[00:09:01] *** sparc-kly has joined #opensolaris
[00:09:55] *** shadeland has joined #opensolaris
[00:10:23] *** ChanServ sets mode: +o sparc-kly
[00:14:59] *** linescanner has joined #opensolaris
[00:16:50] *** deather_ is now known as deather
[00:17:15] *** linescanner has left #opensolaris
[00:17:18] *** laca has quit IRC
[00:17:53] *** sahafeez has quit IRC
[00:17:58] *** kumamoto has left #opensolaris
[00:17:59] *** laca has joined #opensolaris
[00:18:06] <agliodbs> cool, Joyent is doing the next SVOSUG
[00:18:30] <Tpenta> nice
[00:18:42] <Tpenta> I really must start on the next SOSUG, it's just been so damned busy
[00:18:46] *** Gman is now known as GmanAFK
[00:19:03] *** Mazon is now known as mazon
[00:19:28] <agliodbs> Tpenta: well, next up is Thursday
[00:19:33] <agliodbs> join us! be one of us!
[00:19:45] <agliodbs> or is SOSUG somewhere else?
[00:19:45] * Tpenta is in Australia
[00:19:46] <agliodbs> ah
[00:19:51] <agliodbs> S.O. ==?
[00:19:54] <Tpenta> SOSUG is Sydney
[00:20:03] <agliodbs> got it
[00:20:13] <agliodbs> Tpenta: there's a huge PostgreSQL user group there
[00:20:20] <agliodbs> you could do a combined meeting
[00:20:22] <Tpenta> hmmmm
[00:20:30] <agliodbs> the lead for it is on #postgresql right now (swm)
[00:20:35] <Tpenta> I know we had someone in Sydney actually working on the dtrace probes into postgres
[00:20:57] * delewis wonders when Oracle plans on doing something with DTrace
[00:21:32] <Tpenta> I really hop so, Id like to be able to do probes like "oracle:::query-start,oracle:::query-end"
[00:21:38] <Tpenta> for timestamps alone
[00:21:44] <delewis> Tpenta, precisely :-)
[00:21:50] <delewis> that's the *exact* thing I've been wanting to do.
[00:22:02] <delewis> but unfortunately, I haven't really read a lot about Oracle working on such probes.
[00:22:19] <twincest> mysql was bundled way before postgres, where are my mysql probes? :(
[00:22:39] <Tpenta> I believe that tere is a push to get brendan's javascript ones into the firefox we distribute
[00:22:48] <delewis> MySQL was bundled in a different consolidation than Postgres, though, twincest.
[00:22:56] <delewis> /usr/sfw vs. /usr
[00:22:59] <quasi> delewis: oracle probably prefer you using their tuning tools
[00:23:10] * dclarke sips coffee
[00:23:12] <delewis> sort of a difference there on priority and interests :-)
[00:23:34] <delewis> quasi, *what* tuning tools? :-)
[00:23:37] <Tpenta> itr might also be nice to use the query-start and query-end probes to do things like answer "what system calls are made as part of this query?"
[00:23:48] <Tpenta> actually statspack and it's succesor are pretty damned useful
[00:24:01] <delewis> querying some view for arbitrary information isn't exactly preferred to writing a DTrace script to use some probes that an Oracle provider provides :-)
[00:24:47] <delewis> actually, if Oracle did a DTrace provider correctly, it would put some companies that provide Oracle debugging/tuning applications out of business.
[00:25:08] <delewis> there's on particular application (that cost bundles, which I can't seem to recall the  name of at the moment) that is used for debugging Oracle Net sessions.
[00:25:12] <delewis> s/on/a/
[00:26:03] <jteo> mmm
[00:26:13] <Tpenta> there is also the issue about having to provide different versions of Oracle for 10 and pre 10 (although I seem to recall a PSARC case to provide a way to have the same binaries)
[00:28:12] *** compukid has quit IRC
[00:29:00] <boyd> Does anyone know how I can convince an XVR-100 to display out of DVI port instead of the VGA port?
[00:31:00] *** tsoome1 has joined #opensolaris
[00:32:15] <elektronkind> Tpenta: what, adding some wierd ld.so dtrace stub mojo in pre-10, I guess?
[00:33:06] <elektronkind> oracle as a software distribution is on crack anyway.
[00:33:39] <elektronkind> its way of doing things is just completely off the wall by today's standards.
[00:34:11] <delewis> elektronkind, can't say I disagree, but it does work.
[00:34:29] <delewis> I've done enough installs on a wide variety systems where I know the install procedures backwards and forwards. :-(
[00:34:36] <elektronkind> it's the whole relinking thing that baffles me
[00:34:41] *** uncertainty has quit IRC
[00:35:22] <elektronkind> I'm no DBA, but I have been known to install 10g and execute 'startup' and 'shutdown immediate' on TV
[00:35:27] <jteo> elektronkind: oh that. it's bizarre.
[00:35:52] <delewis> DB2 is pretty funky, as well, but it doesn't re-link at install-time :-)
[00:36:28] <jteo> oracle is a dinosaur. a fast and expensive one, but a dinosaur nonetheless.
[00:36:33] *** ibb has quit IRC
[00:36:43] <delewis> dinosaur as far as installers go anyway.
[00:36:55] <delewis> its features are well, quite impressive.
[00:37:53] <elektronkind> right now I'm occupied by trying to figure out how to get one of my department's oracle dbs, which is 8.0.5 running on 2.6 onto our new server, which is 10.2 running on solaris 10.
[00:37:55] <delewis> the spatial and analytical capabilities are extremely nice.
[00:38:00] <jteo> yes. provided you know the magic incantations to tune it.
[00:38:16] <delewis> jteo, I wouldn't quite call it magical.
[00:38:21] <delewis> there's a tuning and a concepts guide.
[00:38:30] <agliodbs> Tpenta: well, one
[00:38:33] <agliodbs> sorry
[00:38:33] <delewis> lots of material, yes.
[00:38:36] <delewis> magic, no. :-)
[00:38:43] <jteo> delewis: :)
[00:41:37] <jteo> intriguing, this alliance.
[00:42:06] <delewis> I wonder what AMD thinks
[00:42:37] <quasi> delewis: they are too busy cashing in on their stock options while they can ;)
[00:42:40] *** mlh has quit IRC
[00:43:00] <delewis> quasi, that and/or wishing they hadn't partnered with Sun.
[00:43:20] <delewis> I'm just hoping Sun doesn't take the Itanium2 route.
[00:43:21] <Tpenta> read the announcement, we are not ditching AMD
[00:43:37] <quasi> delewis: I think it still worked well for amd
[00:44:05] <delewis> Tpenta, no, but you are creating a line of servers during the first half of this year using Intel processors. How large that line will be, I'm not certain.
[00:44:08] <elektronkind> Dear everyone: AMD has not been voted off the island. Thanks, Love Sun
[00:44:28] <quasi> delewis: I sure don't hope that either, but jonathans comments about very large intel boxes are a bit scary
[00:45:04] <elektronkind> I'm sure we'll see things like X8000 intel blades
[00:45:27] <elektronkind> and I bet a few 1u boxes of various socket counts
[00:45:27] <UnixTitan> while amd has not been voted off the island.. the new partnership is a step in that direction.
[00:45:30] <UnixTitan> and that sucks.
[00:45:36] <elektronkind> not necessarily
[00:45:43] <quasi> elektronkind: I don't think they have either - most of all, I think Sun sees this as a way to push solaris, get more hw support and generate a bit of positive press
[00:45:54] *** edwardocallaghan has joined #opensolaris
[00:45:59] <edwardocallaghan> Hi all
[00:46:02] <elektronkind> this doesn't have to be a football game where one side completely wins and the other completely loses.
[00:46:23] <UnixTitan> elektronkind: actually, its more competitive than you think
[00:46:33] <elektronkind> so many people think this is a all or nothing world
[00:46:35] <edwardocallaghan> Some deals going down here : http://search.ebay.co.uk/_W0QQsassZmrwho0QQhtZ-1
[00:46:46] <delewis> elektronkind, true, but it doesn't make any direction very apparent.
[00:46:53] <edwardocallaghan> elektronkind:Hi
[00:47:02] <edwardocallaghan> I really like your music man !
[00:47:02] <elektronkind> hello edward
[00:47:06] <elektronkind> thanks :)
[00:47:09] <delewis> offering two server lines from vendors that produce processors of the same architecture isn't very sensical.
[00:47:28] <quasi> another thing is that AMD is currently behind on technology / performance - Sun would be silly not to approach intel and intel are probably desperate enough for positive press that they will do anything ;)
[00:47:53] <UnixTitan> delewis: not sensical at all.
[00:48:07] *** comay has quit IRC
[00:48:14] <delewis> Fujitsu and sun4u were at least somewhat distinctive in their applications.
[00:48:21] <UnixTitan> quasi: not really.. intel is getting tons of positive press out of the apple deal.
[00:48:30] <edwardocallaghan> I love that Dreamscape mix
[00:48:37] *** tsoome has quit IRC
[00:49:05] <quasi> UnixTitan: that's sort of played out and is old news to wall street
[00:49:21] <edwardocallaghan> What is up with the deal with Intel ?!
[00:49:37] <UnixTitan> amd rocks.
[00:49:51] <edwardocallaghan> I think bad turn  !
[00:50:17] <edwardocallaghan> Sun have been doing everything right so far but this is a mistake !
[00:50:20] <jmcp> why?
[00:50:21] <delewis> personally, I don't what x86 vendor Sun chooses to go with. I understand the need to offer x86-based product lines, but consistency is good in your product line.
[00:50:22] <elektronkind> like jonathan said in his blog, I think Sun is getting more out of this, though. Intel's hawking Solaris now. That's a complete 180 from before. Intel still provides a lot more than microprocessors, and now it's pretty much a guarantee that Solaris can support these other chips far sooner than we used to have hoped.
[00:50:40] <delewis> customers are going to wonder "Should I choose the AMD64 gear?" or "Should I choose the Intel gear?"
[00:50:58] <delewis> and I pity the sales rep that has to inform the customer of the various advantages and disadvantages between the two.
[00:50:59] <quasi> elektronkind: yeah, I think so too
[00:51:08] <delewis> care*
[00:51:23] <jteo> buy whatever you want. but buy it from Sun? ;)
[00:51:30] <delewis> jteo, indeed. :-)
[00:51:48] <Tpenta> There are a lot of customers out there who we can't sell to because of a customer loyalty to Intel, so now we offer intel based boxes too
[00:52:07] <Stric> elektronkind: a funny thing.. I saw an intel 10GbE card at a webshop.. then the same card at sun.. at half the price ;)
[00:52:08] <edwardocallaghan> They are not going to do Xeon workstations are they :p ?
[00:52:13] <Stric> (10GbE cards aren't cheap)
[00:52:14] <elektronkind> delewis: that can confuse the uninitiated, yes, but it's a situation that can be relatively easily addressed, because before today the situation was: <customer> I need/want Intel systems, but Sun only has AMD. I'll look at Dell/IBM/HP instead"
[00:52:34] <elektronkind> Sun's not the only one playing Intel/AMD combo.
[00:52:53] <quasi> also, with the nvidia chipsets and the latest drops in power use for woodcrust, I don't think there's going to be much extra work for sun in providing an intel variant
[00:52:57] <elektronkind> Dell, IBM, and HP have been for a bit now (Dell only recently coming into the AMD fold)
[00:53:04] <jmcp> edwardocallaghan: yes, there will be xeon-based workstations
[00:53:04] *** lasseoe has quit IRC
[00:53:05] *** lasseoe has joined #opensolaris
[00:53:08] *** peteh has quit IRC
[00:53:09] *** darrenm has quit IRC
[00:53:14] <edwardocallaghan> Oh good
[00:53:28] <delewis> elektronkind, that's true, and obviously Sun has been succesful with marketing two competing product lines before (Fujitsu and sun4u).
[00:53:34] * quasi would prefer a core 2 duo based workstation
[00:53:39] <edwardocallaghan> Then I can do my BBQ's on my Workstation in Australia
[00:53:40] <delewis> I just hope it turns out positively.
[00:53:43] <Tpenta> http://www.youtube.com/watch?v=DgKtXKDhsl4&eurl= - ROFL (completely off topic)
[00:54:07] *** peteh has joined #opensolaris
[00:54:11] <jmcp> quasi: you + me + rest of the world
[00:54:26] * Tpenta thought it was earnings today instead of tomorrow, and I saw this email from Jonathan to everyone with te subject "Congratulartions Everyone", just hope he can use the same subject tomorrow
[00:54:29] *** uebayasi has joined #opensolaris
[00:54:32] <elektronkind> now if only Intel would get over it and adopt HyperTransport.
[00:55:20] <Tpenta> who isto say that there are not patent licensing issues to address with hypertransport (and with how intel has treaded AMD in the past, I suspect it would not be cheap)
[00:55:37] <elektronkind> isnt HT a open spec?
[00:55:43] <Tpenta> no idea
[00:55:46] <elektronkind> I might be wrong, but I thought it was
[00:56:22] <edwardocallaghan> When are proper RISC CPU's going to come back !?
[00:56:23] <sommerfeld> I suspect a state of MAD-like deterrence exists between AMD and Intel on patents.  If either side starts patent lawsuits, only the lawyers win...
[00:56:39] <edwardocallaghan> I hate these f**king x86 crap
[00:56:58] <delewis> edwardocallaghan, modern x86 CPUs are pretty much RISC, nowadays.
[00:57:07] <delewis> with CISC instructions implemented in microcode.
[00:57:38] <quasi> sommerfeld: yeah, that could get really ugly
[00:57:50] <edwardocallaghan> I know
[00:58:09] <edwardocallaghan> But I want to see more SPARC
[00:58:14] <delewis> and you will.
[00:58:23] <edwardocallaghan> I look forward to Rock coming soon
[00:58:59] *** hspaans has quit IRC
[00:59:05] <sommerfeld> from hallway conversation: so, if the future of SPARC is Rock, is AMD Scissors and Intel Paper, or is it the other way around?
[00:59:15] * quasi just wants an US IV+ workstation that's not wildly expensive ;)
[00:59:26] <delewis> quasi, dream-on. :-)
[00:59:33] <elektronkind> sommerfeld: :)
[00:59:58] <Plaidrab> So long as they don't announce the tapeout while playing Stone Free
[01:00:08] <quasi> delewis: yeah, that sounds like a good plan (with it being 1am now ;)
[01:00:08] *** comay has joined #opensolaris
[01:00:12] <elektronkind> I'm just happy to think that we can see more intel boxes with "Solaris Inside" logos on them.
[01:01:14] * delewis is still satisified with his UltraSPARC-III gear.
[01:01:20] <elektronkind> of course, my ultimate bet is that, in fantasy OS land, "MacOS X 11" (ha!) will have a sunos kernel.
[01:01:21] <edwardocallaghan> I think it's a mistake by Sun
[01:01:48] <quasi> delewis: IV would be fine, IV+ is probably overkill for my desktop
[01:02:13] <edwardocallaghan> How would a Xeon Desktop compare to a IV+ Desktop ?
[01:02:22] <delewis> the only CPU-intensive workload on my workstation is usually video encoding/decoding.
[01:02:46] <delewis> and I suspect performance would be a lot better if I got around (someday) to porting MPlayer to Sun Studio, but alas.
[01:03:21] <elektronkind> doesn't mplayer use ffmpeg libs?
[01:03:46] <delewis> elektronkind, yes.
[01:03:57] <elektronkind> because I'm pretty certain that I've gotten ffmpeg to compile under sun studio 11
[01:04:17] <delewis> elektronkind, not without major hacking.
[01:04:26] <delewis> steleman has done it
[01:04:44] <delewis> a lot of the inline assembly code (don't ask) depends on how gcc packs structures.
[01:04:53] *** Vratha has joined #opensolaris
[01:04:58] <Vratha> hi
[01:05:12] <alanc> hmm, don't see gisburn or nrubsig here - his dbx bug just came across triage list
[01:05:23] <g4lt-mordant> so ffmpeg ofr sparc is right out then? ;P
[01:05:24] <alanc> CR 6516145 Created P2 opensolaris/triage-queue Please update OS/Net Sun Studio 11 tarball to include the fix for dbx
[01:05:33] <Tpenta> wow, roland is not here looking for bill, that's unusual
[01:05:40] <Vratha> i've got a quick couple questions since i've never used opensolaris before but am thinking about trying it in a VM before installing on a real machine... does it have good hardware support, and does it have a good directory server that windows clients and OS X clients can use?
[01:06:16] <alanc> I thought mike was his favorite stalking target of late
[01:06:30] <Tpenta> apparantly I get asked for when I'm not here too
[01:06:35] <jmcp> Vratha: sun.com/bigadmin/hcl for the hardware compatibility list
[01:07:01] <jmcp> Vratha: as far as I'm aware the supplied Sun Directory Server (or whatever they're calling it today) works just fine with windows and osx
[01:07:12] <Vratha> sweet!
[01:07:19] <Vratha> well i haven't installed opensolaris yet.. i'm downloading it now :)
[01:07:20] *** wraith has joined #opensolaris
[01:07:51] <wraith> Is it posible to run windows inside solaris using qemu or smth else?
[01:08:03] <Vratha> jmcp: but anything compatible with the commercial solaris is compatible with opensolaris?  or are both projects the same now, except one has commercial support options?
[01:09:24] <Auralis> wraith: yes
[01:09:24] <quasi> Vratha: most likely opensolaris has more hardware support than Sol10
[01:09:55] <Vratha> Auralis: you do VMs on solaris?  is there anything like vmware for it?
[01:10:21] <Vratha> quasi: yeah, well i guess that is to be expected since solaris is probably only setup to run on Sun hardware
[01:10:48] <Auralis> qemu for now, a special win4solaris is upcoming, Xen is in the works, and you have solaris zones
[01:10:55] <Vratha> hmm
[01:11:14] <Vratha> well i may be running solaris in vmware when i build my big server in the future, since i want it to manage VMs
[01:11:52] <wraith> so,it is posible to run windows using qemu,yes?
[01:12:03] <Auralis> wraith: yes
[01:12:10] <quasi> Vratha: no, not so
[01:12:15] <wraith> thanks
[01:12:39] <Vratha> quasi: hmm?  to what part of what i said were you referring?  i've said a few things since then
[01:12:41] <edwardocallaghan> You only should use qemu if you are using a SPARC
[01:12:42] <quasi> Vratha: but opensolaris gets to run much of the code that makes it into Sol10
[01:12:43] <alanc> Vratha: solaris x86 supports mostly non-Sun hardware in fact
[01:12:55] <Vratha> oh, nice
[01:13:04] <alanc> since for x86, there's a lot more non-Sun than Sun out there
[01:13:13] <Vratha> yeah
[01:13:40] <quasi> Vratha: so, in many ways opensolaris is just a testing ground for things that become mainline in the long term
[01:13:52] <alanc> hmm...CR 6516146 Created P2 opensolaris/triage-queue Bad kernel fault at addr=0xdeadbf15
[01:13:53] <Vratha> well my main reason for coming to opensolaris is to help manage directory servers that are compatible with windows clients without having to spend $1K on windows server (which i wouldnt do... i'd just use pgina if i had too)
[01:14:02] <alanc> 0xdeadbeef + 16?
[01:14:05] <Vratha> quasi: yeah, sounds a lot like fedora in that sense
[01:14:17] <quasi> Vratha: ick, not quite
[01:14:50] <edwardocallaghan> Fedora try to make out that its stable
[01:15:09] <Vratha> yeah, which it most certainly isn't
[01:15:10] <edwardocallaghan> It only becomes 'stable' after half a year of fixes
[01:15:46] <Vratha> so there any performance benchmarks between opensolaris and linux?
[01:15:55] <quasi> Vratha: a few
[01:16:18] <Tpenta> alanc, sounds like someone doing something bad with a freed pointer
[01:16:23] <quasi> Vratha: but it is very hard, because sun often opts for stability over performance
[01:16:32] <edwardocallaghan> Linux is good for small servers and when your learning but for real work you need Solaris
[01:16:48] <Vratha> yeah
[01:16:57] <Vratha> nah, for real work you can use linux
[01:17:03] <Vratha> large investment banks use it all the tie
[01:17:05] <Vratha> time*
[01:17:06] <edwardocallaghan> The big thing for Solaris is that it scales
[01:17:11] <quasi> Vratha: http://blogs.sun.com/bmseer/ has done quite a lot of benchmarking
[01:17:18] <alanc> Tpenta: probably, but it's a kernel bug, so I leave it to someone like comay to triage
[01:17:21] <sommerfeld> alanc: indeed.  typical when you pass an uninitialized (struct foo *) to someone
[01:17:21] <delewis> Vratha, kind of hard to use Linux for your storage infrastructure.
[01:17:29] <sommerfeld> fault on whatever offset you first touch
[01:17:40] <Vratha> delewis: yes, that i'd agree with.. NFSv4 support blows there right now
[01:17:57] <delewis> Vratha, lack of decent volume management is another.
[01:17:58] <delewis> and lack of a consistent device naming convention is yet another.
[01:18:03] <edwardocallaghan> Linux needs a lot of admins to press reboot after a kernel panic ;0
[01:18:04] <Vratha> yes
[01:18:13] <Vratha> but linux does have computing performance on its side
[01:18:19] <sommerfeld> ndisapi:fake_intrhand+47
[01:18:20] <sommerfeld> d'oh
[01:18:33] <Tpenta> I'd be interested in seeing the crashdump
[01:18:54] <mv> i got this error while building current on Failed to create generic usr archive:   350830 blocks cpiotranslate: usr/lib/fs/xmemfs: no packaging info
[01:18:59] *** pikapika has quit IRC
[01:19:04] <mv> can i ignore it, or is it important
[01:19:11] <Vratha> delewis: does the opensolaris/solaris storage infrastructure even support windows clients properly?  because, as you probably know, samba blows
[01:19:12] <edwardocallaghan> Well my Fedora 6 install is slower then my Solaris install
[01:19:17] <alanc> submitter's email address is in the bug report
[01:19:26] <delewis> Vratha, that's not what  storage infrastructure means.
[01:19:31] <Vratha> edwardocallaghan: fedora 6 was slower than my ubuntu install... they did something bad in that release
[01:19:31] <delewis> and yes, Solaris supports Samba.
[01:19:38] <edwardocallaghan> http://fedoraproject.org/wiki/EdwardOcallaghan no bias
[01:19:47] <Vratha> delewis: you are talking about SAN/NAS, right?
[01:19:54] <sommerfeld> mv: xmemfs was just nuked from ON
[01:19:56] *** UnixTitan has quit IRC
[01:20:00] <delewis> Vratha, correct.
[01:20:13] <Tpenta> ndisapi?
[01:20:18] <mv> sommerfeld: thx, so i just close my eyes :)
[01:20:21] <Vratha> ok, well i thought what you said also included the area of supporting various file systems; sorry
[01:20:25] <edwardocallaghan> RPM's are getting a much needed revise soon
[01:20:38] <sommerfeld> mv: for best results, do a clobber build and start over; it's probably a leftover from before the removal
[01:20:41] <delewis> Vratha, no, I meant volume management, multipathing support, etc.
[01:20:46] <Tpenta> WAG, someone is using a 64 bit windows driver that plays with floating point regs
[01:20:51] <delewis> multipathing on Linux is a PITA.
[01:21:04] <delewis> especially give you don't have a consistent device naming convention on top of it all.
[01:21:09] <sommerfeld> Tpenta: actually my guess would be a NDIS driver which assumes it gets zero-filled memory from the OS
[01:21:09] <delewis> s/give/given/
[01:21:10] <Vratha> delewis: do the tools that come with opensolaris make it easy to get samba working with its directory server and all that?  i've used linux enough to know that i don't want to dick around with something that should be "trivial" for several hours
[01:21:23] <Tpenta> anyone asked for the crashdump?
[01:21:24] <Auralis> what good are doing 20 filesystems, when each either craps out under load, corrups silently data or goes bonkers at the first sign of trouble?
[01:21:25] <sommerfeld> Tpenta: so maybe the NDIS wrapper should be calling kmem_zalloc ..
[01:21:25] <delewis> Vratha, Sambs is the same as it is on Linux.
[01:21:30] <Vratha> :(
[01:21:31] <delewis> Samba, rather.
[01:21:38] <Vratha> well i suppose there is always NFS
[01:21:41] <Tpenta> I can ask for it then put it up on mdb.sfbay
[01:22:00] <wraith> Auralis, and what I need to launch windows?
[01:22:06] <Vratha> just that that kinda sucks because then you don't have the security unless you pay for an NFSv4 windows client
[01:22:09] <wraith> iso image?
[01:22:36] <edwardocallaghan> be back
[01:22:45] <edwardocallaghan> watching sun video :p ...
[01:23:22] <Auralis> wraith: in the optimal case, you wait for win4sol and a windows cd/dvd
[01:24:00] <Tpenta> mail sent
[01:24:25] <wraith> Auralis: but I need it now)))
[01:24:28] <Vratha> quasi: thanks for the link; hehe, it's hard to believe all of it since it's posted on sun.com though :)
[01:25:21] <Auralis> wraith: then grap qemu from blastwave.org/testing
[01:25:26] * tsoome1 is away - bbl - [AcidJazz - Log ON]
[01:25:43] *** compukid has joined #opensolaris
[01:26:18] <wraith> I already done it, now I don't know what to do next))
[01:26:27] <Vratha> delewis: you run any linux VMs under solaris?
[01:26:53] <Auralis> wraith: i would start with reading the docs
[01:27:45] <Tpenta> alanc: also updated the bug to say I've asked so the poor guy does not get inundated with requests
[01:27:56] * Auralis goes back to her movie
[01:31:00] <wraith> ok thx))
[01:31:03] *** wraith has quit IRC
[01:31:20] <Vratha> do any of you run linux VMs under opensolaris or solaris?  i'm just wondering how well they run
[01:32:55] <g4lt-mordant> brandz FTW
[01:36:21] <edwardocallaghan> BrandZ should be easy to setup but have not done it yet
[01:37:23] *** Jiko_ is now known as Jiko
[01:38:46] <Vratha> nice
[01:39:00] *** Kmays has joined #opensolaris
[01:39:12] <Vratha> but does it take advantage of intel's VT-x? :)
[01:39:21] <delewis> Vratha, better.
[01:39:28] <delewis> it just remaps Linux system calls to Solaris system calls.
[01:39:36] <Vratha> haha
[01:39:48] <Stric> a kind of paravirtualization
[01:39:53] <Vratha> yeah
[01:40:10] <Stric> Vratha: vt-x is kinda slow.. a "context switch" takes many many cycles
[01:40:23] <Vratha> i'd kinda like to see a VT-x capable VM thing for opensol.
[01:40:30] <Vratha> yeah
[01:40:53] <Vratha> but the nice thing is it's pretty transparent and doesn't require mods to the guest OS
[01:40:53] <edwardocallaghan> Intel like to do everything with cycles ... American mussel
[01:41:09] <Vratha> actually, does brandz require mods to the guest OS?
[01:41:23] <edwardocallaghan> no no
[01:42:02] <Vratha> edwardocallaghan: oh, then how does it remap the calls?
[01:42:16] <Stric> it doesn't run linux
[01:42:19] <Stric> it just pretends to
[01:42:58] <Vratha> you mean i wouldn't actually run a linux kernel but would just run all the normal linux tools in a zone?
[01:43:09] <Vratha> and when the tools make a syscall they are remapped to solaris calls?
[01:43:15] <Stric> which believes it's running under linux
[01:43:26] <Vratha> ah, makes sense
[01:43:38] <Vratha> best of luck to them with keeping up with all the changes that go on in linux
[01:43:39] <edwardocallaghan> yes that's what it does
[01:43:44] *** nostoi has quit IRC
[01:43:53] *** jsubl2 has joined #opensolaris
[01:43:56] <jmcp> Vratha: only certain kernel and glibc versions are supported
[01:44:05] <Vratha> :-/
[01:45:00] <Vratha> well i'll probably stick with my original plan of putting opensolaris in a VM itself then, hehe
[01:45:22] <jmcp> Vratha: what did you expect? given the attitude of core linux kernel people (gkh!) towards stable in-kernel interfaces, it's easier from an engineering standpoint to pick a version and support that
[01:45:29] *** jacotton has quit IRC
[01:45:38] <Stric> hm.. all this talking got me to try the hw virtualization that just got into linux..
[01:45:49] <Stric> booting the ubuntu edgy installer right now using hw virt
[01:45:59] <Vratha> jmcp: yeah, but it's even easier to just use VT-x :)
[01:46:18] <delewis> Vratha, easier != optimal.
[01:46:22] <jmcp> Vratha: which when brandZ was being developed wasn't actually a reality in production quantities
[01:46:29] <Vratha> delewis: i didn't say it was
[01:46:29] <jmcp> besides, not everybody has vt-x
[01:46:30] <delewis> as Stric said, context switches are very, very expensive.
[01:46:55] <dlg> brandz has advantages over hardware virtualisation
[01:47:05] <delewis> dlg, indeed.
[01:47:07] <Vratha> i know it does
[01:47:10] <delewis> very granular resource control
[01:47:19] <Vratha> but it's not really what i want
[01:47:20] <delewis> and least-privilege support
[01:47:35] <dlg> well, it lets you use linux apps without having to use linux
[01:47:40] <dlg> that sounds great to me
[01:47:44] <Vratha> yeah
[01:47:51] <Vratha> can't argue with that i guess
[01:48:03] <Vratha> opensolaris keeps its own repo of linux apps?
[01:48:28] <edwardocallaghan> Blastware
[01:48:31] <Vratha> anyway, the projects listed on the opensolaris page look pretty spiffy; it looks like they're trying to modernize many parts of it
[01:49:07] <Vratha> are they intending to target the desktop someday too?  because that's what some of these projects look like to me
[01:49:22] *** nostoi has joined #opensolaris
[01:49:53] <Stric> This isn't final software, but hardware virtualization using kvm(.sourceforge.net) is slower than software virt
[01:50:09] <Vratha> yes, i know, but it's more convenient
[01:50:24] <Stric> I'm talking about qemu with or without hw-"accel"
[01:50:39] <Vratha> oh
[01:51:36] <Vratha> main reason i like true VMs using VT-x is because it's a cheap way for me to multiplex my system with any kind of OSes
[01:51:39] <Vratha> which is nice for home use
[01:52:14] <Vratha> because i don't have thousands of bucks to spend on systems at home :)  nor would i want a ton of them
[01:54:42] *** delewis has quit IRC
[01:55:07] <edwardocallaghan> I think more people should be SPARC desktops and put a PCI-x86 PC card in and run there other OS in there
[01:55:30] <Vratha> eh, i like my OS X laptop
[01:55:47] <Vratha> it's superfun yeahyeah rocketship :)
[01:55:56] *** GmanAFK is now known as Gman
[01:56:24] <edwardocallaghan> :p no more PPC
[01:56:31] <Vratha> intel
[01:57:23] <edwardocallaghan> I know :p
[01:57:30] <Vratha> :)
[01:57:31] <jbk> anyone know the name of that netmeeting-like program that comes in nevada?
[01:57:47] <salmandr> jbk: ekiga?
[01:57:50] <Vratha> opensolaris needs more working wifi drivers for 802.11g :)
[01:57:56] *** delewis has joined #opensolaris
[01:58:03] <jbk> that sounds right..
[01:58:04] <Vratha> so does linux too... freakin' card mfgs. not giving out specs
[01:58:19] <edwardocallaghan> Fedora is f**k now, the more updates I apply in the hope that it will fix the random application closing it just gets woes
[01:58:34] <Vratha> edwardocallaghan: that's fedora
[01:58:41] <jbk> anyone know if it does the desktop sharing like netmeeting or sunforum?
[01:58:57] <delewis> Vratha, it should be irrelevant whether or not a peripheral manufacturer chooses to give out specs or not.
[01:59:17] <delewis> that should be required is the operating system providing consistent interfaces that the vendor can use to write a driver.
[01:59:24] <Vratha> delewis: ok, but either way we need drivers
[01:59:27] <edwardocallaghan> I need to get everything off my disks and then I am a SPARC only man
[01:59:50] <alanc> vino lets you give control of your desktop to someone else - it's in the JDS menus as something like "Remote Desktop Access"
[01:59:57] <alanc> (it's VNC under the hood)
[02:00:04] *** peteh has quit IRC
[02:00:38] <jbk> hmm the problem is, netmeeting/windows messenger is our standard :(
[02:00:39] <Vratha> what i'd have to do to have wifi would be to buy a proprietary (well, it might actually have FOSS on it when purchased, but it's irrelevant) that can just be a bridge for a server NIC so that i can use the server as a WAP
[02:00:52] <Vratha> but that's annoying
[02:00:57] <jbk> it's used pretty much for desktop sharing (I'm looking mostly at viewing than hosting)
[02:00:58] *** peteh has joined #opensolaris
[02:01:20] <Gman> edwardocallaghan, did you report bugs with fedora? if you didn't, you only have yourself to blame
[02:01:49] <jbk> sunforum did it, but looks like that's gone (and sparc only)
[02:02:23] <edwardocallaghan> I am on the Fedora team !
[02:02:34] <stevel> that doesn't answer gman's question
[02:02:47] <edwardocallaghan> I can't find the problem applications just close them selfs
[02:03:16] <edwardocallaghan> I thing there's some kind of segment fault in a lib ?
[02:03:23] <edwardocallaghan> I have no clue
[02:03:28] <edwardocallaghan> Still looking
[02:04:47] <Vratha> i wonder if opensolaris will run on my new EPIA CN13000G that's on its way to me
[02:04:56] <Vratha> because that'd be a pimp 16W server
[02:06:47] *** sahafeez has joined #opensolaris
[02:07:26] <sahafeez> so when will sun be shipping intel based servers with EFI and not openfirmware?
[02:08:00] <delewis> sahafeez, they aren't shipping any non-SPARC systems with OpenFirmware at the moment.
[02:08:50] <sahafeez> ah, did not know that. what are they shipping with? BIOS? i have not played with any non-sparc hardware
[02:09:09] <dlg> ugh, efi
[02:09:10] <dlg> ofw > *
[02:09:11] <delewis> sahafeez, yes, with a management-card add-on for the servers.
[02:09:12] <alanc> have to ship with BIOS on x86, or no OS'es will run
[02:09:47] <sahafeez> so sun x86 stuff has BIOS?
[02:09:53] <alanc> absolutely
[02:09:58] <delewis> alanc, there is a BIOS compatibility mode with EFI.
[02:10:05] <sahafeez> Ugh!
[02:10:36] <sahafeez> OFW was the best. i was quite pissed when apple moved to EFI
[02:11:57] <sahafeez> but on the hole, the EFI is not that bad on the mac and i can still do target mode, etc..
[02:12:05] <sahafeez> s/hole/whole
[02:12:34] <jcsmith> if i have two executables, say sun ssh and open ssh both installed in my path, one in /bin, the other in /usr/local/bin is there a better way to have typing ssh default to /usr/local/bin ssh other than alias ssh="/usr/local/bin/ssh"?
[02:12:47] <twincest> i said earlier that intel systems with nice firmware would be nice, but the market isn't there
[02:13:00] <dlg> jcsmith: put the openssh path first
[02:13:05] <twincest> who wants servers that can only run solaris when you already have sparc for that
[02:13:06] <dlg> jcsmith: remove sunssh
[02:13:08] <sahafeez> hum, well they could just use EFI like apple and be done with it.
[02:13:17] <g4lt-mordant> jcsmith, path checks sequentially
[02:13:36] <jcsmith> ehh, i'd like to remove sunssh, how ever sradmin says no and also, i'm not sure if i like the idea of putting /usr/local/bin before /bin in my path
[02:13:37] <darkcmd> isn't there an openfirmware complient bios for x86?
[02:13:40] <sahafeez> mv the sun ssh to ssh-sun
[02:13:47] <Vratha> unf.. solaris express community edition is downloading at 300+ KBps
[02:13:52] <sahafeez> darkcmd yes
[02:13:55] <Vratha> also, solaris express is what i want if i want opensolaris, right?
[02:14:14] <sahafeez> but the reason i brought it up was the sun/intel announcement.
[02:14:21] <sahafeez> so EFI would make more sense
[02:14:44] <twincest> EFI would make more sense because... Sun wants to become an Apple reseller? :)
[02:15:08] <sahafeez> no since the intel/sun deal that was just annouced
[02:15:34] <Vratha> sun is going to start releasing intel boxes?
[02:15:35] <Vratha> interesting
[02:15:38] <darkcmd> xeon sun machines?
[02:15:44] <sahafeez> yes
[02:15:48] <edwardocallaghan> Sun should use this Intel thing and build a x86 laptop range now
[02:15:59] <edwardocallaghan> With bloody serial ports !
[02:16:13] <sahafeez> darkcmd: http://www.eweek.com/article2/0,1895,2086212,00.asp
[02:16:20] *** jafari has joined #opensolaris
[02:16:24] <Vratha> maybe i can buy a cheap sun server when i finally have money... something with 1+TB of disk and 4 GB of RAM
[02:16:58] <stevel> bridge coming back online now'
[02:17:02] * sahafeez is just waiting for 10.5 with zfs for his xserve and looking to by an xserve raid
[02:17:04] <boyd> alanc: ping
[02:17:22] <alanc> boyd: pong
[02:17:27] <Vratha> yeah, but you have to buy client licesnse for xserve, right?
[02:17:29] *** gm152 has quit IRC
[02:17:34] <sahafeez> no
[02:17:38] <Vratha> oh
[02:17:40] *** gm152 has joined #opensolaris
[02:17:42] <sahafeez> unlimited
[02:17:50] <Vratha> eh, i heard it wasn't that great
[02:18:07] <Stric> xserve with just built-in disks is a joke
[02:18:16] <CIA-22> hueston: 6496337 Injecting MEM UE "kiuetl1" on a s10_u3_b9 based OPL FF2 Domain causes cpumem-diagnosis faulted.; 6499304 Need to update cpumem-retire agent for faults that go from SP to Solaris
[02:18:17] <boyd> alanc: Not sure if you'd be the one to ask... do you have any idea how to pass command like options to Xsun during a jumpstart install (specifically -dev)
[02:18:19] <Stric> since on macos (pre-zfs) you can't both mirror and partition a disk
[02:18:23] <Stric> macosx
[02:18:33] <alanc> boyd: that I don't know
[02:18:40] <sahafeez> hum, you can install os x to a mirror set
[02:18:55] <Stric> yeah. but not if you partition it
[02:18:56] *** sommerfeld has quit IRC
[02:18:57] <boyd> alanc: Fair enough... just thought I'd try
[02:18:58] <alanc> been years since I've actually done a jumpstart, and never needed to set Xsun options in it
[02:18:59] <movement> hah CIA is back, thanks stevel :)
[02:19:00] <Stric> and you can't partition a mirror
[02:19:22] <stevel> movement: np. and we didn't even get to test the anti-flood capabilities (since there were 20 changesets, it aggregated them all into one notification)
[02:19:27] <sahafeez> ok. everything sucks if it is not the tool you need for the job.
[02:19:45] <Stric> come on.. a server where you can't have more than 1 filesystem?
[02:20:00] <Stric> (it's not a server unless you mirror disks)
[02:20:10] <sahafeez> and the reason you need more then one?
[02:20:11] <movement> stevel: pity the log message doesn't aggregate all the bugs...
[02:20:18] <Stric> I want data and OS separated
[02:20:27] <Stric> because macosx needs a reinstall every now and then
[02:20:29] <sahafeez> i never have anything more then the os on the built in drives.
[02:20:42] <Stric> and then I want my disk images to stay
[02:20:48] <stevel> movement: yeah, that was a conscious decision.  you're welcome to file a bug and raise discussion on tools-discuss if you'd like the notification to aggregate all the comments
[02:20:53] *** sickness has quit IRC
[02:21:04] <movement> I'll see what others think
[02:21:29] <sahafeez> servers should be netbooted or have a mirror set. and data should be on another box (san, wan, etc)
[02:21:36] <Stric> sahafeez: apple sells them with 250 or 500GB drives.. just for the OS?
[02:21:37] <Gman> heh, looks like american express are using solaris
[02:21:50] <Gman> [judging from dtrace-discuss]
[02:21:51] <stevel> that was actually how i'd written up the initial hook - but i believe it was sch who suggested only having the most recent bug comment
[02:22:28] <darkcmd> XSun isn't open is it?
[02:22:33] <sahafeez> no
[02:22:52] <sahafeez> but that is because the sun graphics guys are evil and we all hate them
[02:22:58] <darkcmd> canlicensing issues?
[02:23:05] <darkcmd> licensing
[02:23:06] <darkcmd> I mean
[02:23:14] <sahafeez> no, it is the other reason.
[02:23:35] <Stric> licensing is one reason, yes
[02:23:39] <movement> stevel: makes that putback look like a pretty serious reworking just to fix some fma stuff :)
[02:23:44] <stevel> hehe
[02:23:51] <alanc> Xsun could be opened (except for two modules) if we cared, but no one cares enough to do the work since we're moving to Xorg
[02:24:09] <Stric> alanc: but no sparc drivers..
[02:24:14] <sahafeez> which leave everyone with one gen old sparc fucked
[02:24:14] <darkcmd> ah so the plan is just scrapiing xsun
[02:24:27] <alanc> but the SPARC drivers are separate from Xsun
[02:24:34] <alanc> and they're what people really want to see
[02:24:41] <Stric> in my world they belong together..
[02:24:44] *** nostoi has quit IRC
[02:25:06] <Stric> an X server without a driver doesn't do much and a driver without an X server doesn't either ;)
[02:25:11] <alanc> sahafeez: maybe, maybe not
[02:25:23] <sahafeez> unless you have a very new sun high end video card in you sparc pretty much figure you will not see support in xorg
[02:25:26] <alanc> my Xorg on SPARC test system is Ultra 10 running m64
[02:25:48] <Stric> alanc: is it less of a joke compared to when running with Xsun?
[02:25:53] <alanc> and Martin Bochning keeps promising the ffb driver for Xorg he did for Martux
[02:25:54] <Stric> that's just slower than .. hm.. something..
[02:26:06] <Stric> ffb being creator3d etc?
[02:26:08] <sahafeez> the m64 is supported under linux, IIRC, so it will work. let me know when the release the specs and drivers for an xvr-1000
[02:26:19] <alanc> oh, on m64 it's slower than Xsun, since there's no acceleration in the current driver
[02:26:33] <Stric> :P
[02:26:38] <alanc> sahafeez: given the SPARC group attitude, probably never
[02:26:51] <alanc> Stric: yes, ffb is Creator3D
[02:26:56] <sahafeez> thus my 1st point about evil and hate ;)
[02:27:48] <darkcmd> scrapingso with xorg how are we going to get x on sparc
[02:28:34] <Stric> alanc: any progress on sunray->xorg?
[02:28:36] <darkcmd> man mman my client is all messed up
[02:28:45] <Stric> darkcmd: we notice :)
[02:28:57] <alanc> Stric: no news to report - they've still got it a bit down the todo list
[02:29:06] <Stric> ok
[02:29:17] *** mlh has joined #opensolaris
[02:30:44] <alanc> darkcmd: all I can say is the Solaris group will be delivering Xorg on SPARC and the drivers are up to the SPARC group
[02:30:58] <alanc> if they don't care about SPARC workstations, that's their choice
[02:31:07] <sahafeez> so sometime after bill gates buys a iMac
[02:31:10] <alanc> otherwise, they'll deliver
[02:32:04] <edwardocallaghan> boy:Hi mate
[02:32:19] <edwardocallaghan> %s/boy/boyd/
[02:32:22] <alanc> though, since they care about new sales much more than EOL machines, old hardware is probably stuck waiting for community drivers anyway
[02:32:27] <boyd> Hey
[02:32:45] <edwardocallaghan> How's things /
[02:32:46] <edwardocallaghan> ?
[02:32:54] <boyd> Not too bad... Busy
[02:33:01] *** schily___ has joined #opensolaris
[02:35:11] <edwardocallaghan> Same
[02:35:26] <edwardocallaghan> What's the deal with this Neptune thing in the new SPARC T2 ?
[02:36:04] <alanc> "We have putback the following into Solaris Nevada ON gate in build 56. Webcams compliant to USB video class spec are supported on Solaris."
[02:36:43] <Stric> edwardocallaghan: fast network
[02:38:06] <g4lt-mordant> alanc, one would hope so, since apparently the only use Sun has for USB/firewire is to hook webcams to it :(
[02:38:59] <alanc> you mean webcams, disks, mice, random other storage devices, keyboards, audio
[02:39:04] <alanc> and probably a few more
[02:39:14] <edwardocallaghan> Oh
[02:39:15] *** jafari has quit IRC
[02:39:20] <Stric> usb cup warmers! :)
[02:39:56] <alanc> yeah - and iPod chargers, which were the only use besides webcams for FireWire in Solaris 8 & 9
[02:40:03] *** salmandr has quit IRC
[02:40:12] <alanc> but then other things were added
[02:40:31] *** yongsun has joined #opensolaris
[02:42:37] *** CryptoDivinity has quit IRC
[02:42:38] <Gman> ooh
[02:42:41] <Gman> wonder if that includes my webcam
[02:43:10] *** simford has joined #opensolaris
[02:43:25] <edwardocallaghan> Will the T1 chip get any cheaper ?
[02:43:35] <edwardocallaghan> and still be sold ?
[02:43:59] <elektronkind> geez technorati is so screwed up
[02:44:00] <elektronkind> There are no blog posts tagged linux.
[02:44:05] <elektronkind> yeah, for sure
[02:44:48] <edwardocallaghan> That would be a good move from Sun I think, ultra low end cheap SPARC systems will help more people start adopting them
[02:45:04] <elektronkind> that's what the T1000 is
[02:45:05] <Error_e^ipi> it's called the ultra60
[02:45:12] <edwardocallaghan> lol
[02:45:21] <Error_e^ipi> yeah, or the T1k
[02:45:32] <Error_e^ipi> $2000 for a 1U sparc? yeah...
[02:45:41] <edwardocallaghan> More apps on the SPARC the more SPARC we are going to see
[02:46:11] <dlg> edwardocallaghan: the machine arch is becoming less and less important these days in terms of running code
[02:46:12] <sahafeez> hum. cheap fast sparc workstation
[02:46:19] <g4lt-mordant> Error_e^ipi, I resemble that remark ;P
[02:46:26] <dlg> how it scales is a different issue
[02:46:36] <dlg> but most people write in those crappy interpreted languages that run anywhere
[02:46:48] <dlg> and the people who code in something like c should know how to do it portably
[02:46:59] <edwardocallaghan> lol
[02:46:59] <alanc> Gman: the list they've tested with is all from Logitech - I'd point you at the heads-up, but the e-mail hasn't made it to opensolaris.org yet - should be in the ON flag days page eventually
[02:47:23] <Error_e^ipi> cheap sparc wouldn't = more apps for sparc
[02:47:32] <Gman> alanc, woo! have one of them :)
[02:47:32] <Error_e^ipi> cheap sparc would just mean sun bleeds money
[02:47:44] <Error_e^ipi> faster than they currntly are, i mean
[02:47:51] <Gman> alanc, i'll catch it for the weekly summary
[02:47:54] <Stric> anyone know if the T1k is any good at being a zfs/nfs file server, compared to an equally priced v240 for example?
[02:48:05] <edwardocallaghan> I still then Sun need to offer more SPARC gear 'cheap'
[02:48:19] <Error_e^ipi> edwardocallaghan: why? what would be in it for them?
[02:48:20] <edwardocallaghan> Not saying its expensive here!
[02:48:30] <Gman> edwardocallaghan, ebay ;)
[02:49:02] <edwardocallaghan> I am saying it would invite more 'oh I'll just buy cheap x86 servers' kind of people no?
[02:49:22] <Error_e^ipi> or sun can just sell those people x86, because the chips are cheaper
[02:49:35] <edwardocallaghan> Gman:ebay, What's that can you give me a link ? :p
[02:49:36] <Error_e^ipi> instead of losing money selling sub-$1000 sparcs
[02:49:57] <Error_e^ipi> and the sub-1 grand sparcs they've sold before are a good way to get people to think sparc is old, slow crap
[02:50:03] <Error_e^ipi> ( blade100 )
[02:50:12] *** schily__ has quit IRC
[02:50:50] <edwardocallaghan> I would still like some one to do a compare between a Ultra40 and Ultra 45
[02:52:21] *** comay has quit IRC
[02:53:06] <Stric> http://blogs.sun.com/jimmauro/entry/niagara_io_architecture_performance <- nice spam at the bottom there..
[02:53:17] <Vratha> hmm
[02:53:34] <Vratha> i'm getting this impression that sun releases better software solutions than IBM, based on what little i saw of tivoli
[02:53:55] <Gman> imagine that :)
[02:55:01] *** Dar has joined #opensolaris
[02:55:36] <alanc> you can't have a blog on blogs.sun.com with comments enabled without becoming a spam target - bsc's pagerank is just too high
[02:56:10] <alanc> later roller releases have gotten better spam protection, but those spam appear to be over a year old
[02:57:18] <Gman> wonder if sun will start running wordpress now to show that we really are serious about our startup program :)
[02:57:33] <Gman> i'd almost move my blog if that happened
[02:59:53] <sahafeez> sometimes i wonder what the point of the sparc at this point. while the T1 makes sense the normal SPARC is not really better then a Power or an Intel.
[03:00:19] <delewis> sahafeez, depends on the application.
[03:00:32] <delewis> SPARC should be able to beat POWER in some workloads.
[03:00:47] <sahafeez> delewis, they are few and far between
[03:00:54] <delewis> sahafeez, not necessarily.
[03:01:00] <delewis> consider any workload that involves vector calculations
[03:01:15] <delewis> given POWER lacks an SIMD instruction set, whereas SPARC has VIS, SPARC should outperform POWER.
[03:01:55] <sahafeez> ah, power has SIMD
[03:02:00] <delewis> no, it does not.
[03:02:09] <delewis> *PowerPC* does.
[03:02:11] <delewis> POWER does not.
[03:02:16] <delewis> that goes for POWER5, as well.
[03:02:57] <sahafeez> checking...
[03:03:12] <Stric> "Using 'Virtual Vector Architecture' ViVA several POWER5 processors can act together as a single vector processor"
[03:03:15] <edwardocallaghan> How long has Solaris been binary compatible for ?
[03:03:16] <Error_e^ipi> does power not have altivec ?
[03:03:21] <delewis> ericr, no.
[03:03:25] <delewis> er, Error_e^ipi
[03:03:32] <delewis> that was an Apple/Motorola addition.
[03:03:32] <Error_e^ipi> hmm, okie
[03:03:45] <alanc> edwardocallaghan: binaries built for SunOS 4 in the late 80's can still run on Solaris today
[03:03:46] <delewis> Stric, yes several.
[03:03:55] <delewis> not one though.
[03:03:57] <Error_e^ipi> i just assumed it's in PPC970, so it should theoretically be in POWER3
[03:04:05] <edwardocallaghan> Thanks
[03:04:08] <Stric> seems like altivec is coming in power6
[03:04:23] <delewis> Error_e^ipi, that's one of the few things that differentiates PPC970 from POWER3 and POWER4.
[03:04:37] <delewis> the only difference is the dcbzl instruction.
[03:04:41] <sahafeez> POWER5 in cell has SIMD
[03:05:04] <delewis> sahafeez, yes, but the POWER5s you drop $30k/proc do not.
[03:05:11] * Gman considers unsubbing from smf-discuss and subscribing to something more useful
[03:05:22] <sahafeez> POWER6 has all the Altivec
[03:05:38] <delewis> sahafeez, took long enough.
[03:06:01] <sahafeez> delewis, if that is your only app them my point is more then made.
[03:06:27] <jbk> well are you talking perf/core or perf/system ?
[03:06:53] <sahafeez> ok, lets look at this way? could you build an E10K with PPC?
[03:06:56] *** MikeE has joined #opensolaris
[03:06:59] <sahafeez> or intel
[03:07:05] <sahafeez> the answer is yes
[03:07:09] * dlg want solaris on itanium
[03:07:40] <Error_e^ipi> heh
[03:07:41] <MikeE> yikes!
[03:07:57] <sahafeez> so the question becomes one of $$. sun adds lots of value in solaris and the design of the larger systems.
[03:08:12] <Error_e^ipi> lol... itanium
[03:08:33] <MikeE> somebody likes HP gear :-)
[03:08:35] <sahafeez> the sparc is not a big value add anymore. it is a hugh cost center that could be replaced with another chip
[03:09:03] <delewis> sahafeez, not if you take into consideration the number of SPARC applications that customers have.
[03:09:08] <MikeE> this doesn't sound like a fun debate/discussion...
[03:09:37] <sahafeez> i just do not feel that sparc has anything to offer anymore...and remember i am a hugh sun supporter.
[03:09:38] * dclarke returns
[03:09:39] <delewis> its much cheaper for a customer to buy the next-generation SPARC system that it is to hire a team or developers and pay them to port your application to another platform.
[03:09:48] <delewis> s/that/than/
[03:10:06] <dclarke> sahafeez : do you feel that I am a big Sun supporter ?
[03:10:23] <sahafeez> you are right. and it is much cheaper to keep using that DOS app that you have been using for years to. it is a stupid argument.
[03:10:32] <sahafeez> of course
[03:10:33] <Error_e^ipi> i'm sure HP would love to see anyone on Itanium too
[03:10:39] <jbk> it's what's keeping people buying itanics
[03:10:44] <jbk> from hp
[03:10:47] <delewis> sahafeez, sorry, but a DOS application doesn't run Wall Street.
[03:10:49] <dclarke> sahafeez : I question the future value of Sparc also
[03:10:57] <jbk> (as it's what you have to do now to run pa-risc apps)
[03:11:02] <sahafeez> itanic was an over reach
[03:11:17] *** adp` has left #opensolaris
[03:11:25] <dclarke> from a pure business perspective there needs to be a separation between hardware production and software
[03:11:33] <Stric> jbk: saw some statement from HP or so that the fastest sparc machine you could buy was an itanic emulating sparc ;)
[03:11:44] *** darkcmd has quit IRC
[03:11:55] <delewis> Stric, I highly, highly doubt that :-)
[03:12:02] <sahafeez> VLWI was ahead of its time and everyone that had a chance of pulling it off (the sr. alpha team) left.
[03:12:09] <delewis> an E25K has a lot more throughput than a large Itanium system does.
[03:12:11] <dclarke> as much as I love Sparc .. the fastest thing I can get my hands on is 1.6 GHz and its not up to the task
[03:12:33] <sahafeez> delewis: and a box from ibm can do the same thing.
[03:12:48] <dclarke> well .. a massively parallel AMD Optreon will achieve the same or better results
[03:12:53] <MikeE> actually no.. the superdome beats the 1.8ghz e25k...
[03:12:54] <delewis> sahafeez, running what operating system?
[03:13:05] <sahafeez> the question is one of $$$ spent by sun vs. the benift
[03:13:08] <sahafeez> aix
[03:13:09] <MikeE> but the e25k clips the IBM 595 right now I think...
[03:13:12] <Gman> oooh, ericsson mail domains showing up on install-discuss
[03:13:23] <MikeE> delewis hpux vs. solaris
[03:13:26] <dclarke> Gman : nice
[03:13:30] <sahafeez> or S/390 or whatever they are calling the mainframe os now
[03:13:47] <edwardocallaghan> If Sun could get Rock out this year then I think SPARC would be really back in the game
[03:13:49] <MikeE> go look at the spec.org rate-tests... pretty fair comparison... (lets not look at floating point :-)
[03:13:52] <dclarke> Gman : I was talking to some guys from Siemens the other day .. same thing .. how to install and configure etc etc
[03:14:01] <Gman> sweet
[03:14:20] <Gman> when will we see ibm? :)
[03:14:20] <MikeE> but who cares how big/fast something runs. will it run MY app at the BEST price, meeting my customer's SLA...
[03:14:29] <sahafeez> the E25K is a great box. no doubt. the question is the $$ sun puts into it worth what they get out or could that money be better spent somewhere else.
[03:14:35] <dclarke> Gman : they are very seriously looking at Linux on AMD Opteron but .. they have been Sparc soooo long that the Solaris guys are fighting for Solaris on AMD Optron
[03:14:48] <MikeE> the 25k is a 10year-old design that is still selling
[03:14:51] <Tpenta> afternoon dennis
[03:14:55] <sahafeez> Rock, ah yes. and it will run DukeNukem' Forever right?
[03:15:00] <dclarke> Tpenta : good day Sir
[03:15:06] <MikeE> it uses the same CPU-boards as the 2900 --> 6900...
[03:15:23] <MikeE> so how much does the e25k cost in development right now? I don't think all that much...
[03:15:28] <MikeE> they're in the bonus round on that box...
[03:16:32] <MikeE> anyone know if there is a preferred method for adding a "virtual IP" to a system that has IPMP setup to provide network multipathing?
[03:17:31] <MikeE> (should you hack up the same /etc/hostname.NIC entry that includes the IPMP settings, or create a random new one with a :#? (and will that then participate in the IPMP group?)
[03:17:54] <jbk> i think docs.sun.com has you do it with hostname.NIC
[03:18:00] <jbk> that's how i've done it
[03:18:02] <jbk> at least
[03:18:03] *** bondolo has quit IRC
[03:18:34] <sahafeez> http://sunsolve.sun.com/search/document.do?assetkey=1-9-70062-1
[03:18:40] <MikeE> the IPMP documentatino is about the worst doc out there... way open for interpretation....
[03:19:05] <dclarke> so .. let me ask a question here : what is TCPIP MTU and how would you rewrite the MTU discovery process for maximum efficiency ?
[03:19:16] <sahafeez> well it is sun. the docs are 1. perfect in everyway. or 2. a compete cluster fuck
[03:19:36] <delewis> sahafeez, try Oracle documentation :-)
[03:19:37] <dclarke> that was one of the questions that I got today ... from the nice people at Google
[03:20:03] <sahafeez> you have to be kidding me? was this an interview question?
[03:20:08] * dclarke just rtying to stop the bickering
[03:20:19] <dclarke> sahafeez :  yes
[03:20:26] <sahafeez> ah. whats the job?
[03:20:38] <Error_e^ipi> i like sun's docs generally
[03:20:39] <dclarke> sahafeez :  mailroom boy I think :-)
[03:20:40] <Error_e^ipi> *shrug*
[03:21:06] * sahafeez turned down a job a google...because i was braindead
[03:21:27] <Error_e^ipi> dclarke: mtu is packet size
[03:21:32] <dclarke> sahafeez :  how many interview questions and interviews did they put you through
[03:21:37] <sahafeez> 7
[03:21:45] <dclarke> Error_e^ipi :  I know MTU .. its sort of packet size
[03:21:54] <dclarke> sahafeez :  I am at number 3
[03:22:21] <sahafeez> pre-ipo then i did not hear from them for 4 months, they ipo'd and then called and made the offer
[03:22:30] <dclarke> Error_e^ipi :  I always remember is as 20 years before Columbus sailed the ocen blue in 1492 he sent a packet at 1472 bytes
[03:22:48] <sahafeez> dclarke: the job was designing a internet backbone
[03:22:49] <dclarke> sahafeez :  damn !
[03:23:06] <dclarke> sahafeez :  sweet position to have .. *before* IPO
[03:23:20] <dclarke> now it may be a pile of paper millionaires in there
[03:23:24] <sahafeez> the way the stock has gone since ipo i would still have made a ton
[03:23:37] <dclarke> sahafeez :  it was like that in my Lotus days
[03:23:57] <jamesd> mtu == maximum transport unit,  its the largest packet a link can support, i think its 1592 for  ethernet....   packets larger will automaticly broke into smaller packets by the network stack.
[03:23:57] <sahafeez> but i had moved the fam, bought a house, etc...so i had to say no.
[03:24:06] <dclarke> sahafeez :  Lotus shares were at $16 a piece and employees were getting them cheaper and the company would toss in an equal amount
[03:24:15] <MikeE> mtu is 1500 typically isn't it...
[03:24:16] <dclarke> jamesd : 1492
[03:24:23] <sahafeez> jamesd: i do not think he really wanted an answer. it was just a question google asked him
[03:24:25] <dclarke> jamesd : 1492 - 20
[03:24:49] <dclarke> actually he wanted a new algorithm for MTU discovery and then response
[03:25:05] <MikeE> hmm... ifconfig -a shows 1500 I guess...
[03:25:10] <dclarke> I simply quoted the ICMP response idea and then used a binary backoff algorithm
[03:25:25] <elektronkind> wow
[03:25:28] <dclarke> MikeE : run snoop .. 20 bytes are overhead
[03:25:39] <elektronkind> what is up with thathuge commit that just went through to onnv?
[03:25:51] <sahafeez> ethernet is 1500. tokenring (16m) is 17914. fddi is 4352. x.25 is 576
[03:25:52] <dclarke> MikeE :  then there is the IP header .. the TCP header .. the ETHER MAC address stuff
[03:26:07] <MikeE> with TCP yes, with LLT not sure..
[03:26:18] <MikeE> right so the payload is 1492?
[03:26:22] <sahafeez> no
[03:26:35] <dclarke> sahafeez :  yes yes .. however the guy wanted to talk about much larger sizes based on congestion response in gig ethernet
[03:26:50] <dclarke> in any case .. it was a fun chat
[03:26:51] <elektronkind> oh, jumbo frames
[03:26:58] <sahafeez> ethernet sucks. wish tokenring won
[03:27:02] <MikeE> aah jumbo frames?
[03:27:09] <MikeE> FDDI haad nice 8k packets
[03:27:14] <sahafeez> dclarke: what was the job?
[03:27:31] <MikeE> gotta run guys... 'nite
[03:27:36] <dclarke> however I pointed out that there was no easy and fast way to calculate the best way to respond from the NIC and a better processor would be needed to do floating point right on the NIC
[03:27:43] <edwardocallaghan> See you guys a bit later
[03:27:46] *** edwardocallaghan has left #opensolaris
[03:27:46] *** MikeE has quit IRC
[03:28:07] <dclarke> sahafeez :  thus far they are talking about a manager in their server side that watches over google.com
[03:28:08] <elektronkind> most (but not all) 1gb ethernet interfaces and all 10gb ethernet interfaces can do what's called jumbo frames... sometimes up to 12k... the switch has to support them too
[03:28:13] <elektronkind> oh drat he's left
[03:28:15] <sahafeez> dclarke: tcp offload? does not work. or does. depending on who you talk to
[03:28:26] <dclarke> sahafeez :  bingo ..
[03:28:27] <sahafeez> 9000 is the default jumbo
[03:28:48] <dclarke> sahafeez :  lucky for me .. I spent a lot of time over the past six months reading CISCO books and reading up on su3ch things
[03:29:29] <dclarke> in any case .. I very quickly went from being overwhelmed to comfortable and I started to stutter .. I was talkign fast and hand waving and doing my own algortihm .. it was fun
[03:29:36] *** pseudoXh4 has joined #opensolaris
[03:30:04] <pseudoXh4> Since Solaris is no longer Slowaris (or so I heard) on x86, how does its performance compare with other OSes, such as the BSDs?
[03:30:07] <dclarke> when I stopped to point out that I sometimes stutter because my mouth is a serial device and my brain parrallel it was a classic congestion problem with mutex locking
[03:30:19] <sahafeez> dclarke: my interviews were not like that. depends on who you talk to. my job was for designing a network
[03:30:22] <elektronkind> you're such a nerd
[03:30:33] * dclarke is a geek .. not a nerd
[03:30:36] <elektronkind> ;)
[03:30:51] <dclarke> sahafeez :  pre-IPO .. that would have been sweet
[03:30:56] <sahafeez> and i am not a prg'r so when they start talking about designing algortihm i just kinda went hum?
[03:31:05] <delewis> pseudoXh4, its mostly called "Slowaris" by people that don't understand why it was called "Slowaris" in the first place.
[03:31:18] <dclarke> okay .. I gotta go get coffee and look at the children .. they have forgotten who I am today .. I have been in front of a computer
[03:31:23] <elektronkind> pseudoXh4: there's some recent numbers regarding that
[03:31:26] <delewis> and Solaris in combination with Sun Studio have a set a large quantity of world records.
[03:31:27] <elektronkind> pseudoXh4: hold on a sec
[03:31:30] <Tpenta> slowaris was coined for Solaris 1 -> Solaris 2.0
[03:31:36] <delewis> (most of which have been verified through by 3rd parties)
[03:31:41] <delewis> like SPEC and so fourth.
[03:31:44] * dclarke back in a while
[03:31:54] <pseudoXh4> elektronkind, thanks.
[03:32:01] <elektronkind> pseudoXh4: http://blogs.sun.com/bmseer/entry/update_solaris_beating_linux_performance
[03:32:26] <delewis> kind of difficult to set a benchmark with Linux when you're using sub-standard compilers.
[03:32:32] <Tpenta> someone going to drop that link on the blinkered personage who mentined sloaris on osnews?
[03:32:55] *** postwait_ has joined #opensolaris
[03:32:56] <delewis> Tpenta, napalm sounds more fitting :-)
[03:33:03] <sahafeez> however they showed me what they had designed and asked me what i thought. about 4 hours later i had torn it to shit and redid most of it. i got the feeling i was talking way over the head of the people there. 2 many PHD eggheads that have never done production and did not realize that just becuase it is suppose to work one way and the vendors told you something did not make it true in the real world. it is kind like believing cisco wh
[03:37:14] <Tpenta> I just responded
[03:38:13] <Tpenta> heh, I worked as an SA and IT manager at a govenrment researcher. The main thing that we were scared of was PhD's who thought they new sysadmin.
[03:40:03] <elektronkind> argh. I have 1 hour to rack, jumpstart and add to my SAN a x4100, and ready for a oracle 10g installation
[03:40:06] <sahafeez> theory and practice are 2 very different things...the biggest sin i have seen is believe the vendor and not understand the concept of you have to manage it after it is built
[03:40:07] <pseudoXh4> So like.
[03:40:16] <pseudoXh4> Anyone know how long it takes to get that free Solaris kit thing?
[03:40:17] <pseudoXh4> :O :D
[03:40:25] <sahafeez> fill out a form.
[03:40:50] <elektronkind> pseudoXh4: ask your postal service :) you can download installation DVD or CDROM iso as well
[03:40:55] <pseudoXh4> Yeah..
[03:41:00] <pseudoXh4> But it's not as cool as being sent one in the mail.
[03:41:13] <pseudoXh4> I downloaded the DVD a few weeks ago but then lost it. :(
[03:41:20] <pseudoXh4> And it took forever.
[03:41:57] <elektronkind> I imagine sun wants to get the requested media into peoples' hands asa
[03:42:00] <elektronkind> asap
[03:42:00] <Error_e^ipi> took me about 2 weeks to get my s10u2 DVD
[03:42:16] <pseudoXh4> Hmm, you're in the US though..
[03:42:17] <Kmays> dclarke: good luck
[03:42:19] <delewis> I received mine within a week or so.
[03:42:25] <Error_e^ipi> pseudoXh4: no, i'm canadian
[03:42:28] <pseudoXh4> I'm in Australia... wondering how long I'd wait to get it.... :\
[03:42:32] <pseudoXh4> Error_e^ipi, close enough... :P
[03:42:43] <delewis> of course, I live in TN about an hour away from a FedEx hub.
[03:42:50] <delewis> the main FedEx hub, that is.
[03:43:00] <pseudoXh4> That's very fortunate... :P
[03:43:24] <elektronkind> yeah, Error_e^ipi is canadian, so it actually took 1.5 U.S. weeks for the DVD to get to him ;)
[03:43:25] <sahafeez> ok, i am now bored with the web.
[03:44:02] <pseudoXh4> Hahah.. I get it. :P
[03:44:54] <Error_e^ipi> :P
[03:45:06] <elektronkind> ok time to sling some hardware around ye olde data centre
[03:45:35] <alanc> sac_nextcase 1, alanc 0
[03:45:58] <alanc> one of these days I'll learn how to file an externally visible ARC case the right way...
[03:46:35] * pseudoXh4 wonders if OS has azalia audio support.
[03:47:14] * sahafeez wonders if there is any beer in fridge
[03:47:23] <Error_e^ipi> what the hell is azalia audio?
[03:47:37] * alanc was wondering the same thing
[03:47:48] <sahafeez> sound like it is an mp3 from al-qaeda
[03:47:48] <pseudoXh4> Uhhh..
[03:47:57] <pseudoXh4> Those whizz-bang HD Audio cards in laptops these days.
[03:48:02] <pseudoXh4> Well, "HD Audio".
[03:48:05] <pseudoXh4> They're pretty shitty.
[03:48:10] <Error_e^ipi> never heard of them
[03:48:14] <pseudoXh4> Like, low volume and stuff.
[03:48:23] <pseudoXh4> They're usually referred to as SigmaTel HD Audio or Intel HD Audio.
[03:48:29] <pseudoXh4> I think Conexant makes them too.
[03:48:30] <sahafeez> now watch as all the fbi agents descend on opensolaris
[03:49:05] <pseudoXh4> But yeah.
[03:49:20] <pseudoXh4> atm I have NetBSD on my laptop... but from the looks of it I have some nice options if I went with OpenSolaris..
[03:49:49] <Error_e^ipi> try it
[03:50:00] *** stevel has quit IRC
[03:50:08] <Error_e^ipi> burn a copy of BeleniX, that'll let you know if everything's likely to work
[03:50:49] <pseudoXh4> Hm, guess I will.
[03:50:50] <alanc> hmm - opensolaris-arc mail archives could be a bit easier to find - had to go to pipermail list of archived lists to find it
[03:50:55] <Gman> OpenSolaris Weekly News #45 - http://mail.opensolaris.org/pipermail/opensolaris-discuss/2007-January/023604.html
[03:51:14] <Gman> alanc, hrm, i could probably fix that
[03:51:37] <sahafeez> bored
[03:51:51] <alanc> was surprised the ARC community discussions link only shows arc-discuss and not opensolaris-arc
[03:52:11] *** yippi has joined #opensolaris
[03:52:29] <Gman> alanc, hrm, seemingly not
[03:52:46] <Gman> discussions is a blank page for me, so there must be some magic foo to tie a mailing list to a project/community
[03:52:59] <alanc> well, Xorg 7.2 ARC case seems to have made it through on the second try - http://mail.opensolaris.org/pipermail/opensolaris-arc/2007-January/000901.html
[03:53:48] *** st has joined #opensolaris
[03:58:22] *** shadeland has quit IRC
[03:58:25] <Gman> anyone know the moderation rules for opensolaris-announce?
[03:58:55] *** shadeland has joined #opensolaris
[03:59:00] <alanc> I just send my source drop announcements there and they magically appear on the list a few hours or days later
[03:59:40] <Gman> yeah, it's a moderated list
[03:59:47] <Gman> and trying to decide whether i should moderate replies to a post or not
[04:01:15] <g4lt-mordant> do they make hte poster or OP look lke an idiot?
[04:01:53] *** shadeland has joined #opensolaris
[04:02:14] <Gman> g4lt-mordant, no, they're reasonable posts
[04:02:23] <Gman> just depends on the rules set for opensolaris-announce, that's all
[04:02:28] <g4lt-mordant> in that casse, there's no fun in it.  I'd say no
[04:02:34] <Gman> :)
[04:03:58] <alanc> do they qualify as an announcement?   or just someone who forgot to fix cc?
[04:04:45] <Gman> forgot to fix the cc i'd imagine
[04:09:48] *** peteh has quit IRC
[04:10:31] *** peteh has joined #opensolaris
[04:11:22] *** Theoden-Nexenta has joined #opensolaris
[04:17:19] *** gm152 has quit IRC
[04:21:56] *** edwardocallaghan has joined #opensolaris
[04:22:13] <edwardocallaghan> Hi
[04:22:16] <jbk> hi
[04:22:20] <edwardocallaghan> boyd: Are you around ?
[04:22:29] <boyd> sorta
[04:22:43] <jbk> more rectangular? :)
[04:23:30] <edwardocallaghan> Collage just sent me a email telling asking me URGENT do you want to confirm accommodation from the letter we sent you?:-/What letter?
[04:23:54] *** st has quit IRC
[04:23:59] <edwardocallaghan> F**king Great Britain can't even get mail to you
[04:24:24] <edwardocallaghan> Kind of worried ?!
[04:25:00] *** gm152 has joined #opensolaris
[04:25:03] <edwardocallaghan> I emailed them back as soon as they emailed me so hopefully it will be fine
[04:26:02] <edwardocallaghan> Why do these thing always happen to me?
[04:28:47] <boyd> Umm... I dunno?
[04:28:49] <boyd> :)
[04:28:50] <CIA-22> Rod Evans: PSARC/2006/714 ld(1) mapfile: symbol interpose definition; 6494214 Refinements to symbolic binding, symbol declarations and interposition; 6500110 *file* could produce better error diagnostics
[04:28:57] <elektronkind> I think I'll beat my record for standing up a new server
[04:29:46] <elektronkind> 15 mins to unpack a x4100, add a fc card to it, rack it, cable it (helps that there happened to be optical cables already there) and now it's jumping.
[04:30:48] <jbk> heh.. i'm lucky if i can get power run in 15 hours :)
[04:30:49] <edwardocallaghan> boyd:What do you suspect will happen?
[04:31:43] <boyd> edwardocallaghan: It sounds like it should be fine.
[04:31:44] <elektronkind> we have these cool new dual-circuit power strips in our racks now, complete with a multimeter for each circuit (measuring volts, amps, and the summed load)
[04:32:17] <jbk> i have the disadvantage of working in a large bureaucracy :)
[04:33:13] <jbk> you don't want to know how long it's taken me to get working terminal server connections for 2 boxes :)
[04:33:52] <edwardocallaghan> OK , When I look at the email my face look like this > =-O
[04:34:04] <elektronkind> the CTS signal on the serial link needs to be proposed before a committee?
[04:34:09] <edwardocallaghan> The mail is so bad here
[04:34:27] *** MikeTLive has joined #opensolaris
[04:34:50] <jbk> almost.. just too stratified, so trying to coordinate the excessive number of groups needed to get the most basic stuff done takes work
[04:35:09] <jbk> though in this case, just the network people being lazy
[04:35:23] <edwardocallaghan> They lose exam papers and do you know what the news paper said, they questioned the guy and he replied "I dumped the mail in the field because I'm dyslexic"
[04:41:16] *** deedaw has quit IRC
[04:45:58] *** alanc is now known as alanc-away
[04:50:06] *** MikeTLiv1 has joined #opensolaris
[04:50:07] *** MikeTLive has quit IRC
[04:55:15] *** ibb has joined #opensolaris
[04:58:24] *** linma has quit IRC
[07:39:12] *** echelog has joined #opensolaris
[07:39:44] <jteo> mlc39: zfs is COW.
[07:39:57] <Vratha> actually, can anyone tell me right now: if you upgrade solaris and it updates the kernel, do your installed nvidia drivers stop working?  or is it more like windows where it just keeps working
[07:39:58] <ibb> computer on wheels??
[07:40:38] <jteo> Vratha: it keeps working.
[07:40:38] <steleman> vratha: your drivers will just keep working
[07:40:38] <Vratha> SWEET
[07:40:38] <Error_e^ipi> Vratha: it will keep working
[07:40:38] <Vratha> that alone is really good news
[07:40:38] <Error_e^ipi> stable ABI
[07:40:38] <Vratha> word
[07:40:39] <Vratha> that's freakin' awesome
[07:40:39] *** mazon has joined #opensolaris
[07:40:40] <Vratha> now that's what a professional unix system is all about :)
[07:40:42] <ibb> vrathat did u have problems with linux with that?
[07:40:58] <Chile`> what's happening with ati dri?
[07:41:06] <Vratha> ibb: oh god yes.  any time the kernel version changed, even withing the same major version, the drivers crapped out and would have to be rebuilt
[07:41:15] <steleman> ati dri is searching for its inner child
[07:41:22] <mlc39> jteo: I'm only vaguely acquainted with copy-on-write.  As a result of ZFS being copy-on-write, are you suggesting that the initial performance shortcomings I speculated do not in fact apply?
[07:41:36] <Chile`> steleman: meaning dead in the water?
[07:41:45] <steleman> chile: no idea honestly.
[07:42:06] <jteo> mlc39: unless you are only doing reads to your pool, the data will be rewritten constantly.
[07:42:14] <Chile`> the only thing I've been able to find on any of the lists is a vague "radeon dri should be ported sometime in the near future" a few months ago
[07:42:15] <jteo> and hence redistributed.
[07:42:22] <Chile`> when intel dri was added
[07:42:55] <steleman> chile: i dont know what ati is doing with their drivers. the situation is not that stellar on linux either (which they supposedly support)
[07:43:29] <Chile`> on linux/bsd the opensource r200/r300 supports full acceleration for 2D
[07:43:36] <ibb> steleman: its horrible on linux, i cant get opengl working on my ati 9000
[07:43:49] <Vratha> hmm, i would think ATI would be supported at some point since AMD is going to integrate their stuff and Sun sells AMD items
[07:43:50] <Chile`> 2d is all I care about here
[07:44:08] <mlc39> jteo: So, the idea of copying all files to new files (in place, one at a time - like cp $FILE $FILE.tmp; mv $FILE.tmp $FILE) would in principle redistribute the data in the background as the copy took place.  But if one of the vdev's was already largely allocated, then there may be greater struggle to find free blocks.  In my situation, my usage is at 50%...
[07:44:29] <jteo> mlc39: so it would write to the other vdev
[07:45:27] <ibb> mic39: 14 drives sounds like a large amount of space
[07:46:10] <Vratha> is Sun supposed to start having ATI drivers since they're going to be integrated on AMD boards?
[07:46:24] <mlc39> ibb: Right now I have 12 drives, and if I had ideally RAIDZ2 setup then that's 10 drives of space.  But with the two RAIDZ2's of 6 each, that's a total space of 8 drives.
[07:47:03] <ibb> that is over a terabyte of actual space, isnt it?
[07:47:14] <mlc39> ibb: Yes.
[07:48:15] <g4lt-mordant> Vratha, uhm, sun DOES have ATI drivers, look at the XVR100
[07:48:26] *** tsoome1 has quit IRC
[07:48:41] <Vratha> g4lt-mordant: oh, my bad
[07:48:58] <Vratha> i'm pretty Sun stupid atm... i had sworn off sun a few years ago after bad times with solaris 8 administration
[07:50:03] <ibb> mic39: they are all stripes no mirroring going on then right on the 12 drives?
[07:50:11] <steleman> well maybe this new deal with Intel will result in 3D drivers for Intel gfx (rumor has it that Intel wants to become a major player in the gfx market)
[07:50:28] <Vratha> sounds good to me: more competition
[07:50:49] <ibb> steleman: rumors might be right as they are working hard for directx10 hardware
[07:50:50] <ofu> is it possible that zpool status hangs during resilvering?
[07:51:09] <Vratha> i should port my userspace device driver framework to opensolaris
[07:51:11] <g4lt-mordant> steleman, doubt it, since any sparc gfx pretty much has to be openfirmware, since sparc doesn't sit too well with peecee cards
[07:51:40] <jteo> ofu: yes.
[07:51:45] <steleman> g4lt: i wasnt talking about Intel gfx on SPARC boxes ;-)
[07:51:52] <mlc39> ibb: No.  The discussion above is saying that I have 12 physical drives.  Have only set up 8 of them on a RAIDZ2, so I have 6 drives' worth of data.  4 drives are now to spare, but I wanted to integrate the other four drives in an elegant way, but I learned that it's not possible to transparently add more drives to an existing RAIDZ2 configuration.
[07:53:12] <ofu> jteo: why? Is this a known bug?
[07:53:15] <ibb> mic39: like they said before, it should include a solution to this problem soon, i would like that to happen as i am trying to move from the ntfs drives that i have
[07:54:11] <Error_e^ipi> being able to turn a lone device or a mirror in to a raidz would i think be relatively easy
[07:54:16] <jteo> ofu: resilvering can consume excessive bandwidth as i recall.
[07:54:39] <mlc39> ibb: I overestimated what ZFS could do in this regard, since the man page didn't freely advertise this limitation (and I read optimistically rather than doubtfully).  But, if it doesn't already do this, it sounds like the implementation would be a bit tricky if it's not done already.
[07:54:40] <ibb> Error_e^ipi: one after anothe it keeps adding on
[07:55:31] *** dlg has left #opensolaris
[07:55:41] <Error_e^ipi> add 2 drives... copy block 1 of the data to one, checksum, write to disk 3, then copy block 2 to disk 2, block 3 to disk 3, checksum, write over block2 on disk 1
[07:55:43] <ibb> mic39: that is the same way i am looking at this zfs, but i should be looking with a bit of question marking
[07:55:48] <Error_e^ipi> and so forth
[07:56:19] <Error_e^ipi> raidz2 i won't speculate on how to do it
[07:56:52] <mlc39> Error_e^ipi: I can see that the non-realtime solution would be easier than allowing a live update.
[07:56:52] <Error_e^ipi> because it involves "oooOoOoO SPookY ComPlex MatheMatics!!!! ooOOoooOOOooo"
[07:56:53] <ibb> Error_e^ipi: i would like to use the raid features that zfs has, so i could add drives remove and on and one, i don't know the true limitation
[07:58:19] *** bougie has quit IRC
[07:59:10] *** boxubi has joined #opensolaris
[08:00:29] <ibb> ok, well good night all
[08:01:33] <mlc39> Hmm.  I'm contemplating having three 4-way RAIDZ (not RAIDZ2).  Then I could tolerate 3 drive failures (best case), but worst case no more than 1...
[08:04:37] <mlc39> Do we know if a cross-vdev file move could potentially skip the decompress-recompress phase if the original file was flagged as compressed?
[08:08:59] <nprice> WTF did they do to the nv55b installer?  it's hauling ass!
[08:11:49] <Error_e^ipi> i haven't installed since build 30
[08:13:31] <Vratha> ok, i've burned my OpenSolaris DVD and will install it sometime tomorrow after my home directory backup is finished
[08:13:51] <Vratha> but does opensolaris come with a package management utility by default?  and don't tell me about blastwave :)
[08:13:58] <Vratha> i'm just curious if sun has its own repo right now
[08:14:08] <Error_e^ipi> nope
[08:14:11] <Error_e^ipi> just use blastwave
[08:14:16] <Error_e^ipi> or sunfreeware
[08:14:16] <Vratha> ok
[08:14:23] <Vratha> which would you recommend?
[08:15:18] <Error_e^ipi> blastwave, it pulls in dependencies
[08:15:23] <Vratha> does opensolaris have anything similar to Windows Update for updating the core system at least?  you know, like in case a new Studio came out or whatever
[08:15:24] <Error_e^ipi> and doesn't install in to /usr/local
[08:15:30] <Vratha> ok
[08:15:37] <mlc39> Hmm.  I wonder if I can use 'zfs send' and 'zfs receive' as a way to skip the decompress-recompress that would otherwise be necessary...
[08:15:44] <Vratha> well maybe opensolaris should add blastwave by default then :)
[08:15:44] <Error_e^ipi> solaris10 does, but not Osol
[08:15:49] <Vratha> ok
[08:15:49] <Error_e^ipi> too much of a moving target
[08:15:51] <Vratha> yeah
[08:16:08] <Vratha> Error_e^ipi: well is it pretty easy to pull the latest code to build and install to update your system at least?
[08:16:19] <Error_e^ipi> oh yeah
[08:16:22] <Vratha> good :)
[08:16:25] <g4lt-mordant> use sunstudio ;P
[08:16:25] <Error_e^ipi> it's a hg checkout then you run a script
[08:16:41] <Vratha> g4lt-mordant: i will :)  i don't plan to use GCC after what i've learned about studio's performance :)
[08:16:52] <razrX_> Vratha: i'd suggest using live upgrade for upgrading sxcr builds
[08:17:01] <g4lt-mordant> Vratha, and osol has BFU
[08:17:05] <Vratha> BFU?
[08:17:14] <g4lt-mordant> blindingly fast upgrade
[08:17:17] <Vratha> razrX_: ok, well is that some utility or something?
[08:17:26] <Vratha> hmm, well i'll have to read the docs i guess
[08:17:34] <jteo> an astute observation.
[08:18:07] <razrX_> Vratha: each sxcr release comes with couple of SUNWlu packages to create, delete, upgrade a current BE (Build Environment) to a new BE
[08:18:08] <Vratha> g4lt-mordant: is live upgrade what you're talking about though?  because i found the doc for that at least
[08:18:22] <Vratha> razrX_: cool :)  well i found the docs for it
[08:18:35] <g4lt-mordant> no, BFU is the script for getting/building ON
[08:18:38] <razrX_> so you upgrade a copy image of your currently running release to a new one and boot off of that
[08:18:56] <Vratha> ah ha
[08:18:59] <Vratha> that's pretty nice
[08:19:10] <razrX_> and you're all doing this while the system is running in multi-user
[08:19:18] <Vratha> yeah
[08:19:25] <Vratha> sounds great
[08:20:03] <razrX_> so only downtime you'll have is when you luactivate the new BE and reboot (don't use reboot(1M) !!) your box
[08:20:20] <Vratha> well i'll read the docs :)  i'm sure it'll tell me how to reboot properly
[08:21:35] <razrX_> once luupgrade and luactivate was successfull use shutdown(1M) or init(1M)
[08:22:00] <Vratha> you mean to completely halt the system?
[08:22:19] <razrX_> in the event the new sxcr build you luupgraded is giving you issues, just luactivate the previous BE and reboot the box again
[08:22:36] <razrX_> no, just use `shutdown -y -i6
[08:22:39] <razrX_> or init 6
[08:22:42] <Vratha> ah, ok
[08:22:44] <razrX_> just don't use reboot
[08:23:21] <Vratha> ok, well i notice the docs are describing what you've said
[08:23:34] <Vratha> so i'm sure i won't mess up since it's in the docs and i won't remember any of this until then :)
[08:23:39] <razrX_> be sure to read them thoroughly
[08:23:44] <Vratha> i shall
[08:23:50] <razrX_> lol
[08:23:57] <jteo> famous last wods.
[08:24:00] <Vratha> but i won't do an upgrade for a few months anyway; i'm kinda busy until then
[08:24:03] <jteo> s/wod/word
[08:26:29] <Vratha> *sigh* ; i live alone and still close the bathroom door
[08:31:19] <Vratha> i also realized i have no DVD drive for the server i plan to use; i'll have to go get one tomorrow :-/
[08:31:31] <Vratha> but that's fine, since backing up my home directories will take several more hours
[08:31:35] <Vratha> and also i'm talking to myself :)
[08:37:24] *** mazon is now known as Mazon
[08:47:33] <mlc39> Okay.  I'm quite confused.  I'm trying zfs send / zfs receive from one ZFS pool to another.  It's generating heavy I/O, but I don't see any data yet in the destination pool.  (So there's no reflection of increases in file space used).  Is it working?  Will I have to wait until the operation completes entirely to see the new free space outputs?
[08:49:09] <jteo> mlc39: wait.
[08:50:03] <mlc39> jteo: I might need to wait a week though.  It's important to know if I have executed the right commands.  For example, it looks like I needed to first create a read-only snapshot to even be able to use zfs send...
[08:50:13] <Vratha> does opensolaris have anything like pressing ^I to see the status of a running app?
[08:50:32] <Vratha> if it does, then maybe he can do that
[08:51:07] <jteo> mlc39: correct. but the readonly snapshot costs nothing
[08:51:58] <Vratha> so with leopard supporting ZFS, i wonder if apple is considering switching over to it :)
[08:53:05] <mlc39> jteo: Yeah, I know it costs no space.  I'm just trying to verify that I'm doing things correctly.  When I enter the zfs receive part of the command, I name the new pool (pool2) followed by a new filesystem name (pool2/test).  Then it says it is receiving the stream into pool2/test@ro.
[08:53:28] <jteo> correct
[08:53:57] <Vratha> heh, it would be nice to get some more feedback though :)
[08:53:58] <mlc39> jteo: So it's normal to not even see an entry for `test' under /pool2 ?
[08:54:46] <jteo> mlc39: you could use the verbose option -v.
[08:54:50] <mlc39> It makes sense, in that the zfs send/receive is probably meant to be an atomic operation.  Yet, I'd hate for it to fail at the end...
[08:55:07] <mlc39> jteo: I did.  I use zfs receive -v pool2/test
[08:55:16] <jteo> not very wordy is it. ;)
[08:57:08] <mlc39> Nope.  When I issued zfs send pool@ro | zfs receive -v pool2/test, it completed after one second -- but there, even though the command looked innocent, it never copied the actual filesystems from pool.  Instead, it just created empty directories for each filesystem under the original pool.
[08:58:31] *** slowhog has left #opensolaris
[08:59:11] <andersmo> mlc39: the zfs command operates on individual zfs filesets in the pool, afaik?
[08:59:54] <jteo> zfs send diskpool/testfs@testsnap | zfs recv -v pool2g/testfs@testsnap
[08:59:54] <jteo> receiving full stream of diskpool/testfs@testsnap into pool2g/testfs@testsnap
[08:59:54] <jteo> received 29.0Kb stream in 1 seconds (29.0Kb/sec)
[09:02:02] <mlc39> jteo: Hmm.  In my setup I omitted the second of your @testsnap in the recv part...  But it seems to default to the same behavior.
[09:03:30] <mlc39> jteo: Supposing I had testfs and testfs2, I speculated that a single command would do the trick.  But I don't think it's possible in that way.  The main thing I'm hoping is that compression is going to be inherited in the transfer.
[09:06:01] <Vratha> psh.. opensolaris doesn't like Parallels Desktop i guess
[09:08:15] *** boxubi has quit IRC
[09:09:05] <Vratha> it lets me choose "Solaris Interactive Text (Console Edition)" and then it does "Using RPC Bootparams for network configuration information." and then it pukes saying it can't find the install software
[09:09:10] <Vratha> which is one the freaking CD
[09:09:11] <Vratha> DVD
[09:09:36] <mlc39> jteo: So you have a /pool2g/testfs directory after your command?  If you no longer care about preserving the snapshot, would you then delete @testsnap on pool2g?
[09:09:45] <jteo> yup.
[09:09:54] <jteo> ls pool2g/testfs
[09:09:58] <jteo> A           testdir     testfile    testprog    testprog.c
[09:10:05] <Vratha> why would opensolaris be saying that?
[09:10:23] <mlc39> Okay.  So the ``deleting'' of a snapshot is hardly destructive.  Just would want to make sure of that.
[09:11:50] *** sparc-kly_ has joined #opensolaris
[09:12:15] *** wezzy has joined #opensolaris
[09:12:32] *** mlc39 is now known as mlc38
[09:12:43] <jteo> time for work. have fun mlc38.
[09:13:07] <mlc38> jteo: Thanks.
[09:13:27] * Vratha decides to try OpenSolaris on a real system to see if it complains or not
[09:13:42] *** wezzy has quit IRC
[09:13:45] <Vratha> it's possible that Parallels Desktop just isn't building a proper VM for OpenSolaris, which is disappointing :(
[09:13:52] <Vratha> but then again it doesn't build a proper VM for minix either
[09:15:11] *** mlc38 is now known as mlc37
[09:16:23] *** mlc37 has quit IRC
[09:17:09] *** Cyrille has joined #opensolaris
[09:17:56] *** uncertainty has joined #opensolaris
[09:20:19] <Vanuatoo> Why is solaris express 1/07 delayed?
[09:23:04] <Vratha> does solaris express boot from EFI now?
[09:26:22] *** dunc has quit IRC
[09:26:37] *** pogma_ has quit IRC
[09:27:17] *** tsoome has joined #opensolaris
[09:28:35] *** tsoome1 has joined #opensolaris
[09:28:59] *** sparc-kly has quit IRC
[09:29:12] *** tsoome has quit IRC
[09:32:09] *** pogma_ has joined #opensolaris
[09:33:08] *** Mazon is now known as mazon
[09:33:46] *** mazon is now known as Mazon
[09:38:56] <Vratha> oh, nice; sun is going to GPL solaris and put a GNU userland in its system
[09:39:03] <Vratha> hopefully we'll see that with opensolaris
[09:41:21] *** Tpenta has quit IRC
[09:41:52] *** MaGre_ has joined #opensolaris
[09:44:39] *** MaGre has quit IRC
[09:44:50] *** MaGre_ has quit IRC
[09:44:59] <Gr|ffous> Vratha, I don't believe that rumour has been confirm
[09:45:00] <Gr|ffous> ed
[09:45:09] <Vratha> ok
[09:47:44] <Cyrille> actually, I think that rumour has been specifically denied: http://blogs.sun.com/richgreen/entry/all_the_news_that_s
[09:48:49] <Vratha> well they didn't deny that they might rely on a GNU userland more
[09:49:22] <Cyrille> he didn't deny anything about the future (who would?), but he said that whatever was printed is just speculation.
[09:49:30] <Vratha> yeah, that's true
[09:49:52] <Vratha> i woudln't mind seeing a more GNU userland though
[09:50:35] <Cyrille> for the moment there's the nexenta distribution for that purpose.
[09:50:52] <Vratha> yeah, but it's in alpha
[09:51:05] <Vratha> i think i'll just stick with the mainline and like it
[09:51:37] <Cyrille> I think you can also add the GNU tools to the basic distribution with packages from sunfreeware.
[09:51:46] *** qdk has quit IRC
[09:51:57] <Vratha> yeah, or blastware maybe
[09:52:01] <Vratha> blastwave*
[09:52:07] *** bzcrib has joined #opensolaris
[09:52:12] <Vratha> i wouldn't really know yet as i haven't gotten opensol installed yet
[09:52:21] <Vratha> i'm working on backing up my existing system first
[09:52:25] <Vanuatoo> does anybody know why solaris express 1/07 is delayed?
[09:53:01] <Cyrille> most likely (and no, I don't know) because some things aren't finished/working as they are expected.
[09:53:28] <Vanuatoo> Cyrille: Do you know any place where I can track the progress?
[09:53:39] <Vanuatoo> or some good bugzilla query?
[09:55:03] <Vratha> boo; the VIA VT6103 needs to work in opensolaris
[09:55:15] <Vratha> otherwise i won't be able to use it on my new mini-itx system :(
[09:55:58] <Cyrille> Vanuatoo: I'm afraid I don't know.
[09:56:15] <Vratha> too bad i wasn't planning for opensolaris when i purchased that system (that hasn't arrived yet)
[09:56:28] <Vanuatoo> Hopefully one day opensolaris will have some great bug system
[10:00:09] <raph_ael> hello
[10:00:42] <Vratha> and since i don't have a system where i can install opensol yet, is it possible to install it without having to install the GUI, or is that some integral part of the system?
[10:01:05] <Vratha> maybe the GUI libs would be ok though so i could run the apps remotely
[10:01:36] <tomww>  Vratha: you want no X11 at all on the system, or just the installation on text-screen ?
[10:01:44] <Vratha> no X11 at all
[10:01:47] <Cyrille> the GUI is definitely not an "integral part of the system".
[10:02:09] <Vratha> Cyrille: is it pretty easy to setup solaris' directory server without the GUI still?
[10:02:14] <tomww> is it to save space or other reasons?
[10:02:17] <Vratha> that's the main thing i worry about
[10:02:24] <Vratha> tomww: i won't have a head on it
[10:02:25] *** dunc has joined #opensolaris
[10:02:27] <Cyrille> directory server?
[10:02:43] <Cyrille> you mean the Sun Directory Server?
[10:02:44] *** eryc has joined #opensolaris
[10:02:45] <Vratha> yes
[10:03:04] <tomww> you could install just the default including X11 an dont fight with dependencies :)
[10:03:16] *** gustav3d has joined #opensolaris
[10:03:26] <Cyrille> you can install it in text mode, then start the server and its administration server, and then install just the console on any (suitable) box and manage it from there.
[10:03:27] <oxygene> Vratha: you can do the initial install in textmode (via serial console, even) - for everything else, you could just X11 over network
[10:03:58] <Vratha> sounds good.. though i'll probably do most things via console anyway :)
[10:04:07] <Vratha> except that directory management
[10:04:25] <Vratha> unless it ends up being really easy which would be a major bonus
[10:04:55] *** Tpenta has joined #opensolaris
[10:04:58] <tomww> Vratha: to leave out X11 is a bit of extra work, so if you don't have space or security reasons, install the whole thing (even without a graphics card)
[10:05:28] <Vratha> ok
[10:10:04] *** alanc_away has quit IRC
[10:10:07] *** alanc-away has quit IRC
[10:10:07] *** yippi has quit IRC
[10:10:30] *** alanc-away has joined #opensolaris
[10:10:37] *** alanc has joined #opensolaris
[10:12:29] *** silk has joined #opensolaris
[10:15:06] <Vratha> oh, hmm
[10:15:17] <Vratha> i thought Sun gave that away for use on solaris, but i think i'm wrong
[10:15:32] <Cyrille> gave what away?
[10:15:39] <Vratha> sun directory server
[10:16:32] <Cyrille> there's something in Solaris with a license for so many entries (don't remember the exact number), some assembly may be required though.
[10:16:41] <Cyrille> it's not in OpenSolaris though.
[10:16:51] <Vratha> ah, ok
[10:16:59] *** mega has joined #opensolaris
[10:17:02] <Vratha> well i guess i'll just do my usual ldap+kerberos setup :-/
[10:17:04] <quasi> Cyrille: I think that restriction is gone nowadays
[10:17:22] <Cyrille> quasi: really?
[10:17:25] <Vratha> only problem is that i can't do AD mappings for windows clients
[10:17:50] <quasi> Cyrille: yeah, I'm pretty sure I've seen it
[10:17:53] <Cyrille> an LDAP server is not an AD server (and some may argue, an AD server is not an LDAP one either)
[10:18:38] <Vratha> AD is pretty much ldap with different schemas
[10:18:45] <Vratha> and builtin kerberos support
[10:19:05] <Gr|ffous> and a departure from the rfcs to enable multimaster
[10:19:18] <Vratha> and probably a few more bells and whistles, but more-or-less a directory server
[10:19:38] <Vratha> Gr|ffous: yeah, that was covered under the "different schemas" :)
[10:20:29] <Gr|ffous> heh, I'm less subtle about that one
[10:21:38] * tsoome1 is away - bbl - [AcidJazz - Log ON]
[10:22:49] <ofu> i cant find any information about the promised 8socket intel systems... what kind of chipset will they use?
[10:23:08] <Gr|ffous> hehe, I think you're several months early there
[10:23:32] <ofu> i hope, netburst is history then
[10:23:39] *** rgl has joined #opensolaris
[10:23:48] <Gr|ffous> besides, are you sure it's 8 socket. I heard "8-way"
[10:24:29] <ofu> (where larger implies greater than 4 socket - and no, I'm not talking about 5 socket).
[10:24:34] <ofu> http://blogs.sun.com/jonathan/entry/but_we_did_not_hug
[10:24:59] <Gr|ffous> ah.. well spotted :)
[10:25:05] <ofu> so i assume 8 sockets, which is pretty interesting
[10:25:11] <Gr|ffous> absolutely
[10:25:13] <ofu> so you can directly compare x4600 to it
[10:26:16] <Gr|ffous> I'm not so sure I feel that easy with supporting Intel yet.
[10:26:20] <Vratha> hmm, is that sun directory server able to act as a replacement to active directory?  i'm just mainly curious
[10:26:29] <Gr|ffous> After months of sun bagging intel, to *now be selling them...
[10:26:52] <Gr|ffous> it just doesn't quite seem right. I wonder how they'll market it, so they don't undercut themselves
[10:27:21] <Gr|ffous> A partnership with Intel makes a lot of sense, but selling servers with Xeons...hmmm
[10:27:28] <ofu> whats the problem? The core-architecture just feels superior to opterons
[10:27:40] <kirma> I really doubt if 4-8 CPU solutions from intel are half as good as Opterons
[10:27:41] <andersmo> "Nothing personal. Just business."
[10:27:43] <andersmo> =)
[10:27:55] <ofu> after years of struggling (selling netburst), the core stuff is better
[10:28:24] <andersmo> single socket core 2 dualcores are pretty decent bang for $$.
[10:28:37] <kirma> they still haven't designed proper SMP bus architecture
[10:28:49] <Gr|ffous> well, not according to Sun's own benchmarks. So either they've been lying, or they have been selling the wrong product all this time. That doesn't 'feel' good at all :)
[10:29:12] <andersmo> Gr|ffous: You can prove anything given the right benchmark. =)
[10:29:19] <ofu> kirma: thats why i suppose intel wants to learn from sun how to build large boxes
[10:29:26] <Gr|ffous> yeah, but you can't then go back on it.
[10:29:27] <andersmo> They've been marketing, just like everyone else. =)
[10:29:34] <Gr|ffous> Not without looking liking an ass.
[10:29:49] <andersmo> Oh yes, you can, it's just even more marketing. =)
[10:29:54] <Gr|ffous> hahaha
[10:30:01] <ofu> and i think, those boxes will be core-architecture with csi-connections
[10:30:04] <kirma> well, they really should use something more NUMAish like hypertransport, and I don't really think they need Sun for that.
[10:30:15] <Gr|ffous> csi?
[10:30:16] <andersmo> Looking like an ass every once in a while is also part of the marketroids' job. =)
[10:30:31] <ofu> Gr|ffous: this new serial point-to-point interconnect
[10:30:32] <kirma> csi?
[10:30:38] <ofu> we can do something like ht, too
[10:30:49] <ofu> content streaming architecture?
[10:31:06] <ofu> Common Serial Interface (CSI)
[10:31:09] <ofu> thats it
[10:31:53] <Gr|ffous> something like infiniband?
[10:31:54] <kirma> "It is expected to release between 2008 to 2009"
[10:32:44] *** uebayasi has quit IRC
[10:33:59] <andersmo> Gr|ffous: more like hypertransport, I figure?
[10:34:24] <andersmo> well, dunno.
[10:34:46] *** simford has quit IRC
[10:35:14] <Gr|ffous> I wonder when the x86 makers are going to discover CMT.
[10:35:26] <Gr|ffous> clearly they finally agree multicore is the way forward
[10:36:07] *** Qapf has joined #opensolaris
[10:36:56] <Qapf> anyone toy around with tuning samba and zfs and tinkering with SO_RCVBUF? im tinkering with it myself and i just wanted to hear if someone else had found a sweet spot for the value
[10:39:14] <Gr|ffous> I have to confess I know nothing about it myself, but then I've never used sparc ;)
[10:39:54] <Gr|ffous> I gather it all works as advertised, not like hyperthreading
[10:40:04] <Vratha> hmmm
[10:40:15] <Vratha> well i found out i can download the sun directory server without any charge i guess?
[10:40:29] <Vratha> but i'd have to pay for the sun enterprise version of that i think... man i'm confused
[10:40:35] *** qdk has joined #opensolaris
[10:41:12] *** razrX_ is now known as razrX
[10:42:38] <Cyrille> no, it seems to be free. What enterprise version are you talking about?
[10:42:55] <Vratha> http://globalspecials.sun.com/servlet/ControllerServlet?Action=DisplayPage&Locale=en_US&id=ProductDetailsPage&SiteID=sunstor&productID=36271100&Env=BASE#
[10:43:08] <Vratha> which is part of http://www.sun.com/software/products/directory_srvr_ee/index.xml
[10:43:35] <Vratha> but either way, it still looks like you have to have active directory in order to support windows clients, which is really annoying
[10:44:16] <Gr|ffous> oh, really? that's a bummer
[10:44:23] <Vratha> well i'm only guessing
[10:44:46] <Vratha> oh, but a site i found mentioned this unfortunate belief to be true :(
[10:44:58] <Gr|ffous> I don't know much about their directory, I would like to know more though
[10:45:51] <Cyrille> if you want to download even from your page you end up in http://www.sun.com/download/products.xml?id=4373c4c2 which states it's free.
[10:46:10] <Cyrille> That other page is most likely the license fee for assurance and support.
[10:46:11] <Vratha> oh
[10:46:16] <Vratha> ah, ok
[10:46:29] <Vratha> well that's always nice then
[10:47:13] <Cyrille> and yes, I'm fairly certain that you can't authenticate windows users on windows clients without an AD.
[10:47:26] <Vratha> you can with pgina
[10:47:29] <Vratha> http://pgina.org
[10:47:35] <Vratha> but you miss a lot of the good stuff ADS offers
[10:47:45] <Cyrille> ^without an AD^with a Sun DS
[10:47:54] <Cyrille> if that makes you feel better ;-)
[10:48:05] <Vratha> heh, well you can with pgina
[10:48:10] <Vratha> just you don't get the good stuff :-/
[10:48:21] <Cyrille> there's good stuff that AD offers?
[10:48:22] <Vratha> MS really knew how to make it so no one else could do that
[10:48:27] *** nostoi has joined #opensolaris
[10:48:37] <Vratha> yeah, like proper roaming profiles, group system policies, etc
[10:49:24] *** broadcast has quit IRC
[10:50:10] *** broadcast has joined #opensolaris
[10:51:31] *** Gman has joined #opensolaris
[10:52:40] *** jopi has quit IRC
[10:53:38] * Gr|ffous waves at Gman
[10:54:12] <Gman> hi Gr|ffous
[10:54:35] <Gr|ffous> settled back in again yet?
[10:54:48] <Gman> slowly settling back in
[10:55:06] *** debugger has joined #opensolaris
[11:02:08] *** agliodbs has quit IRC
[11:05:00] *** nostoi has quit IRC
[11:05:18] *** darrenm has joined #opensolaris
[11:08:14] *** Vratha has quit IRC
[11:09:11] *** bxnp-work has joined #opensolaris
[11:09:25] <bxnp-work> cool just found this channel on the via the website
[11:09:28] <bxnp-work> hi everybody
[11:11:13] <Gr|ffous> hi bxnp-work
[11:11:16] *** rgl has quit IRC
[11:11:31] <bxnp-work> hi Gr|ffous
[11:11:53] <bxnp-work> well, just reading the website and got currious, so i thought i give this channel a visit
[11:12:00] <Gr|ffous> welcome
[11:12:22] <Gr|ffous> the topic isn't all that exciting, but the conversation can be at times
[11:14:05] *** LeftWing changes topic to "Latest builds: SXCR: 55 | ON: 55 | WOOOO!"
[11:14:12] <LeftWing> Now it's 12% more exciting.
[11:14:14] <Gr|ffous> lol
[11:15:12] <Gr|ffous> I must check out what's new in 55
[11:15:35] <cmihai>  Bleeding Internet Exploiter 7. New interface makes self signed certs look like 404
[11:15:46] <LeftWing> haha
[11:17:33] <bxnp-work> so i see that open solaris comes with the kde desktop is that correct
[11:17:49] *** jopi has joined #opensolaris
[11:18:02] <jopi> good morning
[11:18:03] <cmihai> So now I'm getting emails: should I continue even though Iexplore tells me: "NOT RECOMMENDED" in huge red fonts?
[11:18:03] <Gr|ffous> not a heck a lot really, but the looks
[11:18:21] <cmihai> Even the address bar turns red.. sheesh.
[11:18:22] <Gr|ffous> bxnp-work, I don't think so. I think a new community has started, to make that happen though
[11:18:36] <quasi> cmihai: no, this is now #windos-lusers ;)
[11:18:39] *** gallium has joined #opensolaris
[11:18:44] <quasi> s/now/not/
[11:18:58] <bxnp-work> and what is the purpose of ths channel, is this the general opensolaris irc channel
[11:19:12] <Gr|ffous> support and chat about opensolaris
[11:19:27] <Gr|ffous> cmihai, may not really be selling it all that well here...
[11:19:30] <Gr|ffous> bad cmihai
[11:20:24] <bxnp-work> cool Gr|ffous, well i think i come here more often, for now i am going to idle a bit, cause i have to work again
[11:20:33] <bxnp-work> :) thanks for your nice welcome Gr|ffous
[11:20:46] <Gr|ffous> sure :)
[11:21:52] <cmihai> Gr|ffous: sure, but it's a fun part of having Windows clients on your network. Sometimes they have to push teh red button :)
[11:22:29] *** gallium has quit IRC
[11:23:56] *** Gman has quit IRC
[11:26:55] *** yongsun has quit IRC
[11:35:18] *** Gman has joined #opensolaris
[11:35:39] * tsoome1 is away - bbl - [AcidJazz - Log ON]
[11:37:57] <Cyrille> is that the one with "Do Not Press" written below it?
[11:41:25] <cmihai> Uhuh
[11:45:10] *** noyb has quit IRC
[11:46:16] *** yippi has joined #opensolaris
[11:47:23] *** damienc has joined #opensolaris
[11:52:27] *** kb7sqi has quit IRC
[12:00:40] *** Tpenta has quit IRC
[12:00:52] *** Dar has quit IRC
[12:01:40] *** CapRiCoRN^80 has joined #opensolaris
[12:02:03] *** rgl has joined #opensolaris
[12:02:24] <CapRiCoRN^80> hi i m trying to build fftw and getting error below
[12:02:32] <CapRiCoRN^80> checking for C compiler default output... configure: error: C compiler cannot create executables
[12:14:23] *** LordKing has joined #opensolaris
[12:18:30] *** Tpenta has joined #opensolaris
[12:18:35] *** debugger has quit IRC
[12:20:07] *** yippi has quit IRC
[12:21:05] *** Kmays has quit IRC
[12:22:03] <quasi> CapRiCoRN^80: what's the output of which cc ?
[12:24:18] *** ChanServ sets mode: +o Tpenta
[12:24:27] <richlowe> mornin' Tpenta
[12:24:40] <CapRiCoRN^80> . /usr/ucb/cc
[12:24:42] <richlowe> bugger, "evenin'" rather.
[12:25:07] <richlowe> /usr/ucb/cc is not the compiler, but a decade old hack that should have vanished already.
[12:25:10] <Tpenta> rich
[12:25:14] <richlowe> cc should be in /opt/SUNWspro/bin, or gcc is in /usr/sfw/bin
[12:25:54] *** axxl has joined #opensolaris
[12:26:23] *** richlowe has quit IRC
[12:26:52] <CapRiCoRN^80> gcc is in /usr/bin
[12:27:32] <CapRiCoRN^80> . /usr/ucb/cc
[12:29:17] <Cyrille> but that's not the proper C compiler. You can find the Sun Studio one in /opt/SUNWspro/bin provided you've installed it.
[12:30:06] *** richlowe has joined #opensolaris
[12:30:08] <CapRiCoRN^80> no i dont have /opt/SUNWspro/bin
[12:31:06] <quasi> CapRiCoRN^80: /usr/ucb/cc is a shell script that usually fails because there's nothing in /usr/ccs/bin/ucbcc
[12:31:29] <CapRiCoRN^80> ok
[12:31:34] <PerterB> then use gcc, eg "CC=/usr/sfw/bin/gcc ./configure" (depending on what shell you're using)
[12:31:53] <PerterB> or download Studio 11
[12:32:07] <CapRiCoRN^80> bash
[12:32:32] <Cyrille> so CC=/path/to/gcc ./configure
[12:32:42] <andersmo> hash bang slash bin slash bash ? (what's with all the "sh"'s? To compress better? =)
[12:33:21] <CapRiCoRN^80> Cyrille: i used -bash-3.00# CC=/usr/sfw/bin/gcc ./configure
[12:33:30] <Cyrille> and?
[12:33:36] <CapRiCoRN^80> n got error
[12:33:41] <CapRiCoRN^80> checking for Fortran 77 name-mangling scheme... configure: error: cannot compile a simple Fortran program
[12:33:57] <Cyrille> looks like it needs a fortran compiler then.
[12:34:24] <Cyrille> didn't think anyone or anything still did.
[12:35:20] <CapRiCoRN^80> i got g77
[12:35:27] <CapRiCoRN^80> .. /usr/bin/g77
[12:36:25] <Cyrille> look at the configure log then, you might get more info on what's actually failing when compiling the fortran program.
[12:36:46] <CapRiCoRN^80> u r right
[12:36:53] <CapRiCoRN^80> but how can i get rid of it
[12:37:16] <Cyrille> by fixing whatever it is that doesn't work when it tries and compiles the program?
[12:37:45] <Cyrille> can't help you if we don't know exactly what's going wrong (and getting "error: cannot compile" is not "knowing what's going wrong").
[12:38:10] *** broadcast has quit IRC
[12:38:23] *** broadcast has joined #opensolaris
[12:39:42] <CapRiCoRN^80> i can paste u config.log
[12:39:55] <Cyrille> you could do that, in a pastebin please.
[12:40:03] <CapRiCoRN^80> ok
[12:40:21] <CapRiCoRN^80> pastebin address ?
[12:40:29] <cmihai> paste.lisp.org is nice
[12:41:07] <Cyrille> pastebin.ca, what he said, you take your pick.
[12:41:30] <Cyrille> there isn't exactly a shortage of such things.
[12:42:26] <CapRiCoRN^80> Cyrille: it works . after changing my path settings .
[12:42:36] <Cyrille> good
[12:43:01] <CapRiCoRN^80> thx
[12:45:25] <cmihai> The audience is listening.
[12:45:35] <Cyrille> :-)
[12:51:37] *** triplah_ has quit IRC
[12:52:17] *** Gman has quit IRC
[13:03:24] *** eugene_ has joined #opensolaris
[13:04:39] <eugene_> jmcp: hiya
[13:05:31] *** kb7sqi has joined #opensolaris
[13:06:04] <jmcp> hi eugene_
[13:06:07] <jmcp> long time no see!
[13:08:04] <eugene_> jmcp: yeah!
[13:08:24] <eugene_> jmcp: sorry for not contacting you in sydney. the conf kept me so busy.
[13:13:53] *** bengtf has joined #opensolaris
[13:16:27] *** gm152 has joined #opensolaris
[13:17:18] *** broadcast has quit IRC
[13:17:19] *** darrenm has quit IRC
[13:17:19] *** Tpenta has quit IRC
[13:17:41] *** Tpenta has joined #opensolaris
[13:20:44] <quasi> http://blogs.sun.com/roumen/entry/sun_on_youtube ;)
[13:24:32] <jmcp> eugene_: LCA?
[13:24:38] *** mega_ has joined #opensolaris
[13:24:49] <jmcp> eugene_: as long as you caught up with gman and the Humbug folks I won't mind :)
[13:26:59] <Peanut> I don't think I should be watching guys in their underwear from work ;-)
[13:27:25] <Peanut> Quasi: that link needs a 'NSFW' ;-)
[13:27:34] <Cyrille> or watching guys in their underwear period.
[13:28:17] <Peanut> Yup
[13:28:39] *** ChanServ sets mode: +o Tpenta
[13:29:36] *** uncertainty has quit IRC
[13:29:43] <eugene_> jmcp: yes
[13:29:51] <eugene_> jmcp: hehe i met up with gman :)
[13:31:23] <jmcp> cool
[13:31:50] <eugene_> cool guy
[13:31:51] *** qdk has quit IRC
[13:34:34] *** sickness has joined #opensolaris
[13:35:30] <sickness> morning all
[13:35:49] <Stric> Anyone seen any form of benchmark with a T1000/T2000 as ZFS+NFS server? I'm having trouble finding such..
[13:35:59] <quasi> Peanut: damn, picky workplace you have
[13:39:36] *** kloczek has quit IRC
[13:40:27] *** klocze2 has joined #opensolaris
[13:40:48] *** pygi has joined #opensolaris
[13:40:49] <jmcp> Stric: what, like joyent.com?
[13:40:51] <pygi> hello folks
[13:41:04] *** mega has quit IRC
[13:41:23] *** CapRiCoRN^80 has quit IRC
[13:41:58] <pygi> anyone willing to talk about solaris scsi layer? :)
[13:43:35] <jmcp> yes
[13:43:40] <Stric> jmcp: got anything more specific than that? maybe my google karma is bad, but I can't find much ;)
[13:44:09] <jmcp> Stric: sorry :( ... but have you tried Roch's and Richard Elling's blogs? b.s.c/roch and b.s.c/relling
[13:44:33] <jmcp> pygi: depending on your question, I might even be able to answer it intelligently :)
[13:44:50] *** bzcrib has quit IRC
[13:45:05] *** Risky_ has quit IRC
[13:45:33] *** Risky has joined #opensolaris
[13:46:06] <pygi> jmcp, in basics, I'd like to port : http://libburnia.pykix.org/browser/libburn/trunk/libburn/sg-linux.c and this: http://libburnia.pykix.org/browser/libburn/trunk/libburn/os-linux.h
[13:47:03] <jmcp> how far have you got?
[13:47:20] <pygi> jmcp, nowhere, now I'm at the planning phase :)
[13:47:24] <pygi> that's why I'm asking anyway
[13:47:29] <jmcp> ok
[13:47:40] *** qdk has joined #opensolaris
[13:47:47] <jmcp> so ... why spend effort on this port when you could use Joerg's cdrtools and libraries?
[13:47:50] <kimc> is Solaris Volume Manager available for download ?
[13:47:59] <Doc> didnt you just answer your own question?
[13:48:04] <pygi> jmcp, because ... bleh :)
[13:48:25] <pygi> jmcp, just because it's better :)
[13:48:27] <richlowe> Are there any releases still supported that don't ship with SVM?
[13:48:36] <Doc> i'd write the bits on a cd-r with a pen if it would mean i could avoid joerg
[13:48:37] * richlowe thought 8 had, though it may have been unbundled still
[13:48:37] <jmcp> pygi: really? how
[13:48:46] <jmcp> Doc: now, now ....
[13:48:53] <pygi> jmcp, well, it surely isn't better for solaris atm, since no solaris support :)
[13:48:58] <Doc> richlowe: define "ship with"
[13:49:03] <pygi> Doc, hehe :-)
[13:49:08] <Doc> sol 8 and below it was unbundled, but on the CD
[13:49:10] <richlowe> Doc: "on same media as".
[13:49:17] <pygi> Doc, on linux you can, and partially on freebsd as well
[13:49:19] <kimc> i can't find a way to get SVM anywhere
[13:49:27] <jmcp> pygi: so what information are you actually trying to find out?
[13:49:30] <Stric> jmcp: hm... seems to be mostly generic zfs stuff.. what I was interested in was if T2000 is a good or bad choice as file server (other options are for instance v240)
[13:49:33] <richlowe> Doc: since I recalled it being unbundled but forgot when it ceased to be. :)
[13:49:40] <jmcp> Stric: uh ... .no idea
[13:49:45] <Doc> all versions of SVM have shipped bundled
[13:49:51] <Doc> SDS was unbundled tho :)
[13:49:52] <pygi> jmcp, some docs about the scsi layer so I could port it :)
[13:49:56] <Stric> jmcp: thanks anyway though ;)
[13:50:05] <jmcp> Stric: personally, I'd be choosing a box which had the potential for massive io throughput
[13:50:06] <richlowe> Doc: I was also ignoring the name-of-the week stuff. :)
[13:50:19] <Stric> jmcp: which T2000 seem to have..
[13:50:20] <jmcp> pygi: docs.sun.com, look for "Writing Device Drivers"
[13:50:21] <pygi> jmcp, or someone willing to step up and just do it who is more familiar with opensolaris then I am :)
[13:50:39] <pygi> jmcp, will do, thanks
[13:50:53] <Doc> hmm.. you could probably use a pencil and call it a CD-RW
[13:51:02] <Stric> x4500 would be nice, but it's too high up
[13:51:07] <jmcp> pygi: that and the source - usr/src/uts/common/io/scsi
[13:51:40] <pygi> k, will look
[13:55:47] <pygi> thanks :)
[13:55:55] <jmcp> you're welcome
[13:56:04] <jmcp> now if I could only help somebody for real ....
[13:56:10] <pygi> let's hope a port can happen :)
[13:56:28] <pygi> jmcp, heh xD
[13:56:39] <pygi> I'll happily delegate the porting task xD
[13:58:19] *** gm152 has quit IRC
[13:59:13] *** Dimas4u has joined #opensolaris
[14:00:37] <Dimas4u> any brazillian in DA house?
[14:01:18] * jmcp sleeps
[14:02:48] *** pygi has left #opensolaris
[14:06:28] *** uebayasi has joined #opensolaris
[14:12:18] *** BadKarma is now known as iThink
[14:12:31] *** jamesd has joined #opensolaris
[14:12:31] *** ChanServ sets mode: +o jamesd
[14:23:58] *** compukid has quit IRC
[14:26:37] *** compukid has joined #opensolaris
[14:28:02] *** kszwed has joined #opensolaris
[14:33:57] *** bzcrib has joined #opensolaris
[14:34:02] <Stric> Hm.. Anyone DTracey around? If I'd like to see how often for instance fletcher_2_native (in zfs) is called, how would I go about doing that?
[14:35:13] <richlowe> dtrace -nfbt::fletcher_2_native:entry { @[probefunc] = count(); }'
[14:35:15] <richlowe> wait.
[14:35:16] <richlowe> wait more.
[14:35:17] <richlowe> ^C
[14:35:22] *** tsoome1 has quit IRC
[14:35:41] *** tsoome has joined #opensolaris
[14:36:08] <Stric> where is the first ' supposed to go? :)
[14:36:29] <richlowe> before the 'fbt'
[14:36:30] <richlowe> sorry about that.
[14:36:49] <richlowe> well, that gives you a count, not a frequency I guess, too. :)
[14:37:28] <Stric> I think I can calculate frequency given count and time :)
[14:37:29] <Stric> thanks!
[14:37:33] *** tsoome1 has joined #opensolaris
[14:37:54] * Stric is investigating stuff
[14:51:33] *** tsoome has quit IRC
[14:52:04] *** tsoome1 is now known as tsoome
[15:16:47] *** wutzel has joined #opensolaris
[15:28:28] *** MikeTLiv1 has quit IRC
[15:30:26] <myrkraverk> hello ;)
[15:30:37] <myrkraverk> is it possible to source level debug kernel modules?
[15:30:49] <twincest> there is no source level debugger for the kernel
[15:30:55] <twincest> but mdb has some nice helper functions to make it easier
[15:31:15] *** calumb has joined #opensolaris
[15:31:30] <myrkraverk> twincest: oh, k -- it's just that I don't really know intel asm very well ;/
[15:31:41] <myrkraverk> I've been afraid to try mdb ;-P
[15:31:48] <twincest> technically it's AT&T asm :P
[15:31:59] <myrkraverk> twincest: I meant the cpu -- I know mips
[15:32:36] <myrkraverk> but apart from that, how would I debug I module I'm trying to load?
[15:32:46] <twincest> what is wrong with it?
[15:32:59] <myrkraverk> the module? that's what I don't knom ;(
[15:33:14] <twincest> there must be something wrong or you wouldn't be trying to debug it
[15:33:22] <myrkraverk> for some reason, the qotd_2 -- from the tutorial -- fails to load for me ;(
[15:33:38] <myrkraverk> twincest: yes, but I don't know *what* ;(
[15:34:07] <myrkraverk> twincest: afaict, there's no difference between the source I wrote, and the one in the tutorial
[15:34:34] <axisys> can I install b55 on disk 1 while I am running sol 10 on disk 0 on my workstation?
[15:34:51] <axisys> i am looking for something like live install
[15:34:54] <PerterB> yes... live upgrade
[15:35:16] <twincest> LU to b55 on the workstation
[15:35:19] <axisys> so basically i am upgrading disk 1 to disk 1 ?
[15:35:20] <twincest> then move the disk to the other machine
[15:35:37] <twincest> since lu doesn't touch the running system it should work fine
[15:35:38] *** osdummy has joined #opensolaris
[15:35:58] <axisys> twincest, PerterB  thnx guys
[15:37:07] <axisys> twincest: I can pull the b55 os from a jumpstart server thru nfs correct?
[15:37:14] <axisys> while doign lu
[15:37:18] <axisys> doing*
[15:37:38] <twincest> i would assume so, it just needs the contents of the install media
[15:39:45] *** polk__ has quit IRC
[15:44:50] <richlowe> Yes.
[15:47:17] *** CryptoDivinity has joined #opensolaris
[15:49:32] *** mrdeviant has joined #opensolaris
[15:51:49] <Vanuatoo> Will Solaris Express run fine on Core 2 Duo?
[15:52:08] <cmihai> depends
[15:52:19] <cmihai> Proc is fine, other hardware should be in the HCL
[15:52:28] <Vanuatoo> cmihai: I understand
[15:53:05] <Vanuatoo> What image should I use?
[15:53:11] <cmihai> Depends on what you want.
[15:53:17] <cmihai> Solaris Express, SXCR..
[15:53:17] *** uncertainty has joined #opensolaris
[15:53:26] <Vanuatoo> what architecture I mean
[15:53:31] <cmihai> Well, x86 duh.
[15:53:41] <cmihai> You don't have a SPARC that's for sure :)
[15:53:47] <Vanuatoo> Will it run in 64 bit mode like on amd64?
[15:53:56] <cmihai> Core 2 Duo is EM64T and that's AMD64 and that's x86-64 and that's x86.
[15:54:02] <cmihai> And the answer is yes.
[15:54:14] <cmihai> cmsol% isainfo -vvv
[15:54:15] <cmihai> 64-bit amd64 applications
[15:54:23] <cmihai> Same proc.
[15:54:28] <Error_e^ipi> em64t is just intel's implimentation of amd64
[15:54:42] <cmihai> Vanuatoo: so you should be fine.
[15:55:09] <cmihai> Vanuatoo: hell, Sun sells Opteron (AMD64 architecture)
[15:55:14] <cmihai> (servers)
[15:55:25] <mrdeviant> and soon they'll be selling intel chips too
[15:55:32] <myrkraverk> Vanuatoo: it works well for me
[15:56:34] *** nrubsig has joined #opensolaris
[15:56:34] *** ChanServ sets mode: +o nrubsig
[15:56:47] <nrubsig> alanc: ping!
[15:56:48] <Vanuatoo> I have amd64, so it boots 64 bit kernel and all of the userland is 32bit. I wondered if it would be the same with Conroe
[15:57:04] <cmihai> Vanuatoo: virtuall ALL UNIX systems do that.
[15:57:07] <cmihai> MacOS does that.
[15:57:12] <cmihai> Solaris on SPARC does that, etc.
[15:57:21] <nrubsig> jamesd: ping!
[15:57:32] <cmihai> There are limitations to 64 bit systems (more memory usage, etc) so unless you have a justified reason to use 64 bit apps.. don't.
[15:57:46] <cmihai> So it's fine to go with 64 bit kernel, some 32 bit userland and some 64 bit (critical apps)
[15:57:51] <cmihai> Eg: databases, etc.\
[15:57:59] <cmihai> So don't worry, it's fine, there isn't another way.
[15:58:08] <cmihai> And NO, 64 bit apps aren't faster.
[15:58:15] <nrubsig> Can anyone please look at http://www.opensolaris.org/os/project/ksh93-integration/announcements/ and play the human spellchecker for the newest entry ("Making progress...") ?
[15:58:25] <nrubsig> cmihai: that is not true
[15:58:35] <nrubsig> cmihai: 64bit apps can be faster in some cases.
[15:58:40] <cmihai> nrubsig: in most aren't.
[15:58:41] <delewis> cmihai, in the case of AMD64 you have more registers in 64-bit mode.
[15:58:50] <nrubsig> delewis: ping!
[15:58:58] <cmihai> nrubsig: for most userland it really isn't justified.
[15:58:58] <Vanuatoo> Is it possible to instruct my Solaris Express install to boot 32 bit kernel?
[15:59:03] <delewis> and AMD64 has various memory models so that you don't always have to use 64-bit pointers in a 64-bit application.
[15:59:13] <nrubsig> cmihai: http://www.opensolaris.org/os/project/ksh93-integration/announcements/ any comments ?
[15:59:21] <cmihai> Woohoo?!
[15:59:35] <nrubsig> cmihai: I mean the text, syntax, spelling...
[15:59:50] <cmihai> Uh, don't know, I'm not ispell :P
[16:00:00] <mrdeviant> i've never seen so much noise made about a shell that almost no one uses :P
[16:00:01] * nrubsig crosses his eyes
[16:00:17] <Odin-LAP> cmihai: You aren't? Oh, we've been under the wrong impression this whole time, then...
[16:00:25] <Vanuatoo> Please make bash the default shell for every user and root
[16:00:33] <cmihai> Odin-LAP: nah, find some other "ortographic taliban" :))
[16:00:33] <Error_e^ipi> never!
[16:00:39] <Stric> mrdeviant: haven't you heard? everyone should stop using perl, awk, java, python etc and use ksh93 instead
[16:00:41] <quasi> mrdeviant: on solaris it gets used quite a bit
[16:00:42] <cmihai> (It's how they call spelling nazis in Spanish btw) :)
[16:01:03] <Odin-LAP> cmihai: "Orthographic", and it's something the english-speaking world still hasn't learnt about...
[16:01:04] <nrubsig> Vanuatoo: Sure, as soon as Bash becomes posix-compilant, supports multibyte locales properly and stops sucking 4MB per instance
[16:01:05] <mrdeviant> quasi, not by any of the solaris admins i know. in fact, the only admins i know who use ksh are AIX admins
[16:01:41] <quasi> mrdeviant: look at the scripts that come with solaris - ksh all over the place
[16:01:42] <Error_e^ipi> could that possibly be because of the ancient version of ksh that solaris ships with?
[16:01:42] * delewis uses ksh88/ksh93
[16:01:45] <cmihai> And those use IBM's bastardized version of ksh :)
[16:01:46] <nrubsig> Vanuatoo: right now bash is a standard-violating, UTF-8-only and memory hungry junkyard.
[16:01:47] <delewis> then again, I also use AIX.
[16:01:48] * quasi uses ksh
[16:01:50] <cmihai> Most OpenBSD users use pdksh.
[16:01:58] <cmihai> (since it's the default shell on OpenBSD)
[16:02:05] <Odin-LAP> nrubsig: utf-8 only?
[16:02:17] <Vanuatoo> nrubsig: If we follow your logic, JDS should not be used neither
[16:02:31] <Vanuatoo> It's sucking memory and utf-8 aware
[16:02:37] <nrubsig> Odin-LAP: try ja_JP.PCK or cn_ZH.GB18030 with bash. These are non-unicode locales which widespeak usage.
[16:02:42] <Vanuatoo> Same goes to Xorg
[16:03:00] <delewis> Vanuatoo, neither of which are covered under POSIX, unlike a shell, which most Linux users tend to refer to as /bin/sh.
[16:03:02] *** Berny_ has joined #opensolaris
[16:03:04] <nrubsig> Vanuatoo: UTF-8-aware is good, multibyte-locale aware is better
[16:03:24] *** swa_mobil has quit IRC
[16:03:25] *** Berny has quit IRC
[16:03:27] <nrubsig> Vanuatoo: bash doesn't deal well with multibyte locales which are not based on utf-8
[16:03:28] <Odin-LAP> nrubsig: I have had problems with anything other than ASCII in bash...
[16:03:35] <cmihai> Took a while for SXCR: 55 to be released
[16:03:59] <Vanuatoo> in that case make code completion-aware all shells used by default
[16:04:00] <Odin-LAP> nrubsig: Including both ISO 8859-1(5) and UTF-8... 8)
[16:04:07] <quasi> cmihai: it was broken
[16:04:17] <cmihai> Right. Is it "Fixed" now?
[16:04:19] <Vanuatoo> I'm tired typing su and after that bash
[16:04:37] <PerterB> Vanuatoo: alias su='su root -c bash'  sorted
[16:04:58] <nrubsig> Vanuatoo: right now you can use ksh93 as /sbin/sh if you want... :-)
[16:04:58] *** RaD|Tz has joined #opensolaris
[16:04:59] <quasi> cmihai: I think the 55b was supposed to fix a couple of the really bad errors
[16:05:02] <delewis> ah. ha.
[16:05:04] <Vanuatoo> PerterB: thanks
[16:05:09] <Teknix> quasi: like what?
[16:05:09] <cmihai> Or will it be "The Downword Spiral" Now :)
[16:05:10] <delewis> it looks like gcc is not doing simple loop optimizations for SPARC.
[16:05:18] * delewis shakes head in shame.
[16:05:29] <Stric> Vanuatoo: some will complain now, but I've been using tcsh as root shell for many years without problems. This doesn't say there won't be problem for you, but there hasn't been problems for us.
[16:05:34] <quasi> Teknix: see the release notes ;)
[16:05:39] * Odin-LAP wonders when/if Sun Studio gets opened...
[16:05:52] <Stric> There was some issues in solaris 2.4 where some init scripts assumed you had /bin/sh as shell
[16:05:53] <cmihai> quasi: is it fixed now?
[16:05:59] <delewis> given a chunk of multi-dimensional column-wise code gcc is *not* reverse the loops to transform it into row-wise code.
[16:06:08] <delewis> Sun Studio does that :-)
[16:06:21] <delewis> s/reverse/reversing/
[16:06:42] <Odin-LAP> delewis: gcc has as its significant benefit the ability to work in disturbingly diverse environments.
[16:06:45] <cmihai> quasi: doesn't matter I guess, LiveUpgrade ftw :). But still.. is it worth bothering with the svn_55 on site?
[16:06:55] <Odin-LAP> delewis: That, and being free (as in FSF).
[16:07:06] *** uebayasi has quit IRC
[16:07:10] <delewis> Odin-LAP, and its still a compiler largely based on 80s compiler standards.
[16:07:15] *** bzcrib has quit IRC
[16:07:20] <Vanuatoo> Is gcc going to be upgraded to 4.x
[16:07:21] <Stric> delewis: I assume that was mostly for that speccpu02 test or whenever it was..
[16:07:43] *** Mazon is now known as mazon
[16:08:29] *** laca has joined #opensolaris
[16:08:33] <quasi> cmihai: I think it is if you grab 55b
[16:08:34] <Odin-LAP> delewis: I never said otherwise. I was pointing out the two reasons it gets used. ;)
[16:08:51] *** nrubsig has quit IRC
[16:10:47] <delewis> looks like gcc-4.1.1 code is doing the same thing.
[16:11:01] *** edwardocallaghan has joined #opensolaris
[16:11:04] <edwardocallaghan> Hi
[16:11:53] <cmihai> quasi: where's 55b?
[16:12:10] <myrkraverk> Odin-LAP: about the if -- the rumor is "yes"
[16:13:19] <quasi> cmihai: sol-nv-b55b-x86-dvd-iso-a.zip is what it says when I go to download 55 as a dvd
[16:13:35] * cmihai takes a closer look.
[16:13:37] <cmihai> Good point.
[16:14:27] <Odin-LAP> myrkraverk: Which leaves the question of "when"...
[16:14:55] <myrkraverk> Odin-LAP: yes :/
[16:15:40] <Odin-LAP> Huh. Sun-Intel "strategic alliance", you say?
[16:15:42] *** gustav3d has quit IRC
[16:16:07] *** PSG|Tartiflette has joined #opensolaris
[16:16:11] <PSG|Tartiflette> hi all
[16:16:17] <PSG|Tartiflette> i have a question
[16:16:22] <Odin-LAP> What's next, Sun and Apple realise they've been going down a disturbingly similar path and merge?
[16:16:37] <Odin-LAP> PSG|Tartiflette: Do ask it, then. Don't keep it in.
[16:16:52] <PSG|Tartiflette> dtrace is on solaris when we install solaris, or we must install dtrace after ?
[16:16:55] *** kszwed has quit IRC
[16:17:06] <Stric> no extra install
[16:17:07] <cmihai> PSG|Tartiflette: it's in Solaris 10.
[16:17:07] <Error_e^ipi> it's an integral part of solaris
[16:17:16] <cmihai> Like most that isn't JES or whatever.
[16:17:20] <Error_e^ipi> you don't need to install it, and you cant just rip it out
[16:17:35] <PSG|Tartiflette> ok
[16:17:36] <PSG|Tartiflette> thanks
[16:17:37] <PSG|Tartiflette> :)
[16:17:55] <Marv|LG> os x is prettier then solaris :)
[16:18:17] <Vanuatoo> Marv|LG: It does not matter
[16:18:25] <Odin-LAP> Marv|LG: It's also considerably more consumer-oriented than Solaris.
[16:18:28] <Vanuatoo> We've still got ZFS and DTrace
[16:18:35] <Marv|LG> I know I am just being a smart alec
[16:18:36] <quasi> Marv|LG: if you want a pretty os, perhaps you should just switch back to vista
[16:18:37] *** PSG|Tartiflette has left #opensolaris
[16:18:46] <Odin-LAP> Vanuatoo: ... both of which are rumoured to be in Leopard, actually. ;)
[16:18:48] <cmihai> Vanuatoo: no we don't ;)
[16:18:54] <cmihai> Leopard already has ZFS.
[16:18:54] <Odin-LAP> (DTrace confirmed, actually.)
[16:18:59] <cmihai> Heh
[16:19:06] <cmihai> FreeBSD has them also.
[16:19:11] <Marv|LG> dtrace in os x?
[16:19:19] <cmihai> Linux has a port of ZFS via FUSE... etc.
[16:19:20] <Error_e^ipi> Marv|LG: in leopard, yeah
[16:19:23] <Vanuatoo> That comes to the thing that MacOSX is better development platform than Soaris
[16:19:28] <Odin-LAP> Marv|LG: Yes, in the next version.
[16:19:31] <Error_e^ipi> cmihai: i wouldn't trust that with my life
[16:19:39] <cmihai> Error_e^ipi: duh
[16:19:53] <cmihai> Most people still don't trust ZFS.\
[16:19:53] <Odin-LAP> Vanuatoo: They're different things, different purposes...
[16:19:58] <mrdeviant> Vanuatoo, os x is not a better dev environment. it doesn't have mdb or dbx, and it uses the bizarre mach-o toolchain
[16:20:04] <Error_e^ipi> for that matter, i wouldn't trust _linux_ as far as i can throw it
[16:20:15] <cmihai> Let alone some port via FUSE to Linux...
[16:20:16] <Vanuatoo> mrdeviant: Consider I'm only Java developer
[16:20:17] <Cyrille> how far can you throw an OS?
[16:20:27] <cmihai> Vanuatoo: guess who makes Java.
[16:20:40] <Cyrille> then no OS is a better development platform if you're talking Java.
[16:20:45] <Odin-LAP> Cyrille: Depends. Is it on a CD, or is it a hardcopy of the source? :D
[16:20:54] <cmihai> Wouldn't it make more sense for SUN Solaris to be a better development platform for SUN JAVA? :)
[16:21:09] <cmihai> I can throw Ubuntu cd's pretty far.
[16:21:10] <mrdeviant> Vanuatoo, so am i (primarily). os x's java always lags behind other platforms
[16:21:10] <Vanuatoo> cmihai: You probably don't know that Java is 1.5 fast on windows as the desktop compared to solaris
[16:21:11] <cmihai> We do it for sport here :)
[16:21:16] <CryptoDivinity> you can trust linux, linux is our savior.
[16:21:27] <cmihai> Vanuatoo: bullcrap.
[16:21:34] <Cyrille> I don't trust saviours.
[16:21:38] <Vanuatoo> cmihai: Don't be a troll
[16:21:43] <Odin-LAP> Oooooh. Java fight!
[16:21:56] <ibb> the funny thing about zfs is that i had a dream, that i create but not mounted and it did not work just to find out i did not mount it to a location
[16:22:02] <Vanuatoo> I've used both of them and solaris as a desktop is in its infancy
[16:22:05] <Cyrille> "my Java can beat up your Java" "can not" "can too"
[16:22:21] <cmihai> Vanuatoo: you've just thrown a random number at me.
[16:22:22] <cmihai> 15.
[16:22:25] <cmihai> 1.5. PI. Whatever.
[16:22:38] <oxygene> Vanuatoo: switching topics every 3 lines and _then_ calling other people trolls.. nice one
[16:23:02] <Vanuatoo> cmihai: Do you use windows seriously?
[16:23:21] <Vanuatoo> I mean for java devlopment or using NetBeans on windows?
[16:23:21] <cmihai> Vanuatoo: no, I just administer 2000 of them.
[16:23:39] <Cyrille> don't which is worse, managing it or using it ;-)
[16:23:40] <Vanuatoo> Then how do you know that what I'm talking about is bullcrap
[16:23:52] <oxygene> Vanuatoo: give me dtrace/java for windows, then we can talk ;)
[16:24:17] <cmihai> Vanuatoo: since it's really a subjective thing. Java is faster _for you_ on _your_ version of Windows with _your_ version of Java on _your_ hardware.
[16:24:24] <Vanuatoo> Because Sun is making jdk for windows and jdk for solaris does not mean that solaris jdk is better on solaris
[16:24:54] <cmihai> Or maybe you're doing it Get The Facts style compareing it with some ancient SPARC. You really aren't saying. Besides, how does MS make a better devel platform for Java just cause it's faster for you?
[16:25:26] <Vanuatoo> Unfortunately windows makes 90% of the worlds desktops, so  I guess sun should follow the market
[16:25:44] <Cyrille> the underlying system/OS/hardware could be faster for everything, it's hard to compare an application on two different OSs and possibly on different hardware.
[16:25:45] <Vanuatoo> On the server I don't argue but on the desktop java shines only on windows
[16:25:54] <AbeFroman> of those 90%, i bet only .002% develop java
[16:26:03] <Vanuatoo> AbeFroman: I don't think so
[16:26:09] <Stric> I bet it's less
[16:26:16] <AbeFroman> 90% of the market includes people like my parents
[16:26:17] <ibb> Vanuatoo: dont follow the market , lead it
[16:26:31] <Vanuatoo> ibb: tell it to sun
[16:26:56] <ibb> i cant, they can
[16:27:10] <AbeFroman> grrr.  forgot not to feed the troll
[16:28:24] <Vanuatoo> I've been using Solaris Express as my primary java development platform for about a year and unfortunately see not significant improvements in desktop area
[16:28:39] *** bitvector2 has quit IRC
[16:28:44] <mrdeviant> that's because solaris makes a lousy desktop. it's a server os.
[16:28:47] <cmihai> What does "desktop area" have to do with anything... Sigh...
[16:29:24] <Vanuatoo> mrdeviant: People here tell me opposite
[16:29:40] <ibb> that is the point of this channel
[16:29:49] <cmihai> Vanuatoo: most people here are technical geeks.
[16:29:50] <Stric> to tell the opposite? :)
[16:29:51] <Vanuatoo> And I hear everday that solaris is going to be ubiqitous
[16:29:54] <ibb> encourage the use of opensolaris
[16:30:09] <cmihai> You can't expect my (or your needs) to apply to everyone.
[16:30:12] <Cyrille> support the use of opensolaris.
[16:30:33] *** uncertainty has quit IRC
[16:30:33] <Cyrille> we're not going door to door in pairs asking people if they will accept opensolaris as their OS and saviour.
[16:30:34] <cmihai> Or for regular users who only used WinXP to simply start using Solaris as a desktop
[16:30:59] <ibb> but we encourage its use, of course
[16:32:20] <Vanuatoo> This is the bug that is filed 9 months ago. http://bugs.opensolaris.org/view_bug.do?bug_id=6418690
[16:33:41] <Vanuatoo> Messenger is the critical part of the desktop. and the only messenger available in JDS lacks the ability to view your conversation log.
[16:33:56] <ibb> wouldnt this be a gaim bug?
[16:34:10] <Vanuatoo> ibb: On linux everything works fine
[16:34:15] *** mrdeviant has left #opensolaris
[16:34:48] <ibb> i had system freezes, when running 2 intensive encodings at once on dual core amd64
[16:35:11] <cmihai> Vanuatoo: I have (and never had) such issues viewing logs with GAIM on Solaris.
[16:35:40] <Vanuatoo> cmihai: You must be using older version of gaim
[16:35:47] *** calumb is now known as calAFK
[16:35:51] <Vanuatoo> 1.5 or something right?
[16:36:05] <cmihai> I doubt that it's me that's using the older version.
[16:36:11] <AbeFroman> so there's a gaim logviewing bug, and that's going to make you say windows makes a better desktop?
[16:36:36] <Vanuatoo> AbeFroman: No. Don't treat me like a windows fanboy
[16:36:37] <cmihai> GAIM from SXCR 54, 53 and 40 all work fine.
[16:36:51] <AbeFroman> then stop trolling
[16:36:59] <AbeFroman> nothing is perfect.
[16:37:06] <Vanuatoo> That makes me say that little attention is paid to desktop things in opensolaris
[16:37:09] <ibb> AbeFroman: he needs answers, and this is the place
[16:37:28] <ibb> Vanuatoo: opensolaris now uses kde as desktop package
[16:37:33] <AbeFroman> alanc would have something to say about the desktop i'm sure
[16:38:00] <Vanuatoo> alanc is the X maintainer As I know
[16:38:06] <Vanuatoo> gman is JDS guy
[16:38:23] <Cyrille> Vanuatoo: attention is paid, but there are only so many people working on it, and if as some are implying that particular bug isn't as obvious as you make it to be, and it's in an application that is hardly critical, it may not be dealt with very promptly.
[16:38:44] <Vanuatoo> Cyrille: gaim is not critical app?
[16:38:46] <AbeFroman> my apologies to the respective osol contributors
[16:38:59] <cmihai> I'd hardly consider GAIM or logging in GAIM critical.
[16:39:06] <cmihai> And it's NOT happening for me.
[16:39:07] <Cyrille> I wouldn't think the view log function of gaim is, no.
[16:39:12] <cmihai> This bug is either FIXED
[16:39:17] <cmihai> Or happening only to you.
[16:39:36] <cmihai> And nothing is stopping you from upgradeing to the latest SXCR or Vermillion
[16:39:39] <Vanuatoo> If you look at the bug report it's reported against snv_41 and snv_52
[16:39:39] <Cyrille> I could add anecdotally that I never even started gaim on my JDS desktop.
[16:39:52] <cmihai> Or just installing another GAIM from blastwave, pkgsrc, sources or whatever.
[16:40:01] *** tsoome has quit IRC
[16:40:03] <cmihai> (or simply use irssi + bitlbee :P)
[16:40:27] <Vanuatoo> I'm looking forward to sx 7/01 but it's delayed. I don't have the luxury running sxcr releases
[16:40:51] <edwardocallaghan> Try http://www.nexenta.com/ as a Desktop OS and stop complaining
[16:41:12] <Vanuatoo> edwardocallaghan: So it's prohibited to complain?
[16:41:19] <axisys> on ultra 20 no matter which of the two disk i boot from the grub considers it as root (hd0,0,a). is there a way i can choose disk from grub instead of changing in the bios?
[16:41:26] <edwardocallaghan> No
[16:41:35] <edwardocallaghan> But that's not what Solaris is for
[16:41:55] <Vanuatoo> I don't want to use Nexenta. I want to use Solaris from Sun and I want everything to work fine even on the desktop
[16:42:10] <edwardocallaghan> From what I can see, Server builds should not be the same as a Desktop build
[16:42:26] <Vanuatoo> edwardocallaghan: There is no server or desktop build.
[16:42:32] <Vanuatoo> There is one build
[16:43:25] <edwardocallaghan> I know, meaning that Solaris is for the server and the Desktop is only for admin things
[16:43:46] <edwardocallaghan> That's why packages are hard to work with ect..
[16:44:07] <edwardocallaghan> Nexenta solves both problem
[16:44:53] * alanc yawns
[16:45:05] <edwardocallaghan> How ever much I would like to see Solaris on the desktop, its not meant to be there unless you are a hardcore admin/coder and don't want some silly APT thing under the hood
[16:45:14] <edwardocallaghan> Any ways...
[16:45:31] <alanc> hmm, looks like I missed nrubsig
[16:46:19] *** calAFK is now known as calumb
[16:47:09] <Vanuatoo> Sorry for complaining about anoyances about solaris on the desktop. I guess (and you insist) that solaris in not the right choice for java development
[16:47:10] *** sahafeez has joined #opensolaris
[16:47:40] <alanc> Sun would like Solaris to be the right choice for developers
[16:47:59] <alanc> so if it's not, give feedback to desktop-discuss or opensolaris-discuss about what's lacking
[16:48:17] <Vanuatoo> I do time to time but see little progress
[16:48:18] <axisys> anyone use ultra 20 here?
[16:48:35] <twincest> edw: there is a difference between "it's not there yet" and "it's not meant to be there"
[16:48:37] <lasseoe> Vanuatoo: ignore that edward guy, has hasn't a clue what he's on about
[16:48:47] *** uncertainty has joined #opensolaris
[16:49:29] <AbeFroman> i have a u20 but only one disk in it
[16:49:33] <edwardocallaghan> lasseoe:Thanks for your feedback
[16:49:42] <lasseoe> you're welcome.
[16:50:02] <quasi> edwardocallaghan: I tend to agree with lasseoe on that subject
[16:50:08] <axisys> AbeFroman: if you move the disk to second slot can u still boot?
[16:50:16] *** bougie has joined #opensolaris
[16:50:17] <Vanuatoo> Actually I was expecting more constructive dialog here, not so much offense :)
[16:50:30] <alanc> JDS changes usually come in batches around new GNOME releases, so there wasn't a lot of change from nv_43-nv_50 in the desktop, then a whole bunch from nv_51 to nv_54 with the GNOME 2.16 / FireFox 2.0 / StarOffice 8 integrations
[16:50:47] <edwardocallaghan> I did not mean any offense at all, sorry if it sounded like that
[16:50:55] <bougie> hello
[16:51:00] <edwardocallaghan> Hi
[16:51:05] <Vanuatoo> alanc: I have snv_52 but still gnome 2.14
[16:51:12] <alanc> and for real desktop dialog, Gman doesn't wake up for a few hours still - it's 4:49 am in NZ
[16:51:37] <alanc> hmm, I thought 2.16 went in 51 or 52
[16:51:48] <alanc> actually - maybe it wasn't till 53
[16:51:57] <Vanuatoo> 53 IIRC
[16:52:44] <Vanuatoo> If I had time to research dtrace, I would not feel so desperate
[16:52:45] <Stric> 54 has 2.16 stuff at least
[16:53:20] <alanc> yeah  - my office desktop is on nv_53 at the moment, so that's the one I was thinking of
[16:53:28] <Stric> 2.16.1 to be more exact
[16:53:53] <edwardocallaghan> Be back later when both eyes wake up
[16:56:00] <Vanuatoo> alanc: Do you plan to upgrade to Xorg 7.2 when it's ready?
[16:56:21] <alanc> yes
[16:56:37] <alanc> just filed the ARC review for it yesterday in fact
[16:56:45] <alanc> have it almost ready to integrate
[16:56:49] <Vanuatoo> Is it decided what releases of X,Gnome will solaris 11 be based?
[16:57:06] <alanc> the ones that are current when it has feature freeze
[16:57:30] <alanc> they just haven't decided when Solaris 11 release date or feature freeze will be yet
[16:57:31] <Vanuatoo> So is it decided when Feature freeze happen?
[16:57:40] <Vanuatoo> ok, I understand
[16:59:06] <Vanuatoo> I've openoffice 2.1 installed on my sx 12/06 and when I open any excel file in it and then close the program it starts eating some memory and closes after 10 seconds. Is it a known bug?
[17:00:09] *** CryptoDivinity has quit IRC
[17:01:51] <AbeFroman> that happens to me too, but with and ods file
[17:02:06] *** jopi has quit IRC
[17:02:08] <AbeFroman> it takes just long enough to bring up the "force quit" dialogue, but then it goes away
[17:02:18] <Vanuatoo> AbeFroman: Exactly
[17:02:34] *** uncertainty has quit IRC
[17:04:59] <Vanuatoo> alanc: if I open some page in Firefox. for instance (http://www.sun.com/bigadmin/hcl/data/sol/) and do the fast scrolling up and down CPU usage is going high. Is it normal?
[17:06:21] <alanc> I don't know
[17:06:32] <cmihai> Heh, it's "normal" with sun.com :P
[17:06:53] <iThink> haha
[17:07:01] *** iThink is now known as GoodKarma
[17:07:33] <cmihai> It's better now, but the Java stuff on their page really caused spikes in the CPU load :)
[17:09:25] <GoodKarma> my uber-masta-smartphone powered by ...yes you've guessed it... Java crashed last Monday and I overslept....my folks at the office weren't happy to hear that...bummer
[17:09:31] <edwardocallaghan> Same happens on Fedora as well
[17:14:08] <Error_e^ipi> you mean it might be firefox that's full of shitty code?
[17:14:11] <Error_e^ipi> never.
[17:14:13] <Error_e^ipi> </sarcasm>
[17:14:30] <edwardocallaghan> Lynx ;)
[17:15:39] <Error_e^ipi> konq isn't bad
[17:15:44] <Error_e^ipi> opera's good
[17:16:20] *** calumb has quit IRC
[17:16:48] *** calumb has joined #opensolaris
[17:17:08] <edwardocallaghan> Why is it that GUI things are always less stable ?
[17:17:58] <Vanuatoo> edwardocallaghan: Because there are many things involved
[17:18:25] <alanc> because a GUI crash won't cost millions of dollars in downtime
[17:18:45] <Error_e^ipi> except on windows
[17:18:51] <edwardocallaghan> lol
[17:18:54] *** bondolo has joined #opensolaris
[17:19:45] <edwardocallaghan> But I don't get why a web browser should not be stable as anything after all these years ?
[17:20:00] *** calumb has quit IRC
[17:20:09] *** calumb has joined #opensolaris
[17:20:43] *** Theoden-Nexenta has joined #opensolaris
[17:20:52] *** xea has quit IRC
[17:21:36] *** xea has joined #opensolaris
[17:22:19] *** inhonho has joined #opensolaris
[17:26:22] *** eugene_ has left #opensolaris
[17:27:42] *** inhonho has quit IRC
[17:29:10] *** mega_ has quit IRC
[17:30:20] *** calumb has quit IRC
[17:30:29] *** calumb has joined #opensolaris
[17:31:00] *** calumb has quit IRC
[17:31:16] *** calumb has joined #opensolaris
[17:32:12] <alanc> edwardocallaghan: because people keep wanting new features and adding plugins from all sorts of different places that weren't tested with each other
[17:32:13] *** bengtf has quit IRC
[17:32:16] *** polk__ has joined #opensolaris
[17:33:25] <Vanuatoo> I think major problems arise when things need to be crossplatform
[17:36:39] <steleman> browser based webware: the way client-server was not supposed to work :-)
[17:38:11] *** tsoome has joined #opensolaris
[17:38:34] <alanc> when I still ran Mozilla, it crashed much less when I removed the flash plugin
[17:38:45] <AbeFroman> flashblock is your friend
[17:38:46] <AbeFroman> (on firefox)
[17:38:47] <alanc> the later flash plugins for Solaris seem much more stable in Firefox
[17:39:23] <steleman> i have to say that firefox on Solaris is now much more stable than on Linux
[17:39:28] <Vanuatoo> But still flash is the most bloated part of firefox and unfortunately many sites use flash
[17:39:38] <steleman> the latest flash plugin on Linux is crappy
[17:40:07] <quasi> Vanuatoo: acrobat is worse
[17:40:18] <edwardocallaghan> 9 is out for Linux and its only 7 for Solaris, P.S. I hate flash
[17:40:23] <Vanuatoo> I use evince and works great so far
[17:40:27] <edwardocallaghan> No flash for BSD :p
[17:40:40] <alanc> wow - unbloat - the Xorg 7.2 server module was released last night - changes since previous release: 1914 files changed, 77679 insertions(+), 111649 deletions(-)
[17:40:58] <edwardocallaghan> http://www.youtube.com/watch?v=5bPMwAHdJCg&mode=related&search= This is funny I think
[17:42:26] *** mazon is now known as Mazon
[17:43:06] *** dunc_ has joined #opensolaris
[17:46:20] <Vanuatoo> alanc That is a great news :)
[17:48:10] <edwardocallaghan> Will this mean 56 will have 7.2 ?
[17:48:37] <alanc> not unless you have a time machine
[17:48:47] <alanc> 56 ISO's were built last week
[17:48:50] *** sommerfeld has joined #opensolaris
[17:48:50] *** ChanServ sets mode: +o sommerfeld
[17:48:55] <alanc> 57 code freeze has passed for X
[17:49:03] <Vanuatoo> 57
[17:49:04] <alanc> I'm trying to get it in 58
[17:49:25] <Vanuatoo> For me it's sx 3/07
[17:49:28] <alanc> I think most of the deletion was dropping LBX support
[17:49:57] <alanc> now that the world has moved to ssh -X or NX instead
[17:50:22] <cmihai> NX is pretty much Linux centric and propriotary, isn't it?
[17:50:27] *** Vratha has joined #opensolaris
[17:50:39] <Vanuatoo> I use x11vnc to access my office machine from home. Is there any replacement for it
[17:50:40] <Vanuatoo> ?
[17:51:09] <cmihai> TightVNC, RealVNC, ssh -X, X via TCP, Tarantella, ...
[17:51:20] <alanc> cmihai: I thought NX was released under GPL
[17:51:31] <cmihai> alanc: there was a "FreeNX" thing, no idea how that is.
[17:51:36] <cmihai> I still think you need their client or something.
[17:51:38] <alanc> JDS includes Vino
[17:51:43] <cmihai> And it still seems to be Linux-centric.
[17:52:06] <Vratha> hi
[17:52:07] <steleman> FreeNX works on Solaris
[17:52:20] <cmihai> Does it now?
[17:52:21] <cmihai> Interesting.
[17:52:22] <alanc> for Solaris, we'll be recommending ssh -X, with compression for those who need even more bandwidth savings, since that's what we already have shipped
[17:52:33] <Vratha> so the VT6103 NIC doesn't seem to have a driver in opensolaris.. think it'll work anyway?  i don't quite have my new system that has that chipset
[17:52:35] <alanc> not many people used LBX anyway
[17:53:59] <Vratha> like, do any of you have the VT6103 chipset working under opensolaris?
[17:54:05] *** mikaeld_ has joined #opensolaris
[17:54:10] *** mikaeld has quit IRC
[17:54:50] <edwardocallaghan> 56 is out?
[17:54:55] <cmihai> No.
[17:55:04] <edwardocallaghan> Oh I better download it...
[17:55:15] <edwardocallaghan> Oh then when are the ISO going to be around ?
[17:55:44] *** Theoden-Nexenta has quit IRC
[17:57:45] <Vanuatoo> alanc: I like x11vnc because it can connect  to existing display. does ssh -X provide the same functionality
[17:58:02] <alanc> no - that would be vino
[17:58:10] *** Theoden-Nexenta has joined #opensolaris
[17:58:17] <alanc> 56 is being tested internally inside Sun
[17:58:22] <Vanuatoo> alanc: Has it been aded lately?
[17:58:23] <alanc> it's actually behind schedule
[17:58:29] <Vanuatoo> I don't have vino
[17:58:32] <alanc> Vino was added with GNOME 2.14 I thought
[17:58:43] <cmihai> You're on some anciant SXCR then
[17:58:47] <cmihai> Or SX something
[17:58:48] <AbeFroman> what do you use to connect to vino?
[17:59:01] <Vanuatoo> TightVNC
[17:59:04] <Vanuatoo> I ugess
[17:59:21] <AbeFroman> i used vncviewer but it was in /opt/sfw/bin so that's not really standard issue
[17:59:40] <alanc> there's a java client included too
[18:00:14] <Vanuatoo> what is the name of vino server executabel?
[18:00:17] <Vanuatoo> vino?
[18:01:20] <alanc> in the gnome menus it's called something like "Remote Desktop Sharing"
[18:01:56] <cmihai> /usr/bin/vino-preferences
[18:02:01] <Vanuatoo> I don't have the entry and find / -name vino returned null
[18:02:18] <alanc> I see /usr/bin/vino-session and /usr/bin/vino-preferences on nv_53
[18:02:21] <Vratha> so, with sun and ibm having their own directory servers, i wonder what the point of them is really, because if you want to login XP users on a domain, you pretty much have to run active directory
[18:02:28] <Vratha> can anyone help me understand that?
[18:02:35] <AbeFroman> embrace and extend?
[18:02:58] <cmihai> Vratha: there's at least 100 directory servers
[18:03:10] <cmihai> But like it or not, AD won.
[18:03:15] <Vratha> :-/
[18:03:30] <hile_> you use a decent directory/krb5 solution and make the AD domain a child of that that trusts your MIT krb5 domain for auth
[18:03:35] <hile_> at least, that's how it should be
[18:03:46] <Vratha> hmm
[18:03:55] <Vratha> only problem is that i'd have to buy windows server for that, haha
[18:04:00] <Vratha> and i'm a cheapass
[18:04:06] <hile_> and that's my problem how?
[18:04:23] <Vratha> what was the point of your question?
[18:04:28] <hile_> Eventually, I'd have to do the same if I do mor work with SSO,etc
[18:04:38] <hile_> but I sure as hell don't plan on paying retail
[18:05:07] <edwardocallaghan> Just going to move over to my SB2k be back later tonight
[18:05:08] <Vratha> well how do you get the price breaks.. not that i'm gonna buy... i only run a home network with SSO... at work we already have ADS
[18:05:16] <Vratha> and i bet part of our campus is using sun's DS
[18:05:25] <Vanuatoo> alanc: I'm on snv_52, so have to wait for next sx
[18:05:31] <hile_> Vratha, friend works at MS
[18:05:32] <edwardocallaghan> Take care thanks...
[18:05:36] *** stevel has joined #opensolaris
[18:05:36] *** ChanServ sets mode: +o stevel
[18:05:37] <Vratha> hile_: ah
[18:07:03] <hile_> and prior to that, the terms of employers' site licenses were favourable
[18:07:15] <Vratha> yeah
[18:07:24] <Vratha> well our university gets deals with MS i think
[18:07:29] <Vratha> so it's extremely cheap for them
[18:07:50] <hile_> the uni went to (and worked at) is doing exactly what i mentiond as the correct way to do it
[18:08:05] <Vratha> yeah, i'm wondering if our uni. is doing that
[18:08:14] *** deather_ has joined #opensolaris
[18:08:22] <Vratha> we use ADS all over campus for the windows boxes, but we do have some high level sun systems in use somewhere
[18:08:23] <hile_> they have an authoritative MIT KDC or 10 against which everything else talks
[18:08:55] *** edwardocallaghan has left #opensolaris
[18:09:00] <Vratha> do they also have a non active-directory DS that ADS uses?
[18:09:29] <hile_> yeah
[18:09:34] <Vratha> nice
[18:09:47] <hile_> all the AD-specific stuff lives in AD land
[18:09:59] <hile_> everything else it grabs from their LDAP servers
[18:10:04] <Vratha> does your campus also have a lot of solaris workstations or something?
[18:10:08] <hile_> the way i understand it
[18:10:13] <hile_> some dpeartments have some
[18:10:32] <hile_> but they have a metric fuckload of AIX, Solaris and (sadly) loonix for HPC
[18:11:23] <Vratha> why do they have all those unix systems?
[18:12:02] <Vratha> our campus is mainly XP clients with very few linux clients... the sun systems are not connected to the primary SSO systems
[18:12:40] <hile_> surely you've got enough neurons taht i don't have to justify that with an answer
[18:13:01] <hile_> not the least of which is "linux sucks when it comes to storage management"
[18:13:23] <Vratha> no, why do you have them all?  i mean we use windows and linux and freebsd quite heavily on the server side just fine in my department
[18:13:46] <hile_> what do you mean why have them all?
[18:13:57] <Vratha> i mean what are your uses for them?
[18:14:09] <Vratha> i guess now i've learned that storage is one reason
[18:14:28] <hile_> well, honestly, it's using the best resource for the situation...
[18:14:46] <Vratha> which is?  i'm trying to learn what that situation is out of curiosity
[18:14:55] <hile_> for TSM, GPFS, and DCE/DFS (migrating off), AIX wins hands down
[18:15:28] <Vratha> ah
[18:15:33] *** inhonho has joined #opensolaris
[18:15:37] <hile_> other things are on solaris (i don't know the breakdown)
[18:15:40] *** rgl has left #opensolaris
[18:15:51] *** dunc_ has quit IRC
[18:15:54] <Vratha> sounds good
[18:16:05] <Vratha> never used AIX :-/
[18:16:11] <Vratha> but GPFS sounds cool
[18:16:28] <Vratha> solaris have anything like that?
[18:17:12] <hile_> i don't know.
[18:18:18] <hile_> I'm pretty certain IBM Probably has clients/server ported to solaris though, much like they did for DCE/DFS
[18:18:26] <Vratha> hmm, looks like IBM makes GPFS available for linux too though
[18:18:56] <hile_> well just because they make it available doesn't mean it's a good solution by any means.
[18:19:03] <Vratha> yeah
[18:19:29] <Vratha> does AIX have anything like ZFS?
[18:19:46] <hile_> well AIX's LVM is nothing to be laughed at
[18:20:58] <Vratha> man that TSM looks nice
[18:21:10] <Stric> Solaris has been the joke up until now
[18:21:22] <Vratha> Stric: oh?
[18:21:29] <Stric> with volume managers
[18:21:43] <Stric> SVM is just crap compared to AIX' LVM for instance
[18:21:54] <Vratha> Stric: what does solaris have in terms of volume management now?  and yes, i'm new to opensolaris and solaris in general
[18:22:03] <Stric> Vratha: zfs
[18:22:07] <inhonho> Stric: how many schedulers does aix have ?
[18:22:08] <Vratha> oh
[18:22:29] <Stric> inhonho: I thought this was about filesystems
[18:22:30] <hile_> Stric: ZFS assuming that your backup solution supports backing it up
[18:22:33] <inhonho> hehehe
[18:22:49] <inhonho> i wanted to flame a little bit :(
[18:22:50] <inhonho> sorry
[18:22:51] <inhonho> ehhehehe
[18:22:58] <hile_> TSM (last i looked anyhow which was a while ago) supports it in that it'll back it up but strips the acls
[18:22:58] <Stric> and the answer is: no clue
[18:23:14] <Vratha> hile_: do you use the linux systems for high performance computing tasks?
[18:23:17] <Stric> TSM supports Solaris' POSIX draft ACLs I think
[18:23:38] <hile_> i don't
[18:23:41] <hile_> i don't do HPC shit
[18:23:43] <Stric> Pretty sure it does
[18:23:51] *** hile_ has quit IRC
[18:24:24] *** Dimas4u has quit IRC
[18:24:36] <Vratha> hmm
[18:24:54] <Stric> Vratha: linux is very big in hpc
[18:24:55] <Vratha> well sun needs to make a good storage manager if they don't have one already
[18:25:13] <Vratha> Stric: yeah, i'll be using linux for HPC work in a few months at an investment bank
[18:25:21] *** deather has quit IRC
[18:25:32] <Vratha> though i'm sure they use AIX for their storage needs since they love IBM htere
[18:25:42] <Vratha> even if they also love linux there
[18:26:26] *** slowhog has joined #opensolaris
[18:26:56] *** hile_ has joined #opensolaris
[18:27:53] <Vratha> i wonder when IBM will either absorb ZFS into their own AIX and Linux systems or come out with their own 128-bit FS
[18:27:58] <Stric> Vratha: compare stuff in http://blogs.sun.com/erickustarz/entry/creating_a_stripe_in_zfs .. (and the SVM thing is abbreviated, they left out the manual partitioning too)
[18:28:30] <Stric> mirroring is even more commands
[18:28:33] *** Cyrille has quit IRC
[18:29:14] <jamesd> now if you really want complex.. compare  raid5+0..   svm can't even do it and  zfs does in one command line
[18:29:30] <Vratha> hott
[18:29:36] <Stric> I'd like to do mirror(raidz,raidz), but zfs can't do that :(
[18:29:58] <Vratha> on my cheap ass home server that is on the way, i just want a single disk running ZFS :)
[18:30:09] <Stric> unless you create zvols out of the raidz, mirror two of them and then start carving out fs's
[18:31:23] <hile_> me, i just set up luns as mirrored on the arrays (yay for DMX3s) and stripe across them.
[18:31:32] <Vratha> so what kind of storage managers are there for ZFS other than TSM?  or TSM pretty much it
[18:31:40] <Stric> in AIX' LVM, you add a bunch of PVs (physical volumes; disks) to a volume group.. then when you create a logical volume (filesystem), you say if you want it mirrored or such, then it takes care of that part for you
[18:31:42] <hile_> TSM is not a storage manager
[18:31:53] <Stric> TSM is a file backup system
[18:32:07] <Stric> ZFS is a filesystem/volume manager
[18:32:28] <Vratha> ok, well what other backup systems are there?
[18:32:37] <Stric> take your pick
[18:32:50] <Vratha> any ones that fully support ZFS?
[18:33:26] *** uncertainty has joined #opensolaris
[18:33:27] <Stric> no clue
[18:33:58] <Vratha> http://www.opensolaris.org/os/community/zfs/faq/ -- if you scan for "ibm" you find talk about the backup managers
[18:34:14] <Vratha> they either don't support the ACLs or don't support complex ones
[18:34:15] *** yarihm has joined #OpenSolaris
[18:36:51] <richlowe> I thought there was one that now did?
[18:36:56] <richlowe> sadly, I'm forgetting which.
[18:37:07] <Stric> EMC Networker 7.3.2. backs up and restores ZFS file systems, including ZFS ACLs.
[18:37:11] <Vratha> could be; i don't know much about this area really
[18:37:17] <richlowe> Stric: that'd be the one then. ;)
[18:37:20] <Stric> according to that faq.. :P
[18:37:20] <Vratha> oh, sweet
[18:37:23] <hile_> Networker is owned by Sun, isn't it?
[18:37:29] <richlowe> hile_: resold, iirc.
[18:37:38] <Vratha> oh, i missed the first bullet point :)
[18:37:39] <richlowe> under some inobvious and frequently changing name no doubt.
[18:38:01] * Stric used Legato Networker about 10y ago
[18:38:24] <Stric> It just about never got the "I need tapes replaced now" mails right :P
[18:38:36] <Stric> Either it forgot to mail about it, or it spammed you to death
[18:39:44] <Stric> "Oh, right.. I forgot to tell you, but I needed new tapes last week.. since you didn't notice that I never told you, I better send you a mail each minute now"
[18:43:51] <Vratha> hahaha
[18:43:53] <richlowe> I've always liked UPS notifications for that.  "The power is off!" ... "The power is still off!" ... "The power is off yet more!"
[18:43:59] <elektronkind> Networker is owned by EMC
[18:44:11] <elektronkind> Sun resells it as Enterprise Backup Suite
[18:44:44] <elektronkind> EMC bought Legato like two years ago or something like that
[18:45:00] <elektronkind> But Sun's EBS product is completely Networker
[18:45:18] <Stric> richlowe: we found an old manual for a printer.. which had some numeric codes for what was happening.. one was "The printer is busy" and another was "The printer is very busy" .. ehm.
[18:45:45] <oxygene> Stric: "printer is printing" vs. "printer's buffers are all full"
[18:46:04] <Stric> I doubt it had buffers.. matrix printer from early stone age
[18:46:35] *** bxnp-work has quit IRC
[18:46:38] <oxygene> the one I have lying around has 60 bytes or so - one line of text
[18:49:33] <Vratha> so is sun actually making a comeback or is it still tanking?  at least it's seemed like it was tanking in the last few years
[18:51:40] *** Kush- has joined #opensolaris
[18:51:47] *** Kush- has left #opensolaris
[18:52:42] *** CosmicDJ has joined #OpenSolaris
[18:53:13] <oxygene> Vratha: weird question
[18:53:50] <Vratha> i mean i hadn't heard much about their use in recent year... thogh i'd seen quite a few sun workstations in a couple of our labs
[18:54:01] <oxygene> hm.. cute.. some java applet manages to hang firefox (and sometimes X) on b55
[18:54:04] <Vratha> so i was just curious
[18:54:11] <Vratha> i mean their stock isn't worth that much
[18:54:30] <oxygene> given that metric, privately held companies tank on incorporation *shrug*
[18:54:57] <Vratha> i think that publicly traded stock price means something
[18:57:26] <oxygene> it means that the crowd of headless chickens on caffeine (or worse) on some "markets" heard catastrophic rumours.. sometimes these rumours are true, but well, I've seen too many developments on stock without any connection to reality.. *shrug*
[18:57:33] *** jesse-jads has quit IRC
[19:09:02] *** bzcrib has joined #opensolaris
[19:09:59] *** Tpenta has quit IRC
[19:11:07] *** damienc has quit IRC
[19:23:25] <jamesd> is this a known bug on update 3   fbt: NOTICE: cannot instrument return of fd_intr at 1341228: non-canonical return instruction
[19:23:40] <jamesd> this is a fesh full install sol10u3 on a e450
[19:23:44] <jamesd> er fresh
[19:24:59] *** Vratha has quit IRC
[19:27:08] <CIA-22> daemon@tonic-gate: Added tag onnv_57 for changeset 40259348e0bc4a9fff86db134b3b1aa044b6bfc9
[19:27:09] <CIA-22> Craig Mohrman: 6512868 message from 1085341 fix does not need to be I18N
[19:27:09] *** bengtf has joined #opensolaris
[19:27:59] *** bzcrib has quit IRC
[19:28:10] *** bengtf has quit IRC
[19:28:24] *** dunc has quit IRC
[19:32:11] <movement> jamesd: yes
[19:32:54] <jamesd> i guess its not a problem... so just ignore it.
[19:33:06] *** shadeland has quit IRC
[19:34:39] <CIA-22> Jesus Pasallo: 6497646 prtdiag output formats are different on ontario and erie
[19:35:56] <uncertainty> good evening :)
[19:38:21] <uncertainty> mhh my box is telling me that metasync timed out on bootup... does this suggest a defective harddrive?
[19:39:34] *** Tpenta has joined #opensolaris
[19:39:51] *** calumb has quit IRC
[19:55:41] *** alanc is now known as alanc_away
[20:00:08] *** _william_ has joined #opensolaris
[20:01:01] *** dlynes_laptop has joined #opensolaris
[20:02:56] <_william_> hi all
[20:06:29] *** yarihm has quit IRC
[20:11:32] *** pikapika has joined #opensolaris
[20:12:10] <pikapika> hello
[20:14:37] <uncertainty> heya
[20:15:33] *** deather_ is now known as deather
[20:23:50] <nprice> How to make Windows useful: install Xming and export your solaris x11 display to your Windows desktop :)
[20:24:42] *** Theoden-Nexenta has left #opensolaris
[20:25:55] *** tsoome1 has joined #opensolaris
[20:26:00] <nprice> What is CIA-22 displaying?
[20:26:11] <richlowe> ON putbacks
[20:26:19] <nprice> neat
[20:26:45] *** axxl has quit IRC
[20:26:51] <Sporq> any of you know what the ramifications would be in shutting off ipf service?
[20:26:58] <sommerfeld> used to be "onbot" doing that
[20:27:19] <quasi> Sporq: not having ipfilter running
[20:27:37] <Sporq> ipfilter / pfil
[20:27:56] <stevel> sommerfeld: movement changed its name
[20:27:57] *** yarihm has joined #OpenSolaris
[20:27:59] <Sporq> right -- ipfilter is simply installed by default but doesn't actually do anything until you configure, correct?
[20:28:06] <CIA-22> gg81521: 6491115 x64: move pcicfg to usr/src/uts/i86pc from usr/src/uts/common
[20:31:09] <quasi> Sporq: correct
[20:39:45] *** tsoome has quit IRC
[20:41:32] *** LordKing has quit IRC
[20:46:28] *** TrogL has joined #opensolaris
[20:46:45] *** klocze2 has quit IRC
[20:47:24] *** adp` has joined #opensolaris
[20:48:15] <uncertainty> my solaris box is failing with metasync, failing with hal...
[20:48:20] <uncertainty> what does this mean?
[20:50:38] <TrogL> having problems with installer.  My server is multiboot so I don't have one specific disk specified as boot disk.  When I get beyond disk1 it reboots.  when I manually try to boot from the OPenSolaris boot disk it correctly finds the boot device, btu says "cannot open disk5"
[20:51:10] <TrogL> disk5 is my NVALIAS for the boot disk
[20:54:58] *** klocze2 has joined #opensolaris
[20:55:53] *** klocze2 has quit IRC
[20:56:29] *** klocze2 has joined #opensolaris
[20:57:59] *** jesse-jads has joined #opensolaris
[20:59:08] *** RaD|Tz has quit IRC
[20:59:23] *** wutzel has quit IRC
[20:59:49] *** klocze2 has quit IRC
[21:02:26] *** Gman has joined #opensolaris
[21:02:37] <Gman> hey hey hey
[21:02:46] <Error_e^ipi> yo
[21:03:40] *** kloczek has joined #opensolaris
[21:05:00] *** inhonho has quit IRC
[21:05:22] <uncertainty> is there any way to disable metasync on bootup?
[21:07:17] <Qapf> lol, my solaris b55 box is completly unresponsive.
[21:07:31] <Qapf> something is eating cpu
[21:08:11] <tsoome1> solaris should really reserve some resources for root....
[21:08:18] *** tsoome1 is now known as tsoome
[21:08:41] <jmcp> Qapf: probably your font cache
[21:08:53] <jmcp> Qapf: wait a bit longer :)
[21:09:02] <Qapf> jmcp, x is off :P
[21:09:11] <Qapf> /usr/lib/fm/fmd/fmd
[21:09:31] <jmcp> Qapf: but if the service isn't disabled, it'll still run on boot
[21:09:46] <quasi> Qapf: ah, fault management
[21:09:47] <jmcp> svc:/application/font/fc-cache:default
[21:09:55] <Qapf> hrm
[21:10:09] <jmcp> Qapf: is that fmd process chewing a lot of cpu?
[21:10:13] <Qapf> yes
[21:10:20] <Qapf> 56 percent
[21:10:21] <jmcp> Qapf: can you get a shell?
[21:10:24] <Qapf> i have one
[21:10:27] <Qapf> slower than hell
[21:10:27] <jmcp> run "fmdump"
[21:10:28] <Qapf> but its going
[21:10:54] <Qapf> will do, have to fight the lag, so it will take me a bit
[21:11:13] <jmcp> I expect you'll see a heap of errors logged by fmdump
[21:11:19] <Qapf> log is empty
[21:11:22] <jmcp> oh
[21:11:28] <Qapf> fmdump: /var/fm/fmd/fltlog is empty
[21:11:45] <jmcp> strange
[21:12:55] <Qapf> system seems responsive after i issued disable to font cache
[21:14:46] <jmcp> ha!
[21:15:43] <hile_> morning jmcp
[21:15:58] <Qapf> nope
[21:15:59] <Qapf> i lied
[21:16:12] <Qapf> system still going unresponsive
[21:16:30] <Qapf> fmd is skyrocketing
[21:16:35] <Qapf> system stops responding
[21:16:39] <Qapf> it falls, i can to ps again
[21:17:03] <jmcp> hi hile_
[21:17:57] <jmcp> what is in /var/adm/messages?
[21:18:05] <jmcp> (use pastebin.ca or rafb.net/paste)
[21:18:24] <Qapf> uh-oh
[21:18:27] <Qapf> ata errors
[21:18:30] <Qapf> lots of ata errors
[21:19:03] <Qapf> dead drive, and its thrasing the os
[21:19:10] <Qapf> or bad sata converter
[21:20:15] <axisys> just installed b55
[21:20:32] <Qapf> damnit zpool, spit out status and tell me what drive it is
[21:20:34] <axisys> do i have to install sudo or is it somewhere in the OS?
[21:20:59] <Qapf> axisys,  no sudo by default, sunfreeware has it
[21:21:04] <Qapf> i dont think
[21:21:05] <jmcp> back in 5...
[21:21:12] <axisys> sol 10 comes with sudo in companion cd
[21:22:07] *** laca is now known as lacaAFK
[21:22:49] <elektronkind> interesting
[21:23:05] <elektronkind> unlike the original X4100, which had 4 e1000g interfaces
[21:23:16] <elektronkind> the x4100 m2 has 2 nge and 2 e1000g
[21:24:01] <elektronkind> I'm trying to decide if this an afront to my sensibilities
[21:24:23] <Qapf> elektronkind, its an afront to your sexuality, quick rip out the nge chips before they get you
[21:24:25] <AbeFroman> seems to cut against the recent intel announcement
[21:25:04] <Qapf> i have an intel board with an nge chip on it, realtek is doing well for themselves
[21:25:27] <elektronkind> no, my manhood is pretty resilient when it comes to ethernet chipsets.
[21:26:02] <Qapf> wouldn't be suprized if intel decided to stop making the effort for all but the highest end ethernet chips and left the low end to realtek
[21:27:41] *** Odin-LAP has quit IRC
[21:38:32] <richlowe> Hm.
[21:38:47] *** estibi has joined #opensolaris
[21:41:57] *** TrogL has left #opensolaris
[21:43:30] <Qapf> pounds of ata errors being generated and zfs hasnt removed the disk or degraded the array
[21:43:48] <sommerfeld> Qapf: zfs will only remove a disk in the event that an open fails.
[21:43:53] <jmcp> Qapf: is the zpool mirrored?
[21:43:57] <Qapf> raidz
[21:44:06] <jmcp> so it has redundancy
[21:44:11] <Qapf> indeed
[21:44:18] <Qapf> its annoying, because zpool status -v has hung
[21:44:37] <Qapf> and system won't respond to init 5 to kill itself and let me start to play with the hardware
[21:44:38] <richlowe> sommerfeld: it'd be great to see FMA down around the disks and busses to rectify that.
[21:44:40] <sommerfeld> zpool status will sometimes take a while
[21:44:42] <Qapf> because the cpu is spiking
[21:44:51] <sommerfeld> richlowe: indeed.  but baby steps are necessary..
[21:45:26] <richlowe> sommerfeld: are you aware of anything odd being done to $SRC/lib/libc/port/gen/closedir.c recently?
[21:45:45] <richlowe> (after the per-dir locking putback, but before onnv_57 closed)
[21:45:46] <sommerfeld> need a higher level agent to say "too many errors, kick in hot spare"
[21:46:10] <sommerfeld> richlowe: sec
[21:46:42] <Qapf> i would just kill the drive but i dont know solaris well enough to tell what drive is generating all these ata errors :( wanted zpool to tell me
[21:47:02] <Qapf> ahh, zpool degraded itself now
[21:47:48] <sommerfeld> richlowe: hmm.  newest delta in teamware is 1.25 corresponding to the per-DIR locking putback.
[21:47:58] <richlowe> sommerfeld: bugger, I'd hoped something obvious may have happened.
[21:48:13] <richlowe> for some reason the onnv-scm merge is having me hand merge it, and I can't even vaguely see why.
[21:48:16] <sommerfeld> and it's showing same modtime as opendir.c
[21:48:27] <richlowe> we haven't touched it, if nothing odd happened to it, I'm totally confused.
[21:54:00] <axisys> i did not have to install sudo.. since the other disk on u20 has sudo on zfs fs .. i just imported the fs from sol 10 u3 to b55 and walla
[21:56:28] *** UnixTitan has quit IRC
[21:58:24] <tsoome> why sudo if you have a rbac
[21:58:46] *** Odin-LAP has joined #opensolaris
[21:59:03] <axisys> tsoome: i am used to sudo
[21:59:10] <tsoome> too bad
[21:59:22] <tsoome> :P
[21:59:30] <axisys> tsoome: still did not get the hang of it w/ rbac
[21:59:46] <axisys> tsoome: i need to sit down and play with some quick examples some day
[22:00:40] <tsoome> or perhaps I should learn some sudo one day:D
[22:01:04] <tsoome> but definitely not on solaris:D
[22:05:04] *** Plouj has joined #opensolaris
[22:05:36] *** digix_ has joined #opensolaris
[22:05:41] <Plouj> is open solaris licensed under a GPL compatible license?
[22:08:29] <tsoome> man google?
[22:09:34] <Plouj> bah
[22:09:42] <Plouj> it's easier to use wikipedia than the official os site
[22:09:59] <tsoome> surprised?
[22:12:09] <Doc> woot!  3 cent profit for Sun!
[22:12:25] <andersmo> Lemmesee... I went to os.org, then on the first link under "about" - "project overview", which had a "licensing info" resource link in a box on the right side. Not that hard? =)
[22:12:49] *** digix_ has quit IRC
[22:13:00] *** digix_ has joined #opensolaris
[22:13:11] <uncertainty> jmcp: are you there?
[22:13:53] <tsoome> why one should worry if os is licensed with gnu license?
[22:15:26] <sommerfeld> Plouj: Like every other open-source operating system, opensolaris contains code licensed under multiple licenses.
[22:15:32] <Plouj> "a module covered by the GPL and a module covered by the CDDL cannot legally be linked together"
[22:15:50] <Plouj> tsoome: I said "GPL compatible"
[22:15:59] <Doc> plouj: really?  where in the CDDL does it say that?
[22:16:25] <Doc> ohh.. wait..  it's the over-restrictive GPL that doesn't allow it, right?
[22:16:28] <tsoome> Plouj: thats details;)
[22:16:29] <Plouj> Doc: it doesn't. The fsf site says that: http://www.fsf.org/licensing/licenses/index_html#GPLIncompatibleLicenses
[22:16:44] <sommerfeld> Plouj: GPL-licensed programs running on solaris are linked with the CDDL-licensed libc.
[22:17:01] <Doc> yah.. GPL is about as incompatible with other licenses as each release of Linux is with each other
[22:17:09] <tsoome> "free" gnu ...
[22:18:06] <Doc> i suggest finding who wrote the GPL code you're interest in, and talking them into using a non-restrictive license such as CDDL
[22:20:39] <jmcp> uncertainty: yeah, I'm here
[22:20:47] * jmcp returns from taking wife to train station
[22:20:58] *** digix has quit IRC
[22:21:06] <uncertainty> yeah i think the problem i had yesterday with updating is now some kinda... gone ^^
[22:21:39] <jmcp> sorry, which problem?
[22:21:47] * jmcp tries frantically to page-in
[22:21:56] <uncertainty> updating nevada, b49 ^^
[22:22:52] <jmcp> oh
[22:22:57] <jmcp> glad that problem is gone
[22:23:01] <uncertainty> the box just went completely down ^^
[22:23:05] <jmcp> oh
[22:23:09] <uncertainty> hardware failure i think
[22:23:15] <jmcp> :(
[22:23:18] <uncertainty> our hoster is currently checking the disks
[22:23:26] <uncertainty> man i'm so... argh ^^
[22:23:35] <uncertainty> failsafe mode was okay
[22:23:39] <uncertainty> but nothing more
[22:23:48] <jmcp> :(
[22:23:49] <jmcp> sorry to hear that
[22:23:53] <uncertainty> normal system booted, spittet out many warnings about not being able to metasync
[22:23:56] <Odin-LAP> Doc: The CDDL isn't less restrictive than the GPL, I think you'll find. The difference is that the CDDL is supposed to be applied file-by-file, while the GPL applies project-wide. How the hell that works, in terms of legalities, is anyone's guess...
[22:24:15] <uncertainty> (this all took about 15 minutes, after that there was a login prompt, taking input at approx 1 char/sec)
[22:24:26] <uncertainty> and after login a kernel panic followed immediately ^^
[22:26:06] <uncertainty> i want solaris to be able to boot from zfs so bad ^^
[22:26:18] <jmcp> you + me + restofworld
[22:26:32] <uncertainty> ^^
[22:26:55] <uncertainty> is it okay for you if i'm angry on SVM and call it "crap"? ^^
[22:27:25] <delewis> SVM isn't "crap" if you consider the rest of the market.
[22:27:39] <delewis> its incredibly easy to mirror root with SVM than it is with other *cough* operating systems.
[22:28:21] <uncertainty> okay but currently i tend to say that i don't like it :D
[22:28:38] <delewis> uncertainty, trying using Linux storage :-)
[22:28:41] <delewis> s/trying/try/
[22:28:49] <delewis> SVM will be a god-send.
[22:28:51] <uncertainty> but i remember those times when you were only able to do mirroring on freebsd with the old vinum ^^
[22:29:17] <delewis> I liked vinum, actually.
[22:29:23] <uncertainty> yeah i know, linux lvm isnt cool ... but after that bad failure of my box... ^^
[22:29:34] <uncertainty> but geom_mirror is much easier to set up
[22:29:51] <delewis> so is the AIX LVM
[22:29:57] <delewis> one command and you're mirrored.
[22:30:22] <uncertainty> cool ^^
[22:30:41] <jmcp> yay ... SUNW made a profit for q2 of 07
[22:30:59] <uncertainty> cool :)
[22:31:21] <quasi> jmcp: finally!
[22:31:26] <jmcp> quasi: it's been a while
[22:32:17] <Plaidrab> SVM isn't that bad.
[22:32:25] <Plaidrab> Nicer than SDS was
[22:32:53] <uncertainty> whats sds?
[22:33:00] <Plaidrab> Solaris Disk Suite.
[22:33:05] <Plaidrab> SVM's predecessor
[22:33:08] <uncertainty> okay
[22:33:12] <uncertainty> until when was it used?
[22:33:34] <Plaidrab> Oh. Um. I'm sure it's still used, but for most practical answers, probably 2003
[22:33:51] <uncertainty> okay
[22:33:52] <Plaidrab> I haven't seen it personally since 2001.
[22:33:57] <uncertainty> ^^
[22:34:28] <uncertainty> umm... this metasync followed by kernel panic problem i described above, has anyone seen that?
[22:35:38] <Plaidrab> I'm afraid I've never seen an SVM or SDS caused kernal panic
[22:36:04] *** alanc-away is now known as alanc
[22:36:24] <delewis> try booting an encapsulated root mirror with less than the required amount of metadbs.
[22:36:37] <alanc> SDS is used on Solaris 8 and older - SVM replaced it in Solaris 9 and later
[22:37:29] <Odin-LAP> And ZFS replaces SVM in a release of 10? :p
[22:37:36] <Plaidrab> I seem to remember it being available for 8, but I'm not wholly certain there.
[22:37:40] *** lloy0076 has joined #opensolaris
[22:37:49] <delewis> Plaidrab, it was bundled
[22:37:57] <delewis> as a separate package on the installation media
[22:38:51] <Plaidrab> I remember many arguments on the list about which we needed to use for the new mailstore standard, vertitas, or SVM.
[22:40:24] <quasi> veritas is nice if you have too much money
[22:40:50] <delewis> I'd still use SVM for root mirroring if VxVM were around.
[22:40:51] <AbeFroman> heh
[22:41:04] <delewis> VxVM root disk encapsulation sucks.
[22:41:51] <lasseoe> Indeed
[22:42:21] *** adlpaf has quit IRC
[22:42:47] <quasi> delewis: absolutely - especially if you do silly things like mirroring your root disk on san
[22:42:56] <Plaidrab> There were arguiong performance issues on CAN scale male servers.
[22:43:07] <Plaidrab> Which I think are now solely in colorado anyway. heh
[22:44:08] <alanc> need male servers to scale to handle all the "natural male enlargement" spam
[22:44:19] <alanc> 8-)
[22:44:34] <Plaidrab> heh
[22:45:07] <Plaidrab> I rather enjoyed working with SOMS. It was prety well behaved, but we didn't do anything excoctic with it
[22:45:32] <Doc> damn that's strong coffee!!!
[22:46:08] *** bougie has quit IRC
[22:49:40] <nprice> got a stripped down version of solaris running in vmware on my windows workstation, and i export its display to my workstation with xming, and i share my worksatation's files via nfs with windows SFU, it's almost like running OS X haha
[22:51:31] *** nostoi has joined #opensolaris
[22:57:25] <sickness> nprice: cool :)))
[22:57:48] <_william_> gn
[22:57:49] *** _william_ has left #opensolaris
[23:02:06] *** awg has quit IRC
[23:02:37] *** lloy0076 has quit IRC
[23:03:02] *** hell` has joined #opensolaris
[23:04:57] *** Error_e^ipi has quit IRC
[23:08:04] <axisys> i just installed b55 and trying to setup iscsi disks with 5310.. is there some simple doc/steps out there?
[23:08:28] *** hspaans has joined #opensolaris
[23:08:43] <hspaans> woooo?
[23:08:50] <hspaans> g'day, all
[23:09:11] <jmcp> hspaans: hi
[23:09:27] <jmcp> hspaans: it's "w000000t" because SUNW made a profit and beat Wall St estimates
[23:10:09] <hspaans> wtf
[23:10:14] <axisys> in other words how to setup 5310 disks as iscsi targets and initiate iscsi connection  from my b55 running on u20
[23:10:15] *** Gman has quit IRC
[23:10:36] <tsoome> read docs.sun.com?
[23:10:43] <tsoome> ;)
[23:11:22] *** swa_mobil has joined #opensolaris
[23:12:02] <elektronkind> so Sun has demonstrated it knows what Step 2 is.
[23:12:03] <axisys> tsoome: sure :P
[23:12:10] <sommerfeld> elektronkind: heh
[23:12:37] <sommerfeld> axisys: can't help with the 5310 end, but iscsiadm is the tool on the client ("initiator") end.
[23:12:55] <elektronkind> too bad sunw is -0.09 today
[23:13:19] <sommerfeld> i think you want to add a "discovery-address" for the server.
[23:13:22] <elektronkind> were the result announced after closing?
[23:13:26] <sommerfeld> elektronkind: check after-hours numbers
[23:13:35] <sommerfeld> they're back over $6/sh
[23:13:42] <elektronkind> no kidding
[23:13:43] <delewis> wow.
[23:13:48] <elektronkind> excellent, smithers
[23:13:51] <delewis> I remember how long it took to break $5
[23:13:52] <axisys> sommerfeld: thats a start
[23:13:54] <sommerfeld> http://finance.yahoo.com/q?s=SUNW
[23:14:03] <delewis> and now its up to $6 in a much less time.
[23:14:08] <elektronkind> holy cow
[23:14:10] <elektronkind> 6.11
[23:14:13] <elektronkind> day-am
[23:14:19] * delewis falls over
[23:14:26] <elektronkind> last trade: 5.66
[23:14:34] <axisys> sommerfeld: iscsiadm: missing object <-- what is the object?
[23:14:46] *** doof has joined #opensolaris
[23:15:03] <sommerfeld> I think the syntax is something like: "iscsiadm add discovery-address <ip address>"
[23:15:07] <axisys> sommerfeld: so once i start the target (server) with discovery-address .. the initiator will see it?
[23:15:14] <doof> hi
[23:15:17] <tsoome> no
[23:15:22] <quasi> elektronkind: the after hours trading looks good though
[23:15:23] <doof> i try to install  snv_55 in domU and the domain crash. does someone use it  ?
[23:15:41] <tsoome> you should also start the initiator
[23:15:41] <sommerfeld> snv_55 doesn't have xen in it yet
[23:15:58] <axisys> sommerfeld: how do i get the discovery-address.. does not have to be 5310.. just looking for concepts
[23:16:10] <richlowe> Nevada in general doesn't have xen in it yet.
[23:16:15] <richlowe> beyond the direct boot bits.
[23:16:20] <Plaidrab> : will be happy when it's back to what I paid for it
[23:16:21] <doof> sommerfeld: but i want to use it in a domU (HVM)
[23:16:53] <richlowe> Not sure about HVM, movement or dme woud be the people to ask.
[23:16:56] <richlowe> (or xen-discuss@)
[23:17:32] <doof> i use solaris10 x86 in domU and it work well
[23:17:44] <sommerfeld> axisys: so, i think the way it works is that it's *typically* the ip address of the iSCSI target device
[23:19:15] <axisys> sommerfeld: gotcha
[23:22:49] <Qapf> heh, i just noticed, i was lazy and let the solaris installer do all my partitioning, my var is so small i cant install ncurses
[23:23:20] <Qapf> pkgadd can't extract the files temporarly prior to placing them where they belong
[23:23:21] *** gm152 has joined #opensolaris
[23:23:26] <Qapf> yet /export/home has 11gb
[23:24:21] <Qapf> oh well
[23:27:53] *** Ludo_ has quit IRC
[23:29:42] <CIA-22> Craig Mohrman: 6512868 message from 1085341 fix does not need to be I18N (fix lint)
[23:31:18] *** swa_wtf has joined #opensolaris
[23:31:47] <dwc-> back in the black
[23:31:54] * dwc- applauds the sunw empl
[23:32:54] *** swa_mobil has quit IRC
[23:33:35] *** ChanServ sets mode: +o Tpenta
[23:34:51] *** yarihm has quit IRC
[23:36:13] * Tpenta cheers!!! wooooooohooooooo
[23:36:48] *** salmandr has joined #opensolaris
[23:39:17] <Tpenta> holy crap, up 46 cents after hours
[23:41:21] <Plaidrab> Somethign get announced?
[23:41:28] <quasi> Tpenta: going to be very interesting to see how trading starts tomorrow
[23:41:28] <Tpenta> earnings
[23:41:34] <Tpenta> we seriously beat the street
[23:41:41] <Tpenta> $126m profic 3c/share
[23:41:52] <quasi> 3c rather than 0c
[23:42:45] <jmcp> quasi: 0c EPS was what the Street expected
[23:42:54] <jmcp> it is *soooooo* nice to beat the Street :)
[23:42:55] <quasi> jmcp: yep
[23:43:02] *** estibi has quit IRC
[23:43:34] * elektronkind posts a "sun is dead/solaris is dead" article to lxer
[23:43:41] * jmcp snorts
[23:44:37] *** jsubl2 has joined #opensolaris
[23:44:43] <Plaidrab> Hee!
[23:45:04] <Plaidrab> I like the fact that's a $300Mill shift in 4 quarters.
[23:45:47] <Plaidrab> Sadly, Sun will never be what it was when I was hired on again, i think.
[23:45:56] <elektronkind> <wall st.> I see your Schwartz is as big as mine...
[23:46:07] <Plaidrab> :)
[23:46:54] <Plaidrab> At least it's close to what it was the last time I bought any. another quarter like this this year and I bet it'll hit $10 before year end
[23:47:10] <jmcp> what I'd love to know is what Sun's mgmt is planning to buy with the USD700mil that KKR is putting in
[23:48:46] <elektronkind> I thought Sun had bn's of cash on hand
[23:48:51] *** Ludo_ has joined #opensolaris
[23:49:35] <jmcp> yeah ... there's some interesting finance stuff going on
[23:49:40] <jmcp> interesting == I don't understand it
[23:51:23] <elektronkind> you know what I'd like to see one day
[23:51:39] <Plaidrab> Cedega for Solaris?
[23:51:56] <uncertainty> solaris booting from zfs and liveupgrade for zones?
[23:52:05] <Plaidrab> "What do you mean you don't have Creator 3d Drivers?"
[23:52:07] <elektronkind> running some future version of MacOS where Darwin was replaced with Solaris.
[23:52:38] <uncertainty> i tend to say this might not be as far from reality as you think :/
[23:53:01] <elektronkind> I know. I'm kind of giddy about that prospect.
[23:53:02] <Plaidrab> It would have to finally escape it's single/low processorcount "Slowlaris" rep for that to be useful.
[23:53:13] <elektronkind> the sun/intel thing yesterday really cemented it, I think
[23:53:43] <Plaidrab> Aqua on Solaris does make the body need a cold shower.
[23:53:54] <uncertainty> lol ^^
[23:54:01] <elektronkind> yes. "hawt" would be the term
[23:54:12] <uncertainty> i mean...
[23:54:23] <uncertainty> developers keep telling me leopard is really unstable currently
[23:54:39] <Plaidrab> Well, the latest batch of Sun Workstations already look too much like Macs :)
[23:54:43] <uncertainty> so... osx is generally mature, why should it be so unstable, producing kernel locks etc?
[23:54:52] <uncertainty> thats why i like them *g*
[23:55:24] <elektronkind> I'm sure if darwin was replaced we'd hear a leak from the developers who have their hands on the leopard seeds.
[23:56:16] <Plaidrab> Apple is practically the little dutch boy when it comes to leaks
[23:56:31] <Plaidrab> I don't like the new workstations. not Purple enough
[23:56:37] <uncertainty> lol ^^
[23:56:44] *** Kmays has joined #opensolaris
[23:56:46] <elektronkind> you mean not SGI enough? ;)
[23:56:47] *** Error_404 has joined #opensolaris
[23:57:05] <uncertainty> my university just got some neat sgi itanium2 cluster ^^
[23:57:11] <elektronkind> I've got a Octane2 chassis here if you want to try to cram a Ultra 40 into it
[23:57:24] <uncertainty> :D
[23:57:41] <elektronkind> even has the old SGI logo, for cool points.
[23:57:44] <uncertainty> on the house where the cluster is located, there's a chiller on the rooftop
[23:58:07] <uncertainty> last week one guy driving past the building called the fire department because he thought it's burning :D
[23:58:10] <Plaidrab> Nah. Though I like the 02s. Netscape was kind enough to leave one in my office when they escaped the campus. :)
[23:58:11] <elektronkind> solaris on itanium2 is my next bet
[23:58:28] <Plaidrab> I liked the Excaliburs
[23:58:51] <Error_404> elektronkind: how 'bout let's just focus on POWER for now, mmkay?
[23:59:09] <Error_404> since IBM still 1) exists and 2) has customers
[23:59:14] *** a_pavlenko has joined #opensolaris





top