[00:00:16] <CIA-25> lling: 6607671 Able to zpool upgrade a pool to an unknown version number
[00:00:51] *** jolts has quit IRC
[00:00:58] *** jolts has joined #opensolaris
[00:11:27] *** slowhog has joined #opensolaris
[00:14:42] <pitty> i have a funny problem on my x4600, second hw mirrored disk, after a reboot, i lose my vtoc and need to copy it back in for it to be usable? has anyone seen something like this before?
[00:18:06] *** vija1 has joined #opensolaris
[00:20:00] *** Wicky2 has quit IRC
[00:21:42] *** theRealballchalk has joined #opensolaris
[00:21:55] <vija1> vija1: is away
[00:21:59] *** vija1 has left #opensolaris
[00:22:18] *** MikeTLive has left #opensolaris
[00:22:38] *** ichigo has quit IRC
[00:22:49] *** slowho1 has quit IRC
[00:23:08] *** theRealballchalk has quit IRC
[00:24:53] *** theRealballchalk has joined #opensolaris
[00:24:59] *** kadath has quit IRC
[00:25:18] *** jafari has quit IRC
[00:25:40] *** theRealballchalk has quit IRC
[00:30:05] *** theRealballchalk has joined #opensolaris
[00:31:32] *** theRealballchalk has quit IRC
[00:32:58] *** Jondice has quit IRC
[00:35:07] *** theRealballchalk has joined #opensolaris
[00:44:26] *** ashner has joined #opensolaris
[00:46:50] <theRealballchalk> are real-time operating systems suitable for server oriented operations like web-hosting etc?
[00:47:11] <theRealballchalk> i guess what i'm asking is how does qnx compare with solaris minus the tools
[00:47:39] <sahafeez> not used for the same things. not sure it is a good question
[00:47:44] <sommerfeld> in the right context, solaris is (also) an RTOS
[00:48:36] <sommerfeld> (general rule of RTOS's: you need to be extremely careful how you use them.  if you are extremely careful in how you use solaris you can use it as part of a real-time system)
[00:48:43] <e^ipi_> theRealballchalk: why would you want to do that? two completely different design goals
[00:49:14] <sahafeez> but if i want something to shut down my nuke planet when things go bad i will stick with the RTOS
[00:49:24] <sahafeez> i am just sayin..
[00:49:34] <theRealballchalk> i don't know much about real-time os.  but there's a site comparing linux windows and qnx and others
[00:50:32] <sahafeez> bah, there is a site comparing different shapes and shades of poop. does not mean it is worth reading :)
[00:50:36] <sommerfeld> just put the nuke plant control daemon in a dedicated processor set...
[00:50:41] <theRealballchalk> wait why am i even comparing qnx to solaris hahaha
[00:50:49] <jbk> real-time generally means you just want to guarantee that certain things will take no longer than X to run, etc.
[00:51:02] <jbk> (i.e. bounded time)
[00:51:13] <sommerfeld> in other words, in a real time system, late answers are wrong answers.
[00:51:18] <jbk> doesn't necessairly mean faster, but predictable
[00:51:25] <theRealballchalk> oooh that's not good
[00:51:34] <theRealballchalk> i c
[00:51:50] *** ShadowHntr has quit IRC
[00:52:10] *** pitty has left #opensolaris
[00:52:12] <sahafeez> all real time means is that the os will act on something in X time frame and it can be interruped to respone to stuff.
[00:52:33] <jbk> for certain problems that's what you want
[00:53:25] <jbk> control systems can be an example of that
[00:53:30] <sahafeez> yes, nuke planets, plans, etc..
[00:53:34] <sahafeez> planes even
[00:53:35] <sommerfeld> tcp/ip is in some sense a real time problem, albeit with generally loose timing constraints.
[00:54:12] <sommerfeld> (must ack that packet within a few minutes or else peer gives up..)
[00:54:43] <iron_angel> you might call that soft RT...
[00:54:53] <iron_angel> Most OSes can do soft RT after a fashion.
[00:55:49] <sahafeez> softRT bad for nuke plant.
[00:56:26] <sommerfeld> the definition i'm familiar with uses soft/hard real time as attributes of the problem being solved, not the system being used to solve it.
[00:56:29] <sahafeez> i want tachyonOS for that
[00:57:25] <sommerfeld> hard real time -> ghostbusters-don't-cross-the-streams-bad when you're late.
[00:57:48] <sommerfeld> soft real time: "oh well, we glitched the video playback.  we'll get it right next frame"
[00:58:27] <sahafeez> very good alliteration there
[00:58:49] <iron_angel> s/alliteration/metaphor, but yes.
[00:58:50] <sommerfeld> i'd imagine there would be both hard and soft RT problems in a nuclear power plant control system.
[00:59:15] <sahafeez> iron_angel damn my english teacher!
[00:59:29] <iron_angel> the nuke plant isn't even the trickiest thing to get right, either - try something like an anti-missile system!
[00:59:49] *** NikolaVeber has quit IRC
[00:59:50] <sahafeez> yah.
[01:00:08] <sahafeez> sorry chief. missed it by that much
[01:00:34] <iron_angel> heh, yeah :/
[01:01:32] <sommerfeld> yah.  there was a clock drift problem with the patriot missile system software during gulf war I .  (workaround was regular reboots).
[01:01:49] <iron_angel> kluge city!
[01:05:24] <sahafeez> does not matter. never hit a damn thing
[01:05:42] <iron_angel> well, it *did* - just not exactly as intended.
[01:06:04] <iron_angel> On the other hand, PAC-1 was never intended for ballistic missile defense. The fact that it came close is something of a surprise.
[01:06:18] <iron_angel> PAC-3 is designed for that, but it's new.
[01:06:22] <sahafeez> hum, i have read some pretty detailed reports that it never hit a another missile. however was not made to do it so not suprized
[01:06:29] <sbahra> sahafeez, tachyonOS for?
[01:06:43] * sahafeez was in the scudbox at the time hunting..
[01:06:59] <sahafeez> we were talking about real time os's
[01:07:04] <sbahra> QNX has been an interesting read so far.
[01:07:13] *** kadath has joined #opensolaris
[01:10:19] <jbk> wow.. i think that's the fastest response i've ever seen from an email i've sent to an opensolaris list
[01:10:30] <sommerfeld> the decode ASI's thing?
[01:10:35] *** pfa3rh has quit IRC
[01:10:52] <sommerfeld> IIRC some of the ASI space is implementation-specific
[01:10:58] *** Fullmoon has quit IRC
[01:13:47] <sommerfeld> ..so there may not be an unambiguous decode.  but when there is it makes a lot of sense.
[01:14:42] <jmcp> jbk: if it could be unambiguous, I'd like to see it happen by default
[01:14:58] * jmcp goes back to his backport putback
[01:16:57] <jbk> sommerfeld: that could be possible, unfortunately the documentation doesn't do a good job of providing comparisons
[01:17:25] <sommerfeld> I suspect that (for obvious reasons) most of the ASI's you see will be the portable ones.
[01:18:35] <jbk> though the one thing i was wondering about is i do know that certain combination of ld/st instructions will generate data access exceptions -- so there is a question if those cases should still disassemble (I'm thinking yes as it's different than 'not valid', but not 100% sure)
[01:20:35] <jbk> some are also only valid w/ VIS/VISII
[01:21:49] <jbk> and a few are renamed (though appear to retain the same meaning)
[01:22:59] <jbk> ASI_QUAD_LDD_REAL -> ASI_TWINX_REAL
[01:23:55] <sommerfeld> one possibility would to decode as:  lda [%r1 + %r2] 0x80 <ASI_PRIMARY>, %r3
[01:24:15] <jbk> is that still legal syntax?
[01:24:38] <sommerfeld> by analogy to:    call      +0x2cc0e8     <PLT:sleep>
[01:25:17] <sommerfeld> or maybe  lda [%r1 + %r2] 0x80, %r3        <ASI_PRIMARY>
[01:26:16] <sommerfeld> (was root-causing an odd hang in sccli today and it turned out to be an apparently intentional sleep(1000) ....)
[01:26:27] <jbk> hmm true.. if there doesn't have to be a lot of special casing (i.e. checking combinations of instruction + asi values), it'd actually be pretty trivial -- just put the names in a table and use the asi value as the index
[01:27:17] *** sahafeez has quit IRC
[01:27:20] <sommerfeld> (or, really sleep( (x+1) * 1000)
[01:27:34] <jbk> sccli?
[01:28:01] <sommerfeld> management CLI for the 33xx/35xx storage arrays.
[01:28:05] <jbk> ahh
[01:28:28] <jbk> last time i used sun storage was an a3500 connected to an E5500 :)
[01:28:59] <sommerfeld> it can communicate either inband over scsi/FC or out-of-band via IP.  it took a 1000 second nap when the DVD-ROM drive rebuffed its advances.
[01:29:22] <jbk> haha
[01:29:38] <jbk> and i'm guessing it's probably not as obnoxious as emc gatekeepers/vcmdbs?
[01:29:56] <jbk> (the inband communication)
[01:30:05] <sommerfeld> I'm not familiar with EMC storage.
[01:31:07] <jbk> they present pseudo-luns to a host to do their communication, however you cannot write to them (usually), so you will see very frequent complaints on solaris about invalid labels
[01:31:34] <sommerfeld> i believe it can use the regular luns for inband communications to the raid controllers.  with jbods it uses the ses devices.
[01:31:53] <sommerfeld> i just have jbods hooked up right now.
[01:35:54] <sommerfeld> but.. the pseudo-luns thing sounds obnoxious.
[01:36:01] <jbk> i've often wondered about some way to tell the kernel to ignore the lack of labels on certain devices just to shut things up
[01:37:16] <jbk> it is.. mostly because they present themselves as unlabeled scsi-disks that appear to be write-protected (i never bothered to look at how the software was actually manipulating them -- probably using uscsi would be my guess though)
[01:39:06] <sommerfeld> i'd think that they'd be less obnoxious as, say, SES devices, but other os's might not cope so well.
[01:39:51] <jbk> supposedly you can temporairly make them writeable and slap a label on them, but much like making luns from multiple arrays visible to the same fc port, emc claims it's 'bad'
[01:40:18] <wesolows> They should absolutely present SES instead of attaching proprietary functionality to pseudo-LUNs
[01:40:35] <wesolows> presenting SES on existing storage LUNs should be handled by any OS as well; it's part of the standard.
[01:41:30] <jbk> i would agree, however i'm not in a position to make them change their firmware :)
[01:41:46] *** JBeck has quit IRC
[01:42:14] <wesolows> sure you are, just refuse to buy their stuff until they can prove they've read the standards
[01:42:17] *** andy__ has joined #opensolaris
[01:42:20] *** phips|mb has joined #opensolaris
[01:42:29] <jbk> i'm not the one that makes the purchasing decisions
[01:42:42] *** Andy_Pease has quit IRC
[01:42:43] <wesolows> granted, I'm sure their StripperCare(SM) sales force is a big selling point.
[01:42:48] <jbk> :)
[01:43:13] <wesolows> Now with 28% more kickbacks in every bottle!
[01:43:38] <jbk> i wouldn't be surprised if something like that happened at my old job
[01:44:07] <jbk> one of their competitors offered _really, really_ good pricing, and they still went all emc
[01:44:21] <wesolows> No doubt.
[01:44:37] <jbk> i always found it amusing the san architecture there was geared towards maximizing cost
[01:44:47] <wesolows> that's what SANs are for though
[01:45:05] <wesolows> boondoggles and budget black holes with plenty of sleazy sales tactics and kickbacks galore.
[01:45:10] <jbk> you need this box to talk to two different arrays? well you need at least 2 fc ports
[01:45:11] <sommerfeld> reinventing the wheel, ignoring decades of network engineering experiences, etc.,
[01:45:31] <wesolows> With margins in excess of 70% across the board, it's a bit like unregulated gambling: it lends itself to organised crime
[01:46:03] <jbk> just annoying much like the original meaning of 'i' in raid has been changed, they seem to now forget what the 'n' in san was supposed to mean
[01:46:36] <wesolows> The 'I' still maintains its original meaning in 'raidz' and 'raidz2' :-)
[01:46:49] <jbk> true
[01:47:12] <sommerfeld> when the first google hit returns "Fredrickson also said EMC could ?never substantiate? the stripper-and-whipped-cream incident. " you gotta wonder
[01:47:26] <jbk> what annoys me though is i meet people that don't even know what the original meaning was
[01:47:35] *** slowhog has left #opensolaris
[01:47:42] <wesolows> sommerfeld: I don't really wonder; I have little doubt it's all true.
[01:47:56] <wesolows> sommerfeld: The real question is, what *haven't* we heard rumours of?
[01:48:36] <sommerfeld> I meant to continue ".. what else happened"
[01:48:47] <wesolows> indeed :-)
[01:49:45] <wesolows> That's really the only thing I miss about being on the buying end of IT.
[01:50:05] *** Gekkko_ has joined #opensolaris
[01:50:08] <sommerfeld> clearly StripperCare is the only thing that's going to consistently trump the golf course deals.
[01:50:28] <wesolows> Well, that and outright kickbacks, yes.  But people go to jail for that.
[01:54:07] *** Gekkko_ is now known as Gekkko
[01:54:21] *** stevel has quit IRC
[01:55:24] <jbk> out of curiousity is the fact that some of the stuff in ON uses things in /usr/include (as opposed to the versions in workspace) that prevents it from being built on older releases?
[01:58:06] <sommerfeld> there are a number of reasons.  that's merely one of them.
[01:58:15] <flyingparchment> ON depends on all kinds of odd stuff to build, like apache
[01:58:32] <sommerfeld> the big one is dependencies on stuff (headers & libraries) not in ON.
[01:58:52] *** nrubsig has quit IRC
[01:59:06] <sommerfeld> another is our general ethos of self-hosting by developers.  as in "damnit, you better actually be using the bits you're developing"
[01:59:12] *** Fullmoon has joined #opensolaris
[02:00:01] *** nrubsig has joined #opensolaris
[02:00:07] <nrubsig> flash crash
[02:00:13] *** ChanServ sets mode: +o nrubsig
[02:00:35] *** vmlemon has quit IRC
[02:01:24] *** willnorris has joined #opensolaris
[02:10:47] *** _hsilva has joined #opensolaris
[02:11:29] *** Gekkko has quit IRC
[02:13:31] *** alfism has quit IRC
[02:14:17] *** Gekkko has joined #opensolaris
[02:17:41] *** pizdec_ has joined #opensolaris
[02:20:58] *** dunc has joined #opensolaris
[02:21:54] *** Andy_Pease has joined #opensolaris
[02:23:57] *** Megaf|Sono|Dormi has joined #OpenSolaris
[02:24:24] *** __hsilva has quit IRC
[02:25:23] *** alanc is now known as alanc-away
[02:28:10] *** perlmonk has quit IRC
[02:28:34] *** andy__ has quit IRC
[02:29:22] *** Shiv__ has joined #opensolaris
[02:30:52] *** Das_Gekt has joined #opensolaris
[02:34:02] *** Gekkko has quit IRC
[02:35:02] *** pizdec__ has joined #opensolaris
[02:36:51] *** vijay has joined #opensolaris
[02:37:01] *** rcorreia has quit IRC
[02:37:44] *** charlesnw has quit IRC
[02:38:29] *** Shiv__ has left #opensolaris
[02:38:37] *** Shiv__ has joined #opensolaris
[02:39:12] <vijay> i am trying to install Azureus Jar file on opensolaris, but i get this exception
[02:39:13] <vijay> # java -jar Azureus3.0.3.0.jar
[02:39:14] <vijay> Exception in thread "main" java.lang.NoClassDefFoundError: org/apache/commons/cli/CommandLine
[02:39:23] <vijay> anyone could help ?
[02:41:47] <Chipdancer> try #java or #azureus
[02:42:41] <vijay> ok
[02:43:12] <Chipdancer> or you could just use google - the first hit looks promising
[02:43:46] *** Megaf|Sono|Dormi has left #OpenSolaris
[02:45:17] *** Megaf|Sono|Dormi has joined #OpenSolaris
[02:46:15] *** comay has quit IRC
[02:49:37] *** pizdec_ has quit IRC
[02:51:31] *** pizdec__ has quit IRC
[02:53:14] *** Jagoma has joined #opensolaris
[02:53:51] *** phips|mb has quit IRC
[02:57:17] *** johnlev has quit IRC
[03:00:15] <CIA-25> semery: 6231080 Solaris Kerberos needs a cmdline utility ... (fix export source build)
[03:03:37] *** theRealballchalk has quit IRC
[03:04:51] *** Megaf|Sono|Dormi has left #OpenSolaris
[03:05:28] *** hile_ has quit IRC
[03:10:55] *** bengtf has quit IRC
[03:11:06] *** theRealballchalk has joined #opensolaris
[03:11:15] *** sommerfeld has quit IRC
[03:13:07] *** willnorris has quit IRC
[03:14:48] <theRealballchalk> guys what's the unix command to see the MAC addresses of all users on the network?
[03:15:05] <Stric> first ping (or something) everyone, then arp -a
[03:15:06] <hallu_> arp -a
[03:15:23] <Stric> and that's machines, not users.. normally ;)
[03:17:46] <theRealballchalk> ahhhhh ok
[03:17:49] <theRealballchalk> hmm
[03:18:04] <theRealballchalk> ok so how is it that after pinging...it shows up in arp -a?
[03:18:16] *** dlynes_laptop has quit IRC
[03:18:36] <theRealballchalk> thanks
[03:18:38] <theRealballchalk> brb
[03:19:27] *** anders__ has joined #opensolaris
[03:22:48] *** bondolo has quit IRC
[03:24:31] *** anders__ has quit IRC
[03:24:42] *** anders__ has joined #opensolaris
[03:24:50] *** Shiv__ has quit IRC
[03:25:10] *** Shiv__ has joined #opensolaris
[03:32:46] <Stric> theRealballchalk: because it's not needed unless you are trying to contact the machines
[03:33:08] *** bengtf has joined #opensolaris
[03:34:20] <Doc> where can i download Ubuntu Solaris from?
[03:34:50] <Shiv__> You mean Nexenta(GNUSolaris)?
[03:39:49] <wesolows> There's no such product on the market that I know of; if there is, then someone is infringing on one or more companies' trademarks.
[03:39:50] *** theRealballchalk has quit IRC
[03:41:24] <g4lt-sb100> annd more importantly is subject to many lynch mob activities
[03:42:15] *** Mdx4 has quit IRC
[03:43:41] *** Marv|LG has quit IRC
[03:44:17] *** Chihan has left #OpenSolaris
[03:45:31] *** jwk404 has joined #opensolaris
[03:47:02] <jbk> are the gcc shadow builds only done with x86, or are the also done when building ON for sparc as well?
[03:47:08] <wesolows> both
[03:53:47] *** sommerfeld has joined #opensolaris
[03:53:47] *** ChanServ sets mode: +o sommerfeld
[03:54:51] *** duri has quit IRC
[03:55:38] *** sahafeez has joined #opensolaris
[03:55:57] *** victori has quit IRC
[03:56:38] *** victori has joined #opensolaris
[04:03:30] *** duri has joined #opensolaris
[04:06:50] *** perlmonk has joined #opensolaris
[04:17:37] *** Tpenta has quit IRC
[04:20:02] *** Tpenta has joined #opensolaris
[04:24:12] *** ChanServ sets mode: +o Tpenta
[04:24:31] *** Triskelios has joined #opensolaris
[04:26:38] *** sahafeez has quit IRC
[04:31:07] *** Atomic-Punk has joined #opensolaris
[04:39:19] *** Megaf|Sono|Dormi has joined #OpenSolaris
[04:40:13] *** kaiwai has quit IRC
[04:40:52] *** AtomicPunk has quit IRC
[04:51:32] *** Megaf|Sono|Dormi is now known as Megaf
[04:53:42] *** Megaf has left #OpenSolaris
[04:58:15] *** sioraiocht has joined #opensolaris
[04:59:07] *** laca has quit IRC
[05:03:00] *** jmcp has quit IRC
[05:07:16] *** Atomic-Punk has quit IRC
[05:07:44] *** Atomic-Punk has joined #opensolaris
[05:35:04] *** jmcp has joined #opensolaris
[05:35:42] <moazamraja> re
[05:42:03] <e^ipi_> ra
[05:42:53] <moazamraja> heh
[05:45:53] *** e^ipi_ is now known as e^ipi
[05:50:32] *** perlmonk has quit IRC
[06:01:51] <jbk> hmm.. looks like make/dmake won't include multiple files if you do 'include $(VAR)' and $VAR is a list
[06:08:32] *** eboutilier has quit IRC
[06:14:13] *** ShadowHntr has joined #opensolaris
[06:16:48] *** ShadowHntr has quit IRC
[06:30:47] *** eboutilier has joined #opensolaris
[06:30:56] *** dclarke has joined #opensolaris
[06:34:26] *** xorg has joined #opensolaris
[06:34:31] <xorg> hi all
[06:34:37] *** sommerfeld has quit IRC
[06:37:28] *** xorg has left #opensolaris
[06:41:45] *** Sup3rkiddo has joined #opensolaris
[06:45:33] *** SteVe- has joined #opensolaris
[06:45:51] *** vijay has left #opensolaris
[06:47:20] *** sparcdr has joined #opensolaris
[06:47:42] *** SteVe- has quit IRC
[06:47:57] <sparcdr> :o
[06:48:05] *** libkeise1 has joined #opensolaris
[06:48:54] *** dclarke has quit IRC
[06:51:49] *** sparcdr has quit IRC
[06:59:32] <Tempt> boyd: ping
[07:00:26] <boyd> pongm
[07:00:33] <boyd> hmmm parity errors
[07:00:58] <Tempt> We need to catch up and sort out that zfs preso for the next MSOSUG at some point.
[07:02:40] *** libkeiser has quit IRC
[07:03:27] <boyd> Yes, we do.
[07:03:39] <Tempt> The question is .... when?
[07:04:03] <boyd> I realise. I'm thinking
[07:04:25] <boyd> It's about the 17th isn't it?
[07:04:40] <Tempt> err
[07:04:44] <Tempt> something like that.
[07:05:39] <boyd> Well, I'm in melb from now till then... although the actual week of the 17th is not yet confirmed
[07:05:56] <Tempt> ha.
[07:06:25] <Tempt> We could let it slide until the next meeting
[07:06:52] <boyd> Pretty much any weeknight evening could work... I'd need to check of any other events in the family for a given evening.
[07:07:11] <boyd> What about a chat on Thursday arvo/evening to start
[07:07:40] <Tempt> Hmm, thursday-sun are out
[07:07:55] <Tempt> It'll be my time off work, and I doubt I'll be in the mood to really think about Solaris.
[07:08:04] <Tempt> (or do anything other than relax and booze)
[07:08:25] <Tempt> tuesday?
[07:09:12] <boyd> I'm only doing 3 days this week so I'm not in until wed
[07:09:22] <Tempt> aaah
[07:09:58] <boyd> We might be able to do wed, but I'd need to run it past sandy
[07:10:13] <Tempt> Alright, probably sort it closer to the day then.
[07:10:20] <boyd> ok
[07:10:31] <boyd> See my comment about SGD in the QC lounge?
[07:10:33] <Tempt> If all else fails, we could discuss it using some sort of electronic communication medium.
[07:10:38] <boyd> People were looking at me funny :)
[07:10:39] <Tempt> Oh, yes, I caught that one this morning.
[07:10:52] <Tempt> Heh, the crowds not used to fluxbox? :-)
[07:11:02] <boyd> Fluxbox and emacs :)
[07:11:09] <Tempt> haha.
[07:11:12] <Tempt> OH NOES
[07:11:20] <Tempt> emacs! on my hosts! THE HORROR!
[07:11:25] <boyd> Not really the stuff of the minions leaving canberra at the end of a parliamentary sitting
[07:11:58] <Tempt> What were you teaching in Canberra, anyway?
[07:12:01] * boyd wonders if he partly uses emacs because of the type of reaction it gets
[07:12:09] <boyd> Cluster 3.2
[07:12:24] <Tempt> They have clusters in Canberra? :)
[07:12:38] <boyd> Many govt depts.
[07:12:50] <boyd> 2 big ones in particular
[07:13:09] <Tempt> At least one of which shall remain nameless?
[07:13:10] <boyd> But as it turns out.. half the class were from out of town
[07:13:11] <Tempt> :)
[07:13:20] <boyd> Yes, one of them will
[07:13:30] <Tempt> geez, travelling to Canberra to do a training course that could be done syd/mel
[07:13:39] <Tempt> Making sure people don't enjoy their travel?
[07:14:16] <boyd> Well, the funny thing is that people can easily get training funds but can't get travel funds to do the training :O
[07:14:30] <Tempt> That was a common story at Telstra, too.
[07:14:47] <Tempt> I think that might be because purchases come with 'training credits'.
[07:16:04] <boyd> That's part of why Sun now offer the "Fly and learn" thing that includes training, flights accom and 2 meals a day for those in bne, adl and per
[07:17:55] <Tempt> aaah
[07:18:03] <Tempt> Sun needs to knock the meal vouchers up a notch.
[07:18:53] <Tempt> Apparently while the Mitre now accepts the vouchers, you can't actually buy anything more than a bowl of chips with it
[07:18:59] <Shiv__> Hmm...Travelling all the way to canberra...what happens to the TCO of the customer? :)
[07:20:09] <boyd> Tempt: That's a little harsh... they're worth 12 bucks. There are mains at the Mitre for 14. There's also places where you can get a sizable amount of food for 12. , s it's all about choice I suppose
[07:21:08] <Tempt> The sydney deal worked much better with the food court over the road.
[07:22:09] <boyd> Yeah... it all kinda depends where you are... there are only 4 venue in canberra
[07:22:30] <boyd> ... and 3 of them are crappy
[07:23:06] <Tempt> Ouch.
[07:23:29] <boyd> Hey, those blade 2500s keep listing for about 2200
[07:24:08] <boyd> Holy moly... did you see this: http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=160160169249
[07:24:21] <boyd> That price seems out of step with the price for a whole V880
[07:24:45] <Tempt> haha
[07:24:46] <Tempt> Yes.
[07:24:53] <Tempt> vendor is a scammer too.
[07:24:58] <boyd> Heh
[07:25:07] <boyd> Well, family life calls
[07:25:08] <Tempt> Used to list them for low starting bid and canx the auction time and time again
[07:25:10] <boyd> Better go
[07:25:15] <Tempt> have fun.
[07:47:10] *** m0le has quit IRC
[07:47:37] <axisys> i think i am seeing an acctadm telnet bug on my machine right now
[07:47:58] <axisys> adm running acctadm which is telneting to thousands of IPs all over the world
[07:48:12] <axisys> anyway to confirm it is a bug ?
[07:48:17] <axisys> or exploit
[07:48:21] <axisys> or wordm !
[07:48:24] <axisys> worm!!
[07:53:09] <Tempt> ummn
[07:53:12] <Tempt> that sounds very bad
[07:53:36] <Tempt> capture as much state as possible and then you'll have to quarantine the host.
[07:54:14] *** libkeise1 has quit IRC
[07:54:56] <e^ipi> nuke it from orbit
[07:55:01] <e^ipi> (it's the only way to be sure)
[08:04:22] <axisys> i am collecting the ustack() of that process
[08:04:44] <axisys> the command acctadm has no argument
[08:05:07] <axisys> thanx to lofi and external fw to connect the dots
[08:05:16] <axisys> s/lofi/losf/
[08:07:38] <axisys> hopefully this will catch more stuff
[08:07:40] <axisys> dtrace -n 'syscall:::entry /execname == "acctadm"/ { @[ustack()] = count(); }'
[08:14:08] *** e^ipi_ has joined #opensolaris
[08:14:09] *** e^ipi has quit IRC
[08:18:07] *** Sup3rkiddo has quit IRC
[08:18:31] *** Sup3rkiddo has joined #opensolaris
[08:20:35] *** karrotx has joined #opensolaris
[08:22:36] <Shiv__> Which is a good security scanner for solaris (apart from lsof,nmap).
[08:22:55] <Shiv__> I had used nessus but since 2 years nessus3 has been made close source.
[08:23:25] <Shiv__> Wanted to build a good one and put it as part of SFE
[08:28:32] *** charlesnw has joined #opensolaris
[08:28:33] <axisys> how do I get the expected checksum of /usr/sbin/acctadm?
[08:28:49] <Tempt> ask someone with the same solaris release?
[08:29:18] <axisys> pkgchk -l -P /usr/sbin/acctadm does not show it.. I am using 6/06
[08:31:29] <Shiv__> acctadm is part of SUNWcsu
[08:31:31] <Tempt> cksum /usr/sbin/acctadm
[08:31:31] <Tempt> 2491655128      28728   /usr/sbin/acctadm
[08:48:17] *** charlesnw has quit IRC
[08:48:37] *** xushi_ has joined #opensolaris
[08:51:57] *** FBdev has joined #opensolaris
[08:54:23] *** smellyfish has joined #opensolaris
[08:54:32] *** xushi_ has quit IRC
[08:54:41] *** smellyfish has quit IRC
[09:03:24] *** estibi_ has joined #opensolaris
[09:05:09] <Shiv__> Anyone installed builds 70-72?
[09:05:23] <Shiv__> I tried b70, b70a, b70b & b72
[09:05:30] <Shiv__> Faced problem with all of them.
[09:05:40] <Shiv__> Am still with B64a
[09:07:06] *** reflect has joined #opensolaris
[09:07:55] * Shiv__ : Everyone slumbering... :)
[09:11:28] <g4lt-sb100> no, everyonee trying to figure out what the fuck you're on about this time
[09:12:38] *** reflect_ has quit IRC
[09:13:08] <Shiv__> Touchy !?
[09:23:13] *** estibi has quit IRC
[09:23:53] *** nrubsig has quit IRC
[09:26:11] <g4lt-sb100> you gave the channel two minutes to respond.  yet youy call ME touchy?
[09:26:53] <Tempt> touchy.
[09:27:16] <Tempt> Obviously this channel is in need of the great aide to communication invented.
[09:27:19] <Tempt> Beer.
[09:30:03] *** theRealballchalk has joined #opensolaris
[09:30:54] <Shiv__> :)
[09:32:56] * Shiv__ Gets a coffee
[09:35:43] *** theRealballchalk has quit IRC
[09:39:39] <Shiv__> Seems like it will be only B74 that can be installed (http://www.opensolaris.org/jive/thread.jspa?messageID=156632&tstart=0)
[09:40:23] <Shiv__> B70 to B73 each of them have had some issues w.r.t installation
[09:41:48] *** blah2 has quit IRC
[09:46:52] *** echelog has joined #opensolaris
[09:49:18] *** _hsilva has quit IRC
[09:49:18] *** _hsilva has joined #opensolaris
[09:53:55] <palowoda> Shiv__: I ran into the same problem.  The workaround is to put the optical dvd/cd rom on channel 2 and the master boot drive on channel 1.
[09:54:10] <palowoda> Or wait till build 74
[09:54:45] *** Atomic-Punk has quit IRC
[09:56:29] <Shiv__> palowoda: Can you please elaborate on the channel 1/2 part.
[09:56:45] <Shiv__> palowoda: I wonder how it would be in mycase since I am using the iso image with VMplayer.
[09:57:02] *** Atomic-Punk has joined #opensolaris
[09:58:49] <palowoda> Not that anyone considers vmware a valid issue.  but look at http://www.opensolaris.org/jive/thread.jspa?messageID=156632&tstart=0  where jkiel is talking about the problem.  I just ran into the workaround by accident because I had a system that install properly with those releases and a second identical one that did not with that specific error message.
[09:59:01] <palowoda> Once the drives where rearranged it installed.
[10:00:08] *** dmarker has quit IRC
[10:00:13] <palowoda> If you want to complicate debugging the problem by all means use vmware.
[10:00:26] *** dmarker has joined #opensolaris
[10:02:09] *** theRealballchalk has joined #opensolaris
[10:04:08] *** Wicky2 has joined #opensolaris
[10:04:51] <Shiv__> palowoda: Hmm...as per  the thread seems like fix hasnt gone into B74 either. On the system where it is failing, I do not have an option but to use VMware.
[10:05:39] <Shiv__> palowoda: I will wait or I'll try a bfu.
[10:06:17] <palowoda> Well It's marked to be integrated into build 74.  That doesn't mean it'd going to make it.  It affects Sun's U40 machines so I'll bet it's a high priority.
[10:07:07] <Shiv__> palowoda: As per the thread B74 gate is already closed and the fix isnt in.
[10:07:33] <palowoda> I noticed that.  But I at least found a workaround for my problem. :)
[10:09:05] <palowoda> Hardware is a dime a dozen these days.  I don't really care.
[10:09:08] *** theRealballchalk has quit IRC
[10:09:21] *** johnlev has joined #opensolaris
[10:11:41] *** NikolaVeber has joined #opensolaris
[10:11:57] <Shiv__> palowoda: Not for everyone :)
[10:12:47] <palowoda> Shive__: I thought everyone was suppose to be like me.  My dreams a shattered.  Life is not the same. :)
[10:14:20] <palowoda> Ok I'll let you wait till build 75.
[10:14:40] *** Fish- has joined #opensolaris
[10:14:49] <Shiv__> palowoda: He he...
[10:15:12] *** Fish- is now known as Fish
[10:15:16] <Shiv__> palowoda: Hey...your input helped, I changed CD/DVD drive settings in VMware and it worked...hats off....
[10:15:35] <palowoda> Hey it drove me nuts for two weeks.
[10:15:53] <Fish> hello
[10:16:55] *** theRealballchalk has joined #opensolaris
[10:25:54] *** theRealballchalk has quit IRC
[10:26:02] *** Sup3rkiddo has quit IRC
[10:29:47] *** Sup3rkiddo has joined #opensolaris
[10:39:50] *** FBdev has quit IRC
[10:49:16] <Wicky2> b00
[10:49:37] *** Wicky2 is now known as WickedWicky
[10:57:07] *** pizdec_ has joined #opensolaris
[11:01:39] *** Vorbis_w57 has quit IRC
[11:07:14] *** Mdx4 has joined #opensolaris
[11:08:08] *** bengtf_ has quit IRC
[11:10:34] *** libkeiser has joined #opensolaris
[11:11:18] *** pizdec__ has joined #opensolaris
[11:11:56] *** Mdx4 has quit IRC
[11:19:56] *** Pietro_S has joined #opensolaris
[11:27:07] *** pizdec_ has quit IRC
[11:27:08] *** Shiv__ has quit IRC
[11:27:21] *** Shiv__ has joined #opensolaris
[11:28:08] *** blindfish has joined #opensolaris
[11:44:09] *** estibi_ is now known as estibi
[11:58:48] *** nostoi has joined #opensolaris
[12:02:24] *** blindfish has quit IRC
[12:11:08] *** FBdev has joined #opensolaris
[12:24:46] *** Trisk[laptop] has joined #opensolaris
[12:25:41] *** Vorbis_w57 has joined #opensolaris
[12:36:36] *** srirama has joined #opensolaris
[12:38:25] *** Triskelios has quit IRC
[12:39:18] *** srirama has quit IRC
[13:01:40] *** crash__ has quit IRC
[13:02:58] <Doc> [21:02:28][@milliways ~]$ /usr/sfw/sbin/smbd
[13:02:59] <Doc> Abort (core dumped)
[13:03:00] <Doc> hrmm...
[13:03:31] <quasi> nice
[13:04:34] <Doc> only when run as non-root, and it's not suid so it's probably not a security issue
[13:04:37] *** mikefut has joined #opensolaris
[13:04:51] <quasi> conspiracy theorists would probably link that to suns recent dealings with M$ ;)
[13:05:29] <snuff-home> heh
[13:06:46] <Doc> except the latest samba patch is about a year old...
[13:09:18] *** deather_ has joined #opensolaris
[13:10:56] *** Snake007uk has joined #opensolaris
[13:12:25] *** rvalles has quit IRC
[13:16:29] *** kaiwai has joined #opensolaris
[13:26:35] *** deather has quit IRC
[13:28:16] *** pizdec_ has joined #opensolaris
[13:30:52] *** dpn` has joined #opensolaris
[13:31:36] *** nostoi has quit IRC
[13:37:57] *** sparcdr has joined #opensolaris
[13:38:10] <sparcdr> hey :)
[13:38:52] <JWheeler> hi sparcdr
[13:39:03] <sparcdr> hey there been gone like 14 hours :o
[13:39:14] <sparcdr> what's new in the last 24 hours?
[13:39:39] <JWheeler> I've just been lurking sorry, I haven't actually been reading
[13:40:44] *** blindfish has joined #opensolaris
[13:41:35] <sparcdr> ah
[13:41:41] <sparcdr> I guess I could say I am as well >_>
[13:41:48] <sparcdr> well nice to see everyone anyways
[13:41:55] *** mikefut has quit IRC
[13:43:05] <sparcdr> (alanc, JWheeler, cmihai, wesolows, and Tpenta) that is (+others)
[13:43:14] <JWheeler> :)
[13:43:16] <sparcdr> *yawn*
[13:43:26] <JWheeler> I'm fighting compiling issues with mplayer
[13:43:31] <sparcdr> bit early today oh well
[13:43:32] <JWheeler> do you know much about linking?
[13:43:35] <sparcdr> Sun Studio?
[13:43:40] <JWheeler> no, gcc sorry
[13:43:45] <sparcdr> hmm
[13:43:55] <sparcdr> i was checking (GNU toolchain makes pain)
[13:44:08] <sparcdr> what's bricking JWheeler ?
[13:44:16] <sparcdr> pastebin the errors for me
[13:44:19] * JWheeler will run off with his question anyone... since he thinks he understands the concept;
[13:44:20] <sparcdr> might be able to troubleshoot
[13:44:25] <JWheeler> ok
[13:44:52] *** pizdec__ has quit IRC
[13:45:53] <JWheeler> So, I need to build mplayer with a v8 binary, not v8plus. No matter what -mcpu=v8 or -mv8 (the gcc for make it a v8 binary), I'm getting v8+ binaries. So I did a find ./ -exec file {} \; | grep 32-bit | grep V8, and I see that somewhere in there, there is in fact another v8+.o file
[13:45:57] <sparcdr> mplayer in blastwave uses gcc afaik, take a look at their patches if I can't find what's wrong.  I'm hoping it might be a missing library or a conflicting env variable, but probably not if you're talking about linking, it's probably a kludge in the code with pointers (Usually is :P)
[13:46:55] <sparcdr> never had good luck using the latest official sources, it takes a few weeks for it to build right on solaris from experience, sorta like the kde situation
[13:46:58] <JWheeler> I'm guessing that when gcc tries to assemble the binary, there is some kind of lowest common denominator logic going on in there. When it tries to link all the .o's together, it finds that 1 v8+ .o file, and hence has to make a v8+ final binary, despite my pleading for plain v8. Would that follow?
[13:47:09] <sparcdr> oh
[13:47:18] <sparcdr> yeah
[13:47:19] *** crash__ has joined #opensolaris
[13:47:22] <JWheeler> sparcdr: I'm the blastwave mplayer maintainer :)
[13:47:28] <sparcdr> hahaha
[13:47:29] <sparcdr> :P
[13:47:38] <JWheeler> or at least, I will be, when I get this latest build out. The old guy is AWOL
[13:47:47] <sparcdr> mixing bit modes, linking against lowest ABI will break things
[13:48:18] <sparcdr> ive had my share of software ignore CFLAGS etc
[13:48:18] <sparcdr> :/
[13:48:41] <JWheeler> Well I just wanted to check my logic first, which you seem to have validated. So the question becomes, why does 99% of mplayer build with v8 as requested, but only one .o pops out as v8+. That's the heart of my problem - right?
[13:49:02] <flyingparchment> jwh: embedded asm?
[13:49:16] <JWheeler> it seems to be 'vis' stuff
[13:49:20] <flyingparchment> as will set the v8+ flag if it sees a suitable instruction
[13:49:27] <JWheeler> I've not used sparc before, so I don't really know what I'm seeing in here
[13:49:32] * kaiwai does a cheer
[13:49:56] <flyingparchment> gcc -mcpu doesn't affect inline asm of course, because it's not gcc generating the asm :)
[13:50:08] <JWheeler> dsputil_vis.o is the file in question
[13:50:17] <JWheeler> # sun sparc
[13:50:18] <JWheeler> OBJS-$(TARGET_ARCH_SPARC)              += sparc/dsputil_vis.o \
[13:50:18] <JWheeler> #sparc/dsputil_vis.o: CFLAGS += -mcpu=ultrasparc -mtune=ultrasparc
[13:50:31] <flyingparchment> you might want to talk to delewis, he's the mplayer sparc maintainer..
[13:50:39] <JWheeler> (I commented out the last line), which in turn gave me /usr/ccs/bin/as: "<stdin>", line 3224: error: cannot use v8plus registers in a non-v8plus target binary
[13:50:40] <JWheeler> /usr/ccs/bin/as: "<stdin>", line 3226: error: cannot use v8plus registers in a non-v8plus target binary
[13:50:51] <JWheeler> as is the assembly stuff, right?
[13:50:55] <flyingparchment> yes
[13:50:59] <kaiwai> you'll probably need to re-write stuff
[13:51:05] <flyingparchment> in the meantime you can probably just remove the sparc stuff and replace it with the C version
[13:51:06] <JWheeler> yeah, I need to wave down delewis, he's been really helpful to me
[13:51:48] <JWheeler> sorry, can you expand on that last sentenc?
[13:51:53] <JWheeler> +e
[13:52:40] <sparcdr> bah sparc has low enough clockrate :D
[13:53:08] <flyingparchment> JWheeler: this is ffmpeg, right?  all the asm functions also exist in a C version, so it can work on other platforms.  e.g. there's C code, then hand-optimised x86, sparc, ... versions of the same code.
[13:53:20] <flyingparchment> JWheeler: you need to make it compile the C code instead of using the sparc stuff
[13:53:51] <sparcdr> JWheeler, i cant help you because I work with x86_64 mainly
[13:54:18] <sparcdr> maybe try gas to link it?
[13:54:22] <JWheeler> Let me see if I have this all correct in my head :) : There is vis code in here. Vis is the cpu extensions present in v8+ (perhaps like mmx - in x86-speak, which I do understand). Accordingly, vis code needs -mcpu=ultrasparc to work, which is why I'm getting this _one_ v8+ .o file, which in turn is upgrading my final binary to v8+ too?
[13:54:50] <flyingparchment> JWheeler: yes
[13:54:55] <sparcdr> hand link it with offening portions stripped
[13:55:09] <JWheeler> it's libavcodec, which I believe is ffmpeg, yes
[13:55:12] <sparcdr> not literally stripped of course :D
[13:55:14] <flyingparchment> don't hand link it, modify the compile to not use the hand-written SPARC code
[13:55:20] <sparcdr> oh
[13:55:23] <flyingparchment> i've done this before on x86, it's not particularly hard
[13:55:28] <sparcdr> well i wish i could help (No sparcs on site)
[13:55:53] <sparcdr> wouldnt you want to use gnu assembler for gcc programs?  they can link sparc.
[13:56:00] <JWheeler> I appreciate the support at least though sparcdr :)
[13:56:08] <sparcdr> i really wish i could help
[13:56:15] <JWheeler> thanks flyingparchment, lemme see if I can butcher the makefile to suit
[13:56:18] <sparcdr> maybe with x86/x86_64 :) i can
[13:56:27] <sparcdr> ie: blastwave stuff
[13:56:38] <sparcdr> my ultra-20 comes in a few days, ships tomorrow
[13:56:43] <flyingparchment> JWheeler: this solution sucks though, because obviously you won't get the VIS stuff on v8+ systems (which is like, everyone :)
[13:56:50] <JWheeler> sparcdr, I'm not sure on the reasons/history, but the BW gcc using the SUNW as, rather then gas *shrug*
[13:57:00] <sparcdr> flyingparchment, cringe @ no simd-equivalent instructions on sparc :/
[13:57:17] <JWheeler> flyingparchment, no that's fine! It's a BW standard that we build for v8, not v8+
[13:57:49] <sparcdr> well you can always try it, not like gcc is that great anyways, might as well use all slow tools and compilers and provide a slow build of mplayer so we can drive our blood pressure up :P
[13:57:49] <JWheeler> That's why I'm in here trying to get it to run on v8, otherwise I'd build the whole lot with v8+ (and v9) - isaexec style
[13:57:56] <JWheeler> isaexec is a whole different topic though ;)
[13:58:23] *** rvalles has joined #opensolaris
[13:58:25] <JWheeler> yes well.... that's bw as it stands today
[13:58:33] <JWheeler> rest assured though, changes are happening
[13:58:39] <flyingparchment> JWheeler: it's not really fine, it means mplayer will be slower on v8+ systems than it should be
[13:58:52] <flyingparchment> VIS is exactly meant for apps like mplayer :)
[13:59:04] <JWheeler> The mailing list have been explosive over the past few weeks. It really sounds like the migration to solaris 10 as the base standard will be happening soon
[13:59:19] <sparcdr> v8 systems would be lowest common denominator, very slow machines in comparison to say v9 style (UltraII+)
[13:59:28] <sparcdr> so it's not the final solution :)
[13:59:39] <JWheeler> once that happens, dennis is cooking up a very new gcc 4.2, and I imagine the minimum cpu requirements will also be jumping up, v9 maybe?
[14:00:13] <sparcdr> JWheeler, not in the sparc loop much anymore (Aside from remote boxes) so I'll just say yes, because it'd be sense
[14:00:13] <JWheeler> there can't be many v8 machines still out there, running solaris 8, but because there are some, that is the base :/
[14:00:24] <JWheeler> heh :)
[14:00:25] <sparcdr> they dont run mplayer well even with vis
[14:00:34] *** Snake007uk has quit IRC
[14:00:45] <JWheeler> so VIS is like MMX is it? flyingparchment?
[14:00:52] <JWheeler> or perhaps SSE
[14:00:55] <sparcdr> my friend who does sparc work says sse
[14:01:06] <JWheeler> ah
[14:01:34] <JWheeler> all these sun4d/m/u/whoknows is doing my head in
[14:01:45] <sparcdr> not as many simd instructions on a sparc as on x86
[14:02:09] <JWheeler> I'm sure for those in the business, keeping up in realtime over the course of years is easy, but for someone like me that hadn't even seen sparc until 2007; I'm struggling!
[14:02:14] <sparcdr> id assume that vis handles more operations than sse + mmx together, and afaik sse3 replaces mmx completely
[14:02:27] <sparcdr> i first saw a sparc in 2004
[14:03:24] *** bunker has joined #opensolaris
[14:04:00] <sparcdr> brb
[14:04:34] <Stric> sparcdr: mplayer doesn't necessarily mean video either.. it can play various music formats too
[14:04:47] <sparcdr> I am aware, I use it for ogg
[14:04:59] <sparcdr> regardless, most sparc systems ive used have terrible audio
[14:05:00] <sparcdr> :)
[14:05:56] <JWheeler> I can find 2 dsputil.h files. One for sparc - which includes vis code, and another for x86 - which includes mmx code. Hum, me thinks neither is going to cut it here!
[14:06:15] <JWheeler> I'm starting to get the sinking feeling that mplayer isn't going to work on plain v8
[14:06:27] <JWheeler> mind you, what does it do on PPC... hmm
[14:07:08] <sparcdr> it works :)
[14:07:52] <JWheeler> well, I can just comment this dsp stuff out altogether, but I'm thinking that it's going to want that .o file somewhere!
[14:08:06] <sparcdr> just not pre-power4 systems unless mplayer was hacked up by several developers and are older than dirt of course.  you have to draw a line.
[14:08:58] <flyingparchment> JWheeler: there is definitely a C version somewhere.  look harder :)
[14:09:14] * JWheeler squints
[14:09:36] * sparcdr glares
[14:10:56] <sparcdr> meh I can't find it myself
[14:12:06] <JWheeler> I assume the one in the root of libavcodec is the one that you're refferring to, it's just the header of the file that's making me nervous:
[14:12:22] <JWheeler>  * @file dsputil.h
[14:12:22] <JWheeler>  * DSP utils.
[14:12:22] <JWheeler>  * note, many functions in here may use MMX which trashes the FPU state, it is
[14:12:22] <JWheeler>  * absolutely necessary to call emms_c() between dsp & float/double code
[14:12:27] *** perlmonk has joined #opensolaris
[14:12:33] <JWheeler> having said that... it compiled
[14:12:38] <sparcdr> cool
[14:12:43] *** xushi_ has joined #opensolaris
[14:13:41] <JWheeler> right, from the top then
[14:13:44] * JWheeler gmake cleans
[14:13:46] *** FBdev has quit IRC
[14:14:36] <JWheeler> I think the biggest issue, is that I'm trying to build this on a v9 host, so I'm manually hacking the configure script to 'misdiagnose' this into a v8, but I guess it's still too clever elsewhere
[14:15:03] *** FBdev has joined #opensolaris
[14:15:15] *** master_o1_master has joined #opensolaris
[14:18:51] *** kFuQ has joined #opensolaris
[14:19:29] *** kaiwai has quit IRC
[14:20:00] *** dunc has quit IRC
[14:23:35] <JWheeler> hmm, maybe it just assumes that if you have sparc, then you can do vis. I don't see anything in the configure script to check for it, only that you're sparc. Well, patch time I guess then!
[14:24:02] <sparcdr> don't break too much :)
[14:24:25] <JWheeler> It's just used within the BW gar repository, not upstream - it should be OK :)
[14:25:05] <flyingparchment> is there even a sparcv8 fast enough to run mplayer? :D
[14:27:23] *** master_of_master has quit IRC
[14:27:24] *** sparcdr has quit IRC
[14:28:43] *** blindfish has quit IRC
[14:29:00] <JWheeler> I don't know my history very well. When we're talking sparcv8, what are we talking about? 400Mhz?
[14:30:07] *** dunc has joined #opensolaris
[14:32:59] <flyingparchment> i don't think there are any v8 sparcs faster than 200MHz
[14:33:41] <Stric> JWheeler: divide by 4 and you have the fast ones
[14:34:15] *** mikefut has joined #opensolaris
[14:34:18] <flyingparchment> sparc went 64-bit (v8+/v9) pretty early
[14:34:49] <Doc> is columbus day a public holiday in the US?
[14:34:53] *** blindfish has joined #opensolaris
[14:35:16] *** ShadowHntr has joined #opensolaris
[14:35:19] <JWheeler> oh, so was v8 only 32-bit?
[14:35:31] <flyingparchment> yes
[14:40:13] <JWheeler> oh right, a bit change then!
[14:40:28] <JWheeler> they're 32-bit compatable though right? like x86-64?
[14:40:40] <flyingparchment> v9 in 32-bit mode = v8+
[14:41:12] * JWheeler scratches his head
[14:41:29] *** dunc_ has joined #opensolaris
[14:41:36] <flyingparchment> its v8 but it has some more instructions that real v8 cpus didn't have
[14:42:08] <JWheeler> so code that was built for v8s, would be 32-bit. for v8+ it would have always been 64-bit?
[14:42:38] <flyingparchment> no, v8+ code is 32-bit, but it'll only run a v9 cpu
[14:43:23] <JWheeler> so there wasn't a v8+ cpu then?
[14:43:33] <JWheeler> specifically, I mean
[14:43:41] *** jafari has joined #opensolaris
[14:43:51] <flyingparchment> right
[14:44:08] <JWheeler> ah, with you now
[14:44:19] <JWheeler> Hence why the jump from 8 to 8+ is so big
[14:50:22] *** dunc has quit IRC
[14:52:52] *** sparcdr has joined #opensolaris
[14:53:14] *** dunc_ has quit IRC
[14:53:16] <sparcdr> yeah so how's it coming JWheeler ?
[14:54:21] <JWheeler> well, I'm trying my hand at C code (stand back everyone!)
[14:55:45] <JWheeler> #elif defined(ARCH_SPARC) && defined(HAVE_VIS)
[14:55:58] <JWheeler> (I added the defined(HAVE_VIS) bit)
[14:56:13] <Symmetria> urgh anyone running the blastwave version of apache2 and having problems with LFS?
[14:56:24] <flyingparchment> LFS?
[14:56:25] <sparcdr> LFS?
[14:56:32] <sickness> linux from scratch?!? :P
[14:56:37] <JWheeler> basically, the logic was flawed again, it keeps on trying to pull in the vis specific dsputil, rather then the C version that flyingparchment directed me to
[14:56:39] <sparcdr> no sickness lol
[14:56:51] <sparcdr> Lufs?
[14:56:56] <Symmetria> large file support
[14:57:00] <sparcdr> ah
[14:57:04] <sickness> oh...
[14:57:05] <sparcdr> what version of apache?
[14:57:08] <Symmetria> 2.2.4
[14:57:20] <sparcdr> solaris version and arch?
[14:57:29] <cmihai> ... and I thought it was Log-Structured File System
[14:57:33] <Symmetria> SunOS mirror 5.11 snv_62 i86pc i386 i86pc
[14:57:34] <JWheeler> is that valid code? ^^
[14:57:41] <sparcdr> they call it "largefile" btw
[14:57:44] <Symmetria> (79)Value too large for defined data type: access to /ubuntu-cdimage/releases/feisty/release/ubuntu-7.04-dvd-i386.iso failed
[14:57:50] <sparcdr> yes JWheeler
[14:57:51] <flyingparchment> Symmetria: you run nevada on mirror?
[14:57:54] <Symmetria> heh, well the guys in #apache call it lfs *shrug*
[14:58:00] <sparcdr> eh what
[14:58:03] <Symmetria> flyingparchment *nod* because its got ip exclusive zones
[14:58:06] <sparcdr> ubuntu != solaris
[14:58:07] <flyingparchment> ah
[14:58:08] <sparcdr> so what gives?
[14:58:11] <Symmetria> sparcdr errr its a mirror
[14:58:14] <Symmetria> its what its serving
[14:58:16] <sparcdr> ok
[14:58:23] <sparcdr> gotcha
[14:58:30] <sparcdr> how big is the iso?
[14:58:33] <Symmetria> 4gig
[14:58:33] <sparcdr> I'm assuming >2GB
[14:58:38] <Symmetria> dvd iso
[14:58:45] <sparcdr> how may bytes and arch?
[14:58:51] <sparcdr> are you using regular i386 solaris?
[14:59:02] <sparcdr> btw blastwave is not 64-bit aware
[14:59:10] <sparcdr> you'd want to use coolstack or hand compile it
[14:59:20] <Symmetria> its 64bit
[14:59:23] <Symmetria> coolstack?
[14:59:24] <sparcdr> anything over 4gb is probably not gonna work
[14:59:27] <cmihai> Yes
[14:59:28] <flyingparchment> 32-bit apps can access large files
[14:59:29] <sparcdr> csw doesnt have 64-bit apache
[14:59:32] <flyingparchment> as long as they're compiled properly
[14:59:37] <sparcdr> flyingparchment, depends on a few things
[14:59:38] <cmihai> CoolStack is _only_ 64 bit actually.
[14:59:43] <sparcdr> yep cmihai
[14:59:58] <sparcdr> since sun makes x64 systems, and coolstack is specifically tuned for their systems :)
[15:00:07] <sparcdr> but it works on em64t/amd64 alike
[15:00:16] <cmihai> They should have just included the damn thing in Solaris or something.
[15:00:22] <Symmetria> ok where can I find info about coolstack
[15:00:23] <sparcdr> they did
[15:00:26] <sparcdr> he's using blastwave
[15:00:35] <sparcdr> and the included is 32-bit anyways afaik
[15:00:36] <cmihai> Symmetria, http://unixsadm.blogspot.com/2007/08/installing-coolstack-on-solaris-and.html
[15:00:51] <cmihai> Symmetria, it's got some stuff about integrating it with SMF too.
[15:01:03] <sparcdr> but yeah csw is not 64-bit aware, the original largefile implementation made it possible to address just up to 2gb, and later 4gb, no more than 4gb on 32-bit.
[15:01:13] <sparcdr> it makes life own btw, good stuff
[15:01:45] *** sbahra has quit IRC
[15:01:52] <sparcdr> but nowadays should be something like 16gb on 32-bit, I'd hope, blastwave obviously isn't aware
[15:01:55] *** dunc has joined #opensolaris
[15:02:10] <flyingparchment> you don't need largefiles for 2GB files
[15:02:20] <sparcdr> yeah I know flyingparchment
[15:02:21] <flyingparchment> largefile applications can address up to 64-bit files, same as 64-bit applications
[15:02:36] <sparcdr> ufs is 64-bit
[15:02:37] <sparcdr> so yeah
[15:02:43] <sparcdr> blame blastwave
[15:02:52] *** ysss has joined #opensolaris
[15:02:54] <sparcdr> handcompiling if you need php is not very fun either :)
[15:03:06] <sparcdr> zones literally save confusion for me
[15:03:20] <flyingparchment> incidentally, cpio on S10 is not largefile aware, which seems odd to me
[15:03:27] <sparcdr> eh
[15:03:31] <sparcdr> cpio is an archiver lol
[15:03:42] <sparcdr> pbzip2 ftw :P
[15:04:01] <sparcdr> sorry, couldnt resist running that on my former quad proc sparc
[15:04:25] *** Mdx4 has joined #opensolaris
[15:04:26] <flyingparchment> we wrote a distributed bzip2 ;-)
[15:04:31] <cmihai> Well, Apache and PHP from CoolStack are OK and all...
[15:04:37] <sparcdr> basically what pbzip2 is (Parallel bzip2)
[15:04:44] <sparcdr> cmihai, they have problems
[15:04:45] <cmihai> But having to hand write the SMF manifests can be a PITA. ANd using them in zones requires extra work.
[15:04:48] <Symmetria> heh Im installing coolstack now
[15:04:56] <JWheeler> don't blame them, file a bug, so we know about it :)
[15:04:58] <sparcdr> I figured it out with a lot of fiddling
[15:05:07] <sparcdr> i'd rather be an ass JWheeler :D
[15:05:08] <flyingparchment> sparcdr: distributed as in running on multiple machines, not just cpus
[15:05:13] <sparcdr> oh
[15:05:16] <sparcdr> that's sweet flyingparchment
[15:05:33] <flyingparchment> (we compress very large files and have a not of not-very-fast machines)
[15:05:46] <sparcdr> JWheeler, not a bug actually, none of blastwave is 64-bit aware afaik
[15:06:07] <JWheeler> a lot is
[15:06:14] <JWheeler> we do a number of 64-bit libraries
[15:06:15] <sparcdr> compiling all php dependencies against source compiled dependencies causes a crack in your tooth
[15:06:16] <sparcdr> examples?
[15:06:36] <JWheeler> most binaries won't be, unless there is a good reason. MYSQL is available in 64-bit for example
[15:07:06] <flyingparchment> well, regardless of being 64-bit binaries, things should still be largefile enabled if they support it
[15:07:10] <sparcdr> i am aware of that, im not one of those idiots who think everything should be 64-bit
[15:07:16] <flyingparchment> there's really no reason not to, nowadays it just causes problems
[15:07:41] <sparcdr> blame the porn industry >_>
[15:08:56] <sparcdr> JWheeler, I've always said that only math and server oriented things need to have 64-bit versions
[15:09:22] *** Shiv__ has quit IRC
[15:09:56] <JWheeler> I'm sure a lot of code would still be 16-bit if someone didn't decide that we needed to move on though
[15:09:59] <sparcdr> ill just go away now, not like im helping anyone
[15:10:15] <sparcdr> JWheeler, heh bad thought, no matter how frugal the program is
[15:10:19] <JWheeler> presumably there was a performance hit going from 16-bit to 32-bit, way back when?
[15:10:36] <sparcdr> yeah I bet, during the pentium days when they switched over all the dos code
[15:10:38] <flyingparchment> in some cases it was actually faster, because people had all kinds of hacks to extend the 16-bit address space
[15:10:41] <JWheeler> I say embrace 64-bit, wherever you can
[15:10:50] <JWheeler> sure it might be bigger, and slower, sometimes
[15:10:57] <sparcdr> JWheeler, unless the program doesnt address files more than 8gb
[15:11:00] <JWheeler> but it's progress people! :)
[15:11:01] <flyingparchment> like overlays on the PDP-11, where you had to optimise to make sure that code was in the same overlay as the code it called
[15:11:06] <sparcdr> haha progress, sure
[15:11:15] <JWheeler> Besides, your 64-bit CPU is probably fast enough to suck up the differences :)
[15:11:15] <Symmetria> hrm
[15:11:23] <sparcdr> need I mention WOW64
[15:11:23] <Symmetria> something wrong with sun's download system or something
[15:11:33] <sparcdr> Symmetria, happens
[15:11:41] <Symmetria> tried to download coolstack from 3 different points and its just hanging *sigh*
[15:11:51] <sparcdr> they also hardcode most their jscript for firefox specifically
[15:11:52] <JWheeler> sparcdr, well there is a good example there, look a the size of installation for that steaming pile of ....
[15:12:07] <JWheeler> supporting bits of both results in confusing library duplication
[15:12:10] <sparcdr> that's not from being 64-bit
[15:12:16] <sparcdr> even home 32-bit versions are 8gb
[15:12:29] <JWheeler> I think we can all agree that 64-bit is well and truely ready, I say run one or the other
[15:12:36] <sparcdr> both
[15:12:37] <tomww> anyone familier with IP_MULTICAST on Solaris? I'm looking for references how to initialize it correctly (pulseaudio looks like it doesn't)
[15:12:38] <sparcdr> LP64
[15:12:41] <JWheeler> transitioning across both for years and years is nasty
[15:12:51] <sparcdr> so is the world
[15:13:06] <sparcdr> you dont hear me asking people to make trash recepticles 64-bit either
[15:14:46] <JWheeler> each to their own I guess
[15:14:58] <JWheeler> I ran gentoo for years, and really liked that it was 'native' 64-bit
[15:15:10] <JWheeler> but even there, you have /lib32
[15:15:13] <sparcdr> native, meh, minus a few apps
[15:15:21] <sparcdr> same goes with irix, 10 years back
[15:15:22] *** ashner has quit IRC
[15:15:38] *** deather_ is now known as deather
[15:15:52] <sparcdr> windows is the reason 64-bit adoption is stagnated, the reason for another cycle of hackery
[15:15:53] <JWheeler> I think it's great that we have all these cpus that can do 32 & 64-bit, but it's really stretching the life of 32-bit out
[15:16:09] <sparcdr> it's gonna be like that until 2012
[15:16:14] <JWheeler> Yes, well Winxp-64 just didn't work
[15:16:20] <sparcdr> my est until more than 70% have 64-bit
[15:16:25] <sparcdr> it worked, crapily
[15:16:38] <JWheeler> 2003-64 is better, I use it wherever possible (in windows only environments, naturally)
[15:16:44] <sparcdr> vista 64-bit works pretty well considering the legacy bagage they finally dropped
[15:16:47] <JWheeler> and I hear that vista is 'better'
[15:16:52] <sparcdr> it is
[15:17:02] <JWheeler> yeah, so hopefully we'll start seeing some progress on that front
[15:17:03] <sparcdr> i know from experience, wow64 on vista doesnt make me cringe much
[15:17:10] <JWheeler> heck, isn't exchange 2007 64-bit only?
[15:17:14] <sparcdr> in due time
[15:17:15] <sparcdr> yes it is
[15:17:25] <JWheeler> ...first good thing they've done!
[15:17:30] <sparcdr> the eval versions are available in 32bit, but only 64-bit is supported
[15:17:40] <sparcdr> ie: in enterprise
[15:17:56] <sparcdr> smb's can still get away with running on 4gb ram
[15:18:21] <sparcdr> a lot of their apps are only 64-bit actually
[15:18:51] <sparcdr> Project Server 2007 i think is only 64-bit, and a lot of high end game servers, like Blizzard's for WOW and probably the Steam Content Distribution system
[15:19:08] *** dunc has quit IRC
[15:20:21] <sparcdr> brb
[15:21:18] <cmihai> http://unixsadm.blogspot.com/2007/09/bsd-is-gpl-reloaded.html - Guess the people who think BSD is licensed GPL really understand what GPL is all about, heh?
[15:21:26] *** sparcdr has quit IRC
[15:21:52] <cmihai> Makes me wonder why I should even bother getting certified.
[15:22:39] *** jamesd_ has quit IRC
[15:23:04] *** dunc has joined #opensolaris
[15:23:19] <JWheeler> Whatever happened to that wifi driver on the LKML?
[15:23:26] <JWheeler> Theo seemed pretty upset about it all
[15:23:29] <JWheeler> who won in the end?
[15:25:02] <cmihai> Won?
[15:25:24] <cmihai> The driver was removed.
[15:26:06] <JWheeler> The legal debate
[15:26:47] *** eboutilier has quit IRC
[15:27:51] *** pizdec__ has joined #opensolaris
[15:31:04] <JWheeler> did it conclude that dual licensed code can be relicensed only under the GPL?
[15:34:27] <cmihai> I didn't follow anything beyond the removal point.
[15:34:37] <cmihai> Theo said the code goes, and the code went from the CVS tree. End of story.
[15:35:15] <JWheeler> from the linux CVS tree?
[15:35:25] <cmihai> No, from the OpenBSD CVS tree.
[15:35:45] <JWheeler> Even though it started off as BSD code?
[15:35:50] <JWheeler> wow.... that sucks
[15:35:59] <JWheeler> one way to make a statement though, I guess
[15:36:08] <cmihai> Best way to permanently end all disputes.
[15:36:37] <JWheeler> Yeah, but they loose both ways now
[15:36:38] <JWheeler> pity
[15:37:25] <Tempt> errr
[15:37:26] <Tempt> hang on
[15:37:44] <Tempt> OpenBSD code was illegally re-licensed as GPL only, and it gets pulled from OpenBSD?
[15:37:58] <Tempt> That is remarkably shit, and makes me want to kick a GPL advocate.
[15:38:49] *** jamesd has joined #opensolaris
[15:38:49] *** ChanServ sets mode: +o jamesd
[15:39:10] *** squallbsr has joined #opensolaris
[15:39:45] <cmihai> No.
[15:40:01] <cmihai> http://undeadly.org/cgi?action=article&sid=20070406104008
[15:40:23] <Tempt> Haha, you aren't going to trick me that easily. I recognise the domain name and have no desire for reading flames today.
[15:41:00] <cmihai> ?
[15:41:45] <h3sp4wn> You are confusing two different issues
[15:42:56] <squallbsr> can anybody help me figure out why when I try to boot the kernel, it reboots my computer with a General Fault.
[15:43:17] *** jamesd has quit IRC
[15:43:21] <squallbsr> I can boot into the debugger\
[15:43:24] *** pizdec_ has quit IRC
[15:44:19] <squallbsr> I want to be able to figure out where the problem is happening, which module, so that I can report the bug with useful info
[15:45:32] <flyingparchment> type $c
[15:46:36] *** mikefut has quit IRC
[15:47:20] <squallbsr> I step all the way to the line before the system reboots, $c returns nothing
[15:48:08] <squallbsr> it happens on this ASM:  0x50c9 lgdt 0x52a0
[15:48:29] <JWheeler> come on mplayer!!
[15:49:12] <JWheeler> Tempt, I agree completely
[15:49:55] *** jamesd has joined #opensolaris
[15:49:56] *** ChanServ sets mode: +o jamesd
[15:50:33] <squallbsr> I'm also on build 70
[15:50:48] <JWheeler> hey... I think I finally have it!
[15:51:15] *** pizdec_ has joined #opensolaris
[15:53:09] <Tempt> Man, my Windows roaming profile has blown out to nearly 2Gb
[15:53:21] <Tempt> No longer logging into that host is taking a looong time these days.
[15:54:46] <WickedWicky> *burp*
[15:56:43] *** reflect_ has joined #opensolaris
[15:57:59] <Tempt> Actually, I really don't need it to roam my profile. Is there a way to tell Windows *not* to pull in the profile for a specific user?
[15:59:02] <JWheeler> yup
[15:59:09] <JWheeler> Well you can tell it not to roam
[15:59:17] <Tempt> How?
[15:59:28] <JWheeler> I don't have a windows box handy... here goes
[15:59:44] <JWheeler> Right Click my computer, profiles?
[15:59:56] <Tempt> okay, let me try.
[15:59:57] <JWheeler> properties first sorry
[16:00:00] <Tempt> This is 2003 server ...
[16:00:10] <JWheeler> Right click my computer, properties, profiles tab?
[16:00:17] <JWheeler> yup, that's what I use
[16:00:28] <JWheeler> if it's not there, it might be under the advanced tab?
[16:00:33] <Tempt> No profiles tab
[16:00:41] <Tempt> aaah
[16:00:42] <Tempt> I see it.
[16:00:54] <JWheeler> choices are roaming, or local, as I recall
[16:01:13] * JWheeler cringes, doing windows support in #opensolaris
[16:01:26] <JWheeler> sssh, don't tell anyone, I'm really an undercover agent of the evil empire!
[16:01:47] <Tempt> haha
[16:01:57] <Tempt> I've only got this VM to publish a couple of applications for SSGD
[16:02:42] *** trip_ has joined #opensolaris
[16:03:36] <Tempt> Ooh, thanks, that worked a treat.
[16:04:18] <Tempt> man, #opensolaris, where people can help you with switches, routers, oracle, bsd, windows, linux, and sometimes even solaris.
[16:04:20] *** reflect has quit IRC
[16:04:21] <Tempt> Amazing!
[16:04:45] <WickedWicky> haha
[16:04:56] <JWheeler> cool :)
[16:05:12] * JWheeler goes back into the closet
[16:05:23] <squallbsr> While booting, after the version banner at the top of the screen, the kernel reboots the computer with a general fault and a BAD TRAP.  I am in kmdb, I have found the line of ASM that reboots the box, $c returns nothing.
[16:05:37] <squallbsr> I need some help so that I can file a bug report that has useful information in it.
[16:05:51] *** cmihai has quit IRC
[16:05:52] <JWheeler> squallbsr, you're already way over my head sorry :/
[16:06:25] *** pizdec__ has quit IRC
[16:06:37] *** pizdec__ has joined #opensolaris
[16:06:58] <squallbsr> Anybody know where/who can help me get this fixed?
[16:08:43] *** sioraiocht has quit IRC
[16:08:52] *** pizdec_ has quit IRC
[16:10:17] *** pizdec_ has joined #opensolaris
[16:10:40] <JWheeler> :/ sorry
[16:11:01] <WickedWicky> I had the saturday morning of my life by the way
[16:11:12] <squallbsr> JWheeler, I thank you for helping
[16:11:25] <WickedWicky> some bright light decided it'd be a good day and morning to "tidy up" the fiber cables in one of our 19" racks
[16:11:46] <JWheeler> lol, poor wicky
[16:12:03] <Tempt> oh dear.
[16:12:07] <Tempt> "tidy up" the fibre
[16:12:10] <WickedWicky> yes
[16:12:11] <WickedWicky> like
[16:12:15] <Tempt> That's gold. Did they use lots of cable ties?
[16:12:20] <Tempt> Nice 90 degree bends?
[16:12:30] <Tempt> tight-tight-tight cable ties at that?
[16:12:50] <WickedWicky> grab them, put a what's the name... those plastic things you use to keep things together, once you tightend them you can only cut them
[16:13:55] <Tempt> yes, cable tie
[16:13:58] <Tempt> Oh, that's marvellous
[16:14:01] <Tempt> Have they been fired yet?
[16:14:21] <WickedWicky> well, HE is pulling new fibre as we speak
[16:14:36] <WickedWicky> fortunaly we have an A and a B side for network and fibre
[16:15:18] <WickedWicky> well hey, I guess these things happen
[16:15:20] <WickedWicky> *shrug*
[16:16:07] <Tempt> They shouldn't.
[16:16:16] <Tempt> I'd be revoking someone DC access already.
[16:16:17] <WickedWicky> true
[16:16:22] <Tempt> and giving them the chewing out of a lifetime.
[16:19:23] *** jmcp has quit IRC
[16:19:38] *** dpn` has quit IRC
[16:20:04] <squallbsr> WickedWicky, now do you think the afore mentioned monkey is capable of pulling the new fibre without screwing it up?
[16:21:34] <WickedWicky> I hope so for my coworkers, cause I aint doing anything today
[16:21:45] <WickedWicky> nor do I think I will be allowed in the datacenter with my current state
[16:23:29] * WickedWicky took some kick ass headache killers
[16:25:37] *** FBdev has quit IRC
[16:26:49] *** pizdec__ has quit IRC
[16:27:05] *** pizdec__ has joined #opensolaris
[16:28:02] *** Shiv__ has joined #opensolaris
[16:28:25] *** trip__ has joined #opensolaris
[16:28:32] *** pizdec_ has quit IRC
[16:31:54] *** kaiwai has joined #opensolaris
[16:31:58] <kaiwai> hi :)
[16:32:07] *** pizdec_ has joined #opensolaris
[16:32:16] <kaiwai> anyone got any experience with gtkpod?
[16:32:22] *** Snake007uk has joined #opensolaris
[16:32:58] <squallbsr> kaiwai, on linux I have used it a little bit
[16:33:01] *** Snake007uk has quit IRC
[16:33:05] <kaiwai> on Solaris mainly
[16:33:27] <squallbsr> kaiwai, it can be finicky, good luck
[16:33:35] *** cmihai has joined #OpenSolaris
[16:35:00] <Tempt> hang on
[16:35:07] <Tempt> I thought you ditched your iPod?
[16:35:44] *** kloczek has quit IRC
[16:36:32] <kaiwai> Tempt: bought it off my brother;
[16:36:32] *** kloczek has joined #opensolaris
[16:36:41] <kaiwai> it works, mounts, gtkpod doesn't compile though
[16:36:42] <Tempt> haha.
[16:36:47] <kaiwai> from the pkgbuild svn
[16:36:50] <Tempt> amarok
[16:36:53] <Tempt> there's a blastwave pkg
[16:37:36] <kaiwai> yeap, tempted to do that
[16:39:40] <kaiwai> does blastwave have a cvs?
[16:41:34] *** kaiwai has quit IRC
[16:46:03] *** trip_ has quit IRC
[16:48:25] *** pizdec__ has quit IRC
[16:51:29] *** pizdec_ has quit IRC
[17:03:26] <JWheeler> svn
[17:03:47] *** reflect has joined #opensolaris
[17:05:15] <JWheeler> I need to check the status of that software, I want to be able to use it myself
[17:07:24] *** gm135 has joined #opensolaris
[17:08:27] *** blindfish has quit IRC
[17:11:31] *** reflect_ has quit IRC
[17:17:22] <WickedWicky> i tried to install amarok yesterday
[17:17:31] <WickedWicky> got the entire KDE with it
[17:24:21] *** Misha has joined #opensolaris
[17:24:32] <Misha> Does OpenSolaris works out of the box with Xeons?
[17:24:41] <Doomshammer> yep
[17:25:02] <JWheeler> you should have had kde_base and kde_libs
[17:25:07] *** sfire||mouse has joined #opensolaris
[17:25:11] <JWheeler> and konqueror
[17:25:28] <JWheeler> are you sure it was the _entire_ kde?
[17:25:45] <WickedWicky> it was a lot more than I need for amarok, lemme put it that way
[17:25:55] <flyingparchment> The physical processor has 2 virtual processors (1 3)
[17:25:56] <flyingparchment>   x86 (chipid 0x3 GenuineIntel family 6 model 15 step 6 clock 2000 MHz)
[17:25:56] <flyingparchment>         Intel(r) Xeon(r) CPU            5130  @ 2.00GHz
[17:25:58] <JWheeler> are you talking about CSWamarok?
[17:26:03] <WickedWicky> oui
[17:26:31] <JWheeler> ?
[17:26:37] <WickedWicky> that's yes, in french
[17:26:43] <JWheeler> oh
[17:26:46] <WickedWicky> :P
[17:27:00] <JWheeler> ok. Well I'm the CSWamarok maintainer, so I'd like to know more about this
[17:27:08] <JWheeler> I install it a lot, and I only get those three come down
[17:27:14] <WickedWicky> the thing is
[17:27:19] <flyingparchment> remember kdebase does include an awful lot too
[17:27:21] <WickedWicky> that blastwave follows all dependencies
[17:27:22] <WickedWicky> so
[17:27:27] <WickedWicky> amarok dpeends on kde_base and kde_libs
[17:27:29] <JWheeler> sure, a whole bunch of other stuff comes down, but not kde specific stuff
[17:27:35] <WickedWicky> kde_base depends on 438943984343984309 other things
[17:28:48] <JWheeler> 32, to be exact
[17:28:56] <JWheeler> first level dependancies, anyway
[17:28:58] <flyingparchment> openldap!!!!
[17:29:14] <flyingparchment> JWheeler: you guys need to learn how to split into -bin and -libs packages like linux
[17:29:15] <JWheeler> runtime
[17:29:16] <WickedWicky> I love blastwave
[17:29:28] *** bunker has quit IRC
[17:29:33] <WickedWicky> but sometimes it does too much
[17:29:35] <WickedWicky> same with PHO
[17:29:38] <WickedWicky> PHP
[17:29:44] <WickedWicky> you want a simple AMP solution
[17:29:46] <JWheeler> flyingparchment, that's the openldap runtime, not the whole thing
[17:29:49] <WickedWicky> you end up with openldap on your server
[17:29:50] <tomww> you might want to look at the SFEgtkpod.spec
[17:29:57] <flyingparchment> JWheeler: last time i used it, i got /etc/init.d/openldap and everything
[17:30:00] <flyingparchment> JWheeler: maybe that changed now..
[17:30:09] <JWheeler> It changed quite recently
[17:30:12] <flyingparchment> ah.  good
[17:30:15] <Misha> flyingparchment, Doomshammer: Cery nice, thanks
[17:30:24] <JWheeler> we're trying, help and suggestions are welcome
[17:30:35] <JWheeler> you'd be amazed at how few active maintainers there are
[17:30:42] <WickedWicky> I saw the list :D
[17:30:45] <JWheeler> I think we do a pretty good job all things considered
[17:30:46] <flyingparchment> Misha: there is one catch: older version of Solaris 10 don't support newer Xeon CPUs, you'll get a panic on boot.  but current solaris 10 and opensolaris have no problem.
[17:30:50] <WickedWicky> you do
[17:30:53] <WickedWicky> it's not that
[17:30:55] <JWheeler> I'm here building mplayer at 3:30am!
[17:31:00] <tomww> http://tinyurl.com/2sbseb  (SFEgtkpo.spec)
[17:31:04] <WickedWicky> but sometimes I get more than I anticipated for
[17:31:04] <JWheeler> on a saturday night!
[17:31:22] <WickedWicky> on a sun4m with 64M of RAM and 4GB harddisk you just want what you ask for
[17:31:27] <JWheeler> WickedWicky, Yes, it is a problem. We are doing our best :)
[17:31:31] <WickedWicky> I know :F
[17:31:33] <WickedWicky> :D
[17:31:40] <WickedWicky> my typing isnt on its best today
[17:31:43] <flyingparchment> i thought about becoming a csw maintainer but then i stopped using csw.  :P
[17:31:52] <flyingparchment> JWheeler: my most annoying peev about csw is you put config in /opt/csw/etc
[17:32:01] <flyingparchment> JWheeler: that completely fucks sparse zones
[17:32:10] <flyingparchment> (it should be in /etc/opt/csw or so)
[17:32:34] <JWheeler> We very intentional don't verge outside of /opt/csw unless absolutely necessary
[17:32:42] <flyingparchment> in this case it is absolutely necessary
[17:32:46] <flyingparchment> and also 'how things are done' on solaris
[17:32:59] <JWheeler> BW is still primarily a solaris 8 shop, zones is something that needs to be considered
[17:33:05] <flyingparchment> packages install into /opt/foo, but config files go into /etc/opt/foo - that's what /etc/opt is for
[17:33:34] <JWheeler> BW is also optimised to be nfs shared, so just the one dir helps
[17:33:46] <JWheeler> I agree that you shouldn't be nfs sharing daemons though
[17:33:49] <flyingparchment> you don't want to NFS share your config files!
[17:34:43] <WickedWicky> JWheeler, can the current amarok package in BW handle MTP devices?
[17:34:54] <flyingparchment> what's mtp?
[17:35:00] <WickedWicky> MPT
[17:35:10] <flyingparchment> what's mpt? :)
[17:35:15] <WickedWicky> Play for sure
[17:35:17] <flyingparchment> (other than a scsi controller)
[17:35:20] <flyingparchment> oh.  ew
[17:35:23] <WickedWicky> like, what the new creative vision:m uses
[17:35:37] <WickedWicky> amarok should be able to work with libmtp
[17:35:55] *** dunc has quit IRC
[17:36:00] <WickedWicky> (if its compiled with mtp support of course :D)
[17:36:20] <JWheeler> WickedWicky, no it can't
[17:36:25] <WickedWicky> eeps
[17:36:35] <JWheeler> I'd like to ipod support in there first, but it is something I'd like to add
[17:37:04] <JWheeler> right at the moment, I'm working on getting 1.4.7 out there, and then I'll have another look at MPT and ipods
[17:37:05] <WickedWicky> cool :D I'd be more than happy to test if you're looking for a tester
[17:37:15] <JWheeler> I can assure you, just building amarok took me a long time!
[17:37:20] <JWheeler> cool, I'll remember that :)
[17:42:16] *** squallbsr has quit IRC
[17:45:14] *** Shiv__ has quit IRC
[17:45:27] *** reflect_ has joined #opensolaris
[17:45:49] *** sponix has joined #opensolaris
[17:48:55] *** trip_ has joined #opensolaris
[17:49:31] *** Biovore has quit IRC
[17:49:33] *** mikefut has joined #opensolaris
[17:51:37] <Tempt> WickedWicky: You should try installing SSGD on your SS5
[17:51:43] <Tempt> WickedWicky: See how long it takes to fire up
[17:51:51] <Tempt> WickedWicky: I'm betting five days.
[17:52:04] <Tempt> WickedWicky: Err, SS20, isn't it? yeah, install it on that.
[17:52:24] *** reflect has quit IRC
[17:52:39] *** Misha has quit IRC
[17:53:28] *** polk__ has quit IRC
[17:58:20] <axisys> this is a hack I got from brendang on how to catch the port from syscall::connect:entry
[17:58:23] <axisys> this->port = ntohs(*(short *)copyin(arg1 + sizeof (short), sizeof (short)));
[17:58:36] <axisys> can anyone explain how that gives the port number?
[17:58:56] <axisys> waiting for brendang to appear so I can ask him later otherwise
[17:59:10] <flyingparchment> i'm guessing it copyin's a struct
[17:59:21] <flyingparchment> and takes the second (?) short-sized member
[17:59:27] <flyingparchment> then does ntohs to get it in host order
[18:00:37] <axisys> flyingparchment: so host order has port number in it? any doc (may be opensolaris src browser) where I can read more examples?
[18:00:48] <flyingparchment> the port number is in host order
[18:00:52] <flyingparchment> it's probably copying in a struct sockaddr_in
[18:01:02] <flyingparchment> sorry, i mean the port number is in _network_ order
[18:01:07] <flyingparchment> ntohs converts it to host order
[18:01:41] <axisys> flyingparchment: pardon my ignorance but what is a network order?
[18:01:53] <flyingparchment> axisys: "network order" is another way of saying "big endian"
[18:02:04] <flyingparchment> axisys: "host order" is the host's native order, which is little endian on x86 or big endian on sparc
[18:02:21] <axisys> ahh
[18:02:22] <flyingparchment> (it refers to the order of bytes in a binary number - 1234 or 4321)
[18:02:37] <flyingparchment> it's called network order because the standard order on the internet is big endian
[18:02:43] <flyingparchment> so that's what sockaddr_in and such use
[18:03:20] <axisys> where does the port number hides ? in connect() ?
[18:03:41] <flyingparchment> the second argument to connect() (arg1) is a sockaddr*
[18:03:42] <axisys> arg1 of connect()?
[18:03:50] <flyingparchment> for AF_INET (Internet sockets) that's really a sockaddr_in*
[18:04:12] <axisys> flyingparchment: uh huh
[18:04:23] <flyingparchment> so that code is taking the sin_port member from the sockaddr_in
[18:04:27] <axisys> flyingparchment: it makes lot more sense now
[18:04:30] <flyingparchment> copyin() is used when you have a pointer and want to access the data behind it
[18:04:37] <flyingparchment> it copies the data from the user program to dtrace
[18:05:09] <axisys> copyin() is diff from copyinstr() which i used before
[18:05:19] *** snuff-home has quit IRC
[18:05:19] *** snuff-work has quit IRC
[18:05:19] *** snuff-away has quit IRC
[18:05:19] *** mlh has quit IRC
[18:05:27] <axisys> for dtrace that is
[18:05:31] <flyingparchment> i assume copyinstr copies a string, using the terminating \0
[18:06:15] <axisys> flyingparchment: these are soooo informatives.. thanks soooo much man!!
[18:06:23] *** trip__ has quit IRC
[18:08:38] *** mikefut has quit IRC
[18:09:17] *** asyd_ has joined #opensolaris
[18:09:27] *** asyd has quit IRC
[18:09:42] *** asyd_ is now known as asyd
[18:11:49] *** vmlemon has joined #opensolaris
[18:11:49] *** JWheeler has quit IRC
[18:11:56] *** vijay has joined #opensolaris
[18:12:59] <axisys> flyingparchment: u r correct
[18:13:02] <axisys> int  connect(int  s,  const  struct  sockaddr   *name,   int namelen);
[18:13:24] <axisys> second argument *is* const struck type
[18:14:00] *** Mdx4 has quit IRC
[18:16:52] *** ChanServ sets mode: +o asyd
[18:18:29] *** zooko has joined #opensolaris
[18:18:46] *** cla has quit IRC
[18:18:50] <zooko> Anybody want to help me port my open source distributed filesystem to Solaris?
[18:18:53] <zooko> http://allmydata.org
[18:18:59] <zooko> v0.5 of this project already ran on Solaris.
[18:19:17] <zooko> Now v0.6 is ready to release, but one of the unit tests fails.  I suspect it might be something to do with struct packing...
[18:19:19] * zooko investigates.
[18:19:45] <flyingparchment> yet another distributed filesystem?
[18:19:55] <zooko> Eh, maybe I shouldn't call it that.
[18:20:00] *** vsrinivas has joined #opensolaris
[18:20:08] <zooko> It focusses on quite a different part of the design space than network filesystems do.
[18:20:31] <vsrinivas> hi, has anyone tried running SXCE on a k6?
[18:20:44] <zooko> I usually avoid calling it a filesystem, actually, because its file update semantics are so limited.
[18:20:49] <zooko> How about "distributed storage grid"?
[18:21:14] <libkeiser> distributed files system is a loaded phrase. most people think you're referring to afs/dfs when you say that
[18:21:51] <flyingparchment> zooko: mogilefs?
[18:22:35] <zooko> flyingparchment: nice!  Thanks for the reference!
[18:22:54] <flyingparchment> i'm kind of dubious about mogile because a) it was written by lj and b) it's written perl and c) it uses mysql
[18:22:58] <flyingparchment> but the idea seems neat
[18:24:09] *** JWheeler has joined #opensolaris
[18:24:27] <WickedWicky> can I ask a question without getting lynched?
[18:24:35] * zooko adds it to http://allmydata.org/trac/tahoe/wiki/RelatedProjects
[18:24:47] <WickedWicky> how come every grub-using Linux distro I download can boot/install via USB-CDROM
[18:24:50] <WickedWicky> but opensolaris cant?
[18:25:53] * Tempt lynches WickedWicky
[18:26:07] <vsrinivas> also, can a zone on an SXCE system take advantage of 3d acceleration from a graphics card?
[18:26:23] <flyingparchment> vsrinivas: not exactly, because zones can't run X servers
[18:26:33] <vsrinivas> oh, I didn't realize that
[18:26:36] <vsrinivas> unfortunate
[18:26:40] <flyingparchment> vsrinivas: a zone can connect to an accelerated X server but you won't have DRI or shared memory, so it won't be very fast
[18:28:02] <vsrinivas> it'd be nice to be able to isolate an X server in that way, oh well
[18:28:22] *** sfire||mouse has quit IRC
[18:29:19] *** Misha has joined #opensolaris
[18:29:20] <Tempt> hmmmm
[18:29:20] *** vijay has quit IRC
[18:29:34] <Tempt> Can't even hand the /dev/fb.. to the zone?
[18:29:38] <Tempt> That's a bummer.
[18:29:59] *** Misha has quit IRC
[18:30:57] <zooko> flyingparchment: I agree that the basic idea of mogilefs looks good.
[18:31:14] <zooko> It is pretty similar to my allmydata.org project, as far as I can see in a minute's glance.
[18:31:23] <flyingparchment> Tempt: i don't think so.  (of course i could be wrong - never tried it)
[18:31:27] <zooko> allmydata.org uses erasure coding (Reed-Solomon codes) instead of replication.
[18:31:39] <flyingparchment> zooko: is that what RAID5/6 uses?
[18:35:19] *** pizdec_ has joined #opensolaris
[18:37:09] <vsrinivas> what if one were to use xen instead? would you then be able to have a guest run an X server and take advantage of acceleration (by giving it the pci device or VMGL?)
[18:37:27] <flyingparchment> i don't know how xen works in solaris
[18:37:33] <flyingparchment> of course it's xVM here, not Xen :D
[18:37:41] <vsrinivas> okay, xVM, then =)
[18:37:54] <vsrinivas> (wasn't the branch called xen here too?)
[18:38:03] <flyingparchment> yeah, but they productised it as xvm
[18:38:58] <vsrinivas> oh, okay
[18:39:42] <flyingparchment> (because xensource wouldn't let them call it 'xen' for some reason)
[18:39:47] <vsrinivas> haha
[18:40:05] <vsrinivas> to be honest, xVM sounds cooler
[18:40:58] <zooko> flyingparchment: the erasure code that we use is a generalization of RAID5/6.
[18:41:26] <zooko> RAID5 is that you have N drives, any N-1 of them have to work.
[18:41:31] <zooko> With RAID6, any N-2 of them have to work.
[18:42:13] <zooko> With general Reed Solomon, you can choose your N between 1 and 256, and your other number K, between 1 and N, such that any K of them have to work.
[18:42:15] <zooko> Make sense?
[18:42:25] <flyingparchment> nod
[18:43:04] *** sartek has joined #opensolaris
[18:43:05] <zooko> So compared to replication, it offers more points on the trade-off curve between overhead of transmission and storage versus reliability.
[18:43:37] <flyingparchment> man, i want an ultra 20 now
[18:43:44] <flyingparchment> shame i have no money and live on welfare
[18:44:41] <vsrinivas> you could get an ultra 30 and pretend you have a more powerful model =)
[18:45:53] <flyingparchment> i actually have a pretty decent workstation (2 1.2GHz Athlons, 1GB) but it keeps crashing for no reason.  which is pretty annoying.
[18:46:28] <zooko> If anybody is interested in the issue of porting allmydata.org to Solaris, here is the trac ticket:
[18:46:31] <vsrinivas> meh. I have a k6-2 400mhz, 640mb of ram. I've never tried sxce on it though...
[18:46:32] <zooko> http://allmydata.org/trac/tahoe/query?status=new&status=assigned&status=reopened&groupdesc=1&group=component&milestone=0.6.0&order=priority
[18:46:40] <flyingparchment> i run ubuntu on my desktop.  ;)
[18:46:52] <zooko> Time for me to cook breakfast.
[18:46:53] <Tempt> OH NOES
[18:47:04] <vsrinivas> I run debian, but I'd love to try sxce; anyone ever try it on a k6? ever get it working?
[18:47:13] <flyingparchment> it should work fine
[18:47:19] <flyingparchment> RAM is more important than CPU to solaris
[18:47:31] <vsrinivas> (its somewhere between 586 and 686, that's why I was curious)
[18:47:55] <flyingparchment> i ran S9 on a dual P133 with 64MB RAM, but i don't think that'll run s10 :(
[18:48:28] *** dunc has joined #opensolaris
[18:48:37] <SunTzuTech> I run S10/11 on some 700-1000Mhz PIII, though 768MB of ram gets a bit thin after mozilla/firefox runs for a while
[18:48:40] <flyingparchment> still have s7 CD-Rs lying around too, heh
[18:53:16] *** sponix has quit IRC
[18:57:37] <h3sp4wn> Is there anywhere I can get a jumpstart target that installs SXDE (CE doesn't read my sata dvdrom atm) but non of the gnome stuff (I intend to use pkgsrc) I want X, CDE, base and all the development headers - and sunstudio (or a good way to deselect everything once its installed and without breaking stuff)
[18:57:57] <h3sp4wn> I intend to use pkgsrc for the oss stuff I want (which is mainly cli anyway)
[18:58:45] *** pizdec_ has quit IRC
[18:58:52] *** pizdec_ has joined #opensolaris
[19:02:38] *** sponix has joined #opensolaris
[19:10:37] *** pizdec__ has joined #opensolaris
[19:14:30] <holcomb> h3sp4wn: you need to set up a jumpstart profile that deselects the JDS cluster
[19:14:50] <holcomb> go into your install media and look for a file called .clustertoc, in there there's the name of the cluster with all that stuff in it
[19:14:58] <holcomb> (don't remember it off hand)
[19:15:11] <holcomb> then in your profile do something like cluster: SUNWCgnome delete
[19:15:14] <holcomb> i think
[19:18:53] *** reflect has joined #opensolaris
[19:19:50] *** switch has quit IRC
[19:20:06] *** switch has joined #opensolaris
[19:20:23] *** dosiu has quit IRC
[19:20:24] *** charlesnw has joined #opensolaris
[19:22:14] *** asarch has joined #opensolaris
[19:22:25] <asarch> Hi everyone, how is it getting along?
[19:22:30] <asarch> How's the weather?
[19:22:32] <zooko> Great!
[19:23:16] <flyingparchment> it's a bit cloudy here
[19:23:21] <asarch> Cool!
[19:24:08] <asarch> Why there are two channels about Solaris?
[19:24:15] <asarch> #Solaris and #OpenSolaris
[19:24:29] <flyingparchment> #solaris is for evil proprietary people
[19:24:30] <flyingparchment> we are open!
[19:24:42] <WickedWicky> what he said
[19:25:33] <asarch> Oh, I see..
[19:25:51] *** dosiu has joined #opensolaris
[19:26:06] *** reflect_ has quit IRC
[19:26:13] <h3sp4wn> holcomb: Thanks alot
[19:26:38] <asarch> Which one do you recommend for learning the Solaris way? Solaris Express
[19:26:41] <asarch> Damn!
[19:27:05] <asarch> Solaris Express Community Edition or Solaris Express Developer Edition?
[19:27:08] *** pizdec_ has quit IRC
[19:27:16] <flyingparchment> don't use SXDE for anything
[19:27:22] <asarch> Why?
[19:27:31] <Symmetria> hrm coolstack apache2 *REALLY* doesnt like being restarted under heavy load
[19:27:45] <flyingparchment> because it's crappy and pointless
[19:27:46] <Symmetria> it has a bad habit of zombiefying processes
[19:27:53] <flyingparchment> it's just an old version of sxce
[19:28:06] *** oninoshiko has joined #opensolaris
[19:28:11] <asarch> I downloaded the "Solaris 10 8/07". Is it a good version of Solaris?
[19:28:29] <zooko> I use Nexenta.
[19:28:32] <flyingparchment> that's the current solaris 10
[19:28:39] <flyingparchment> it's good if that's what you want
[19:29:01] <asarch> Nexenta == Solaris 10 8/07?
[19:29:11] <flyingparchment> no, nexenta is an opensolaris distribution
[19:29:19] <flyingparchment> like SXCE
[19:29:28] <flyingparchment> opensolaris ~= solaris 11
[19:29:29] <zooko> The cool thing about Nexenta is that the installer did exactly what I wanted by default: it detected my two IDE hard drives, set them up in mirrored mode and used ZFS for the fs.
[19:29:29] <oninoshiko> the latest SXCE has more recent work, if that is good or bad is a matter of perspective
[19:29:52] <zooko> The bad thing is that there are apparently very few devs working on Nexenta, and many small bugs.
[19:30:01] <zooko> It reminds me of Debian circa 1996 or so.
[19:30:30] <Symmetria> *sigh* I rebooted a machine a few minutes ago...
[19:30:30] <zooko> The other good thing about Nexenta is, of course, that you can use apt for package management.
[19:30:33] <Symmetria> I hope it comes back
[19:30:39] <asarch> Everybody is developing for Ubuntu right now
[19:30:52] <asarch> /
[19:31:03] <zooko> That was essential to me, as it allowed me to start running Solaris without having to first commit to learning a new administration tool.
[19:31:17] <oninoshiko> zooko, the goodness of useing apt for package management is also a matter of perspective :p
[19:31:20] <asarch> So, Solaris 10 8/07 == Solaris 10 and OpenSolaris ~= Solaris 11, right?
[19:31:21] *** pizdec__ has quit IRC
[19:31:27] <flyingparchment> asarch: yes
[19:31:36] <asarch> Oh, I see...
[19:31:37] <flyingparchment> asarch: basically, opensolaris is the source code of what will become solaris 11
[19:31:41] <zooko> oninoshiko: I love apt pretty well.  But more-over, I already know how to use apt, so I can continue to use it without spending my time learning something new.
[19:32:08] <asarch> That's the part it was missing from me
[19:32:13] <flyingparchment> asarch: SXCE, Nexenta, etc. are opensolaris distributions
[19:32:18] *** Marv|LG has joined #opensolaris
[19:32:25] <flyingparchment> asarch: Solaris 11 (when it's released) will also be an opensolaris distribution
[19:32:43] <flyingparchment> SXCE is the beta version of solaris 11
[19:32:46] <cmihai> asarch, Nexenta is NOT Solaris, it's Ubuntu with a OpenSolaris kernel + networking (ON) and some basic tools.
[19:33:17] <cmihai> And really, double checking in both #*Solaris channels is kind of lame.
[19:33:41] <asarch> ?
[19:33:48] <asarch> I though you were out
[19:33:56] <asarch> That's why I came here
[19:33:57] <asarch> :'(
[19:36:13] <Tempt> The Nexenta concept might be a nice toy, but I don't see it making it anywhere near a production environment in the next five years.
[19:36:19] <cmihai> Really, stop wasting time and grab SXCE 72. it's what you want to run on your laptop. It also includes the wireless driver, so you don't need to do any additional confiugration.
[19:36:23] <flyingparchment> hmm, a remanufactured U20 with 2GB RAM, 2.4GHz CPUs and an FB1400 is only $1,470
[19:36:32] <cmihai> Make sure you read: http://unixsadm.blogspot.com/2007/08/liveupgrade-sxce.html when you install it.
[19:36:36] <flyingparchment> that's pretty nice
[19:36:37] <zooko> When I first needed to talk to some Solaris folks (about how to compile the python interpreter on Solaris), I tried the #solaris channel and was greeted with astonishing rudeness.
[19:36:53] <zooko> The interesting thing was, people were rude in response to my initial message, which was something like "Greetings, people of #solaris!".
[19:36:59] <zooko> In other words, they were being rude solely for the fun of it.
[19:37:09] <Tempt> Yep, that'd be #solaris.
[19:37:14] <Tempt> Very different tone to here.
[19:37:22] <cmihai> You have to create a /altroot partition if you want to use LiveUpgrade. Sure, you can run Solaris 10, SXDE and SXCE on different LiveUpgrade environments, and switch between them at boot time (and keep your configuration and home), but it all depends on what you need Solaris for.
[19:37:27] <zooko> In contrast, an actual Solaris dev helped me understand the problem in the Python header files in #opensolaris, so I've subsequently submitted a patch to the Python devs to make Python compile out-of-the-box on Solaris in the future.
[19:37:33] *** SeSe has quit IRC
[19:37:44] <flyingparchment> lots of unix channels are filled with people who look for ways to mock someone instead of help them
[19:37:45] <Tempt> The primary function of #solaris is to allow jaded sysadmins to bitch to each other. It's an important role, but not good for newbies.
[19:37:51] <flyingparchment> often because they're distussion channels, not help channels
[19:37:54] <zooko> I understand.
[19:38:02] *** gm135 has quit IRC
[19:38:05] <asarch> See you later
[19:38:10] <Tempt> Try asking a question in #linux sometime.
[19:38:24] <flyingparchment> i can only imagine what kind of fearsome kiddies #linux is filled with...
[19:38:25] <zooko> I'd rather not!
[19:38:28] <oninoshiko> then shouldnt it be named #jaded-sysadmin-bitching
[19:38:30] <cmihai> Try asking a Ubuntu question in #Debian :P
[19:38:33] <flyingparchment> (although #linuxhelp wasn't bad when it started)
[19:38:45] *** asarch has quit IRC
[19:38:46] <zooko> I follow the Linux kernel dev process by reading LWN.net, and I'm often reminded of the aphorism about watching sausage being made.
[19:38:51] <Tempt> The bit that gets me about #linux is how quickly the nmaps and nessus runs come in after asking a question
[19:38:57] <oninoshiko> really we could eliminate about 40 or so channels by adding that one
[19:39:11] <Tempt> You really need a cloak before going near it.
[19:40:16] <zooko> Bye for now!
[19:40:19] <oninoshiko> bye
[19:40:50] <flyingparchment> hmm, i wonder if i could convince sun to give me a free computer.  didn't jamesd do that?
[19:41:06] <g4lt-sb100> you also have to remember the #solaris->#opensolaris transition acts as somewhat of a crapfilter for #opensolaris.  how many times has  #opensolaris had to explain wht athte -d in pkgadd does?  that's a regular occurence in #solaris
[19:42:00] <Tempt> argh
[19:42:01] <oninoshiko> ok, in that case a RTFM is not rude, its just obvious
[19:42:32] <Tempt> People take offence quickly.
[19:43:25] *** m0le has joined #opensolaris
[19:43:33] <g4lt-sb100> so yes, we're less polite in #solaris, politeness tends to be hammered out by when you ARe polite, you're still a bastard
[19:44:40] <oninoshiko> i have only once asked about flags in a program... followed promptly by submitting a bug report that order the flag MUST be in should be in TFM
[19:45:36] <g4lt-sb100> speaking of thankless jobs, I get to start my paying one now, so expect me to get even MORE jaded, as if that were possible
[19:45:59] <oninoshiko> atleast your getting paid
[19:46:04] <Tempt> Indeed.
[19:46:14] <Tempt> You get paid to listen to drooling morons
[19:46:18] *** reflect_ has joined #opensolaris
[19:46:58] <g4lt-sb100> well TV tech support, so that goes without saying
[19:47:50] *** Fish- has joined #opensolaris
[19:47:57] <g4lt-sb100> yes "turn it on" HAS been a valid troubleshooting technique for me in the past
[19:47:58] *** Wicky has joined #opensolaris
[19:48:21] <oninoshiko> how about "plug it in"
[19:48:22] *** WickedWicky has quit IRC
[19:48:27] *** Wicky is now known as WickedWicky
[19:48:31] <g4lt-sb100> that too
[19:49:03] <oninoshiko> or "please remove yourself from the gene pool"
[19:49:36] <oninoshiko> AKA "please give me a reason to nominate you for a darwin award"
[19:50:31] *** xtrondo has joined #OpenSolaris
[19:53:48] *** reflect has quit IRC
[19:56:21] *** SeSe has joined #opensolaris
[19:56:22] *** snuff-work has joined #opensolaris
[19:56:33] <cmihai> g4lt-sb100, the most common for me has to be: Read the error message, not just paste it here.
[19:56:36] *** Fish has quit IRC
[19:56:42] <cmihai> "Caps locked is pressed"
[19:56:45] <cmihai> (no shit)
[19:59:29] <flyingparchment> isn't it odd how 'pressed' and 'depressed' mean the same thing?
[19:59:59] * oninoshiko gives the Caps lock key prozac
[20:02:24] *** billbrasky has joined #opensolaris
[20:04:03] *** sartek has quit IRC
[20:11:44] *** bengtf_ has joined #opensolaris
[20:13:10] *** nostoi has joined #opensolaris
[20:16:18] *** Jagoma has quit IRC
[20:16:54] *** reflect has joined #opensolaris
[20:18:57] *** billbrasky has quit IRC
[20:22:13] *** duri has quit IRC
[20:22:24] *** duri has joined #opensolaris
[20:24:26] *** reflect_ has quit IRC
[20:24:43] <flyingparchment> does the U20 have an on-board IDE connector?
[20:27:20] *** Mdx4 has joined #opensolaris
[20:28:17] *** kadath has quit IRC
[20:29:09] <WickedWicky> The system is coming up, please wait.
[20:29:43] <Berny> .oO(booting wicky)
[20:29:59] <WickedWicky> | / - | / -
[20:30:09] <WickedWicky> ops, forgot the \
[20:30:10] *** mikefut has joined #opensolaris
[20:31:01] <WickedWicky> flyingparchment: I dunno, I thought it was SATA?
[20:31:09] <flyingparchment> yeah it comes with sata drives
[20:31:14] <flyingparchment> but a lot of boards also have ide connectors..
[20:33:42] <delewis> flyingparchment: the DVD-ROM drive in the Ultra 20 is iDE.
[20:33:44] <delewis> IDE, rather.
[20:33:48] <flyingparchment> aha
[20:33:56] <flyingparchment> so it must at least have one channel
[20:33:59] <flyingparchment> thanks
[20:34:02] <delewis> sure
[20:34:05] <flyingparchment> (i have a 250G IDE drive i'd rather not waste :)
[20:34:43] *** xtrondo has quit IRC
[20:36:06] *** alanc_away has quit IRC
[20:37:23] *** Gman has quit IRC
[20:39:49] *** tek-ops has joined #opensolaris
[20:39:59] *** program has joined #opensolaris
[20:40:18] <tek-ops> damn, I just realized I dont have any pci-e x4 slots available, back to the drawing board
[20:40:33] *** Trisk[laptop] is now known as Triskelios
[20:41:14] *** program has quit IRC
[20:42:05] <tek-ops> so how about zfs on a bunch of USB 2 disks, maybe?
[20:45:13] <tek-ops> is there anyway to support hot-plugging of SATA drives running in IDE mode?
[20:45:50] <Triskelios> no, the ata driver is not hotplug capable
[20:45:55] <tek-ops> hmmm
[20:46:18] <tek-ops> well, would it at list be able to give LED notificaiton of which drive had failed?
[20:46:55] <tek-ops> all I'm saying is, every day that goes by that this file server remains linux saddens me
[20:46:58] <FireflyST> what's going on?
[20:47:05] *** vsrinivas has left #opensolaris
[20:47:23] <tek-ops> I'm back to the drawing board with my file server
[20:47:47] <tek-ops> I was going to use this card: http://www.newegg.com/Product/Product.aspx?Item=N82E16816118058
[20:48:02] <tek-ops> however I only have pci-e x1 slots avaiable
[20:51:59] <tek-ops> hmmm
[20:52:16] <tek-ops> I guess I could pull the video card and go with a pci-e x1 video...
[20:52:36] <tek-ops> the drive failure led notification is part of the SCSI subsystem?
[21:00:58] *** billbrasky has joined #opensolaris
[21:08:05] *** billbrasky has quit IRC
[21:10:48] *** ShadowHntr has quit IRC
[21:16:10] *** deather is now known as unisex
[21:19:51] *** Mdx4 has quit IRC
[21:19:56] *** pizdec_ has joined #opensolaris
[21:22:57] *** oninoshiko has quit IRC
[21:24:33] *** trip__ has joined #opensolaris
[21:24:50] *** polk__ has joined #opensolaris
[21:28:54] *** pizdec__ has joined #opensolaris
[21:36:24] *** Andy_Pease has quit IRC
[21:36:28] *** mikefut has quit IRC
[21:36:42] *** Andy_Pease has joined #opensolaris
[21:41:43] *** trip_ has quit IRC
[21:45:23] *** pizdec_ has quit IRC
[21:45:46] *** pizdec_ has joined #opensolaris
[21:51:13] *** pizdec__ has quit IRC
[21:51:23] *** pizdec__ has joined #opensolaris
[21:56:32] <vmlemon> Hi, is there an obvious way of spotting a ZFS volume in df output (other than the arbitrary volume name)?
[21:58:02] <PerterB> not really, but you could use df -F zfs, or mount
[21:58:23] <flyingparchment> other filesystems have arbitrary device names, so that doesn't really help
[21:58:37] *** Drone has quit IRC
[22:00:10] <vmlemon> I'm assuming that they have an identifiable /dev node, but I don't have access to a Solaris installation to check
[22:00:26] <flyingparchment> no.
[22:00:32] <vmlemon> OK
[22:00:35] <vmlemon> Thanks
[22:00:36] <flyingparchment> there is no /dev node associated with a zfs filesystem
[22:02:00] <PerterB> The device name will start with the associated zpool name, so if you have a list of zpools on the system you can do it that way
[22:02:00] <tek-ops> vmlemon, do you have to use df?  what's your end goal?
[22:03:14] <PerterB> having said that, I don't think there's anything to prevent people creating zpools with the same name as other filesystems (eg ctfs)
[22:03:33] <vmlemon> It's part of a component of a Perl script that prints the amount of free space on a volume, at the moment it looks at the device node name in the df -h output
[22:04:00] *** ShadowHntr has joined #opensolaris
[22:04:05] <Triskelios> vmlemon: you could just run df -h <mountpoint>
[22:04:10] <jbk> can you not also look at /etc/mnttab?
[22:04:15] <Triskelios> vmlemon: I don't see why a perl script is necessary
[22:04:29] <vmlemon> It's just a single function of the script, that said
[22:04:34] <vmlemon> (not the whole script itself)
[22:05:17] <vmlemon> Thanks anyway, though
[22:06:37] *** Drone has joined #opensolaris
[22:07:44] *** pizdec_ has quit IRC
[22:08:31] <tek-ops> so, I assume theres still much current active development on the AHCI device driver, is the mailing list the only way to keep up to date or is there a progress tracking site that's publically accessable?
[22:12:58] *** xushi_ has quit IRC
[22:13:43] *** pizdec_ has joined #opensolaris
[22:13:44] <jbk> heh
[22:13:51] <tek-ops> ?
[22:13:56] <jbk> i'm tempted to reply to this email on osol-discuss
[22:14:04] <jbk> about lucopy doing 1byte reads
[22:14:20] <tek-ops> havent see it, sorry
[22:14:23] <jbk> asking if perhaps it was the same programmers that wrote the solaris jvm :)
[22:17:33] <jbk> for years, the jvm did the same thing for all file i/o
[22:18:36] <tek-ops> what exactly does it mean to be an observer of a project?
[22:21:03] *** mustang has quit IRC
[22:28:44] *** unisex is now known as deather
[22:29:30] *** pizdec__ has quit IRC
[22:30:15] *** pizdec_ has quit IRC
[22:32:48] <tek-ops> wow, I dont see any blog posts for the device driver community since 2005
[22:40:09] <quasi> maybe they all left sun to join a blog free convent? ;)
[22:40:47] <tek-ops> hey, as long as development is still happening, i'm PUMPED! :D
[22:44:49] <flyingparchment> hmm, Solaris troff produces PostScript output that ps2pdf can't handle
[22:45:17] <flyingparchment> ERROR: /undefined in x
[22:45:47] * Triskelios still doesn't know why the fingerprint auth project is missing the fingerprint reader driver downloads
[22:45:54] <cmihai> So use nroff
[22:45:57] <cmihai> Or groff or whatever.
[22:46:11] <flyingparchment> i don't want to produce terminal output and groff isn't installed
[22:46:15] * flyingparchment copies file to a Linux system
[22:48:31] *** andy__ has joined #opensolaris
[22:49:13] *** Andy_Pease has quit IRC
[22:49:58] *** Bart_M has quit IRC
[22:50:00] *** Bart_M has joined #opensolaris
[22:50:21] <trochej> necromorphina/1390990545/ech
[22:52:10] <trochej> :/
[22:52:16] <trochej> Wrong window, wrong paste :)
[22:52:48] <jbk> uugh i'm in linux upgrade hell at the moment :)
[22:53:59] *** billbrasky has joined #opensolaris
[22:54:04] <trochej> Whats so hellish in upgrading Linux? :)
[22:54:27] <trochej> [d]
[22:55:09] <jbk> haha
[22:55:27] <jbk> even talking to a guy that used to work for redhat, it has around a 20% success rate :)
[22:55:43] <jbk> even when the environment is reasonably controlled
[22:56:15] <tek-ops> sadly I know an Oracle DBA that has to upgrade RHEL 2 to 3, then do the oracle upgrade, then go to RHEL 4
[22:56:24] <tek-ops> i wished him the best of luck, hahahaha
[22:57:59] <jbk> you're probably better off just doing a reinstall
[22:58:28] <tek-ops> 2 and migrate the db each time... I'll tell him to concider that
[22:58:31] <jbk> i seriously think zfs / + clones for upgrade (dunno if it'll still be called live upgrade) will be a _huge_ advantage against linux
[22:59:00] <tek-ops> whatever it's called, LU is sick
[22:59:15] <tek-ops> I hope it's truly here to stay
[22:59:29] *** estibi has quit IRC
[22:59:39] *** cmihai has quit IRC
[22:59:47] <jbk> sun just needs to do a better job advertising it..
[22:59:50] <jbk> yeah
[23:00:04] <jbk> used it at my last job for 2.6->8 and 8->10 upgrades
[23:00:18] <jbk> works reasonably well, though veritas throws some complexity in the mix
[23:00:52] <jbk> but even then, the big issue was that w/ veritas encapsulating your boot disks, you needed a lot more internal disks, or lose redundancy during the upgrade process
[23:00:57] <jbk> zfs of course fixes all of that
[23:01:07] *** jwk404 has quit IRC
[23:01:52] <tek-ops> I've only dealt with UFS root partition(s)
[23:02:03] * Triskelios needs to migrate his laptop to zfs root soon
[23:02:24] <tek-ops> is zfs root going to be in solaris 10 soon?
[23:02:38] *** cmihai has joined #OpenSolaris
[23:04:55] *** vijay has joined #opensolaris
[23:05:00] *** vijay has left #opensolaris
[23:05:09] *** vijay has joined #opensolaris
[23:09:58] <jbk> no idea, they still need to get newboot working on sparc
[23:10:13] *** kFuQ has quit IRC
[23:10:35] <quasi> isn't it still sort of hackish in ON?
[23:11:09] <jbk> yes
[23:11:22] <jbk> you have to install to ufs, create your pool + fs, then copy over
[23:15:25] <Triskelios> does seperate root + data pools still make sense on a laptop?
[23:15:36] <Triskelios> the disk isn't exceedinly fast...
[23:16:10] <g4lt-sb100> yes, until zfs root is implimented
[23:16:36] <Triskelios> well, it is on SX
[23:16:44] <g4lt-sb100> I'm using seperate root+data on mine, only because zfs root no worky
[23:17:46] <g4lt-sb100> well, let me rephrase, the only system I'm willing to test zfs root on is my sparc, so until I've played with it on something I can lose, the answer is no zfs root for you
[23:19:23] <Triskelios> considering I have to do Dead Upgrade (TM) right now, it&s not too much of a problem for me to just back up the root slice to external storage and replace it with a zpool
[23:19:48] *** tulcod has joined #opensolaris
[23:19:58] *** tulcod has left #opensolaris
[23:20:42] <zooko> I have ZFS root on my Nexenta on Athlon64.
[23:20:51] <zooko> By the way, I am a total newbie to Solaris, but I really want this:
[23:21:05] <zooko> http://blogs.sun.com/levon/entry/python_and_dtrace_in_build
[23:21:30] <zooko> Does the phrase "build 65" mean that I have to install a new kernel or a new "ON" or something in order to get dtrace to report on Python stacks for me?
[23:21:51] <Triskelios> b65 means SXCE and SUNWPython
[23:22:04] <Triskelios> Python is not part of ON afaik
[23:22:48] *** rmorse has joined #opensolaris
[23:23:05] <jbk> however, i'm guessing the dtrace bits (which is part of ON) that allow for the integration require b65 or later
[23:24:18] <jbk> zooko: unlike linux, with solaris the kernel + core userland bits (i.e. libc + a some other stuff) are part of ON and generally upgraded as a whole
[23:24:23] <Triskelios> jbk: I don't think any ON changes were involved, it was probably a patch adding DTrace probes
[23:25:11] <jbk> well i suspect the python stack frame support w/ dtrace requires it
[23:25:44] <jbk> if you just want arbitrary interpeter probe points, that probably doesn't require b65
[23:26:31] *** dosiu has quit IRC
[23:27:44] <zooko> Thanks for the answer.
[23:27:46] <tek-ops> I think I'm going to hold off on that for a while
[23:28:15] <tek-ops> but I'm cureious, would LU just replicate the zfs on the same zpool?
[23:28:41] *** dosiu has joined #opensolaris
[23:30:05] *** _alek_ has joined #opensolaris
[23:31:22] *** SirFunk has quit IRC
[23:32:10] *** SirFunk has joined #opensolaris
[23:32:21] <zooko> Okay, the open source project allmydata.org passes all unit tests on Solaris now.
[23:32:54] <zooko> The problem turned out to be a flaw in the unit test code, which flaw was obscured by the fact that on all the other operating systems that we test on, listing a directory always happened to return one subdirectory before another.
[23:33:16] <zooko> Whereas on ZFS, apparently, it sometimes returned the other one, which revealed the flawed assumption in the unit test code that there was only one kind of subdirectory present.
[23:33:23] <zooko> http://allmydata.org/buildbot/
[23:33:39] <zooko> ^-- unfortunately no buildslave for Solaris yet, so unit tests are not run automatically for Solaris
[23:34:07] *** jwk404 has joined #opensolaris
[23:39:04] *** rmorse has quit IRC
[23:40:09] <Triskelios> tek-ops: I'm not sure what the status of ZFS support in LU is, but the idea is that will just work on a clone
[23:40:32] <tek-ops> ahhh, cool cool
[23:43:01] *** peemus has joined #opensolaris
[23:45:19] *** peemus has quit IRC
[23:48:05] *** dosiu has quit IRC
[23:52:47] *** dosiu has joined #opensolaris
[23:53:14] <jbk> well there's a bit more than just cloning /
[23:53:25] <jbk> though how much might depend what comes out of the packaging effort
[23:53:49] <jbk> as right now, the installer as well as LU has to know a lot of dirty details about differences between various versions of packages
[23:58:56] *** e^ipi has joined #opensolaris
[23:58:56] *** e^ipi_ has quit IRC
[23:59:52] *** sioraiocht has joined #opensolaris





top