my own apache unless there is a simple way to recompile apache w/ mpm=event
[20:19:11] <axisys> quasi: question for #apache ? :P
[20:19:48] <quasi> no recompiling - you'd have to go back to the start
[20:20:01] <axisys> quasi: k
[20:20:12] <icon> axisys: someday there will be :)
[20:20:19] <icon> at least if ports ever gets out of the sandbox
[20:20:30] <axisys> icon: :-)
[20:21:56] <icon> what the hell
[20:22:01] <icon> i cant find the sxcr iso's
[20:24:01] <icon> bah, there it is
[20:29:04] *** dunc has joined #opensolaris
[20:29:44] *** TBCOOL has joined #opensolaris
[20:30:09] *** icon has quit IRC
[20:38:13] *** karrotx has quit IRC
[20:38:54] <TBCOOL> sweet.. the solaris box panicked yet again..
[20:39:41] *** karrotx has joined #opensolaris
[20:47:26] <dunc> is anyone able to help me with a strange SATA disk / controller problem?
[20:51:04] *** trs81 has quit IRC
[21:25:24] *** nwf has joined #opensolaris
[21:27:13] *** darkcmd has quit IRC
[21:31:10] *** Azureus has quit IRC
[21:31:19] *** DrewMarin has left #opensolaris
[21:33:04] *** darkcmd has joined #opensolaris
[21:40:15] *** LeftWing has quit IRC
[21:42:22] *** LeftWing has joined #OpenSolaris
[21:42:47] *** yakov has joined #opensolaris
[21:42:48] <yakov> hi
[21:46:54] <yakov> guys, my rtls realtek detected as rtls1 but i have only 1 card so it should be rtls0  - all other config depends on rtls0 AFAIU
[21:47:07] <yakov> how can i change rtls1 to be rtls0 back?
[21:50:01] *** jafari has joined #opensolaris
[21:51:18] <yakov> i.e. when i have /etc/hostname.rtls0 - network/physical tells me that it misconfigured
[21:52:08] <jafari> hello all, does anyone have a better site than this one -> http://www.rite-group.com/rich/solaris_nat.html to set up NAT, i didnt everything possible on my Netra X1 from this site and i cant seem to get my network online?
[21:53:35] *** LeftWing has quit IRC
[21:53:38] *** LeftWing has joined #OpenSolaris
[21:54:05] <cneira_> bigadmin ?
[21:58:58] *** Fish has joined #opensolaris
[21:58:58] <jafari> what about bigadmin
[22:00:15] <jafari> i did a search on bigadmin for nat and it gave me that same link
[22:01:34] *** coffman has quit IRC
[22:06:42] *** pseudo_ has joined #opensolaris
[22:12:28] <yakov> darn, typed sys-unconfig :)
[22:12:51] <yakov> i hope it help....
[22:16:16] *** Fish- has quit IRC
[22:17:10] *** LeftWing has quit IRC
[22:17:12] *** LeftWing has joined #OpenSolaris
[22:17:37] *** Xh4 has quit IRC
[22:22:54] *** stevel has joined #opensolaris
[22:22:54] *** ChanServ sets mode: +o stevel
[22:27:18] * timeless tries to figure out what http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/cmd/zoneadmd/vplat.c#3620 means
[22:33:58] *** _dreams_ has joined #OpenSolaris
[22:35:52] <timeless> does opensolaris support usb ethernet?
[22:35:56] * timeless finds bolthole.com
[22:37:09] <Peanut> All praise Google ;)
[22:39:08] *** slowhog has joined #opensolaris
[22:42:07] *** trs81 has joined #opensolaris
[22:48:32] *** yakov has quit IRC
[22:50:56] *** stevel changes topic to "Latest builds: SXCR: 51a (sparc) / 51 (x86) | ON build: 51 | ON nightly: 20061030"
[22:52:18] *** lisppaste3 has quit IRC
[22:53:10] <Tpenta> steve, are you going tyo do a nightly drop?
[22:53:27] <stevel> yeah - it just finished, i'll post it in a few mins
[22:53:47] <Tpenta> mines done too, I'll wait for yours so I can verify my file list matches
[22:54:57] *** lisppaste3 has joined #opensolaris
[22:55:05] *** stevel changes topic to "yeahyLatest builds: SXCR: 51a (sparc) / 51 (x86) | ON build: 51 | ON nightly: 20061103"
[22:55:30] *** sahafeez has joined #opensolaris
[22:57:07] *** GmanAFK is now known as Gman
[22:57:17] <stevel> tpenta: posted
[22:57:18] <Gman> stevel, in korea? :)
[22:57:26] <stevel> gman: yup. jetlagged :)
[22:57:31] <Gman> sweet..
[22:57:32] *** rodrickbrown has joined #opensolaris
[22:58:10] <sahafeez> ugh. i hate korea. spent to much time there. give me japan any day ;)
[22:58:32] * Gman never been to korea, would love to go sometime
[22:58:38] <stevel> seems nice so far
[22:58:51] *** Somethingelse has quit IRC
[22:59:00] <sahafeez> have fun ;)
[23:00:09] <Gman> stevel, written your talk yet? :)
[23:00:26] <stevel> gman: it was already written for me ;-)
[23:00:29] <stevel> thanks to fvdl and sch
[23:00:37] <Gman> fvdl?
[23:00:38] <stevel> i just had to make very minor revisions
[23:00:41] <sahafeez> ugh. windows srv at work i just rebooted has not come back up ...
[23:00:46] <stevel> s/Hahn/Lau/g
[23:00:47] <stevel> :-P
[23:00:50] <stevel> fvdl = frank van der linden
[23:00:55] <Gman> oh right..
[23:01:05] <Gman> i hope there's pictures!
[23:01:14] <stevel> nope
[23:01:29] <stevel> it's a talk on building and deploying opensolaris.  not really a picture-friendly topic
[23:01:33] <Gman> nod
[23:05:35] <Error_404> you could have a a series of pictures
[23:05:58] <Error_404> step 1: downloading the CD: guy sitting at desk tapping pen against forhead
[23:06:21] <Error_404> step 2: burning the CD: disk drive with red light
[23:06:30] <sahafeez> dont forget the dont drag and drop on sparc slide
[23:06:36] <stevel> step 2: check out ON source from the mercurial repository
[23:06:45] <stevel> step 3: wonder why you keep getting failures
[23:06:54] <stevel> step 4: go onto IRC and bitch at stevel
[23:07:01] <Error_404> lol
[23:07:46] * stevel wanders off in search off some breakfast and coffee
[23:08:36] <sahafeez> just remeber stevel: Te audire no possum. Musa sapientum fixa est in aure.
[23:09:08] *** vmhobbes has quit IRC
[23:22:36] *** vmhobbes has joined #OpenSolaris
[23:24:21] *** Sieghard has joined #opensolaris
[23:26:55] <Sieghard> somehow i still get at least 1 panic a day
[23:28:14] <timeless> can someone explain something about the snv kernel for me?
[23:28:21] <timeless> i'm looking at two functions and i'm confused
[23:31:13] <timeless> http://pastebin.mozilla.org/1200
[23:31:54] <Error_404> i don't suppose we can pull -nd closed bins off the Hg repo, no?
[23:32:27] *** axisys has quit IRC
[23:37:31] <stevel> error_404: not yet
[23:37:32] *** trs81 has quit IRC
[23:37:43] <stevel> i've been trying to think why i shouldn't put closed-bins in the Hg repository
[23:37:54] <Error_404> *shrug*
[23:38:03] <stevel> it seems to make sense to
[23:38:35] <Error_404> it already comes as closed/i386-nd
[23:38:55] <Error_404> if $USER wants it, it'd have to be deliberate
[23:42:21] <richlowe> stevel: Hey.
[23:42:29] <stevel> hey rich
[23:42:47] <richlowe> stevel: did you see the other mail too?
[23:43:07] <stevel> possible?
[23:43:09] <stevel> possibly even
[23:43:19] <richlowe> Yeah, I don't remember if I sent it or not. ;)
[23:43:22] <richlowe> that's why I'm asking.
[23:43:26] <timeless> richlowe, did i do something wrong while reading the onv code http://pastebin.mozilla.org/1200
[23:43:29] <stevel> the cmp/diff one right?
[23:43:36] <richlowe> bugger, I did send it, huh.
[23:43:41] <richlowe> I mean, uh, yeah, that one!
[23:44:11] <richlowe> stevel: I'm more interested in how the FMA thing could have happened.
[23:44:17] <stevel> i can't do cmp, because of the unexpanded ident keywords
[23:44:21] <richlowe> stevel: if you happen to know off-hand, that'd be good, but don't investigate.
[23:44:24] <richlowe> stevel: get -k
[23:44:33] <richlowe> diff you're going to have to teach it to ignore ident stuff, and that gets messy fast.
[23:44:49] <richlowe> get -kp, rather.
[23:44:52] <stevel> yeah that's true. i was trying to avoid doing an sccs call for each file, but given the size of (most) putbacks, that's not too big a deal
[23:46:01] <richlowe> but I'm more concerned with the (first) FMA thingy.
[23:46:11] <stevel> yeah - that's what i'm most concerned about too
[23:46:14] <richlowe> the second one I can guess explanations for.
[23:46:21] <richlowe> the first, is... well, from my take on it impossible.
[23:46:29] <stevel> and i'm worried it's elsewhere in onnv-gate too
[23:46:34] <richlowe> Well, I can't check that.
[23:46:42] <richlowe> but if it is, you're lucky, onnv-gate builds and works. :)
[23:46:59] <stevel> yeah :-P
[23:47:00] <stevel> lucky me
[23:47:06] <richlowe> stevel: Hey, it's a beta.
[23:47:11] <richlowe> anyone not expecting it to break should have been. :)
[23:47:22] <richlowe> anyone who *knows you* and wasn't expecting it to break deserves what they got, too ;)
[23:47:36] <stevel> hahahaha
[23:47:37] <stevel> ouch
[23:48:07] *** LeftWing has quit IRC
[23:48:10] *** LeftWing has joined #OpenSolaris
[23:48:25] * dwc- looks at the topic
[23:48:34] <dwc-> "stevel changed the topic of #opensolaris to: yeahyLatest"
[23:48:41] *** stevel changes topic to "Latest builds: SXCR: 51a (sparc) / 51 (x86) | ON build: 51 | ON nightly: 20061103"
[23:48:44] <stevel> whoops
[23:48:53] <stevel> it's early
[23:48:55] <stevel> i'm jet lagged
[23:49:43] <dwc-> how long you gonna be there?
[23:49:46] <stevel> 8 days
[23:50:02] <dwc-> how much of it will be work?
[23:50:11] <stevel> richlowe: hrm. the SCCS history for usr/src/lib/fm/topo/libtopo/common/mod.c is definitely weird
[23:50:20] <sahafeez> stevel, like kimchi?
[23:50:28] <stevel> dwc-: most of it. i'm going to the DMZ on thursday, but otherwise monday-friday is work
[23:50:35] <richlowe> stevel: danek's fault then!
[23:50:37] * richlowe didn't say that.
[23:50:41] <richlowe> shit.
[23:50:42] <stevel> sahafeez: kimchi is great :)
[23:50:50] * stevel fwds irc log to dduvall
[23:50:57] <dwc-> is the DMZ a big tourist attraction?
[23:51:01] <richlowe> oh, the 'shit' was unrelated.
[23:51:06] <stevel> dwc-: it is for me :)
[23:51:27] <dwc-> don't get kidnapped and dragged across the line
[23:51:39] <stevel> i'm not planning on it
[23:51:39] <sahafeez> take a shot @ the n. happens all the time
[23:51:49] <dwc-> some people will be upset if you're not here to do the next drop ;)
[23:51:54] <stevel> and if i they do, i'll fight them off.... with my fountain pen and my usb flash drive
[23:52:11] <stevel> tpenta can do it ;-)
[23:52:14] <richlowe> "Don't touch me, man.  I'll break your webapps!"
[23:52:51] <stevel> richlowe: http://pastebin.ca/238299
[23:52:58] <richlowe> dwc-: the drops right now are just closed bins.
[23:53:09] * timeless frowns
[23:53:10] <richlowe> dwc-: Tpenta can roll those, I'm not sure he's done whatever source-fu though.
[23:53:20] <timeless> i'm trying to build and load a usbether module on my snv_49
[23:53:27] <richlowe> timeless: given how much I've confused myself so far today, you'd be better asking someone else.
[23:53:31] <timeless> i've built it, and i have an elf thing
[23:53:34] <sahafeez> you need power to run webservers - http://en.wikipedia.org/wiki/Image:Korean_peninsula_at_night.jpg
[23:53:35] <stevel> as near as i can tell, the bridge brought in 1.1.1.1 as the initial delta instead of 1.2
[23:53:39] <timeless> richlowe: ok,
[23:53:47] <timeless> stevel? :)
[23:54:03] <stevel> since those 3 lines of difference are the three lines i'm missing
[23:54:08] <richlowe> stevel: Hm, I thought you normally collapsed the initial CodeMgr delta?
[23:54:17] *** trs81 has joined #opensolaris
[23:54:21] <richlowe> the branch in there seems bogus.
[23:54:21] <stevel> "you" being me?
[23:54:26] <stevel> or "you" being ON developers in general?
[23:54:29] <richlowe> stevel: "you" being "everyone"
[23:54:29] <stevel> timless: hrm?
[23:54:43] <stevel> richlowe: yeah, 1.1/1.1.1.1/1.2 should have been collapsed into one initial 1.1 delta
[23:54:49] <richlowe> normally, 1.1 would have CodeMgr gunk involved.
[23:54:51] <timeless> stevel, did i do something wrong while reading the onv code http://pastebin.mozilla.org/1200
[23:54:54] <dunc> can anyone help me with a weird SATA problem? I have the controller described here:- http://www.sun.com/bigadmin/hcl/data/sol/components/details/801.html     and it's totally fine with 1 disk on either channel, but as soon as I connect 2 at once then it just hangs during "Configuring Devices" stage
[23:55:10] <richlowe> stevel: Yeah, I'd be tempted to see if danek knows what the hell is happening there.
[23:55:19] <stevel> i'll definitely ask him
[23:56:05] <richlowe> stevel: who do we bug other than you, btw? :)
[23:56:42] <Gman> $64 million question :)
[23:56:55] *** sahafeez has quit IRC
[23:57:13] <stevel> bug about what?
[23:57:16] <stevel> drops?
[23:57:22] <richlowe> No, broken tonic stuff.
[23:57:28] <richlowe> normally we have easy access to you... but you're in Korea.
[23:57:34] <stevel> timeless: certainly looks that way, but i'm not well versed in the USB code
[23:57:35] <richlowe> and the rest of the tonic folk hardly ever talk to us. :)
[23:57:51] * stevel likes to slum it ;-)
[23:58:01] <timeless> stevel: what should i do?
[23:58:08] <timeless> note that one of the other directory sets is less confused
[23:59:05] <stevel> hrm. well, SCM-related issues: garypen is probably the best person to ask.  delivery issues re: closed-bins, tpenta.
[23:59:22] <richlowe> hey, I got it right twice.
[23:59:25] * richlowe takes a small bow
[23:59:50] <richlowe> stevel: actually, the thing with mod.c is, if you ended up with 1.1 and 1.1.1.1, how did you lose 1.2?





top