[00:00:06] <stevel> if anything happens in the TW->Hg bridge thing... well.. sorry, you're SOL :-P nobody else wants to even look at that
[00:00:10] <richlowe> (I could see the branch delta# confusing you, but...)
[00:00:27] <richlowe> Hey, I've asked for a copy of it, so that's not entirely true.
[00:00:36] <richlowe> just nobody else who could actually *do* anything.
[00:00:37] <stevel> richlowe: because i use patch.  so i think what happened was that my bridge pulled 1.1.1.1 instead of 1.2, and then when 1.3 came out - it just patched it with the code diff
[00:00:38] *** cneira_ has quit IRC
[00:01:08] <stevel> speaking of which. i've been meaning to propose a project and put that into an SCM repository on os.o
[00:01:11] <richlowe> stevel: well, for our purposes 1.1.1.1 == 1.1, they arrived at once.
[00:01:16] <richlowe> then 1.2, then 1.3
[00:01:27] <richlowe> so I could understand you going 1.1->1.2->1.3 (thus missing the actual *file content*, initially)
[00:01:41] <Gman> hrm, interesting that the search on opengrok is now picking up other gates too
[00:01:48] <stevel> gman: yeah - it picks up all gates
[00:01:51] <Gman> that could get quite confusing :/
[00:01:54] <stevel> you have to restrict it with file-path
[00:01:57] <timeless> gman: it's helpful in this case
[00:01:59] <stevel> not ideal, i know. :(
[00:02:07] <richlowe> Hm.
[00:02:14] <richlowe> stevel: do you not actually pay attention to delta numbers?
[00:02:14] <timeless> it showed me code that doesn't trigger an infinite loop :)
[00:02:18] <richlowe> (or do you guess, or something?)
[00:02:18] <Gman> timeless, yeah, not arguing that it's not useful
[00:02:22] <stevel> richlowe: just bug gman or tpenta.  they always hang out here and they can call people up internally if things go wrong ;-)
[00:02:31] <stevel> richlowe: i just pull random delta numbers out of my ass
[00:02:32] <stevel> :-P
[00:02:43] <Gman> just a little confusing knowing which files are which
[00:02:52] * boyd just came in. What is stevel planning on getting hit by a bus?
[00:03:01] <stevel> boyd: not planning on it. but entirely possible
[00:03:06] <richlowe> stevel: well, my thing here is, I thought the diffs you were working from were done by the gate, and not yours.
[00:03:07] <stevel> i have been known to not look both ways
[00:03:12] <richlowe> stevel: if you're just feeding those through...
[00:03:13] <stevel> richlowe: they are
[00:03:18] <boyd> stevel: Dude, you *gotta* do that
[00:03:22] <richlowe> a rebuild won't do shit for you, because your origins are broken.
[00:03:24] <stevel> which is why i'm now trying to find the diff to see if maybe the gate missed it
[00:03:29] <stevel> in which case, i will totally blame danek
[00:03:33] <richlowe> stevel: I won't.
[00:03:40] <richlowe> I think i'll blame Glynn.
[00:03:42] <richlowe> he doesn't get blamed much.
[00:03:43] <Gman> ;)
[00:03:44] <stevel> sounds good
[00:03:45] <Gman> yeah right.
[00:03:46] <stevel> i'll +1 that
[00:03:48] <boyd> Especially in a country where they drive on the other side to what you're used to :)
[00:04:00] <stevel> boyd: hehe
[00:04:18] <richlowe> You can't believe how often I've screwed that up :\
[00:04:31] <richlowe> look right, then hear the horn and jump clean out of your skin.
[00:04:41] <stevel> i totally drove on the RHS in a tesco parking lot and got my ass chewed out by the parking attendant
[00:04:46] <timeless> stevel: so... what should i do?
[00:04:53] <stevel> literally he asked me "What's wrong with you??"
[00:05:03] <stevel> i yelled back "sorry, it's okay... i'm american"
[00:05:11] <stevel> timeless: ping on opensolaris-code?
[00:05:55] * timeless wonders what that is (google time)
[00:06:02] <stevel> mailing list
[00:07:04] <stevel> richlowe: no, the gate got it right
[00:07:23] * stevel is perplexed
[00:07:57] * stevel tries to reproduce the action of the bridge
[00:08:48] <richlowe> Well, that's fun.
[00:08:59] <richlowe> not as fun as whoever gets to rebuild the project gates is going to have though :)
[00:09:16] <stevel> yeah that'd be me probably
[00:09:17] <stevel> :-P
[00:09:27] <richlowe> stevel: well, I'm still trying to come up with a better way than the one I proposed.
[00:09:36] <richlowe> I'm just not sure if the Parent ID's in export would screw you.
[00:09:43] <richlowe> ... though you could always edit them, if you were evil.
[00:09:59] <stevel> nah.
[00:10:02] <stevel> i may be stupid and clumsy
[00:10:04] <stevel> but not evil
[00:10:09] <stevel> (deliberately anyway)
[00:10:22] <richlowe> well, it's not an *invalid* thing to do, exactly.
[00:10:32] <richlowe> really, you'd switch old-id for new-id, they'd refer to the same change.
[00:15:47] <stevel> aw crap
[00:15:51] <stevel> it is danek's fault
[00:16:15] <stevel> the log shows the contents of 1.2, but says its 1.1
[00:16:52] *** Sieghard has quit IRC
[00:18:29] <stevel> though the weird thing is that should have brought in an empty file (since that's what 1.1 is) instead of bringing in 1.1.1.1
[00:19:02] <richlowe> yeah, that's what I was attempting to say above, in the sentence with all the delta#'s in it.
[00:19:17] * richlowe needs more caffeine to be coherent.
[00:19:26] <stevel> ah so you did
[00:19:37] * stevel needs more caffeine too
[00:20:49] <timeless> ok, simpler problem :)
[00:21:15] <timeless> swift# modload /tmp/usbether/usbnet/usbnet
[00:21:16] <timeless> can't load module: Invalid argument
[00:21:25] * timeless goes to read the source for modload
[00:22:01] <stevel> perhaps we can jointly blame cindi, dduvall, and stevel
[00:22:48] <richlowe> stevel: So, do you really handle create's as a diff from nothing?
[00:22:53] <richlowe> if you *copy* on create, that would explain a bunch.
[00:22:58] <stevel> no, creates are done with an SCCS get
[00:23:03] <richlowe> you may well copy on create, apply 1.1.1.1 (which would do nothing), as 1.2
[00:23:05] <richlowe> and then...
[00:23:28] <stevel> it should have done an SCCS get -G mumble -r 1.1
[00:23:34] <richlowe> ah.
[00:23:41] <richlowe> Bugger.
[00:23:50] *** jafari has quit IRC
[00:24:17] <richlowe> stevel: I don't think you could blame danek or cindi though.
[00:24:23] <richlowe> stevel: timh did the putback with the branch delta...
[00:24:34] <richlowe> or rather, the delta is his, who knows about the putback.
[00:24:36] <stevel> but cindi did the final putback to onnv-gate
[00:24:45] <stevel> pbchk should have caught that
[00:24:55] <Tpenta> it shoudl have, yes
[00:24:58] *** sahafeez has joined #opensolaris
[00:25:01] <richlowe> both pbchk's should have, I thought.
[00:25:04] <Tpenta> that would be the delta check
[00:25:06] <richlowe> (both the gate trigger, and wx)
[00:25:16] <stevel> so it's timh's fault for doing the delget, cindi's fault for not doing the pbchk, and dduvall's fault for not properly catching it.  clearly, i'm absolved here.
[00:25:23] <richlowe> though given my luck with wx-ish things lately...
[00:25:32] <richlowe> stevel: hah.
[00:25:51] <Tpenta> well almost steve
[00:25:54] <stevel> the only thing i'm guilty of is trusting me co-workers
[00:25:56] <stevel> i'm the victim here
[00:26:08] <Tpenta> if it *can* be done, the bridge should know how to handle it, I would have thought
[00:26:15] <Tpenta> or at least complain about it loudly
[00:26:16] <stevel> ssshhhh
[00:26:23] <richlowe> Tpenta: Hm, theres an issue there.
[00:26:36] <richlowe> Tpenta: what if stevel is fussier than danek?
[00:26:43] <richlowe> Tpenta: do you force danek to do what steve asks, or do you diverge? :)
[00:26:50] <stevel> i have high standards
[00:27:18] <richlowe> Tpenta: for instance, in this case if there was a putback on top of it, danek wouldn't be able to undo.  So the bridge would *have* to take it some time, it can complain while doing it, but it has no choice...
[00:27:55] <stevel> yeah!
[00:27:57] <stevel> what he said!
[00:28:14] <Tpenta> the point that I was making is that you have to harden code that accepts input from somewhere you don't have control
[00:28:20] <elektronkind> whoa
[00:28:35] <richlowe> Tpenta: well, that's true too, yes.
[00:28:48] <richlowe> I still can't see exactly how this came about though.
[00:28:55] <elektronkind> I accidentally put my mouse over the gnome workspace selector and hit my scroll wheel. Didn't know it did that.
[00:29:09] <richlowe> elektronkind: it treats any button press as a switch.
[00:29:12] <richlowe> elektronkind: it's most annoying.
[00:29:14] <Tpenta> nice isnt it
[00:29:22] <Tpenta> I sometimes scroll around my workspaces like that
[00:29:24] <stevel> richlowe: yeah, i'm not sure how this happened either. which bugs me
[00:29:30] <Tpenta> as long as you are aware of it, it's nice
[00:29:50] <richlowe> stevel: the fake 1.1 isn't with a message I've seen before.
[00:29:55] <richlowe> though obviously my experience is limited.
[00:30:00] <stevel> as is mine
[00:30:09] * richlowe is used to the new file-type message.
[00:31:55] *** nwf has quit IRC
[00:33:38] *** drio has quit IRC
[00:38:05] *** bunker has quit IRC
[00:44:00] <timeless> stevel: ok, silly post posted
[00:48:17] <Tpenta> nd bins pushed
[00:49:22] *** Fish has quit IRC
[00:54:21] *** nbkk6fo_ has joined #opensolaris
[00:54:22] *** rodrickbrown has quit IRC
[00:54:35] *** piwi has joined #opensolaris
[00:55:11] *** stevel has quit IRC
[01:01:12] <Tpenta> http://blogs.sun.com/tpenta/entry/20061103_flag_day_non-debug_encumbered
[01:06:54] *** neoxed has joined #OpenSolaris
[01:11:11] * elektronkind idly wonders how sunsolve patch db is faring
[01:11:33] *** dunc has quit IRC
[01:33:11] *** jafari has joined #opensolaris
[01:40:40] <jafari> hello all
[01:41:02] <jafari> after i tried setting up nat/ipfilter
[01:41:21] <jafari> and it havent worked for me so i reverted bsck to my old setting
[01:41:35] <jafari> now pkg-get doesnt work
[01:41:36] <jafari> Resolving ibiblio.org... failed: node name or service name not known.
[01:42:10] <jamesd> did you restart it?
[01:43:02] <jafari> yea
[01:43:23] <jafari> should i disable svcadm disable network/pfil
[01:43:40] <jafari> and svcadm disable ipfilter
[01:45:21] <jamesd> possibly
[01:45:51] <jamesd> not sure that disable, clears the old rules...
[01:59:16] *** sahafeez has quit IRC
[02:01:19] *** Cybernd has quit IRC
[02:08:16] *** qu1j0t3 has joined #opensolaris
[02:11:46] <qu1j0t3> hi, anyone have trouble serving NFS from linux to Solaris client, e.g. NFS based install?
[02:12:13] <qu1j0t3> I can use the NFS server, but after a minute or two, it stops responding, until I restart it, and so on.
[02:12:32] <qu1j0t3> I've tried various options such as increasing number of nfsd processes, not advertising tcp ...
[02:13:33] <qu1j0t3> this is e.g. at media #2 of Solaris 10 install on SPARC (E450)
[02:13:37] <qu1j0t3> media #1 was real CD
[02:13:58] *** OnkelSchorsch has joined #opensolaris
[02:16:31] *** Behi has joined #opensolaris
[02:16:33] *** slowhog has left #opensolaris
[02:16:56] <Behi> hi
[02:17:15] <Behi> does anybody know where can I find a prebuilt solaris 10 or opensolaris image for vmware?
[02:20:44] *** Behi has left #opensolaris
[02:21:44] *** qu1j0t3 has left #opensolaris
[02:24:58] <dwc-> jafari: ipf -F a; ipf -F S
[02:25:22] <dwc-> (flush all rules, flush state table)
[02:28:04] <jafari> bash-3.00# ipf -F a; ipf -F S
[02:28:05] <jafari> bash-3.00# /opt/csw/bin/pkg-get install nmap
[02:28:29] <jafari> Resolving ibiblio.org... failed: node name or service name not known.
[02:28:30] <jafari> ERROR: could not get catalog file
[02:28:41] <jafari> ever since i tried configuring nat
[02:28:48] <jafari> pkg-get stop working
[02:29:03] <dwc-> oh... you have nat too
[02:29:15] <dwc-> man ipnat
[02:29:20] <dwc-> and look at the flush/clear rules there too
[02:29:23] <jafari> i removed ipnat.conf
[02:29:47] <dwc-> ipnat -C or something like that -- check the man page
[02:29:48] <jafari> i removed pass out all; pass om all form ipf.conf
[02:29:53] <jafari> ok
[02:29:57] <timeless> can i debug kernel module loading w/o halting my system?
[02:30:02] <jafari> check the man page for?
[02:30:10] <dwc-> whatever the option is to flush the rules
[02:30:17] *** aiki_kid has joined #opensolaris
[02:30:30] <jamesd> timeless, you can do it with dtrace  and  fbt probes
[02:30:40] <jafari> bash-3.00# ipnat -C
[02:30:41] <jafari> 0 entries flushed from NAT list
[02:30:42] * timeless googles fbt
[02:30:43] <jafari> still the same
[02:31:04] <jafari> i rebooted
[02:31:05] <dwc-> does ipnat -l show you no rules and no current fwds?
[02:31:06] <aiki_kid> mplayer cdda:// is the cmd for palying cds with mplayer in solaris?
[02:31:16] <jafari> what else could be soping it
[02:31:24] <jafari> i can do a nsloopup against google
[02:31:25] <jafari> that works
[02:31:53] <jafari> bash-3.00# ipnat -l
[02:31:54] <jafari> List of active MAP/Redirect filters:
[02:31:54] <jafari> List of active sessions:
[02:32:42] <jamesd> timeless, http://docs.sun.com/app/docs/doc/817-6223/6mlkidljt?a=view
[02:32:43] * jafari doesnt want to have to reinstall the os :(
[02:33:10] <dwc-> is nscd running?
[02:33:12] *** sahafeez has joined #opensolaris
[02:33:13] <dwc-> nscd -i hosts
[02:33:16] <jafari> route -u -d ipv4-forwarding ; i have done
[02:33:21] <jafari> no luck
[02:34:22] <jafari> ps -elf|grep nscd;pgrep nscd
[02:34:22] <jafari> 0 S root 117 1 0 40 20 ? 480 ? 18:37:27 ?
[02:34:22] <jafari> 0:00 /usr/sbin/nscd
[02:34:22] <jafari> 117
[02:35:17] * jafari is starting now to like nat on solaris
[02:35:22] <jafari> not*
[02:36:25] <jafari> guess am to use to nat/pf
[02:36:28] <jafari> on openbsd
[02:36:41] <jafari> this box been running for 5 years now with no troubles
[02:37:42] <jafari> just wanted to upgrade from OpenBSD 3.2 (GENERIC) #25: Thu Oct 3 19:51:53 MDT 2002
[02:37:52] <jafari> to see how solaris nat can handle load
[02:39:08] <jafari> openbsd 4 is out and am still running 3.2
[02:39:09] <jafari> lol
[02:39:26] <jafari> guess if it works why bother it hahah
[02:45:43] <timeless> ok, i'm doing something wrong
[02:45:55] <timeless> i found dtrace_for_dev.html on dev.sun.com
[02:49:35] <timeless> but it isn't working :(
[02:49:50] <twincest> anyone know of a test suite for http proxies?
[02:50:04] *** jmcp has quit IRC
[02:51:06] <Tpenta> gman: once I've finished dealing with calls today, I'll go you one better and gather a gcore with full text of that  acrobat issue andthen log a bug
[02:52:21] <jamesd> twincest, you could use  ab that is part of apache to test the proxy
[02:52:36] <twincest> i'm doing that now, but it's not very thorough
[02:53:30] <twincest> i have a couple of browser to test with as well, but ideally i'd like something i can do specific regression testing with
[02:54:22] *** sahafeez has quit IRC
[02:56:07] <timeless> swift# dtrace: failed to compile script ./trace_usbnet.sh: line 17: probe description :fbt::entry does not match any probes
[02:56:58] <jamesd> fbt:::entry   {}
[02:56:59] <richlowe> you have one colon too few.
[02:57:02] <richlowe> fbt:::entry
[02:57:03] <richlowe> 3 colons.
[02:57:04] <jamesd> should work ..
[02:57:10] <timeless> ah
[02:57:25] <richlowe> <provider>:<module>:<func>:<probe> :)
[02:57:37] <jamesd> you should also  at a predicate to limit the amount of data it generates
[02:57:53] <richlowe> but that's no fun! :)
[02:58:03] * timeless goes off to try to figure out where it sends the output
[02:58:11] <richlowe> stdout, if you get any.
[02:58:21] <jamesd> to stdout by default
[02:58:32] <richlowe> sod it, I'll let jamesd answer, he types faster :)
[02:59:02] <timeless> swift# /usr/sbin/dtrace -s ./trace_usbnet.dt &;modload /tmp/usbether/usbnet/usbnet;fg
[02:59:02] <timeless> [2] 11082
[02:59:02] <timeless> can't load module: Invalid argument
[02:59:02] <timeless> [2]  - running    /usr/sbin/dtrace -s ./trace_usbnet.dt
[02:59:35] <richlowe> dtrace -s /trace_usbnet.d -c 'modload ...'
[03:00:41] <timeless> swift# /usr/sbin/dtrace -s ./trace_usbnet.dt -c 'modload /tmp/usbether/usbnet/usbnet'
[03:00:42] <timeless> dtrace: script './trace_usbnet.dt' matched 60110 probes
[03:00:42] <timeless> can't load module: Invalid argument
[03:00:42] <timeless> dtrace: pid 11089 has exited
[03:01:01] <timeless> well, that's progress, i suppose :)
[03:01:06] *** boyd_ has joined #opensolaris
[03:01:11] <richlowe> then you aren't actually firing any probes.
[03:03:00] <timeless> ok
[03:03:06] <timeless> now i get a lot of stuff
[03:04:38] <timeless> http://pastebin.mozilla.org/1203
[03:04:51] <timeless> but i suspect tracing modctl isn't such a good idea :)
[03:06:29] <richlowe> more than that, surely?
[03:06:52] <richlowe> and tracing modctl should be fine.
[03:07:07] <richlowe> at least, one of the options to the dtrace tests, spams the system with modunload's while they run :)
[03:08:12] * timeless tries to parse that line
[03:08:21] <timeless> what do you mean more than that surely?
[03:08:32] <timeless> that's really all the output i got
[03:08:42] <richlowe> timeless: the stuff you pastebin't stops midway through tnf_thread_create
[03:08:48] <richlowe> "pastebin'd"
[03:08:51] <timeless> oh
[03:08:54] <timeless> hrm, well er
[03:08:55] <jamesd> timeless, you asked dtrace to trace the entry of every kernel function that is called.
[03:09:06] <timeless> pastebin probably didn't appreciate the full content length
[03:09:15] * timeless tries to figure out how to 'export' that file
[03:09:20] <richlowe> jamesd: It doesn't look that way.
[03:09:31] <richlowe> jamesd: it looks like everything called on behalf of whatever called modctl
[03:09:51] <timeless> i guess the pastebin had a buffer limit :)
[03:10:02] <richlowe> at least, I'd expect it to be a lot messy and a lot less valid seeming with regarding to the -F bits, otherwise.
[03:10:12] <delewis> # useradd
[03:10:13] <delewis> ld.so.1: useradd: fatal: libc.so.1: version `SUNW_1.22.2' not found (required by file /usr/sbin/useradd)
[03:10:16] <delewis> fun!
[03:10:30] <delewis> (S10u2/SPARC with all patches installed)
[03:10:30] *** uebayasi has joined #opensolaris
[03:10:43] * timeless takes a moment to advertise a competing product
[03:10:49] <timeless> i asked it to connect to the internet, so it rebooted
[03:10:58] <timeless> ok, it wouldn't sell very well if i advertised it that way :(
[03:11:49] <timeless> it's a good thing they shipped it!
[03:12:36] *** setuid has joined #opensolaris
[03:13:20] <setuid> Hi there
[03:13:46] *** nwf has joined #opensolaris
[03:14:29] <setuid> Does anyone have some problems with ACPI on SX b46 (x86) ?
[03:18:51] *** boyd has quit IRC
[03:19:04] <timeless> ok, http://webwizardry.net/~timeless/solaris/trace.log (233k)
[03:19:22] *** setuid has quit IRC
[03:19:45] <timeless> has anyone made any cute java(or other) tree viewer?
[03:19:57] <timeless> something that lets you (un)hide subtrees
[03:20:32] <timeless>   1  <- kobj_zalloc                              -1624272740352
[03:20:36] <timeless> interesting number :)
[03:21:12] *** setuid has joined #opensolaris
[03:21:22] <timeless> sort trace.log|tail -110
[03:21:28] <timeless> seems to show the 'toplevel' functions
[03:23:05] *** setuid has left #opensolaris
[03:23:39] <richlowe> chime might let you.
[03:23:44] <richlowe> never played with it though.
[03:25:14] * timeless assumes dtrace can output retvals in addition to args
[03:25:35] <jamesd> yeap
[03:26:19] <delewis> what's the best way to find what patch modified a certain file?
[03:26:32] <delewis> I don't see any relevant fields in contents
[03:26:54] <timeless> hrm, trace(arg1) is supposed to display the function's return value
[03:26:57] <timeless>   1  <- mod_load                                             22
[03:27:08] * timeless concludes 22 isn't a happy value
[03:27:41] <richlowe> well, that'd be EINVAL yeah, but you knew that...
[03:27:42] <jamesd> you have to probe on return rather than entry  to get the return value
[03:27:59] <richlowe> jamesd: you're basing all this on one error message though...
[03:28:02] <timeless> richlowe: um, i know that because you told me
[03:28:34] <timeless> thankfully the guide i'm following is giving me arg1 output already :)
[03:28:37] *** setuid has joined #opensolaris
[03:28:40] * timeless lucked out as dtrace guides go
[03:28:47] *** setuid has left #opensolaris
[03:32:09] *** drio has joined #opensolaris
[03:32:44] <timeless> ok, i've decided i want to messs with modload!moddebug
[03:33:01] * timeless goes to figure out how to make dtrace maliciously set that to -1
[03:33:38] <richlowe> sadly, kobj_load_module returns EINVAL if practically anything goes wrong for it.
[03:33:40] <jamesd> timeless, can't happen...
[03:33:44] <richlowe> (beyond the file not being there)
[03:33:57] <jamesd> timeless, dtrace's first rule is do no harm.
[03:34:04] <timeless> jamesd: grumble
[03:34:11] <timeless> so at this point i need a real debugger?
[03:34:11] <richlowe> if you want to set moddebug, use mdb.
[03:34:30] <timeless> ok, am i still in userland at this point?
[03:34:35] <timeless> i presume i am
[03:35:18] <richlowe> and you probably want to set moddebug to 0x40000000
[03:36:26] <richlowe> echo "moddebug/W 40000000" | mdb -kw
[03:36:57] <timeless> does that dump my whole system into the kernel debugger?
[03:37:16] * timeless is kinda in X and such
[03:37:38] <timeless> yes, i know, bad me, trying to do kernel work from X
[03:37:43] <richlowe> small-k attaches via /dev/kmdb
[03:37:49] <richlowe> big-K stops the kernel.
[03:38:17] <timeless> ok, interesting
[03:38:27] <timeless> swift# echo "moddebug/W 40000000" | mdb -kw
[03:38:27] <timeless> moddebug:       0               =       0x40000000
[03:38:27] <timeless> swift# modload /tmp/usbether/usbnet/usbnet
[03:38:27] <timeless> can't load module: Invalid argument
[03:38:56] <richlowe> log messages?
[03:39:11] <timeless> much better :)
[03:39:58] <timeless> http://pastebin.mozilla.org/1208
[03:40:22] <timeless> Nov  5 04:24:35 swift genunix: [ID 558463 kern.notice] usbnet not an elf module for this ISA
[03:40:27] <timeless> i guess that's the problem
[03:40:58] *** piwi has quit IRC
[03:41:25] * timeless doesn't understand the lines that say "path is"
[03:41:37] <timeless> is that trying to list my $PATH environment? :)
[03:41:42] <timeless> (not that it matters here, just strange)
[03:43:30] <timeless> /tmp/usbether/usbnet/usbnet:    ELF 32-bit LSB relocatable 80386 Version 1
[03:43:33] <timeless> /kernel/drv/nge:      ELF 32-bit LSB relocatable 80386 Version 1
[03:43:41] * timeless is clearly missing something
[03:44:59] <richlowe> and the machine is not amd64?
[03:45:08] <timeless> oh, yeah, it is
[03:45:26] <richlowe> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/uts/intel/amd64/krtld/kobj_isa.c#44
[03:45:29] <richlowe> :)
[03:45:38] *** esproul__ has joined #opensolaris
[03:46:01] <timeless> hrm, "file" sucks
[03:46:14] <timeless> well
[03:46:23] <timeless> how can i find out if something is amd64?
[03:46:31] <timeless> i'm assuming /kernel/drv/nge is amd64
[03:46:40] <timeless> is it unreasonable for me to expect file to tell me that? :)
[03:46:55] <Doc> no, /kernel/drv/amd64/nge is amd64 (or something like that)
[03:47:14] <timeless> ah
[03:47:21] <timeless> doh :), thanks
[03:47:46] <timeless> i guess i want -xarch=amd64a
[03:47:55] <Doc> hmm.. i'm beginning to think i should have used the X4200 rather than the V40z :(
[03:48:28] <richlowe> timeless: use the flags ON uses.
[03:48:31] <Error_404> i thought of buying an x4100
[03:48:37] <timeless> richlowe: um
[03:48:38] <Error_404> then i realized i have no money
[03:49:06] <Doc> i've got one at work, but i'm going to lose it in the next week or two, so i'm using my v40z instead
[03:49:31] *** esproul_ has quit IRC
[03:50:30] <OnkelSchorsch> hm. did you get it via the try&buy program?
[03:53:02] <OnkelSchorsch> sorry. I was just wondering how long it takes till a try&buy box gets shipped, since I am waiting for a box myself
[03:53:05] <OnkelSchorsch> :(
[03:53:22] <timeless> richlowe: um, i have access to the source browser, and i've read a couple of pages about building, but i don't have space for a build
[03:53:35] <timeless> it's 5am on a sunday :)
[03:54:28] <richlowe> I was meaning use the same cflags, that's all.
[03:54:47] <timeless> yeah, i understand that, i'm just not sure how to find that from where i am
[03:55:00] <timeless> obviously it's somewhere in the source
[03:55:11] <timeless> but i suspect searching for 'cflags' will give me a headache :)
[03:55:16] <twincest> hmm, fixing missing userland options sounds like an interest project
[03:56:23] <richlowe> yes!
[03:57:19] <twincest> wonder how fast sx is under vmware :)
[03:58:29] <richlowe> according to gisburn, not very.
[03:58:38] <jamesd> twincest, its decent if you have enough ram and fast cpu ...
[03:58:43] <Doc> is anything fast under vmware?
[03:59:12] <jamesd> Doc, if you have  a   large config  u20 with 2GB of ram.. its a decent experience.
[04:00:42] <timeless> richlowe: cool, progress!
[04:00:55] <timeless> now i just have undefined symbols :)
[04:01:10] <timeless> /tmp/usbether/usbnet/usbnet_amd64: undefined symbol 'usb_lookup_ep_data'
[04:01:36] * timeless goes grep'ing through /kernel
[04:02:13] <timeless> well that didn't work at all
[04:02:17] <richlowe> usba
[04:02:28] <timeless> swift% find /kernel/drv|xargs grep -l usba_
[04:02:33] <timeless> returned nothing
[04:02:51] * timeless wonders if grep doesn't like binary files
[04:04:02] <timeless> the compiler doesn't approve of usba_
[04:04:07] <timeless> "usbnet.c", line 1417: warning: implicit function declaration: usba_client_detach
[04:04:18] <richlowe> No, I was saying that that's the module that symbol is in...
[04:04:30] <timeless> oh, sorry :)
[04:04:41] <richlowe> link it with '-dy -N misc/usba' (I think)
[04:05:54] <timeless> none of kernel/(misc|kmdb)/amd64/usba(10) seems to have client_attach
[04:06:22] <timeless> or lookup_ep
[04:09:12] <richlowe> > nm /kernel/misc/amd64/usba | grep usb_lookup_ep_data [326]   |               82064|                 191|FUNC |GLOB |0    |3      |usb_lookup_ep_data
[04:09:25] <richlowe> ...but with a line break in there.
[04:10:02] <timeless> so grep doesn't like binaries ?  silly me
[04:12:11] <timeless> oh, duh
[04:12:17] <timeless> this thing doesn't have 5.11 handling
[04:14:12] <timeless> much better :)
[04:14:31] <timeless> relocation error: R_AMD64_32: file /tmp/usbether/usbnet/usbnet_amd64: symbol : value 0xfffffffff0773915 does not fit
[04:14:38] * timeless goes to read about relocations
[04:15:01] <nbkk6fo_> Dear roderick,
[04:15:01] <nbkk6fo_> Congratulations! Your home is about to be upgraded to the power of Verizon FiOS Internet. Soon you'll be able to download purchased music in a flash, watch flawless streaming video and get the edge in online gaming. Here is a summary of your FiOS Internet order:
[04:15:01] <nbkk6fo_> o Scheduled Installation Date -- 11-09-2006
[04:15:02] <nbkk6fo_> o Scheduled Technician Arrival Window -- 08:00-17:00(Hrs)EST
[04:15:03] <nbkk6fo_> o Verizon FiOS Internet Service Order Number -- NY00001600907
[04:15:06] *** nbkk6fo_ is now known as rodrickbrown
[04:15:09] * timeless tries -KPIC
[04:15:51] <g4lt-U60> rodrickbrown, how many times are you going to post that crap?
[04:16:12] <rodrickbrown> it was an accident but when did I post it ?
[04:16:23] <rodrickbrown> g4lt-U60, jealous ? :-P
[04:17:06] <jamesd> rodrickbrown, wow you have a lot of work todo...  a keg for the  party on having such awesome bandwidth,  expresso machine to stay awake to enjoy in... 2 TB  NAS to store the p0rn^Wdata you will download.
[04:17:15] <g4lt-U60> not particularly, I know the owners of an empty building with an OC48
[04:18:12] <rodrickbrown> jamesd, send me the 2TB of nas :-P
[04:18:37] <timeless> ok, progress, now my system hung :)
[04:18:47] <timeless> which i guess means it managed to load the buggy driver :)
[04:18:52] <jamesd> rodrickbrown, you have plenty of money to buy it.. i have seen pics of your house...
[04:19:43] <timeless> heh
[04:20:26] * timeless should add a permanent menu option for booting w/ kernel debugger support
[04:23:28] <timeless> oh fsck
[04:23:36] <timeless> so, developing in /tmp on unix isn't a good idea :)
[04:25:12] <Error_404> obviously
[04:31:23] <timeless> eep
[04:44:17] *** OnkelSchorsch has quit IRC
[04:44:55] <timeless> ok, page fault :)
[04:45:10] *** bobbyz has quit IRC
[04:46:54] *** fik has quit IRC
[04:59:30] *** hile_ has joined #opensolaris
[05:01:20] *** sahafeez has joined #opensolaris
[05:02:03] <timeless> hrm, the mdb doc i have mentions $p, but the live mdb i'm using doesn't have it
[05:02:59] <timeless> i have a stack frame and want to see lals and sim
[05:07:44] *** esproul_ has joined #opensolaris
[05:10:11] *** esproul__ has quit IRC
[05:14:44] *** ericr_ has joined #opensolaris
[05:15:48] *** dakrone has joined #opensolaris
[05:17:22] *** hile_ has quit IRC
[05:24:01] *** rodrickbrown has quit IRC
[05:31:50] <Error_404> hey, neat... this build looks like it might actually work for once
[05:32:42] *** aiki_kid has quit IRC
[05:34:05] <Error_404> oh, maybe not
[05:41:30] *** Gadzooks has quit IRC
[05:48:21] *** movement has quit IRC
[05:53:30] *** dakrone has quit IRC
[05:53:39] *** lacaAFK has quit IRC
[05:58:37] *** mrdeviant has joined #opensolaris
[06:00:16] *** Triskelios has joined #opensolaris
[06:02:21] *** mrdeviant has quit IRC
[06:02:32] *** rodrickbrown has joined #opensolaris
[06:05:06] *** dakrone has joined #opensolaris
[06:08:42] *** error404 has joined #opensolaris
[06:09:01] <error404> I must say, I'm quite sick of this machine locking up like this
[06:09:44] <error404> it's rather obnoxious
[06:17:05] <twincest> i really need a device that can hold all my dvds
[06:17:16] <twincest> hm, or a larger HD
[06:18:01] <error404> throw them in to a big enough array
[06:19:04] <timeless> richlowe: so how do i find the cflags on n uses?
[06:19:17] <timeless> so far i've learned about -xmodel=kernel
[06:19:38] <timeless> and -KPIC and of course -xarch=amd64a
[06:21:46] *** Error_404 has quit IRC
[06:22:13] *** Error_404 has joined #opensolaris
[06:26:09] *** loke has joined #opensolaris
[06:26:36] <elektronkind>  xmodel=kernel?
[06:28:26] *** mayday_jay has joined #opensolaris
[06:28:41] *** error404 has quit IRC
[06:31:23] *** darkcmd has quit IRC
[06:31:32] *** sahafeez has quit IRC
[06:43:36] <noyb> I'm thinkin' benr fell off the planet...
[06:45:42] <timeless> elektronkind? yes?
[06:45:53] * timeless soghs
[06:46:07] <timeless> could someone just tell me what the kernel cflags are for on amd64?
[06:46:26] *** vmhobbes has quit IRC
[06:46:55] <noyb> timeless: I know I can't.
[06:47:19] *** vmhobbes has joined #OpenSolaris
[06:47:19] <timeless> i can't figure out how to google for it :(
[06:48:00] <noyb> lay out a couple of search strings you're looking for, and I'll try to decipher it
[06:48:37] <timeless> hrm
[06:48:44] <timeless> http://src.opensolaris.org/source/xref/onnv/onnv-gate/usr/src/Makefile.master
[06:48:47] <timeless> maybe that's what i want :)
[06:50:24] <noyb> do you want to know what command line arguments to use for compiling OpenSolaris on amd64?
[06:50:39] <noyb> args for CC that is... ?
[06:50:43] <timeless> i'm trying to build a kernel driver for opensolaris amd64
[06:50:51] <timeless> i need both CFLAGS and LDFLAGS
[06:51:18] <timeless> i've collected a bunch, but so far while i can load my module, it page faults/panics when it tries to access local data
[06:51:31] <noyb> ok, that *would* be hard to filter the millions of hits from google.
[06:53:27] <noyb> Tpenta: you're my best guess for who may be most qualified to help timeless to build his kernel module for amd64.
[06:54:06] <Tpenta> i'm actually kinda busy at the moment
[06:54:17] <noyb> I'm pretty confident I've never built a kernel module.
[06:54:17] *** rodrickbrown has quit IRC
[06:54:38] *** rodrickbrown has joined #opensolaris
[06:54:42] <noyb> Tpenta: no problem.  I was just trying to match people to problems.  sorry.
[06:54:49] <Tpenta> :)
[06:55:14] <timeless> i have cflags: -KPIC, -xarch=amd64a -xmodel=kernel
[06:55:37] <noyb> timeless: I'll look for resources on this, see what I find.
[06:56:02] <timeless> thx
[06:56:47] <timeless> sorry, i just really don't have space or justification for a full on world
[06:57:02] <timeless> and i'm not even sure it'd help
[06:59:30] *** EdLin has joined #opensolaris
[07:05:44] <twincest> i don't suppose anyone knows off hand if vmware server can emulate a serial console?
[07:06:30] <timeless> google seems to think so
[07:06:41] <jamesd> yes... that is what i read, though i have never done it
[07:06:52] <twincest> thanks, i'll have to rtfm a bit
[07:06:57] <timeless> socat
[07:06:58] <twincest> emulating a vga console for solaris seems a bit pointless :)
[07:07:22] <timeless> heh
[07:08:07] <timeless> emulating vga for an xserver is pointless to imo. use ssh x11 forwarding
[07:08:29] <twincest> i don't plan to use X, just wanting a platform for ON dev work
[07:08:59] * timeless nods, .. personal+unrelated opinion
[07:10:54] * timeless wonders if noglobals matters
[07:13:13] * timeless ponders filinga bug on open grok
[07:16:26] <twincest> hmm, the docs claim you need full distribution, but i'm fairly sure some of it's not needed to compile ON - apache2 for example?
[07:16:56] <Doc> hmm.. UK are finally relaxing their airport carry-on restrictions a little
[07:17:15] <timeless> there was some  explanation for just building kernel or whatever
[07:17:16] <elektronkind> I'm rockin' my first lx brandz
[07:17:29] <loke> doc: in what way?
[07:17:51] <timeless> i replaced my lx brandz w a nexenta bbrand :)
[07:18:11] <Doc> one bag allowed. liquids less than 100 mls allowed
[07:18:28] <loke> doc: they didn't even allow one bag before?
[07:18:33] <timeless> actually i think centos stopped booting too but i don't care
[07:18:54] <Doc> no, nothing at all
[07:18:59] <loke> doc: and I thought it was annoying when I went back from India a few weeks ago... At least they let me carry my backpack. No way I'm checking in my laptop
[07:20:31] <Doc> i've done 12 flights on multiple airlines in the past few weeks... other than one occasion where they confiscated my 2 oz container of sunscreen (because the container wasnt in a plastic bag, even though it was on the only container i had!) i didnt have any trouble
[07:21:12] <elektronkind> that one ziploc bag could prevent terrorism!
[07:21:22] *** EdLin has left #opensolaris
[07:21:30] <loke> doc: but did you travel to/from the UK?
[07:21:48] <loke> the UK is now on my list of countries I refuse to travel to :-)
[07:21:55] <Doc> well, they did search one of my bags once. the guy searching it is trying to be all strict and rigid as he did it, and the guy next to him and i were chatting about my new sony vaio - was quite humerous the contrast :)
[07:22:06] <Doc> loke: nah..  australia, us and south america only
[07:22:23] <loke> Only place I travelled since december last year was India
[07:22:33] <loke> and that was only on a 48 hour trip :-)
[07:23:12] <Doc> heh.. my last trip to india was only about 15 hours in the air, and that was because i went the long way (via singapore)
[07:23:27] <loke> doc: from where? Australia?
[07:23:54] <Doc> yah
[07:24:09] <loke> doc: I'm based in Singapore though, so the trip isn't that bad... except for the fact that the Singapore Airlines flight to Mumbai arrives at around midnight, and the flight back is around midlight too
[07:25:10] <loke> the flight back gives you a very short night, since the flight time of about 4,5 hours plus 3,5 hours time difference means that you arrive at 8 in the morning. Just abouth right to be able to go to the office...
[07:25:15] <loke> that's pain
[07:25:44] <Doc> yah.. i've done that flight (well, the equiv one from delhi to sin anyway)
[07:26:29] <Doc> had a full day in delhi, caught that flight, had a full day in singapore, then caught the midnight flight to sydney...  slept from before takeoff until just before landing (in business class too :)
[07:28:05] <loke> doc: yeah. business class is nice
[07:28:12] <loke> what did you do in delhi?
[07:28:42] <Doc> not much.. was the starting/ending point for a trip through north india, nepal, tibet and pakistan
[07:35:37] <loke> ah
[07:35:43] <loke> thought it was business
[07:36:08] *** bank has joined #opensolaris
[07:36:55] *** dakrone has quit IRC
[07:36:59] *** sala has joined #opensolaris
[07:37:59] *** dakrone has joined #opensolaris
[07:38:21] <sala> hum... i wanna use xfce4 as wm, and i have installed xfce4 via pkg-get,
[07:39:01] <sala> but i can't find it when i login on X.
[07:39:17] <sala> my os is solaris10 x86
[07:39:31] <sala> and i have installed xfce4_dtlogin .
[07:40:00] <sala> what shall i do next?
[07:40:03] <sala> thx!
[07:40:37] <timeless> so any luck finding the kernelmodule compiler flags ? :(
[07:41:08] *** salamanders has quit IRC
[07:49:55] <bank> how can i upgrade mysqlclient on solaris to support newer version connection?
[07:50:10] <LeftWing> Install a newer one?
[07:51:20] <bank> yes, I try by install mysql5 on this zone. but It still said Client does not support authentication protocol requested by server; consider upgrading MySQL client
[07:51:33] <bank> on another zone I have mysql 5.
[07:51:47] <bank> so I think ... I need upgrade MysqlClient ....
[07:52:10] <bank> install mysql5 isn't help this
[07:52:19] <LeftWing> Are you sure you're running the correct copy of mysqlclient ?
[07:52:24] <elektronkind> solaris comes with mysql 4.0.x clients and libraries, and I think you need at least 4.1.x to talk to a mysql 5 server
[07:53:04] <bank> yes .. but how do I know. In fact I am installing phpmyadmin. and it show sql error
[07:53:15] <bank> that state consider upgrading mysql client.
[07:53:23] <bank> so I install mysql5 on this zone.
[07:53:30] <LeftWing> Well that depends on what MySQL libraries you built PHP against.
[07:53:51] <bank> so I need to rebuild again?
[07:54:08] <LeftWing> Well it depends on what you built it against in the first place.
[07:54:14] <bank> ./configure --prefix=/tank/local/php-5.2.0 --with-apxs2=/usr/apache2/bin/apxs --with-config-file-path=/etc/apache2/php.ini --with-mysql=/usr/sfw
[07:54:19] <bank> * / usr/sfw
[07:54:32] <LeftWing> Well the MySQL in /usr/sfw that comes with Solaris is MySQL 4.0
[07:54:33] <LeftWing> .X
[07:54:42] <LeftWing> Where did you install MySQL 5?
[07:54:54] <bank> default location.. may be .. opt/mysql
[07:55:12] <LeftWing> Well try rebuilding with --with-mysql=/opt/mysql
[07:55:34] <bank>  /opt/mysql/mysql/bin/mysql
[07:55:47] <bank> ok thank you.
[07:56:35] <Doc> tpenta: got a copy of b51_x86 CD isos anywhere handy?
[07:57:12] <Tpenta> I think I burned a b51 dvd, not sure where I put it
[07:58:06] <Doc> hmm.. suppose i could use dvd
[07:58:35] <Doc> screw it, i'll just do it tomorrow
[07:58:40] <Tpenta> :)
[07:59:00] <LeftWing> Zounds.  DIMM3 has failed.
[07:59:14] <Doc> or maybe i'll just stick with windows xp
[07:59:48] <noyb> timeless: did you find this: http://blogs.sun.com/eschrock/entry/adding_a_module_to_opensolaris
[08:00:04] <Error_404> it's funny when people "threaten" with using windows
[08:00:10] <noyb> timeless: also: http://www.mail-archive.com/opensolaris-help at opensolaris dot org/msg00871.html
[08:00:33] <Error_404> eg, "make it have ${feature}" 'no.' "maybe I'll just use windows" 'okay, what do i care?'
[08:00:40] <noyb> timeless: there are a few names in there which may have better answers than me.  Sorry.
[08:00:43] <Doc> problem is that if i install solaris on this thing the battery will probably last about an hour, which makes it moderately useless
[08:02:23] <Doc> hmm.. tad under 3,000 photos in 5 weeks - not bad
[08:08:09] <bank> LeftWing:When I install everything I do in the same manner like .. gunzip, tar-xvf , ./config --prefix=location to install , make , make install
[08:08:32] <bank> so when I uninstall  I can just rm -rf those directory?, Or I just copy this folder to other zone?
[08:09:08] <bank> are there any other file distributed in other location?
[08:13:33] <noyb> timeless: http://www.opensolaris.org/os/community/on/devref_toc/devref_3b/;jsessionid=0A11475A583BE83A1E6D12E91B88770E
[08:14:12] <noyb> timeless: ok...  that's about it for me.  I'm probably no help to you from here.
[08:15:33] <noyb> timeless: FYI, my google search was: build a kernel module site:OpenSolaris.org CFLAGS
[08:17:19] *** anthony79 has joined #opensolaris
[08:18:29] *** kleppari has quit IRC
[08:20:28] <bank> hello I got the same error as this http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6362386
[08:34:01] *** bank has quit IRC
[08:40:41] *** blueandwhiteg3 has joined #opensolaris
[08:44:55] *** bank has joined #opensolaris
[08:45:45] <bank> I upgrade both libxml2 and zlib by use package from sunfreeware
[08:45:52] <bank> but still got this http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6362386 .
[08:46:16] <bank> do I need to upgrade gcc?
[08:47:38] *** robj_ has joined #opensolaris
[08:49:58] <loke> bank: I'd never recommend anyone to use sunfreeware. Their packaging pretty much stinks
[08:50:29] *** sahafeez has joined #opensolaris
[08:51:22] <bank> :( oh
[08:51:34] <sahafeez> finally have a 100% for sure answer why the toshiba dvd sd-m1401 drives in the older suns will not read burned dvd 90% of the time ..
[08:52:42] <sahafeez> while we all think that the a dvd is 4.7gb, turns out that this model will only read 3.9gb dvd-r disk
[08:53:47] <sahafeez> which is the authoring type at the time of release. the disk are $6 a piece
[08:54:04] <bank> loke: I don't know how can I solve.
[08:54:23] <loke> bank: I'd compile it all from cratch
[08:54:24] *** rodrickbrown has quit IRC
[08:54:31] *** rodrickbrown has joined #opensolaris
[08:54:43] <Error_404> or blastwave
[08:54:52] <Error_404> speaking of which, where's dclarke?
[08:54:53] <loke> bank: two advantages: 1) no dependency on /usr/local, which always is a mess when having a shared /usr for zones
[08:55:31] <loke> bank: and 2) no dependency on gcc. Why sunfreeare uses gcc for everything I just can't understand. Particularly a problem for C++ libraries (which are not compatible wih the Sun C++ compiler)
[08:57:15] <bank> I always do that loke, until I got the problem during MAKE
[08:57:16] <sahafeez> what does blastwave use?
[08:57:34] <delewis> I know they use gcc for some packages
[08:57:41] <delewis> MPlayer for instance won't build with Sun Studio
[08:57:48] *** dakrone has quit IRC
[08:57:54] <sahafeez> well you should fix that ;)
[08:57:55] <oxygene> bank: www.freshmeat.net/projects/pmpkg might help you with php5.. download the tarball, unpack it somewhere, enter www/php5 in it, run "../../share/pmpkg install", wait (on solaris 10 run the same in library/libxml first)
[08:58:02] <loke> delewis: mplayer kinda sucks on several levels. but yes, it's full of gcc-isms
[08:58:09] <oxygene> bank: it assumes you have sun studio installed, but can be set to gcc with little work
[08:58:12] <delewis> loke: I know. I maintain it. :-)
[08:58:29] <loke> delewis: haha :-) ok
[08:58:30] <delewis> the code is actually quite standard.
[08:58:39] <delewis> the probably is the use of various gcc extensions
[08:58:42] <sahafeez> <=- is pissed he has to buy a new scsi dvd rom
[08:58:47] <delewis> sahafeez: no
[08:58:49] <delewis> just buy an acard device
[08:58:56] <sahafeez> acard?
[08:58:58] <delewis> then you can use any IDE/ATAPI devices as a SCSI device
[08:59:11] <sahafeez> hum. will it boot on a sparc
[08:59:18] <delewis> of course it will
[08:59:24] <delewis> storage devices don't require FCodes
[08:59:27] <delewis> just the controllers, themselves.
[08:59:28] <sahafeez> cool.
[08:59:29] <sahafeez> done.
[08:59:39] <delewis> you can pop a DVD-RW in your SPARC system and have it work as a SCSI device :-)
[09:00:13] <sahafeez> i am piss now that i know the reason the 2 sun/toshiba dvd i have here will not work
[09:00:19] <delewis> http://www.acard.com.tw/english/fb01-product.jsp?idno_no=23&prod_no=AEC-7720U%20(Support%20ATAPI%20Device)&type1_title=SCSIDE%20Bridge&type1_idno=2
[09:00:52] <sahafeez> will it fit in to my u80, hum
[09:01:01] <delewis> I know people that have got them into U2s
[09:01:04] <delewis> which are far more cramped. :-)
[09:01:26] <Error_404> my u2 is unfortunately dieng
[09:01:30] <Error_404> dying
[09:01:37] <Error_404> stupid language
[09:01:42] * delewis just got a U2 in
[09:01:45] <delewis> first time I've used one :-)
[09:01:51] <delewis> very nice systems.
[09:01:56] <delewis> quite fast, too
[09:02:04] <sahafeez> i had one. sold it.
[09:02:11] <sahafeez> got the u60
[09:02:21] <sahafeez> then sold that and got the u80
[09:02:26] <delewis> considering dropping a QFE and 2x36GB drives into it and making it my router, httpd, smtpd, pop3d, bind, etc.
[09:02:28] <Error_404> delewis: mine's got  a bad ram socket
[09:02:34] <delewis> Error_404: ouch.
[09:02:43] <Error_404> FMD keeps spitting up with bad memory blocks
[09:02:50] <delewis> sure it's not a bad DIMM?
[09:02:55] <sahafeez> FMD?
[09:03:00] <bank> oxygene:install is not a valid command
[09:03:01] <delewis> sahafeez: fault manager daemon
[09:03:04] <sahafeez> ah
[09:03:06] <bank> does it mean I need to set to gcc?
[09:03:41] <Error_404> delewis: certain of it, i've shuffled them around all over the place
[09:03:48] * delewis would never consider getting rid of his Sun hardware
[09:03:53] <delewis> I've still got my SS5 :-)
[09:04:05] <delewis> SS5, U2, SB1000, and soon to be E4500
[09:04:34] <bank> oxygene: would you mind to tell how to set pmpkg to use gcc ...
[09:05:28] <oxygene> argh...
[09:05:42] <oxygene> bank: sorry... that tarball is broken - wait a second
[09:06:13] <bank> ah..
[09:06:41] * oxygene uploads new one
[09:06:50] <delewis> hmm, annoying.
[09:06:53] <bank> oxygene: how do you think with jboss web?
[09:06:59] <delewis> there's no direct links to the Solaris 8 page anymore
[09:07:03] <delewis> had to google for it. :-)
[09:07:04] <sahafeez> i still have my ss5, heck it has a pcmica sbus card
[09:07:13] <oxygene> bank: never worked with jboss
[09:07:23] <delewis> sahafeez: nice
[09:07:31] <delewis> though, I doubt it can even utilize all of the pcmcia bus
[09:07:36] <sahafeez> i have no use for it but i do not want to toss it
[09:07:43] <bank> just feel tired with ..php
[09:07:44] <oxygene> bank: http://openbios.org/~oxygene/Projects/pmpkg/dist-2006-11-05.0907.tar.bz2
[09:08:01] <sahafeez> i had it running openbsd with an intel pccard 802.11 as an ap
[09:08:34] <delewis> hehe, Solaris 8 is tiny.
[09:08:49] <delewis> slightly over 500MB download.
[09:09:48] <sahafeez> were can i get it?
[09:10:06] <sahafeez> i will use the ss5 as a jumpstart box.
[09:10:13] <sahafeez> since i cannot use my x86 box
[09:10:24] <bank> oxygen: thank you. but it still ../../share/pmpkg install
[09:10:33] <bank> not a valid command?
[09:11:06] <delewis> http://www.sun.com/software/solaris/8/
[09:11:35] <sahafeez> cannot get 6 or 7 tho
[09:11:41] <delewis> no
[09:11:48] <delewis> 6 and 7 were not free, in any way.
[09:12:03] <delewis> 8 was free if you were a hobbyist for n number of CPUs
[09:12:10] *** loke has quit IRC
[09:12:24] <delewis> 9 was free if you had under 4 CPUs
[09:12:51] <delewis> and there was no 6, either. :-)
[09:13:11] <bank> ../../share/pmpkg only this it is downloading. ..
[09:13:20] <twincest> i'm sure i remember there being some form of free download for 7
[09:13:58] *** loke has joined #opensolaris
[09:14:11] <sahafeez> 2.6
[09:14:58] *** blueandwhiteg3 has quit IRC
[09:15:16] *** bank is now known as bank__
[09:17:02] <twincest> ooh, CDE 1.7
[09:19:11] <sahafeez> give me 2.51 with openwin
[09:22:45] *** Ezechial_ has joined #opensolaris
[09:25:51] *** sahafeez has quit IRC
[09:26:38] *** sahafeez has joined #opensolaris
[09:28:28] <Doc> 6 never existed, and 7 was never free
[09:28:54] *** mrdeviant has joined #opensolaris
[09:29:36] <loke> doc: you referring to CDE 2.6?
[09:29:40] <loke> i mean 1.6
[09:29:49] <Doc> solaris
[09:30:20] <loke> doc: Solaris 2.6 existed
[09:30:35] *** mrdeviant has quit IRC
[09:30:44] <Doc> did i say it didnt?
[09:30:51] <loke> and Solaris 7 was free on SPARC machines up to 8 CPU's (or was it 4)
[09:30:58] <loke> IIRC
[09:31:06] <Doc> no, that was Solaris 8
[09:31:09] <Auralis> 7 was not free
[09:31:26] <Auralis> 7 edu however did cost only like 20 bucks
[09:34:34] *** sala has quit IRC
[09:37:22] <twincest> hm.. seems the vmware pcn vs. solaris bug got fixed
[09:37:53] *** robj_ has quit IRC
[09:41:56] <loke> Auralis: ah. maybe it was so
[09:43:32] <delewis> loke: prior to Solaris 8 there were two versions of each release -- a regular one and a server version.
[09:43:35] <delewis> they were *not* free
[09:43:45] <delewis> especially, the server version.
[09:46:18] <loke> delewis: there were not
[09:46:31] <loke> there was only one release.
[09:46:43] <loke> there was different licensing though, but that was merely a contract thing
[09:47:35] *** uebayasi has quit IRC
[09:48:01] <oxygene> bank__: sorry.. another tarball (tarball script was broken) http://openbios.org/~oxygene/Projects/pmpkg/dist-2006-11-05.0947.tar.bz2
[09:48:23] *** anthony79 has quit IRC
[09:59:06] *** sahafeez has quit IRC
[09:59:14] *** Fish has joined #opensolaris
[09:59:40] <Fish> hello
[10:03:04] *** pseudo_ has quit IRC
[10:04:31] <twincest> is the Hg repository stable enough for developing from?
[10:07:32] *** PosixC has joined #opensolaris
[10:07:33] *** PosixCompliant has joined #opensolaris
[10:08:37] *** PosixCompliant has left #opensolaris
[10:10:07] *** alanc has quit IRC
[10:10:08] *** Gman has quit IRC
[10:10:59] *** alanc has joined #opensolaris
[10:11:05] *** Gman has joined #opensolaris
[10:11:10] *** sahafeez has joined #opensolaris
[10:12:58] <PosixC> I am trying to install SXCR 50 on t2000 and from the ok prompt I run boot cdrom ; the boot of the cd starts and I get a panic ; first there is a message
[10:12:59] <PosixC> ok Rebooting with command boot cdrom
[10:12:59] <PosixC> then
[10:12:59] <PosixC> WARNING: px1: ino 0x4 blocked
[10:12:59] <PosixC> I don't know what this message means and how to treat it
[10:13:11] <PosixC> after it immedaitely a panic:
[10:13:11] <PosixC> uata: atapi_reset_drive
[10:13:11] <PosixC> uata: ata_timeout_func
[10:13:11] <PosixC> uata: ghd_timer_newstate
[10:13:11] <PosixC> uata: ghd_timer_poll
[10:13:12] <PosixC> uata: ghd_timeout_softirq
[10:13:15] <PosixC> any idea what can it be ?
[10:13:20] <PosixC> what is uata ?
[10:13:45] <Doc> what happens if you try and boot a solaris 10 CD on it?
[10:15:37] <PosixC> Doc, I try to fetch some media; Are you saying that the media is no good ?
[10:16:05] <Doc> could be bad media. could be a dodgy DVD drive
[10:16:08] <PosixC> If I can mount this media on a different machine - does it says something
[10:16:29] <PosixC> because this media is brand new (philips)
[10:16:38] <Doc> there's some dodgy code in the uata reset code that will panic a system when it really shouldnt - i've come across that one before
[10:17:14] <PosixC> I know what is "ata"; what is uata? die drivers  for cdrom/dvd?
[10:17:28] <Doc> but it's always been triggered by faulty hardware in my experience
[10:17:38] <Doc> yah.. it's the driver for the DVD
[10:17:59] <PosixC> Doc, ok , I will try to talk with the dealers
[10:18:23] <Doc> try a sol10 CD first
[10:19:24] <PosixC> again: any clue: if I can mount this DVD from a different machine - does it says something? I had verified DVD md5
[10:19:34] <PosixC> and the md5 was ok
[10:23:00] *** mayday_jay has quit IRC
[10:26:54] <delewis> PosixC: have you verified the media, itself?
[10:29:02] <sahafeez> the sun dvds are fucking dumbass stupid about what dvd media you burn to. they do not all work
[10:29:11] <PosixC> delewis, NO ; I will do it
[10:29:36] <delewis> sahafeez: who boots off DVD, anyway :-)
[10:29:47] <PosixC> sahafeez, on a different t2000 with a different media I had prepated it worked first time!
[10:29:51] <delewis> I download the DVD image, loopback mount it, and run setup_install_server then do a jumpstart.
[10:30:01] <sahafeez> well untill i get my solaris 8 up on my ss5 for a js, i do
[10:30:02] <delewis> much more efficient.
[10:30:07] <PosixC> I had tried with SUN solaris 10 original DVD now
[10:30:14] <PosixC> I got the following
[10:30:21] <delewis> PosixC: Solaris 10 FCS did not support T2000/T1000
[10:30:24] <PosixC> rebooting with commnad : boot cdrom
[10:30:29] <delewis> I think that was S10u1
[10:30:56] <PosixC> it has 3/05 solaris 10 ; is it good or not ?
[10:31:01] <delewis> PosixC: nope
[10:31:06] <delewis> you want S10u1 or S10u2
[10:31:15] <delewis> 3/05 did not support Niagra
[10:31:29] <PosixC> delewis, thnks ; ok, so how should check the media ?
[10:31:39] <delewis> PosixC: digest -a <cdrom device>
[10:31:42] <delewis> er
[10:31:46] <delewis> digest -a md5 <cdrom device>
[10:31:50] <delewis> if you're on a Solaris system that is
[10:31:57] <delewis> or use md5sum (possibly, gmd5sum)
[10:32:12] <PosixC> delewis, can I run md5sum on /dev/xxx from linux ?
[10:32:17] <delewis> sure
[10:32:34] <PosixC> ok, trying...
[10:32:53] <sahafeez> there is just something wrong when an local rysnc is no faster then the one i am running at the sametime to a box 800 miles away
[10:33:17] <sahafeez> s/rysnc/rsync
[10:35:46] <PosixC> delewis, I am a bit ignorant anout it : should the /dev/hdd be mounted ?
[10:36:03] <PosixC> (/dev/hdd is of course the cdrom device on that linux box)
[10:36:06] *** Auralis has quit IRC
[10:36:30] *** Auralis has joined #opensolaris
[10:37:21] <delewis> PosixC: no, it does not need to be mounted.
[10:37:34] <PosixC> can it be mounted?....
[10:37:39] <delewis> no
[10:37:42] <PosixC> thnks
[10:37:57] <delewis> well, it can be mounted, but you don't want to use md5sum on the mount point
[10:38:02] <delewis> just the actual block device
[10:39:55] <PosixC> ok; Just curious , always wanted to learn more about this: so I understand that md5sum on /dev/hdd will be correct also when mounted but NOT correct when running "md5sum /mnt/cdrom"?
[10:39:58] <PosixC> is it so ?
[10:40:41] <delewis> PosixC: /dev/hdd (or whatever cdrom device) is an ISO filesystem
[10:40:52] <delewis> doing md5sum on /dev/hdd is just like doing an md5sum on an ISO that's laying on your disk.
[10:40:59] <delewis> there's *zero* difference, except if the media is bad.
[10:41:04] <PosixC> I see
[10:42:52] <PosixC> ok, md5summing now... in case it will be good - can I try to connect TOSHIBA DVD from x86 which works ; or does the t2000 needs something special ?
[10:43:27] <delewis> PosixC: you should be able to swap out the drives
[10:43:41] <delewis> drives, themselves do not require Fcode of any sort -- only controllers
[10:44:24] <PosixC> don't know what is Fcode ;l am quite new to silarsi
[10:44:32] <PosixC> s/silaris/solaris
[10:44:45] <delewis> it has nothing to do with Solaris, but with SPARC (more specifically, with OpenBoot)
[10:46:08] <bank__> try to build php on solaris for a day. but it always end sadly
[10:46:32] <bank__> ld: fatal: relocations remain against allocatable but non-writable sections collect2: ld returned 1 exit status *** Error code 1 make: Fatal error: Command failed for target `libphp5.la'
[10:46:33] <PosixC> this md5sum /dev/hdd does not seem to finish...
[10:46:50] <delewis> PosixC: it will whenever it reaches EOF
[10:47:13] <Doc> did you try #php ?
[10:47:25] <Doc> remember the road/locked car analogy?
[10:47:35] <PosixC> delewis, it seems to be more than twice/three times  slower than md5sum on hd
[10:47:39] <PosixC> hardisk
[10:47:46] <delewis> PosixC: of course it is
[10:47:54] <delewis> you're reading from a DVD
[11:57:31] *** echelog has joined #opensolaris
[11:59:34] *** mega has quit IRC
[12:01:14] *** lloy0076 has left #opensolaris
[12:03:13] *** rivanvx has left #opensolaris
[12:04:17] *** setuid has joined #opensolaris
[12:04:20] <dwc-> I got my apache2 in /usr/apache2 at install time... I don't think it was on any extra cds
[12:07:23] *** loke has quit IRC
[12:09:17] * quasi could think of several reasons not to use blastwave ;)
[12:09:25] <setuid> why
[12:09:55] <quasi> mostly the dependency hell
[12:10:12] <setuid> I use just their mirrors
[12:10:23] <quasi> I tried svn once and that needed something like 40+ packages
[12:10:31] <setuid> oh
[12:10:46] <setuid> so maybe you have to try portaris
[12:11:01] <setuid> I'm going to try it too
[12:11:18] <quasi> I'm quite happy building my own software
[12:11:24] <setuid> ok
[12:12:01] <setuid> I still don't have enought experience to build bigger projects
[12:13:20] <setuid> I can't find the on-src tarball
[12:13:27] <setuid> on opensolaris.org
[12:13:35] *** jmcp has joined #opensolaris
[12:13:58] <quasi> try the download link ;)
[12:14:06] <setuid> yeah I know
[12:14:15] <setuid> but three are just bfu archives
[12:14:23] <setuid> and some non-free packages
[12:15:18] <setuid> b>x86 Platform Files
[12:15:37] <setuid> no on-src
[12:16:12] <setuid> I'm just newbie but i remebmer that it was there
[12:16:49] <setuid> sorry
[12:16:55] <setuid> there is the link
[12:27:31] <tassieboy> hey guys
[12:28:21] <setuid> hi
[12:28:32] <jmcp> good evening
[12:28:33] <tassieboy> I'm from good 'ol down south (Australia) and have been offered a position in a university to experiment with the Open Solaris system for an Operating Systems course
[12:28:40] <jmcp> great!
[12:28:47] <setuid> yeah
[12:28:52] <jmcp> tassieboy: UTas?
[12:28:59] <setuid> I'm on the f**** high school
[12:29:06] <tassieboy> more or less playing around and seeing what are possible changes to the kernel for the students
[12:29:12] <jmcp> tassieboy: ah
[12:29:13] <tassieboy> no, Uni of Qld
[12:29:15] <setuid> where they now just M$ Windows
[12:29:22] <tassieboy> lived in Tas most my life though
[12:29:23] <jmcp> tassieboy: even better... that's my alma mater
[12:29:48] <tassieboy> but yeah, I don't know much about Open Solaris
[12:30:01] <jmcp> tassieboy: you've come to the right place
[12:30:03] <setuid> I'm too newbie
[12:30:10] <tassieboy> I've just spent the last hour or so reading Wiki etc. finding out about everything easy to learn
[12:30:13] <setuid> but I have 6years experience with Linux
[12:30:14] <tassieboy> it's looking really interesting so far
[12:30:20] <jmcp> so you've already found genunix.org and opensolaris.org, right?
[12:30:25] <jmcp> which subject @ UQ, btw?
[12:30:32] <tassieboy> I've found open solaris and wiki
[12:30:42] *** setuid has left #opensolaris
[12:30:46] *** setuid has joined #opensolaris
[12:31:00] <tassieboy> also found a really useful blog entry explaining the differences between Open Solaris, Solaris Express, Solaris Express Comunity Edition
[12:31:20] <tassieboy> my experience in Unix goes back to first year uni (I'm 3rd year now)
[12:31:21] <setuid> It's not so difficult
[12:31:34] <jmcp> tassieboy: gosh you're young :)
[12:31:43] * jmcp waxes his aged whiskers .......
[12:31:44] <tassieboy> the subject is Operating System Architecture
[12:31:54] 
[12:31:59] <setuid> I'm younger
[12:31:59] <tassieboy> I am young :P
[12:32:00] <jmcp> tassieboy: which department?
[12:32:05] <tassieboy> feeling younger and younger
[12:32:13] <jmcp> comp3301
[12:32:16] <tassieboy> it's part of the software engineering department in AITT
[12:32:18] <tassieboy> *ITEE
[12:32:23] <tassieboy> yeah, 3301
[12:32:45] <tassieboy> I've been offered an awsome opportunity by a lecturer
[12:32:50] *** setuid has left #opensolaris
[12:33:11] <tassieboy> to work over summer (summer here atm), to tutor the course, and to possibly be the Open Solaris Sun Ambassador for UQ
[12:33:14] <jmcp> tassieboy: go for it
[12:33:53] <tassieboy> I'm planning on throwing myself in ;) biggest thing that's happened career wise by far
[12:34:22] <tassieboy> but yeah, I've been watching this room for the last couple of days a bit (and will be a lot in the future I imagine)
[12:34:35] <tassieboy> so I wanted to say g'day and introduce myself
[12:34:53] <jmcp> tassieboy: welcome
[12:34:54] <tassieboy> and see if there was any particular resource you'd recommend for getting started
[12:34:57] <tassieboy> thanks :)
[12:35:30] <jmcp> tassieboy: in your case, http://www.opensolaris.org/os/community/on and look at the "Development Process" and "Developer's Reference" links on the lhs
[12:36:13] <jmcp> tassieboy: I graduated from UQ in 1995
[12:36:26] <tassieboy> what degree did you do?
[12:36:29] <jmcp> BA :)
[12:36:49] <tassieboy> and you're working in the IT industry?
[12:36:50] <jmcp> well, it was really a mishmash of maths, physics, modern history
[12:37:06] <jmcp> yeah.... was working for Sun until they retrenched me in September
[12:37:19] <tassieboy> bugger - where abouts are you working?
[12:37:22] <jmcp> Sydney
[12:37:36] <jmcp> reporting to Colorado for the last 12 months ... software engineer
[12:37:47] <tassieboy> that's awesome
[12:37:53] <jmcp> yeah!
[12:37:56] <tassieboy> I'm doing my first real software work this summer
[12:38:03] <tassieboy> I checked out IBM Tivolli on the Gold Coast just the other day - it's so cool
[12:38:11] <tassieboy> looks like an awesome place to work
[12:38:30] <jmcp> think carefully about what sort of cultural aspects you want in your employment
[12:38:46] *** bank has joined #opensolaris
[12:38:53] <tassieboy> what do you mean?
[12:39:08] <jmcp> well, fr'instance, IBM and Sun have *very* different engineering cultures
[12:39:29] <jmcp> Telstra compared to Optus compared to Three
[12:39:30] <tassieboy> I'd imagine so
[12:39:50] <bank> :) jmcp
[12:39:55] <jmcp> bank: hi!
[12:39:58] <tassieboy> I was planning on working in the UK when I graduate
[12:40:19] <bank> tassieboy:where do you come from?
[12:40:31] <tassieboy> I'm from Brisbane, Australia
[12:40:34] <jmcp> tassieboy: best advice I can give you on that is to make sure you line up at least a 6 month contract *before* you get there
[12:40:47] <tassieboy> where about in Asia-Pacific are you?
[12:40:54] <bank> yes
[12:41:01] <bank> I am in bangkok
[12:41:06] <tassieboy> hello :)
[12:41:37] <tassieboy> there was a company that said they'd sponsor me if I got through all the needed recuitment stuff
[12:41:40] <tassieboy> so that was pretty promising
[12:41:59] <bank> are you going to graduated this year?
[12:42:13] <tassieboy> but after seeing the atmospher at IBM gold coast I'm thinking of looking at the IBM in UK or possibly even stay here a bit
[12:42:16] <jmcp> tassieboy: get it all in writing
[12:42:27] <LeftWing> Aiiee! IBM!
[12:42:30] <tassieboy> I'm graduating next year - only 12 months left to go :)
[12:42:38] <tassieboy> haha, LeftWing - not an IBM fan?
[12:42:43] <bank> Bachelor degree?
[12:42:48] <LeftWing> tassieboy: Not especially, no.
[12:42:49] <jmcp> tassieboy: Sydney (where I am atm) is mainly focused on services, but Brisbane is much more software engineering / development focused
[12:42:50] <tassieboy> Softare Engineering
[12:43:07] <bank> tassie boy me too!1
[12:43:10] <bank> me too!!
[12:43:15] <tassieboy> :)
[12:43:21] <bank> I will on this march..
[12:43:32] <tassieboy> how long ago/till you do graduate?
[12:43:45] <bank> this is my fourth year
[12:43:48] <bank> years
[12:43:58] <bank> last semester on 4th years.
[12:44:04] <tassieboy> I can't wait to be that close to graduating
[12:44:21] <tassieboy> I have exams and two semesters to go
[12:44:38] <LeftWing> Bah, semesters.  I have years to go. =P
[12:44:51] <quasi> LeftWing: IBM is so much "fun" if you like bureaucracy to an extent that would even scare the french ;)
[12:45:06] <LeftWing> quasi: Heh.
[12:45:15] <bank> tassieboy. SDLC UML QA is your kind?
[12:45:27] <tassieboy> quasi: haha, I was suprised when I went down and saw the IBM work place
[12:45:39] <tassieboy> quasi: they seemed really friendly and cool
[12:45:52] <LeftWing> I imagine that's what they want you to think. ;P
[12:45:55] <tassieboy> I was interviewed by a manager and a programmer without shoes :)
[12:46:02] <bank> hey don't forget to put me in your luggage
[12:46:21] <quasi> tassieboy: some places are worse than others, but do expect to spend about half of your time on "paperwork"
[12:46:33] <tassieboy> yeah, I get that feel a bit too
[12:46:52] <jmcp> all places have paperwork / process overhead
[12:47:00] <jmcp> some handle it more efficiently than others
[12:47:01] <tassieboy> bank: I actually really like the UML side of things, have you heard of MDA?
[12:47:23] <LeftWing> At least having process is better than not having, I guess.  But efficient process is a good thing.
[12:47:24] <tassieboy> bank: when I say really like I mean, they have their place ;)
[12:47:30] <quasi> LeftWing: http://sjov.osteklokke.dk/dilbert_ibm.png really says it all - if anything, it has just gotten even more so since 97
[12:47:51] <bank> yes MDA
[12:48:09] <LeftWing> Haha.
[12:48:14] <LeftWing> Ah, Dilbert
[12:48:24] <tassieboy> quasi: I'm going to have to pass that one on, very amusing
[12:48:59] <jmcp> tassieboy: this is the stuff I was working on most recently for Sun  http://www.jmcp.homeunix.com/roller/resources/jmcp/Getting_to_know_the_SAN_stack.pdf, and I also did a pre-release ZFS demo http://www.jmcp.homeunix.com/roller/resources/jmcp/ZFS_SOSUG17oct2005_preso.pdf
[12:49:08] <bank> lol oh my god
[12:49:31] <tassieboy> oh, I was just reading up on ZFS
[12:49:38] <jmcp> bank: wassup?
[12:49:48] <bank> dilbert cartoon
[12:50:07] <bank> OH JMCP .. YOU MUST WORK FOR SUN ...
[12:50:16] <jmcp> bank: I *used to* work for Sun
[12:50:33] <bank> I know I know
[12:50:35] <bank> :P
[12:50:37] <jmcp> bank: when Sun laid off 5000 staff in August/September, I was one
[12:50:38] <jmcp> ha
[12:50:51] <bank> oh .... bad
[12:50:56] <jmcp> it's ok
[12:50:56] <LeftWing> Did they actually get up to 5000?
[12:50:57] <tassieboy> why did they lay off so many?
[12:51:08] <bank> yes ... 5000 oh nooo
[12:51:17] <jmcp> LeftWing: well, I believe 5000 were notified.... the actual end-of-employment date varies due to local laws
[12:51:25] <LeftWing> tassieboy: That's a relatively small number of people for a large multinational company to lay off these days. =P
[12:51:32] <jmcp> tassieboy: to get costs in line... and partly to satisfy Wall St
[12:51:33] <LeftWing> jmcp: Arr. Fair enough.
[12:51:37] <jmcp> LeftWing: 14%
[12:51:52] <tassieboy> jeez
[12:51:53] <jmcp> tassieboy: for years Wall St was calling for layoffs approaching 30%
[12:51:55] <jmcp> yah
[12:52:00] <jmcp> wall st sucks arse :(
[12:52:10] <bank> Wall Street.
[12:52:21] <Stric> 14% would mean like 8 people where I work :P
[12:52:28] * jmcp nods
[12:52:53] *** bank is now known as bank__
[12:53:15] <tassieboy> haha, the place I'm working at over summer would be approx. 0.7 people
[12:53:52] <bank__> working abroad ..
[12:54:03] <jmcp> tassieboy: I used to work @ UQ Library, too
[12:54:46] <tassieboy> jmcp: that's awesome - I was thinking about getting a part time job there
[12:55:17] <tassieboy> jmcp: but then got to focused on uni and hanging out that I didn't have enough time :P
[12:55:32] * jmcp shrugs
[12:55:50] <jmcp> I worked 1/2 time in UQL in my final year, then found a fulltime spot when I graduated
[12:55:57] <tassieboy> jmcp: do you know Philip Machanick?
[12:56:01] <jmcp> nope
[12:56:02] <bank__> are there any possibility for me to working abroad?
[12:56:09] <jmcp> he's the lecturer for comp3301 I see
[12:56:12] <tassieboy> yeah :)
[12:56:22] <tassieboy> well... he wasn't last semester but he supervised it
[12:57:58] <bank__> jmcp : why don't you visit me a bangkok
[12:58:05] <bank__> :P
[12:58:07] <jmcp> bank__: because I'm not travelling that way any time soon
[12:58:12] <jmcp> but when I do, I will be sure to visit you
[12:58:17] <jmcp> & drdoug
[12:58:27] <bank__> drdoug?
[12:58:38] <bank__> the person whom you ever told me?
[12:58:43] <jmcp> bank__: yes
[12:58:59] <jmcp> he used to live in the same city as Brendang and LeftWing, now lives in Bangkok
[12:59:10] <LeftWing> He certainly does.
[12:59:15] <LeftWing> On his way to Japan.
[12:59:21] <LeftWing> (He'll never get to Japan.)
[12:59:44] <bank__> well up until now I never see him
[12:59:53] <bank__> does he online somewhere?
[13:00:06] <LeftWing> He comes online in spurts every now and then.
[13:00:12] <jmcp> online here from time to time
[13:00:20] <LeftWing> He has a blog... http://solaristhings.blogspot.com/
[13:00:24] <bank__> solaris man ?
[13:00:39] <bank__> ah i see
[13:00:42] <bank__> solaris man again
[13:01:34] <bank__> At least he wrote about the coup
[13:03:01] <tassieboy> jmcp: but with ZFS "fly" - amusing ;)
[13:03:18] <jmcp> tassieboy: yes :) ... and the best part is, it really is that simple
[13:03:34] <bank__> tassieboy
[13:03:57] <bank__> which kind of job are you going to apply? on software engineering role?
[13:04:54] *** setuid has joined #opensolaris
[13:04:54] <tassieboy> looking at getting into coding, high level to start off with
[13:05:07] *** vmhobbes has quit IRC
[13:05:14] <tassieboy> wouldn't mind moving into embeded systems etc.
[13:05:45] <tassieboy> but end-goal is to work as a contractor operating a home-business
[13:06:15] <tassieboy> I'm not really sure what's out there yet - more or less really liked computers and html so dived into SE
[13:07:14] <jmcp> ugh
[13:07:21] <jmcp> html is a markup language, nothing more
[13:07:27] <LeftWing> heh
[13:07:48] <tassieboy> haha - yeah, but it was interesting and thus my move to IT
[13:08:01] <tassieboy> but now I'm a C, Java and PHP junkie
[13:08:10] <jmcp> tassieboy: so what sort of embedded systems are you interested in?
[13:08:17] *** setuid has left #opensolaris
[13:08:32] <tassieboy> jmcp: I'm not too sure - all I know is I really liked the low level C coding I've done so far
[13:08:50] <jmcp> tassieboy: how low level has it been? what sort of architectures? what sort of devices?
[13:09:19] <tassieboy> jmcp: atmel, most of it had to include ASM inserts - building things for Team Projects at uni
[13:09:26] <jmcp> ok
[13:09:33] <LeftWing> Atmel's AVR 8-bit RISC?
[13:09:44] <tassieboy> jmcp: the last was a remote thermal logger which communicated via RF to a receiver plugged into a computer via USM
[13:09:53] <tassieboy> jmcp: um, USB
[13:10:02] <jmcp> nice
[13:10:08] <tassieboy> LeftWing: yeah, I'm a sucker for the ATtiny13
[13:10:13] <LeftWing> heh
[13:10:18] <LeftWing> I just ordered an ATmega32 development board.
[13:10:24] <tassieboy> LeftWing: only 8 pins (6 usable) but a whole lotta power
[13:10:45] <LeftWing> I've used some of Microchip's PIC16* chips in about that footprint.
[13:11:04] <tassieboy> I've heard bad things about PIC but never used them myself
[13:11:12] <tassieboy> had a look at a datasheet once and grimaced
[13:11:19] <jmcp> tassieboy: if you're interested, have a look at the PPC port project pages on opensolaris.org, and wander around power.org too
[13:11:19] <LeftWing> Anything in particular?
[13:11:31] *** setuid has joined #opensolaris
[13:12:28] <tassieboy> LeftWing: more or less that the datasheets were all pretty horiffic and that they are not supported well - though the internal cystals are supposed to be awesome or something like that
[13:12:58] <tassieboy> jmcp: yeah - I saw some things in that but was a little overwhelmed - the only low level kernel stuff I've done was with Minix
[13:13:24] <jmcp> tassieboy: hell, you've got a long way to go then with OpenSolaris!
[13:13:43] <jmcp> tassieboy: there's a web interface to all that's been opened so far at http://cvs.opensolaris.org
[13:13:46] <LeftWing> tassieboy: Have you done any Real-time Operating System work?
[13:13:46] <tassieboy> jmcp: I know - that's what so scary :)
[13:13:50] <jmcp> which has *wicked* search engine
[13:14:17] <tassieboy> jmcp: haha - I've had a quick go and read something about it being dubbed "Wicked fast"
[13:14:39] <jmcp> tassieboy: it's even better that it's a true description
[13:14:46] *** vmhobbes has joined #OpenSolaris
[13:15:00] <bank__> bank is run away and back to his java world
[13:15:05] <tassieboy> LeftWing: I haven't done much OS work at all appart from a few quick things to Minix (implemented immediate files, changed the scheduler, something else)
[13:15:06] <jmcp> soft!
[13:15:08] <jmcp> :)
[13:15:24] <LeftWing> Fair enough.
[13:15:30] <tassieboy> bank: I don't blame you :P
[13:15:35] <jmcp> I don't mind java at all. I use it from time to time, but I'm more on the hardcore C and kernel side
[13:16:07] <tassieboy> jmcp: I use java for anything that needs to be done quickly and/or graphically but I'm a true C fan
[13:16:09] <LeftWing> I had a play with Micrium's MicroC/OS-II in the last week or so.  We had a piece of lab equipment to build for a final design project.
[13:16:21] <tassieboy> jmcp: though ashamedly I've never really programmed in C++
[13:16:35] * LeftWing is hanging out for SunSPOTs though.
[13:16:47] <jmcp> tassieboy: I avoid c++ whereever possible
[13:17:01] <LeftWing> tassieboy: Don't be ashamed.  C++ is dirty.
[13:17:11] <jmcp> dammit, if I've fucked up the memory allocation then I want to know about it asap rather than waiting for some crazy GC to come along
[13:17:16] <LeftWing> Real-time Embedded Java is the future. =)
[13:17:35] <jmcp> http://research.sun.com/techrep/2006/abstract-156.html "Writing Solaris Device Drivers in Java"
[13:17:37] <tassieboy> LeftWing: what a dirty combination of words :P
[13:18:01] <LeftWing> tassieboy: Device control is practically what Java was built for! =P
[13:18:02] <tassieboy> LeftWing: what were you using MicroC/OS-II for?
[13:18:12] <bank__> hooley
[13:18:23] <tassieboy> wasn't Java built originally for PayTV?
[13:18:38] <jmcp> tassieboy: no, toasters
[13:18:39] <quasi> jmcp: that's just sick and twisted! ;)
[13:18:42] <LeftWing> tassieboy: I built a piece of lab test equipment for a Uni project at the start of this week.  It did function generation and some frequency counting and AM/FM signal processing stuff.
[13:18:51] <jmcp> quasi: I haven't worked up the courage to read it yet :)
[13:19:15] <quasi> jvm in the kernel
[13:19:31] <tassieboy> LeftWing: you would have been awesome to talk to in the last few months then - I just finished FSK RF Transmitter receiver firmware
[13:19:54] <tassieboy> LeftWing: I programmed the receiver with a frequency counter
[13:19:56] <jmcp> tassieboy: while you're thinking about ZFS, have a look at http://research.sun.com/minds/2006-0928 - interview with the two architects behind it
[13:20:11] <LeftWing> tassieboy: Cool.
[13:20:23] <tassieboy> jmcp: will do, thanks :)
[13:20:33] <tassieboy> I'm trying to learn as much as I can to impress the lecturer tomorrow
[13:20:53] <LeftWing> Do you have to give a demo?
[13:20:53] <tassieboy> that and figgure out exactly how much time I can commit over the holiday while working full time aswell
[13:20:55] *** setuid has quit IRC
[13:20:58] <jmcp> tassieboy: I can vouch for the fact that both Bonwick and Moore are fantastic to meet in person... we managed to schedule the April SOSUG meeting around their holiday in Aus :)
[13:21:41] <tassieboy> LeftWing: the demo for RF was done a few weeks ago - didn't work so much but I managed to impress the lecturer with a webhosting feature written in Java built into part of the product
[13:21:52] <tassieboy> LeftWing: that part was on the PC of course ;)
[13:22:08] <LeftWing> tassieboy: Of course.  Imagine if it was on the embedded device, though!
[13:22:12] * LeftWing waves his arms at SunSPOTs.
[13:22:40] <tassieboy> Tomorrow I've got to convince the lecturer that I'm the man for the job work the Open Solaris stuff - there are awesome possibilities from doing the work
[13:22:47] <tassieboy> some pretty great contacts I imagine
[13:23:08] <tassieboy> this lecturer knows something about everything, and everything about OS
[13:23:59] <tassieboy> so I'm going to have to know everything that I'm saying and know where it came from to seem... valid
[13:24:10] <LeftWing> What sort of business are you looking at operating?
[13:25:00] *** setuid has joined #opensolaris
[13:25:51] <setuid> Does anyone have any informations about Ati 3d acceleration under solaris?
[13:27:15] <bank__> erm . # /usr/bin/java I got Error occurred during initialization of VM Could not reserve enough space for object heap
[13:27:45] <bank__> it may java problem. brb
[13:27:52] *** bank__ has left #opensolaris
[13:28:08] *** taz has joined #opensolaris
[13:28:14] <taz> damn d/c
[13:28:51] *** setuid has left #opensolaris
[13:29:31] <taz> there are a lot of idle people here in the room
[13:30:33] <jmcp> taz: oi! I resemble that remark
[13:31:05] *** bank has joined #opensolaris
[13:31:15] <PerterB> oh, I thought he mayber mispelled "idol"
[13:31:40] <LeftWing> Really, at one time or another, aren't we all just idling.
[13:32:25] <Stric> Lots of people idle while sleeping for instance
[13:35:37] <Peanut> *nod* good morning everyone :)
[13:35:50] *** tazza has joined #opensolaris
[13:35:51] <bank> good evening
[13:36:03] <Stric> good mid-day
[13:36:12] <tazza> I'm starting to think that maybe Opera isn't the best IRC client :P
[13:36:29] <PerterB> not so much, no :)
[13:36:57] <tazza> I now have 3 IDs in here :)
[13:37:08] <jmcp> tazza: you can get better kitchen sinks
[13:37:34] <tazza> I'm a huge Opera fan atm
[13:37:47] <tazza> it's got built in IRC, bit torrent, etc.
[13:37:55] <tazza> only browser that passes the ACID2 test
[13:38:08] <LeftWing> Opera killed my inner child.
[13:38:14] <tazza> haha
[13:38:24] * tazza sighs
[13:38:45] *** uebayasi has joined #opensolaris
[13:38:52] <LeftWing> Who cares if it passes the ACID2 test.  I care if it passes the "works like Mozilla" test. ;P
[13:38:58] <LeftWing> Firefox usually passes that test.
[13:39:14] <tazza> haha... it doesn't have the magic of Mozilla
[13:39:45] <tazza> if it had plugins it'd be perfect
[13:40:04] <LeftWing> heh
[13:40:08] <tazza> but have a look at the Opera skin and then have a look at the FF2.0 skin and you'll see that it's just a poor Opera take
[13:40:28] <LeftWing> Yes, I can see that it is and I wish they'd go back to the FF1.5 skin -- it wasn't shite. =P
[13:40:46] <tazza> I was using the Brushed theme anyway
[13:40:58] <tazza> FF has some awesome skins
[13:42:52] *** tassieboy has quit IRC
[13:42:59] <tazza> yay - only one to go
[13:43:03] *** tazza is now known as tassieboy
[13:45:51] *** LordKing has joined #opensolaris
[13:49:57] *** taz has quit IRC
[13:53:33] *** bunker has joined #opensolaris
[13:57:38] *** loke has joined #opensolaris
[13:58:29] *** vmhobbes has quit IRC
[13:58:37] <LeftWing> Well, Mercurial is easy to build and use.
[14:00:48] <tassieboy> what are you setting that up for?
[14:01:15] <jmcp> tassieboy: mercurial has been chosen as the SCM tool of choice for OpenSolaris
[14:01:26] <jmcp> Sun presently uses teamware and SCCS to manage source trees
[14:01:37] <jmcp> but teamware isn't open source
[14:02:15] <tassieboy> fair enough
[14:02:45] <jmcp> tassieboy: there's a tools community on opensolaris.org which has a whole heap of info on the reasons behind it all
[14:02:59] <jmcp> and evaluation reports on the other options that were considered too
[14:03:18] <tassieboy> wow - there is so much information available on everything hey
[14:03:32] <jmcp> yup
[14:03:38] <tassieboy> and it all intertwines somehow :P
[14:04:16] <jmcp> :)
[14:04:33] <jmcp> there's a *lot* of history to get your head around too
[14:05:06] <jmcp> and engineering processes too
[14:05:17] <tassieboy> the wiki page for Open Solaris helped a lot with the history
[14:05:30] <jmcp> which are incredibly important to get correct, otherwise you can't possibly deliver a product of its scope
[14:05:32] <tassieboy> a basic cap over the most important parts
[14:06:12] <Yamaraj> Is HAL/Tamarack(sp?) now fully functional in SCXR:51 release, or some bits are still missing?
[14:07:11] <jmcp> Yamaraj: iirc some bits are still missing, I think Gman suggested snv_53 for delivery
[14:08:05] <Yamaraj> snv_53 will have GNOME 2.16.x I presume?
[14:08:10] <jmcp> dunno
[14:08:13] <jmcp> I hope so
[14:09:27] <bank> Doc:
[14:09:45] <bank> Doc:Would you mind to describe the assumption of road and lock car again?
[14:09:46] <Yamaraj> If power management and other laptoppy stuff gets integrated into Solaris, a lot of people, including me, will appreciate it.
[14:10:07] <bank> If we can not open the lock car then blame the road is wrong?
[14:14:10] <bank> jmcp: have a nice dine
[14:14:12] <bank> bye
[14:14:15] <jmcp> good night
[14:14:20] *** bank has quit IRC
[14:15:26] *** Yamaraj has quit IRC
[14:20:47] *** dunc has joined #opensolaris
[14:21:31] <tassieboy> jmcp: that was a pretty inspiring article ;)
[14:22:13] *** vmhobbes has joined #OpenSolaris
[14:45:23] *** dunc has quit IRC
[14:45:39] <tassieboy> night all
[14:45:42] *** tassieboy has quit IRC
[14:48:04] *** dunc has joined #opensolaris
[14:49:49] *** laca has joined #opensolaris
[14:50:55] *** Darwin has joined #opensolaris
[14:55:15] *** bougie has joined #opensolaris
[15:04:42] *** craree has joined #opensolaris
[15:05:10] * craree hello everybody !
[15:05:19] *** craree has quit IRC
[15:07:19] *** PosixC has joined #opensolaris
[15:14:56] *** eboutilier has joined #opensolaris
[15:40:43] *** Darwin has quit IRC
[15:48:20] *** setuid has joined #opensolaris
[15:48:42] <setuid> God evening
[15:48:47] <setuid> sorry
[15:48:48] <setuid> Good
[15:49:17] <setuid> Is there any way to change root shell?
[15:49:34] <setuid> I have set csh as default
[15:49:41] <setuid> and want to have bash
[15:50:03] <quasi> like editing /etc/passwd ...
[15:50:16] <setuid> thanks I will try it
[15:50:29] <setuid> I heard that there exist chsh
[15:50:38] <setuid> but didn't find it
[15:50:44] <setuid> anywhere
[15:52:36] *** setuid has left #opensolaris
[15:52:38] *** deather__ has joined #opensolaris
[16:04:23] *** Somethingelse has joined #opensolaris
[16:10:19] *** deather_ has quit IRC
[16:10:43] *** eboutilier has quit IRC
[16:16:22] <PerterB> Quick sanity check: I have a PC running 20u2, root and swap slice mirrored with SVM plus a mirrored zpool.  I should be able to install Nevada onto an existing, unused unmirrored slice, mirror it up later and fix up Grub to dual boot between the two versions, right?
[16:17:39] <PerterB> uh, "running 10u2" :)
[16:18:01] <dunc> i reckon PerterB yes
[16:18:41] <PerterB> I reckon too, just wondering what gotcha I missed :)  (I know I need to not let Nevada upgrade the zpool format)
[16:21:10] <dunc> i have an update btw, i tried my card and disks in another box and it's ok
[16:21:55] <dunc> but.. I only had 1 sata disk connected at install time, and the 1 sata drive ended up being c0d0 wheras my system disk was that usually
[16:21:58] <PerterB> running solaris?
[16:22:04] <PerterB> aha
[16:22:13] <dunc> so, i reckon 2 devices are trying to become c0d0 somehow
[16:22:43] <dunc> so i'm gonna try on the machine they're destined for later on, with same thing, 1 disk at install time
[16:23:26] <dunc> what's this about  upgrading zpool format btw?
[16:23:29] <quasi> sounds a bit like a raid controller getting in the way
[16:23:38] <PerterB> reconfig boot should sort that kind of thing out
[16:23:57] <dunc> quasi: i can't see any way of enabling or disabling raid on this
[16:24:22] <dunc> there is a "press f3 now" thing at boot, but i can't seem to make it do anything
[16:24:57] <dunc> potentially it's just something about this box
[16:25:11] <dunc> i will report back 2nite
[16:25:24] <dunc> in the meantime i'm off to change the water pump on my van
[16:25:40] <PerterB> sounds fun
[16:26:21] <dunc> i've already put the suspension strut back on today, i've discovered how you're actually supposed to do it i think, rather than what I was doing :)
[16:26:50] <PerterB> often the way :)
[16:32:05] *** LordKing has quit IRC
[16:41:51] *** darkcmd has joined #opensolaris
[16:48:25] *** lopa has joined #opensolaris
[16:50:35] <lopa> how can I enable sendmail ?
[16:51:40] <Peanut> svcadm enable sendmail
[16:52:19] <lopa> ummm
[16:53:13] <lopa> by default is it enable?
[16:54:00] <Peanut> Depends on how new your build is, maybe it got disabled in the 'secure by default' changeover.
[16:54:48] <Peanut> 'svcs sendmail' should tell you what status it has right now.
[17:02:44] <sickness> i'm back
[17:12:53] *** lopa has quit IRC
[17:14:34] <jamesd> uhhh our sickness is back call a doctor immediately
[17:15:08] * PerterB sprays some disinfectant around the channel
[17:15:10] <sickness> :)
[17:15:16] *** rhys has joined #opensolaris
[17:27:56] *** bougie_ has joined #opensolaris
[17:28:46] <PosixC> how can I list all the files belonging to a package named SUNWpackNaeme?
[17:29:06] <PosixC> s/SUNWpackNaeme/ SUNWpackName
[17:29:35] <lasseoe> man pkgchk
[17:29:46] *** georg_e_s has joined #opensolaris
[17:33:41] <rhys> so is solaris a user operating system?
[17:34:05] <delewis> rhys: it's a full time-sharing, multi-user operating system.
[17:34:06] <rhys> as in, put it on my lappy and play music, network tools, browse the web, the like?
[17:34:52] *** loke has quit IRC
[17:35:34] *** georg_e_s has left #opensolaris
[17:35:48] <nbkk6fo_> rhys, depends some would argue that they are better OS's that's geared for that type of work load
[17:35:53] <nbkk6fo_> but yes Solaris can be a useable desktop
[17:35:58] <nbkk6fo_> I personally pefer OSX
[17:36:14] <rhys> i just wonder about application avaliablity
[17:36:31] <delewis> application availability?
[17:36:46] <delewis> Solaris complies with POSIX/UNIX2003, etc.
[17:36:52] <PerterB> yup, there are definitely applications available :)
[17:36:53] <delewis> if the application is written properly
[17:37:02] <rhys> so anything i can run on linux, i can probably run on solaris?
[17:37:05] <delewis> of course it's able to run on your "laptop", should you choose to install Solaris.
[17:37:06] *** koolniczka has joined #opensolaris
[17:37:10] <delewis> rhys: most likely, yes
[17:37:14] <nbkk6fo_> rhys, stickly speaking for a desktop OS your going to be very limited on application support
[17:37:25] <delewis> but the majority of software written for Linux is broken shit.
[17:37:26] <nbkk6fo_> less so than linux, and macosx
[17:37:44] <delewis> nbkk6fo_: no
[17:37:58] <nbkk6fo_> delewis, stop sipping the coolaid
[17:38:09] <nbkk6fo_> for a USE DESKTOP experience linux and macos is alot more usable
[17:38:12] <nbkk6fo_> in terms of commerical support
[17:38:13] <delewis> kool-aid? I've been using Solaris on my primary workstation for well over 2 years now.
[17:38:16] <nbkk6fo_> for that type of applications
[17:38:42] <delewis> nbkk6fo_: sorry, I do real work on my desktop -- document composition, development, and the occasional video watching.
[17:38:46] <delewis> Solaris does all of that quite well
[17:38:47] <nbkk6fo_> delewis, and compared to MacOSX and say linux you find better app support on Solaris ?
[17:38:54] <delewis> so I fail to see how you can call it "limited" for the desktop
[17:39:02] <delewis> unless you're one of those new-fangled, peecee gamer types.
[17:39:10] <delewis> nbkk6fo_: define app support
[17:39:11] <delewis> you haven't yet
[17:39:15] *** bougie has quit IRC
[17:39:19] <delewis> so I'm quite perplexed about your measure for comparison.
[17:39:35] <rhys> things like this though, my lappy has a mach64 videocard... xorg7.1 has that pretty easily for me, 3d support driver. im just wondering if solaris works on odd things like that
[17:39:36] <nbkk6fo_> delewis, lets see does Maya run on Solaris? or say photoshop? :-)
[17:39:55] <delewis> nbkk6fo_: one could've asked that question about Linux several years ago
[17:39:56] <Auralis> lets see you affording maya :)
[17:40:00] <delewis> and received similar responses
[17:40:08] <delewis> stop being hypocritcal with respect to Solaris.
[17:40:09] <nbkk6fo_> delewis, yes but were are talking about today
[17:40:14] <nbkk6fo_> that' has no barring on the issue
[17:40:20] <delewis> nbkk6fo_: I think it does.
[17:40:41] *** lopa_ has joined #opensolaris
[17:40:48] <delewis> is there a Photoshop Linux port? I think not.
[17:40:58] <nbkk6fo_> delewis, let me guess you drive a Mac Truck to the grocery store also ?
[17:41:01] <rhys> gimp
[17:41:07] <Auralis> gimp is a joke
[17:41:09] <delewis> nbkk6fo_: I don't drive, period.
[17:41:48] <nbkk6fo_> rhys, anyway depending on what you need to do you're going to be very limited when it comes to day to day desktop use if you use tons of commerical apps
[17:42:05] <delewis> and regardless, if there is some Linux application that won't run natively or build properly on Solaris, you might still be able to use BrandZ (provided you're on x86_32 or x86_64)
[17:42:09] <nbkk6fo_> they just dont exist on Solaris because that's not really the mark Sun or ISV's cator to on Solaris
[17:42:34] <nbkk6fo_> mark/market
[17:42:35] <delewis> nbkk6fo_: please, stop spreading this nonsense unless you can provide case examples.
[17:43:01] <rhys> i always figured solaris as a server OS, dedicated to running a base set of applications. i was just curious when i saw "opensolaris"
[17:43:10] <nbkk6fo_> delewis, stop being silly and just adimit the fact
[17:43:22] <nbkk6fo_> can you run itune son Solaris?
[17:43:31] <delewis> nbkk6fo_: I'll admit that you're an uninformed twit.
[17:43:37] <delewis> that remedies the problem.
[17:43:38] <PerterB> rhys: you're forgetting sun started out as a workstation manufacturer
[17:44:08] <delewis> and *still* manufactures UNIX workstations.
[17:44:24] <delewis> I'd say 25 years is a good run.
[17:44:27] <nbkk6fo_> and workstation OS and a desktop OS isnt the same thing
[17:44:28] <nbkk6fo_> fyi
[17:44:41] <delewis> nbkk6fo_: it is from my perspective.
[17:44:54] <delewis> there are some of us that actually use our computers for productive purposes.
[17:45:03] <rhys> i see the same thing nbkk6fo_. workstation is still rather dedicated to a handful of tasks.
[17:45:08] <PerterB> sure it is, but I agree Solaris isn't an appropriate desktop OS for many applications
[17:45:21] <nbkk6fo_> PerterB, thank you
[17:45:21] <delewis> *sigh*
[17:45:30] <delewis> if Linux is appropriate for a desktop environment, then so is Solaris.
[17:45:37] <delewis> I fail to see the distinction.
[17:45:46] <nbkk6fo_> delewis, linux has ISV support in many areas that Solaris does not
[17:45:48] <PerterB> Linux isn't an appropriate desktop OS for many applications either
[17:46:13] <delewis> as for OS X, you're riding on a 15-year-old innovation, called NeXTSTEP.
[17:46:23] <delewis> PerterB: if you're willing to accept that, great.
[17:46:27] <delewis> our friend, nbkk6fo_, here is not.
[17:46:55] <nbkk6fo_> if anyone wants a practical UNIX desktop environment today I would tell them to go with MacOSX 1) or 2) Linux and to a limited extent Cygwin on Win32
[17:47:04] <delewis> nbkk6fo_: OS X is not UNIX, first of all.
[17:47:13] <nbkk6fo_> UNIX-Like happy
[17:47:32] <PerterB> The BSD bit of it is more Unix than Linux will ever be ;)
[17:47:33] * Auralis thinks OSx desktop sucks, it has just some good apps
[17:47:47] <delewis> well, let's see what desktop tasks I can do on my Blade 1000 (which is SPARC, and thus even has *less* desktop support than x86_32 or x86_64) would.
[17:47:54] <elektronkind> who cares if it isn't "UNIX" proper. It at least acts like it and that it is predictable in that sense.
[17:48:22] <delewis> I can run Bittorrent, an office suite, encode and decode video and audio, surf the web, communicate (via GAIM or Thunderbird, etc.)
[17:48:31] <delewis> I'd say that matches the desktop bill fairly well.
[17:48:36] <delewis> and I have an imaging suite.
[17:48:45] <delewis> so, what on earth is missing?
[17:48:55] <nbkk6fo_> same can be said about someone who drives a 18 wheeler to work and someone who drives a 2 door compact car
[17:49:00] <nbkk6fo_> still which one is more practical for everyday use ?
[17:49:03] <delewis> elektronkind: it was only in 10.4 that OS X received asynchronous IO.
[17:49:05] <nbkk6fo_> that's the whole debate
[17:49:14] <nbkk6fo_> not that Solaris cant function as a useable desktop
[17:49:17] <delewis> nbkk6fo_: I fail to see the validity of your analogy.
[17:49:19] <elektronkind> ah. well, it has it now, so what's the gripe?
[17:49:30] <elektronkind> it didn't have pthreads until 10.2
[17:49:44] <delewis> elektronkind: it's seriously lacking in the multi-threaded department.
[17:49:59] <nbkk6fo_> delewis, possibly because you've only been using Solaris for 2 years?
[17:50:04] <nbkk6fo_> I've been using UNIX for the past 11 years
[17:50:28] <nbkk6fo_> and I know for a fact Solaris does not make a solid desktop environment compared to other offerings out today
[17:51:02] <delewis> nbkk6fo_: would you like an award?
[17:51:16] <Somethingelse> can you play quake on solaris?
[17:51:18] <elektronkind> nbkk6fo_: could you quantify that statement?
[17:51:23] <Auralis> Somethingelse: yes
[17:51:26] <delewis> Somethingelse: quake1 and quake2
[17:51:31] <Somethingelse> sweet :)
[17:51:35] <delewis> even on SPARC :-)
[17:51:36] <Somethingelse> how about 3 and 4?
[17:51:45] <Auralis> pkg-get -i quake :)
[17:51:47] <delewis> Somethingelse: not that I know of :-(
[17:51:52] <Somethingelse> i play em on linux...so i guess they should work on solaris
[17:51:59] <Somethingelse> a, no good, i want 3 and 4
[17:52:08] <delewis> bah. bzflag is all anyone needs. :-)
[17:52:11] <elektronkind> nbkk6fo_: is it the out-of-the-box software that doesn't make Solaris a desktop OS? It is something else that's missing?
[17:52:13] <lopa_> how much space do I need for solaris/opensolaris?
[17:52:13] <lopa_> as a minimum?
[17:52:14] <LeftWing> I would argue that Maya is as specialised as "Workstation OS".
[17:52:21] <elektronkind> bzglag++
[17:52:24] <nbkk6fo_> elektronkind, vendor support
[17:52:25] <Somethingelse> not that i'd use solaris for gaming, but just for the purpose of this conversation :)
[17:52:28] <delewis> elektronkind: I've been trying to get him to quanity his statement for the last 10 minutes.
[17:52:30] <elektronkind> we keep a bzflag server running in out office
[17:52:34] <delewis> he's unable to argue properly and uses invalid analogies.
[17:52:41] <delewis> thus, I gave up on him.
[17:52:51] <LeftWing> I wouldn't use anything apart from Windows as a gaming platform these days.
[17:52:54] <lopa_> anyone?
[17:52:57] <Auralis> Solaris sparc has Houdini, who needs maya if you can get Houdini, plus there is blender
[17:53:20] <delewis> Auralis: who actually uses Maya on their "desktop," anyway?
[17:53:23] <LeftWing> I've seen some great stuff done with blender.
[17:53:26] * delewis recalls the previous workstation/desktop distinction
[17:53:27] <nbkk6fo_> delewis, I dont see how hard its to understand, Vendors dont see Solaris as a viable OS to port their desktop applications which has limited Solaris for ages for breaking out like say MacOSX etc..
[17:53:55] <elektronkind> matlab, mathematica, SAS... really if you're going to use a sparc/solaris desktop, it ain't for piddling around on interwebnetz with mozilla 100% of the time.
[17:54:01] <delewis> nbkk6fo_: vendors? you're talking about specialization, which implies a workstation-like nature.
[17:54:06] <LeftWing> Microsoft won't port their Office suite, that's true.
[17:54:15] *** uebayasi has quit IRC
[17:54:17] <Somethingelse> LeftWing: i get way better performance in Doom3 and Quake4 in linux then i do in Windows.  i'm talkin double the framerates
[17:54:21] <Auralis> LeftWing: i do my stuff in blender on solaris
[17:54:24] <LeftWing> Somethingelse: Well, I don't.
[17:54:42] <delewis> I think we're all masochistic to certain degrees.
[17:54:49] <Somethingelse> well, you probably have an ATI video card :)
[17:54:51] <delewis> there are some of us that use Solaris as a 'desktop' or whatever term you want to use
[17:55:00] <delewis> and there are some who aren't masochistic to that degree.
[17:55:00] <Somethingelse> or something other then nVidia
[17:55:01] <jamesd> who needs a user interface, there is blender that has none... ;-p
[17:55:02] <LeftWing> Heh, I use Sun Rays as my home computing device.  So that's somewhat masochistic. =)
[17:55:05] <delewis> personally, it makes no difference to me.
[17:55:12] <nbkk6fo_> delewis, that's you
[17:55:23] <Somethingelse> delewis: i think Workstation is the term your looking for ;)
[17:55:26] <nbkk6fo_> but for most people they require more than their desktop environment than Xchat Mozilla and Gaim
[17:55:26] <delewis> Solaris does the things I need to do, and I'm completely satisfied as a desktop/workstation-user.
[17:55:40] <nbkk6fo_> ^out
[17:55:47] <LeftWing> nbkk6fo_: What, like StarOffice as well? =P
[17:55:50] <delewis> Somethingelse: I use my Sun Blade 1000 as my primary "workstation," so I think it would be safe to also term it as my desktop.
[17:56:11] <Somethingelse> sure :)
[17:56:29] <LeftWing> Workstation isn't really a viably separable term from Desktop these days.
[17:56:36] <delewis> but as far as multimedia goes -- Solaris is just as adequate as Linux would be.
[17:56:50] <delewis> mplayer works and builds on Solaris, so what more do you need? :-)
[17:56:52] <nbkk6fo_> delewis, this isnt a Solaris vs Linux debate
[17:56:59] <Somethingelse> how is sun blade?  i was thinkin about getting one...but don't know if it's worth it for me.  i think i'm better off sticking to x86 since i probably won't ever take advantage of what SPARC has to offer
[17:56:59] <nbkk6fo_> its a usability issue
[17:57:08] <LeftWing> So much shite is available from Blastwave at the press of a button for the great unwashed.
[17:57:22] <nbkk6fo_> Somethingelse, they are horrible stay away from the Blade 1k-2k models
[17:57:30] <nbkk6fo_> over priced and very slow
[17:57:35] <delewis> nbkk6fo_: what!?
[17:57:36] <Somethingelse> gotcha...was thinking about the 2k
[17:57:38] <PerterB> let's see... Reason, Ableton Live, Sound Diver, Logic... no real Solaris (or Linux) equivalents for any of those
[17:57:43] <delewis> the SB1000 and 2000 models are the most well-constructed workstations
[17:57:46] <delewis> that Sun has *ever* made.
[17:57:57] <Auralis> SB1k rocks
[17:58:01] <lopa_> how much space do I need for solaris/opensolaris? as a minimum
[17:58:07] <nbkk6fo_> delewis, no one cares about well constrcuted it could have been hand crafted in titanium i would care less
[17:58:14] <nbkk6fo_> I want performance in a workstation
[17:58:16] <delewis> nbkk6fo_: SB1000s were actually faster than x86 hardware when they arrived on the market
[17:58:19] <delewis> for almost *every* task.
[17:58:20] <nbkk6fo_> which the sb line really lacked
[17:58:22] <LeftWing> lopa_: Depends how much of it you want to install.  My Solaris 10 server setup installs about 1.5GB
[17:58:23] <nbkk6fo_> which killed it
[17:58:24] <elektronkind> PerterB: don't you dare mention Sound Diver! It's perpetual beta state for OSX doesn't count! ;)
[17:58:30] <Somethingelse> lopa_: i recommend at least 6 gigs...for base install
[17:58:30] <nbkk6fo_> unless you need sparc no one bought them
[17:58:32] <delewis> nbkk6fo_: how did the SB1000 fail you?
[17:58:34] <nbkk6fo_> they were a huge disaster to sun
[17:58:48] <lopa_> is sendmail ON on default on solaris btw?
[17:58:49] <PerterB> elektronkind: fair point, the last time I had a working copy of it was on Windows :)
[17:58:51] <nbkk6fo_> delewis, overall performance
[17:58:54] <delewis> nbkk6fo_: sorry, but you really don't sound like someone that's old enough to have used UNIX for "11 years"
[17:58:55] <Auralis> lopa_: yes
[17:59:00] <Somethingelse> 8 gigs if you plan to turn it into a XAMP system
[17:59:03] <delewis> you're making a load of un-qualified statements.
[17:59:19] <nbkk6fo_> FCAL disks were nice but in the end it was cheaper to get a JBOD and raid across ATA than pay $1k each for 36gig FCAL drives
[17:59:20] <lopa_> Auralis: how can i check that please?
[17:59:31] <delewis> nbkk6fo_: FC-AL is cheap as dirt, nowadays
[17:59:35] <delewis> for 1Gb/s drives, anyway
[17:59:43] <delewis> (which are faster than U160 and thus faster than ATA shit)
[17:59:44] <Auralis> lopa_: what solaris version?
[17:59:47] <Somethingelse> lopa_: ps -ef | grep sendmail
[17:59:48] <nbkk6fo_> delewis, not when I owned a sb
[17:59:54] <Somethingelse> if you get a result, it's on
[17:59:57] <lopa_> Auralis: solaris 10
[18:00:04] <delewis> nbkk6fo_: then how on earth can you say they're terrible?
[18:00:13] <Auralis> svcs -a |grep -i sendmail
[18:00:16] <delewis> they're very well-constructed workstations, and then they arrived on the market they blew all the x86 offerings out of the water
[18:00:17] <LeftWing> ATA is the most disgusting disk interface.
[18:00:19] <Somethingelse> that works too :)
[18:00:42] <elektronkind> sendmail is on by default but think it's only the submit daemon that listens on loopback
[18:00:51] <delewis> plus, if you could afford a Sun Bladde 1000 new in 2000
[18:00:55] <nbkk6fo_> delewis, when I use to pay about $15k for sb1k's and later 2k's
[18:01:00] <delewis> I don't think you'd be griping about the cost of FC-AL drives
[18:01:05] <nbkk6fo_> I was about to get 4x the perf. on HP workstations
[18:01:06] <nbkk6fo_> for my developers
[18:01:15] <delewis> nbkk6fo_: so why the fuck are you griping about the cost of FC-AL drives?
[18:01:35] <nbkk6fo_> I mentioned an area which sucked sb io subsystem and cpu
[18:01:48] <nbkk6fo_> do you remember the original CPU's that shipped with the first SBx000 ?
[18:01:55] <delewis> IO? the Sun Blade 1000 has better IO that most x86 systems, nowadays.
[18:02:04] <delewis> nbkk6fo_: 600MHz UltraSPARC-IIIs that were non-CU
[18:02:17] <elektronkind> PerterB: yeah, I'm very miffed about eMapple dropping Sound Diver :/
[18:02:22] <delewis> (I've got a pair of 750s in my SB1000)
[18:02:31] <PerterB> likewise
[18:02:41] * delewis wishes he could upgrade to the 900MHz CU variants :-)
[18:02:43] <PerterB> I blame Apple
[18:03:05] <Somethingelse> so what do you guys think about Microcrap and Novell partnership?
[18:03:18] <LeftWing> Indifferent.
[18:03:22] <delewis> Somethingelse: irrelevant. that market does not concern me.
[18:03:23] <Somethingelse> is opensource done as we know it?
[18:03:30] <LeftWing> It certainly isn't.
[18:04:19] <PerterB> if Microsoft feel they own patents that Linux or other FOSS are infringing, it could get interesting
[18:04:29] <LeftWing> They're pushing money around in the name of PR and under the impression that each will reap bigger rewards than the other.
[18:04:33] <nbkk6fo_> delewis, if the system was such a hit as you *claim* why did sun kill it after 2 years? :-)
[18:04:39] <nbkk6fo_> dude the system was a disaster
[18:04:59] <nbkk6fo_> its one of the reason why everyone on wall st. had mass exodus to Linux on HP for their devs.
[18:05:02] <delewis> nbkk6fo_: that's the usual life-time of a Sun workstation.
[18:05:25] <nbkk6fo_> delewis, they killed the whole line
[18:05:26] <delewis> the Ultra 2 and previous Sun workstation line (the ones before the Ultra 25 and Ultra 45) lasted that long, as well.
[18:05:37] <delewis> nbkk6fo_: no, there's still the SB150.
[18:05:41] <delewis> (I think)
[18:05:45] <delewis> actually, no that's probably dead.
[18:05:54] <delewis> regardless, that's the lifetime of a Sun workstation line.
[18:05:55] <nbkk6fo_> delewis, look at it in terms of CPU ie. the SB used their flag ship CPU at the time
[18:06:06] <nbkk6fo_> today SUn's workstations are all lowend chips and x86 chips
[18:06:07] <delewis> nbkk6fo_: and again, it still blew most x86 offerings out of the water at the time.
[18:06:18] <nbkk6fo_> the highend SBline was killed believe it or not
[18:06:34] <delewis> high-end?
[18:06:35] <nbkk6fo_> delewis, i'm pretty sure Sun lost millions but what you say
[18:06:42] <delewis> nbkk6fo_: doubtful.
[18:06:49] <delewis> as I said, that's the lifetime of a Sun workstation product line.
[18:06:55] <nbkk6fo_> delewis, in terms of chip. ie sun now only sells sun blades on cheap IIIi chips or x86
[18:06:58] <delewis> you can go to sunsolve and check the introduction and EOL dates.
[18:07:16] <delewis> Ultra 2 was introduced in 1997 and lasted until about 1999 or 2000.
[18:07:27] <nbkk6fo_> ultra 2 was a decent system
[18:07:30] <delewis> and I don't think anyone with half-a-brain would dispute that the Ultra 2 was a decent workstation.
[18:07:40] <nbkk6fo_> we were discussing Sun Blade x000
[18:07:43] <delewis> nbkk6fo_: then by your metric, it must not've been if it only lated ~ 2 years.
[18:07:44] <nbkk6fo_> which I stated sucked
[18:07:47] <PerterB> I do... it was a _lovely_ workstation :)
[18:08:00] <delewis> nbkk6fo_: and like most people capable of arguing, I'm applying your same argument to other areas to make it sound silly.
[18:08:03] <delewis> and indeed, it is.
[18:08:22] <nbkk6fo_> delewis, i've only been discussing the Sun Blade x000 which was Sun's highend workstations
[18:08:26] <nbkk6fo_> which lasted 2 years
[18:08:35] <delewis> nbkk6fo_: the Ultra 2 was fairly high-end when it was introduced
[18:08:41] <nbkk6fo_> if it was such a big success Sun would have continued to put it out
[18:08:46] <nbkk6fo_> OMFG dude
[18:08:51] <nbkk6fo_> THE SUNBLADE 1000 and 2000
[18:09:03] * delewis sigh
[18:09:04] <delewis> this is going nowhere.
[18:09:31] <nbkk6fo_> you keep bringing up non relevent issues, yes the ultra 2 was a phenominal product,
[18:09:44] <delewis> nbkk6fo_: alright, let me lay it for you.
[18:09:49] <nbkk6fo_> The SBx000 pretty much destoryed SUn in the highend workstation space
[18:09:53] <delewis> you say the Sun Blade 1000 must've "sucked" because it only lasted 2 years.
[18:09:56] <Somethingelse> k, i'm gonna go for lunch now.  be back in bit
[18:10:05] <delewis> I provided a counter-example, by mentioniong the Ultra 2 which lasted a similar time
[18:10:11] <delewis> do you see the contradiction?
[18:10:24] <delewis> but alright.
[18:10:39] <nbkk6fo_> delewis, Ultra 2 was a little too late
[18:10:46] <nbkk6fo_> still a decent workstation
[18:10:58] <nbkk6fo_> personally I think better than the sbx000 but still not good enough
[18:11:02] <nbkk6fo_> which is why it was also axed
[18:11:23] <nbkk6fo_> i'm not trying to discredit sun here on their work but the SBx000 lined was horrid
[18:11:31] <delewis> so the Sun Blade 1000 design initially sucked, right? Why did Sun decide to create a new workstation, which was identical to it? a.k.a Sun Blade 2000.
[18:11:48] <nbkk6fo_> Marketing.
[18:11:57] <nbkk6fo_> "changing the name drop a faster chip lets see if anyone takes notice"
[18:11:58] <delewis> if the Blade 1000 line *truly* sucked, as you say, then I find it unlikely that Sun would've wasted their time re-marketing it under the name Sun Blade 2000 (with higher spec'd processors)
[18:12:17] * delewis puts a sign in front of nbkk6fo_
[18:12:21] <delewis> there, I give up.
[18:13:49] <Auralis> the SBx000 where great machines, still are, the problem is that x86 realy overtook sparc shortly after the us3 chips where aviable
[18:13:53] <nbkk6fo_> stop being narrow minded Sun isnt the end all, be all. How do you plan to have a non bias opinion if you never venture to the other side of the grass
[18:14:13] <delewis> nbkk6fo_: actually, I migrated from x86_32 to SPARC several years ago.
[18:14:19] <Auralis> at the time of release of the sb1k it did spanked every other workstation on the marked
[18:14:24] <delewis> (and I'm in the process of ridding myself of all x86 hardware at the moment)
[18:14:31] <delewis> that I own, anyway.
[18:14:36] <delewis> so I'm far from non-biased.
[18:14:53] <delewis> the first computer I used was x86-based, and throughout most of my lifetime, I've used x86 systems.
[18:14:59] <delewis> (which is unfortunate, but true)
[18:15:22] <delewis> s/non-//
[18:15:35] <nbkk6fo_> delewis, would you rather a system running say Core 2 Duo or a IIIi ?
[18:15:55] <delewis> nbkk6fo_: I'd rather have an UltraSPARC-III (preferably CU)
[18:15:58] <delewis> not an 'i'
[18:16:03] <delewis> (which Blade 1000s did not use)
[18:16:16] <nbkk6fo_> HAHA a IIICu ? over a Core 2 Duo ?
[18:16:51] <delewis> nbkk6fo_: absolutely. I'd rather have an elegant RISC design, than 30 years of consumer-minded cruft.
[18:17:40] <Auralis> i would take a quad socket opteron or a niagara2 box please
[18:17:43] <nbkk6fo_> so performance die size, heat hold no bearings ?
[18:17:48] <nbkk6fo_> my god man
[18:18:07] <delewis> Auralis: traitor :-)
[18:18:16] <delewis> (for mentioning the Opteron, that is) :-)
[18:18:17] <LeftWing> die size by itself is a non-event.  Performance and Heat-dissipation are at least quantities of merit.
[18:18:35] <delewis> and irrelevant for the desktop/workstation.
[18:18:42] <nbkk6fo_> LeftWing, small die you can easily compat more transistors and get more perf. theoritically
[18:18:48] <nbkk6fo_> so yes die size makes a huge difference
[18:18:59] <nbkk6fo_> add an i/o controller, large l2 cache etc..
[18:19:00] <LeftWing> nbkk6fo_: Yes but it's a means to an end -- again die size *by itself* is a non-event.
[18:19:00] *** rhys has left #opensolaris
[18:19:07] <delewis> nbkk6fo_: that's the "Intel" way of doing it.
[18:19:10] <delewis> large L2 cache? hah.
[18:19:34] * delewis wonders what nbkk6fo_'s definition of "large" is
[18:19:47] <LeftWing> Probably at least a megabyte!
[18:19:55] <delewis> LeftWing: 512K, don't strain the mind.
[18:20:02] <LeftWing> Bloody CISC.
[18:20:11] <delewis> well, not really
[18:20:16] <delewis> there's CISC done right, like the VAX
[18:20:20] <LeftWing> heh
[18:20:23] <delewis> and there's CISC done wrong, with a consumer in mind.
[18:20:25] <delewis> like x86
[18:20:26] <LeftWing> or CISCish on RISCesque. =P
[18:20:28] <nbkk6fo_> delewis, time to head back to assembler 101
[18:20:41] <nbkk6fo_> x86 is more like RISC these days than CISC
[18:20:54] <LeftWing> ... that was his point?
[18:21:05] <delewis> nbkk6fo_: to the assembly programmer, no it is not.
[18:21:24] <delewis> (which makes your reference to 'assembler 101' worthless)
[18:21:38] <LeftWing> x86 assembly is still an explosion of instructions.
[18:21:56] <elektronkind> geez, we're arguing about cpu arch now?
[18:21:57] <delewis> hey, they at least dropped hardware segment management in AMD64..
[18:21:59] <delewis> that's improvement.
[18:22:21] <LeftWing> elektronkind: Arguing implies that either side will ever back down. =P
[18:22:26] <delewis> elektronkind: yes, we've been all over the board.
[18:22:29] <nbkk6fo_> x86 is unbeatable
[18:22:32] <LeftWing> haha
[18:22:33] <nbkk6fo_> not even intel can kill it :-)
[18:22:35] * delewis chokes
[18:23:12] <LeftWing> It's a foolishly prolonged design that will have to die one day, in order for innovation continue.  It just might take a while.
[18:23:22] <nbkk6fo_> i doubt it will ever die.
[18:23:29] <delewis> it's a wonder I propiertary architecture like x86 has made it this far.
[18:23:31] <nbkk6fo_> x86 has killed so many along the way
[18:23:33] <delewis> s/I/a/
[18:23:40] <delewis> nbkk6fo_: yes, and that's a shame.
[18:23:48] <andersmo> Hey, look at mainframes. They're not dead yet. Backwards compatibility is king for many customers...
[18:23:55] <Error_404> blame windows....
[18:24:05] <nbkk6fo_> no blame price
[18:24:08] <elektronkind> heh. we still run a pair of HP3000's.
[18:24:21] <elektronkind> long live MVS. or something like that.
[18:24:37] <nbkk6fo_> no one cares about efficency and elegance if they can use somehting else for a fraction of the cost
[18:24:41] <LeftWing> nbkk6fo_: The price is because of the volume of sales as much as anything.
[18:24:46] <nbkk6fo_> yes
[18:25:01] <LeftWing> So you can blame the volume of sales on something else.  The whole thing doesn't root at "cost".
[18:25:27] <Error_404> if MS came out tomorrow and said "we've decided to support SPARC", there'd be a huge order for ultra25's the day after
[18:25:45] <LeftWing> Error_404: I think it'd have to be more forced -- "We're moving to SPARC"
[18:25:52] <nbkk6fo_> yes by why would they do that?
[18:25:55] <nbkk6fo_> what's the incentive for them to
[18:26:19] <Error_404> nbkk6fo_: getting off a 30 year old train wreck of an architecture?
[18:26:22] <LeftWing> The point isn't that they won't, it's that they could -- it's an academic suggestion.
[18:26:30] <delewis> nbkk6fo_: that question is irrelevant. What has been said, though, proves why x86 has gained such a large market share.
[18:26:41] <nbkk6fo_> Error_404, x86 isnt that bad as most people make it seem
[18:26:48] <delewis> nbkk6fo_: yes, it is actually *that* bad.
[18:27:01] <delewis> AMD64 just chooses just to expose the user to the cruft by default.
[18:27:03] <LeftWing> These days, everybody uses Windows on x86 because everybody uses Windows on x86. ;P
[18:27:05] <nbkk6fo_> delewis, Sun sells it
[18:27:06] <delewis> not*
[18:27:28] <delewis> nbkk6fo_: companies need to survive, and Sun is doing a better job at AMD64 than anyone else at the moment.
[18:27:42] <delewis> plus, they're trying to capitalize on a market that needs performance at a low cost.
[18:27:49] <LeftWing> They picked a chip that was a better x86 than x86. ;)
[18:27:52] <delewis> they're still dumping money into SPARC R&D and production
[18:27:54] <delewis> namely with Niagra.
[18:28:35] <nbkk6fo_> Sun's pretty much told everyone to get off low end sparc and use x64 and go with sparc for the treaded and large db apps
[18:28:38] <delewis> personally, I'd rather have an elegant design, than performance to the n-th degree.
[18:28:38] <nbkk6fo_> which makes sense
[18:29:21] <LeftWing> Some people still prefer the more solid design of SPARC systems.
[18:29:29] <delewis> not to mention OpenBoot, period.
[18:30:05] <LeftWing> If you want to build a compute cluster, sure go and buy a stack of small Opteron nodes.
[18:30:12] <nbkk6fo_> well in my field of work its all about speed and time to market
[18:30:12] <sniffy> delewis, Niagra is a very pretty and very useful product.
[18:30:24] <nbkk6fo_> we will use what ever is the best in those reguards
[18:30:25] <delewis> sniffy: I never disputed that.
[18:30:34] <nbkk6fo_> which is why we have such a large linux/x86 pressense today
[18:30:37] <sniffy> delewis, Well I never said you did :)
[18:30:48] <LeftWing> nbkk6fo_: You should look into Dell products.
[18:30:56] <nbkk6fo_> we use HP :-P
[18:31:05] <LeftWing> Yes but why not use Dell?
[18:31:10] <delewis> there's nothing like dumping a bunch of half-ass solutions onto your clients.
[18:31:15] <nbkk6fo_> better pricing and better hw from HP
[18:31:40] *** mega has joined #opensolaris
[18:31:45] *** gm152 has joined #opensolaris
[18:31:58] <delewis> 1. Linux 2. x86 3. profit
[18:32:36] <nbkk6fo_> we've been deploying more and more T2000's really nice systems
[18:32:58] <LeftWing> Are you running Solaris on them?
[18:33:01] *** mega has quit IRC
[18:33:04] * delewis would still like to have a Niagra2 with real VIS
[18:33:05] <delewis> :-(
[18:33:11] <nbkk6fo_> yes ofcourse
[18:33:16] <delewis> actually, I'm not sure the Niagra2 will have VIS.
[18:33:20] <delewis> well, hardware VIS, anyway.
[18:33:34] <LeftWing> delewis: Niagra3 perhaps? =P
[18:33:35] <nbkk6fo_> VIS ? hypervisor ?
[18:33:52] <delewis> nbkk6fo_: no, Visual Instruction Set.
[18:33:55] <nbkk6fo_> ahh
[18:33:56] <delewis> it's an SIMD instruction set.
[18:33:59] <nbkk6fo_> yeah
[18:34:00] <delewis> similar to MMX for x86.
[18:34:23] <nbkk6fo_> I really like Sun's roadmap for the future Victoria Falls is very appealing
[18:34:42] <nbkk6fo_> 2 way box with 8 drives, 128 threads 32gb of mem
[18:34:56] <nbkk6fo_> err 2u
[18:35:10] <Error_404> speaking of throwing cheap crap in to a rack because you're too broke to afford anything better, the tyan transport gt20's got an nforce4 ultra board in it... how likely is it to be well supported?
[18:36:03] <LeftWing> Speaking of ... questions... does Xsun allow the use of kerberos as an Xsecurity source?
[18:36:17] <LeftWing>  xhost +krb:<user>@<realm>   etc
[18:36:33] <Error_404> speaking of questions, eh?
[18:36:39] <LeftWing> =)
[18:36:45] <LeftWing> It was worth a try.
[18:41:12] *** schily has joined #opensolaris
[18:45:55] <Error_404> to answer your question, i have no idea
[18:47:09] <LeftWing> I'm pretty sure it isn't supported, which is a shame.  Kerberos is great.
[18:47:10] *** lopa_ has quit IRC
[18:52:03] <Error_404> I don't suppose i can throw in an SX:CR cd & do an upgrade if i've previously bfu'ed?
[18:52:32] *** logic_ has quit IRC
[18:54:33] *** nbkk6fo__ has joined #opensolaris
[18:54:33] *** nbkk6fo_ has quit IRC
[19:01:01] <Error_404> failing that, can i migrate zones to a fresh install given that they're installed on to a ZFS volume?
[19:03:43] <nbkk6fo__> depends on how you migrate it remember their are files fomr /etc that need to be copied over all
[19:05:04] <Saltsa> how is device remove from zfs pool working?
[19:06:10] <Error_404> um... working fine
[19:06:20] <Error_404> oh, you're not referring to my drives
[19:07:34] *** nwf has quit IRC
[19:24:11] *** yarihm has joined #OpenSolaris
[19:28:31] *** Cybernd has joined #opensolaris
[19:39:02] *** vmhobbes has quit IRC
[19:40:59] *** bougie_ has quit IRC
[19:44:12] <pixie_> Error_404 tyan gt20 2865?
[19:44:57] <pixie_> prettymuch the same board as the original x2100 and the ultra20
[19:45:00] <pixie_> works fine
[19:45:09] <pixie_> no sataframework support yet though
[19:48:21] <myrkraverk> is it possible to talk to tip with kermit?
[19:48:44] *** bougie has joined #opensolaris
[19:49:49] <asyd> S 17
[19:49:51] <asyd> oups
[19:53:27] *** anthony79 has joined #opensolaris
[20:02:57] *** kleppari has joined #opensolaris
[20:13:27] *** dclarke has joined #opensolaris
[20:15:48] *** bougie has quit IRC
[20:16:46] *** bougie has joined #opensolaris
[20:21:51] *** dakrone has joined #opensolaris
[20:25:55] *** paul has quit IRC
[20:26:47] *** nwf has joined #opensolaris
[20:39:11] *** lorka has joined #opensolaris
[21:00:04] *** Ezechial_ has quit IRC
[21:05:30] *** Ezechial_ has joined #opensolaris
[21:22:22] *** dakrone has quit IRC
[21:23:19] *** anthony79 has quit IRC
[21:23:52] *** deather__ is now known as deather
[21:24:52] <djgregor> before I get way into this, I
[21:25:14] <djgregor> before I get way into this, I'll ask here: anyone done (or know of) any bash command-line programmable completion scripts for ZFS?
[21:25:37] <djgregor> I.e.: so one can do this: zfs list <tab> and get a list of pools/filesystems/etc.?
[21:26:00] <sickness> heh interesting question... seems like the grub shell feature
[21:26:41] <dunc> bash completion rocks
[21:27:07] <delewis> hey, ksh88 had tab completion before bash did ;-)
[21:27:22] *** lorka has quit IRC
[21:27:23] <dunc> djgregor: i don't know of any sorry, my main experience of it is in gentoo, where many packages allow you to USE it
[21:27:31] <djgregor> I have it basically working for listing pools, filesystems, and snapshots, but it could stand to be more smart (i.e.: only showing one level at a time at the filesystem level)
[21:27:39] <dunc> delewis: I don't doubt it, I've just never strayed from bash really
[21:27:58] <dunc> cvs is one of my favourites
[21:28:19] <dunc> rm file ; cvs remove TAB  and it only gives me file as an option
[21:28:36] <delewis> djgregor: that method of tab completion wouldn't work as the shell wouldn't be ZFS aware
[21:28:43] <delewis> though, you could do something with the ZFS command
[21:29:01] <delewis> similar to how zonecfg works and use libtecla (which is what Solaris ships, instead of readline)
[21:29:10] <delewis> sounds like a valid RFE if it isn't already one.
[21:29:24] <dunc> RFE ?
[21:29:30] <delewis> request-for-enhancement
[21:29:37] <twincest> del: the idea of custom completions is that you make the shell zfs aware
[21:29:38] <dunc> oh right, cool
[21:29:40] <delewis> http://bugs.opensolaris.org
[21:29:55] <delewis> twincest: sounds like bloat that doesn't belong in the shell.
[21:30:04] <delewis> the zfs userland utils should have completion rather than the shell if you want that functionality.
[21:30:10] <twincest> well there isn't a 'zfscfg', where else should it go?
[21:30:11] <delewis> not to mention it's shell-dependent.
[21:30:18] <delewis> twincest: zfs or zpool
[21:30:43] <twincest> where in zfs or zpool is there a possibility of doing tab cmpletion?
[21:31:03] <delewis> twincest: whatever funciontality djgregor requested, obviously.
[21:31:39] <twincest> del: he wants "zfs list <tab>" to show a list of zfs pools.  how can the 'zfs' command do that?
[21:31:48] <twincest> s/pools/filesystems/
[21:32:13] <oxygene> by having a standard protocol to do it - unfortunately, there is none
[21:32:22] *** dunc has quit IRC
[21:32:57] <delewis> twincest: ex: you run zfs, and you get a prompt similar to zonecfg, then you can tab to get the various commands you want, like list, and then it inspects to see what list expects (which is a filesystem) and tab completes that.
[21:33:20] <oxygene> delewis: that's not the same
[21:33:57] *** yarihm has quit IRC
[21:34:11] <twincest> i don't want to have to type 'zfs\nlist /pool/\t\n^D' when i could just type "zfs list /pool/\t\n"
[21:36:38] <delewis> have gisburn implement it in ksh93.
[21:36:38] <Tpenta> well that's the only explanatin that I can think of that fits that crash anyway
[21:36:39] <delewis> :-)
[21:36:49] <delewis> he's feature-hungry.
[21:37:08] <Tpenta> oops wrong window
[21:39:52] *** lorka has joined #opensolaris
[21:40:25] <twincest> hm, how much disk space is required to build ON on x86, roughly?
[21:41:03] <dclarke> 3GB
[21:41:06] <dclarke> 4GB to be safe
[21:41:33] <dclarke> don't bother trying to build entirely in RAM either
[21:41:36] <dclarke> its CPU bound
[21:41:37] <delewis> SPARC is a bit worse I would imagine :-)
[21:42:00] <delewis> given it has to build 50 different kernels
[21:46:08] <delewis> my Ultra 2 is going to hate me for installing JES on it, I think.
[21:46:44] <dclarke> you put JES on an Ultra 2 ?
[21:46:46] <elektronkind> a u2 with 2x400mhz cpus will be able to take it in stride
[21:47:00] <dclarke> I have an 2 x 200 MHz unit here
[21:47:15] <dclarke> I think thats just out of the question
[21:48:21] <delewis> dclarke: 2x300MHz processors here with 1.5GB of memory
[21:48:46] <dclarke> hmmm ... drink a 6 pack of good beer and then give it a try
[21:49:00] <dclarke> that should numb the pain a bit
[21:49:21] <delewis> actually, I think it's handling it a bit better than my 2x866MHz PIII Xeon did with 1GB of memory.
[21:49:43] <delewis> which had one of those finnicky Adaptec SCSI controllers.
[21:49:59] <dclarke> Adaptec ?  finnicky ?
[21:50:02] <dclarke> really ...
[21:50:12] <delewis> dclarke: they're fairly cheap SCSI controllers.
[21:50:19] <delewis> I've always gotten terrible performance from them on Solaris
[21:50:26] <delewis> at least the U160 variants
[21:50:31] <delewis> I haven't tried the U320 models
[21:50:32] <dclarke> I think the old Adaptec AHA-1540CP was the bees knees
[21:50:51] <dclarke> and old Solaris on x86 worked great
[21:50:56] <delewis> my Blade 1000 with internal FC blows my PIII Xeon away when it comes to disk IO
[21:51:25] <dclarke> my Blasde 1000 here has a fried internal FCAL controller
[21:51:32] <dclarke> I had to use an external SCSI disk
[21:51:39] <delewis> and I would be willing to bet the Ultra 2 would blow it away, as well.
[21:51:44] <delewis> dclarke: ouch.
[21:52:36] * delewis is planning on having fun with ZFS when his A5200 /w 22x18GB drives arrives
[21:52:52] <dclarke> how many fibre ports ?
[21:53:03] <dclarke> you may not be really stunned with performane there
[21:53:19] <dclarke> I clocked 170MB/sec of actual IO on dual A5200's
[21:53:25] <dclarke> with SVM stripes
[21:53:30] <delewis> initially, two, but I'll upgrade to 4 at some point.
[21:53:46] <dclarke> going for mpxio ?
[21:53:48] <delewis> and I've got a Brocade Silkworm and a couple of 1Gb/s Emulex LP8000s coming in, too.
[21:53:59] <dclarke> hold it .. is the SAN Foundation kit included these days ?
[21:54:04] <delewis> dclarke: on the systems that have HBAs that are mpxio-capable
[21:54:10] <delewis> the socal HBAs on the E4500 aren't :-(
[21:54:19] <delewis> dclarke: yes
[21:54:20] *** nbkk6fo__ has quit IRC
[21:54:21] <dclarke> in that case .. yes .. you will have fun
[21:54:25] *** nbkk6fo__ has joined #opensolaris
[21:55:23] * delewis wonders why Sun bothered to use copper FC on the Blade 1000 for external
[21:55:35] <djgregor> delewis: I *have* "zfs list <tab>" giving me a list of ZFS pools
[21:55:42] <dclarke> it was a fad at the time
[21:55:47] <delewis> it's actually cheaper to buy a PCI HBA than to bother getting all of the adapters to convert copper to optical
[21:55:48] <dclarke> like Sparc notebooks
[21:55:49] <djgregor> in bash >= 2.0.4, you can extend the tab completion functionality
[21:56:03] <delewis> dclarke: SPARC notebooks would be a fad of mine if I could afford one. :-)
[21:56:12] <dclarke> yeah ..
[21:56:18] <djgregor> so I have a bash shell function that gets invoked when tab completion is attempted for arguments of the "zfs" command
[21:56:20] <dclarke> tadpole pretty much rocks in that department
[21:56:21] <twincest> nfs mount: mount: /onnv: Not owner
[21:56:24] <twincest> is this some NFSv4 thing?
[21:56:45] <dclarke> twincest:  now what have you done ?
[21:57:01] * dclarke looks at twincest suspiciously
[21:57:10] <twincest> i did 'mount /onnv' :) apparently vers=3 makes it work though
[21:57:15] * twincest doesn't understand nfsv4
[21:57:27] <dclarke> me neither
[21:57:31] <dclarke> but it works for me
[21:57:34] <dclarke> :-)
[21:57:36] <delewis> likely your domain is screwed up
[21:57:41] <twincest> well the server is linux, so who knows
[21:57:44] <delewis> ah
[21:57:49] <dclarke> ah ha !!!
[21:57:52] <delewis> in that case, it's definitely the fault of the Linux server.
[21:58:01] <delewis> (which barely has a working NFSv4 implementation)
[21:58:09] * dclarke throws Solaris Admin book at twincest
[21:58:30] <nbkk6fo__> netapp engineers very excellent linux based NAS appliances
[21:58:46] <delewis> they also have a set of in-house NFS patches
[21:58:47] <hali> netapps ar NOT linux based
[21:59:01] <nbkk6fo__> yeah they are
[21:59:08] <hali> they run ontap last time i checked
[21:59:11] <delewis> whatever the case is, they have their own NFS implementation.
[21:59:14] <nbkk6fo__> what do you think ontop is ?
[21:59:37] <hali> an internally developed operating system for storage devices
[21:59:58] <nbkk6fo__> its a linux kernel
[22:00:08] <nbkk6fo__> yes heavily modified but its still based on Linux
[22:00:39] <hali> i strongly doubt that... but i can't say i know for sure
[22:00:52] <hali> ontap was around way before linux became mainstream
[22:01:01] <nbkk6fo__> its no secret ask any netapp engineer
[22:01:35] * delewis pulls his personal netapp engineer out of his pocket
[22:02:11] * hali submits a now ticket
[22:02:20] <delewis> I just recall that Netapp attends the various NFS compatibility conferences whenever they're held.
[22:02:47] <delewis> which would imply they don't use the Linux implementation. :-)
[22:03:02] <nbkk6fo__> I would hope so they co-wrote the spec for nfsv
[22:03:03] <nbkk6fo__> 4
[22:05:21] <djgregor> "Although NetApp's OS looks nothing like BSD we certainly did use a good bit of code from the Net/2 and BSD-lite releases, back when that was being mostly funded by Berkeley and DARPA."  http://toasters.mathworks.com/toasters/15515.html
[22:08:16] <nbkk6fo__> how old is that because I heard that its heavily modifed linux from a netapp engineer
[22:08:25] <nbkk6fo__> not sure how crediable his word is but that was a few months back
[22:09:31] <richlowe> twincest: it's probably a domain name mismatch.
[22:09:32] *** jwtodd has quit IRC
[22:09:51] <twincest> even if they do run linux, their NFS implementation must be fairly different
[22:10:14] <Doc> yes - it actually works!
[22:10:22] <nbkk6fo__> never denied that :-) if anyone can do NFS it would have to be Sun and NetApp as they cowrote the spec
[22:10:32] <sickness> but there should be gpl patches for them somewhere
[22:10:41] <Doc> cowrote is a bit of a strong word
[22:10:57] <sickness> isn't it possible that they distribute a modified linux kernel but not the patches in source form?
[22:11:01] <twincest> rich: look, i actually checked out the source,  maybe i'll even do the sca :)
[22:11:02] <Doc> Sun wrote the spec, netapp was on the task force that turned it into a real spec
[22:11:07] <oxygene> sickness: nfs could be userland
[22:11:52] <sickness> oxygene: uhm, and you are telling me that an high performance embedded storage appliance os does nfs in userland? boh, I don't know...
[22:11:52] *** Ezechial_ has quit IRC
[22:12:24] *** Ezechial_ has joined #opensolaris
[22:12:28] <oxygene> sickness: see qnx, it can be done
[22:12:28] <delewis> if it's done right, it certainly could be.
[22:12:39] <sickness> if the source isn't around they are either using *bsd or some in home developed os, or violating the gpl, I suppose :P
[22:12:49] <sickness> uhm
[22:13:03] <jmcp> morning all
[22:13:08] <sickness> ok so netapps are all doing nfs in userland? sounds strange to me but yeah, impossible is nothing :)
[22:13:20] <hali> netapps don't really have userland
[22:13:22] <delewis> sickness: Mach had a virtual memory implementation (partially) in userland :-)
[22:13:23] <oxygene> sickness: just a theory
[22:13:36] <oxygene> delewis: mach also suck(ed|s)
[22:13:39] <sickness> =)
[22:13:42] <delewis> oxygene: :-)
[22:13:43] <djgregor> nbkk6fo__: it's from last year, and (assuming the email wasn't forged) from a founder and EVP of Netapp
[22:13:43] <hali> looking at that toaser post above they say they ported ping and ifconfig to "kernel code"
[22:14:08] <delewis> scary.
[22:14:17] <delewis> what processor do Netapps use, anyway?
[22:14:18] *** Ezechial_ has quit IRC
[22:14:29] <delewis> I would imagine something non-x86-based.
[22:14:39] <dclarke> a zilog Z80
[22:14:44] <sickness> dclarke: LOL
[22:14:44] <delewis> dclarke: :-)
[22:14:46] <dclarke> over clocked
[22:14:47] <hali> intels
[22:14:50] <oxygene> z80b?
[22:14:54] <hali> normal intel cpu's now... used to be alpha
[22:15:02] <delewis> hali: that must've changed recently, as they couldn't have been shipping that years ago and charging what they do. :-)
[22:15:05] <delewis> ah, thought so.
[22:15:12] <delewis> sort of Cisco did, I suppose
[22:15:27] <delewis> used M68ks and converted to Celerons, and then pawned those off for $50k :-)
[22:15:37] <sickness> "call them stupid" ;P
[22:15:49] <oxygene> istr that there are 6510 clones that run ops in 1cycle (ie. 2-4times as efficient), at 50mhz (ie. 12-50 times as fast as the original one)
[22:15:56] <oxygene> maybe something similar exists for z80
[22:16:21] <hali> i think our 840's costed something like $750k and had pIII procs :)
[22:16:32] <delewis> jeez.
[22:16:40] <delewis> at least IBM gives you a decent processor when you pay that much for the Shark hardware.
[22:16:57] <delewis> you basically have a little pSeries (or two if you go with the 800) in your cabinet.
[22:17:08] <sickness> oxygene: oh no that's bad, I just imagine the c64 games being too fast and unplayable :P
[22:17:14] <delewis> the lower-end DS-models come with Xeons, nowadays.
[22:17:24] <delewis> (dual in the DS4800)
[22:17:32] <hali> at least they're not i960's )
[22:17:47] <delewis> I've seen enough of those on fakeRAID controllers. :-)
[22:18:07] <Gman> hey hey hey
[22:18:09] <hali> remember this LSI Logic metastore system from a few years ago... probably $120k and the storage controller used a pair of i960s
[22:18:10] <delewis> the Sharks and DS4800s have separate "engines" for XORs, IIRC.
[22:18:31] <delewis> aside from the Xeons or POWERs
[22:19:05] <twincest> funny, closed-bins contains shell scripts
[22:19:08] *** jafari has quit IRC
[22:19:44] *** Ezechial_ has joined #opensolaris
[22:20:11] *** jafari has joined #opensolaris
[22:22:42] *** dunc has joined #opensolaris
[22:24:36] *** [tfb] has quit IRC
[22:29:20] *** udos has joined #opensolaris
[22:33:00] *** Ezechial_ has quit IRC
[22:36:23] <oxygene> sickness: imagine those c64 apps being faster by a factor of 200.. now compare this to "modern" software.. ;)
[22:37:00] <sickness> eheh :)
[22:37:01] <dclarke> I have tried in vain to play DESCENT II on Windows XP with a modern processor
[22:37:08] <dclarke> its just .. not working
[22:37:17] <dclarke> on a P90 and DOS it runs fine
[22:37:40] <dunc> heh
[22:37:54] * dunc bigs up C64 s
[22:38:03] <dunc> C64s
[22:38:13] <Somethingelse> i have descent 1 going fine on a p4 on SuSE
[22:38:27] <dclarke> with QEMU ?
[22:38:33] <dclarke> or with Bochs ?
[22:38:36] *** lorka has quit IRC
[22:38:37] <dclarke> or ?
[22:38:40] <dclarke> VMWare ?
[22:38:48] <Somethingelse> suse 8...vanilla install
[22:38:55] <oxygene> dclarke: the source is available
[22:39:08] <oxygene> oh, and dosbox should be able to run d1, too
[22:39:14] <dclarke> the source to Descent ?
[22:39:19] <Somethingelse> yea man
[22:39:20] <dclarke> oh .. decsent I
[22:39:22] <Somethingelse> it's out there
[22:39:37] <dclarke> I want the video controlled missiles in Descent II
[22:40:06] <sickness> dosbox is also good for games sometimes
[22:40:10] <oxygene> I think d2 sourcewas opened, to
[22:40:16] <oxygene> +o
[22:40:39] <Somethingelse> http://www.descent2.com/ddn/#soruces
[22:40:45] <Somethingelse> go down to source codes
[22:40:46] <delewis> oxygene: do you need media with all of the maps/textures?
[22:41:01] <oxygene> delewis: I think so
[22:42:33] *** nettie has joined #opensolaris
[22:42:44] *** nettie has left #opensolaris
[22:44:02] <twincest> hmm, all these -DOPTERON_ERRATUM_## when building seem a bit hackish
[22:48:39] *** razrX has quit IRC
[22:53:44] *** razrX has joined #opensolaris
[22:57:31] <dclarke> once upon a time .. I knew of a way to get a list of all the memory that the kernel was using
[22:57:48] <dclarke> my machine is here draging along with all avail RAM chewed up because of ZFS .. I think
[22:57:48] *** logic has joined #opensolaris
[22:57:53] <dclarke> but I am not sure
[22:59:43] *** Somethingelse has quit IRC
[23:00:01] <dclarke> now its time for a really really off the wall question
[23:00:04] <jmcp> dclarke: echo "::kmastat"|mdb -k
[23:00:10] <dclarke> ah !
[23:00:13] <dclarke> thank you ..
[23:00:16] <jmcp> you're welcome
[23:00:23] <dclarke> let me check and see where all my RAM went
[23:00:24] <jmcp> so where's that wall you want your question to fall off?
[23:01:21] <dclarke> well .. you recall th dtgreet graphic in Solaris 8 and 9 ?
[23:01:27] <jmcp> vaguely
[23:01:34] <jmcp> I've pretty much tried to block that out
[23:01:48] *** piwi has joined #opensolaris
[23:02:31] <dclarke> http://www.blastwave.org/dclarke/RAM_check.log
[23:03:10] <sickness> zfs_znode_cache is 6279168 I think it's not much :)
[23:03:10] <dclarke> ZFS is just killing this machine
[23:04:00] <dclarke> $ swap -s
[23:04:01] <dclarke> total: 208576k bytes allocated + 6840k reserved = 215416k used, 1414856k available
[23:04:06] <dclarke> plenty of swap
[23:04:09] <dclarke> or so it says
[23:04:10] <twincest> sickness: the zio_buf_* look quite large though
[23:04:18] <twincest> iirc that's where zfs eats memory
[23:05:20] <sickness> twincest: lemme see (I was looking at my machine not dclarke's :)
[23:05:46] <dclarke> hit reload on http://www.blastwave.org/dclarke/RAM_check.log
[23:05:55] <sickness> omg it's pretty big here too!!!
[23:05:55] <dclarke> at the top there is my zfs list etc ..
[23:05:59] <dclarke> snv_46 box
[23:06:17] <dclarke> sickness : ZFS should come with a warning
[23:06:34] <dclarke> You will need to be this tall to use this ride.
[23:06:39] <dclarke> that sort of thing
[23:08:33] <sickness> I'm on a snv_50
[23:08:34] <jmcp> dclarke: what are the specs on this on E-class box that you're running it on?
[23:08:51] <dclarke> E-class ?
[23:08:55] <dclarke> whats E class ?
[23:09:12] *** ShadowHntr has joined #opensolaris
[23:09:19] <jmcp> well I see sbus* and fas* in the cache listing....
[23:09:24] <sickness> http://www.rafb.net/paste/results/4zyhHb95.html <- what scares me a bit is the last line :/
[23:09:25] <dclarke> oh ..
[23:09:33] <dclarke> well its an old Ultra 2
[23:09:37] <jmcp> near enough :)
[23:09:40] <dclarke> not Enterprise class at all
[23:09:58] <jmcp> yeah
[23:10:01] * jmcp shrugs
[23:10:12] <sickness> but the box is running fine
[23:10:12] <dclarke> zio_buf_131072            131072   2904   2950 386662400   1140155     0
[23:10:23] <dclarke> try throwing some load at it
[23:10:52] <dclarke> go to http://www.blastwave.org/dclarke/crucible/
[23:10:59] <dclarke> download the binary at the bottom
[23:11:14] <dclarke> run that with it pointed at some ZFS based filesystem
[23:11:18] <dclarke> then check those stats
[23:11:24] <dclarke> while its running
[23:12:05] <sickness> heh
[23:12:05] <sickness> lemme see...
[23:12:32] <dclarke> also .. I nore that the primary buffers seem to be zio_buf_K where K= 512, 16K, 64K and 128K
[23:13:56] <sickness> is it ok if my test box is an amd64 with only 512mb ram?
[23:14:11] <delewis> 512MB of memory seems a bit tiny for an AMD64 system :-)
[23:14:29] <dclarke> thats just .. wrong
[23:14:46] <dclarke> I have 512MB of RAM on my five year old Pentium II 400MHz box
[23:15:03] <dclarke> with Solaris 10 it is just running ..
[23:15:16] * delewis has 640MB of memory in his IBM 7025-F40 RS/6000 from 1995..
[23:15:30] <delewis> with a 166MHz PowerPC 604e :-)
[23:15:31] <darkcmd> solaris requires 512mb ram?!
[23:15:36] <dclarke> oooh ... thats big big bucks back then
[23:15:37] <twincest> dark: 256
[23:15:47] <dclarke> baloney
[23:15:48] <darkcmd> hmm, can you get away with less, with no GUI?
[23:15:49] <dclarke> 512M
[23:16:02] <dclarke> don't even bother unless you have 512M
[23:16:05] <delewis> darkcmd: I suppose
[23:16:06] <dclarke> dead minimal
[23:16:07] <twincest> dark: 256 is required to load the installer ramdisk
[23:16:12] <delewis> but 512MB is really the minimum.
[23:16:15] <delewis> and 1GB is usable
[23:16:18] <delewis> anything above is nice ;-)
[23:16:24] <dclarke> 1GB is plenty
[23:16:27] <delewis> personally, I prefer 1.5GB to 2GB for a workstation.
[23:16:30] <dclarke> 4GB is real real nice
[23:16:38] <darkcmd> why does it require 256 to run the installer
[23:16:44] <dclarke> 32GB and you are swimming and can perform super human tricks
[23:16:47] <delewis> darkcmd: the initial ramdisk for one
[23:16:53] <delewis> and the installer is written in Java.
[23:16:58] <sickness> eheh
[23:16:58] <sickness> ok but why spend more it was running so fine doing only filesharing :)
[23:16:58] <sickness> isn't over spec still a bad thing nowadays? :)
[23:16:58] <sickness> in the past it was...
[23:16:59] <sickness> I have no gui
[23:17:00] <twincest> hmm, my vmware system is a bit confused about timekeeping
[23:17:00] <dclarke> 64GB and you bend the laws of physics and can move faster than light
[23:17:01] <sickness> and no programs running on it, but nfs and samba
[23:17:06] <twincest> 'vmstat 1' is showing about 3 lines per second
[23:17:20] <sickness> dclarke: should I dare to try your test anyway? =)
[23:17:25] <darkcmd> most operating systems I've ran don't require that much?
[23:17:26] <dclarke> twincest : well known bug with VMWare
[23:17:32] <delewis> darkcmd: well, Solaris does
[23:17:33] <sickness> dclarke: is it better on a non compressed fs or on a compressed one? I read the warning but it doesn't tell what's a better choice
[23:17:34] <twincest> any workaround?
[23:17:37] <delewis> Solaris 8 and 9 were pretty thin
[23:17:38] <dclarke> sickness : how much disk space do you have ?
[23:17:45] *** bougie has quit IRC
[23:17:49] <darkcmd> dclarke, how much could I get away with using an alternative installer to install it
[23:17:50] <delewis> but with the introduction of Grub in Solaris 10u1 which needs an initial ramdisk
[23:17:55] <delewis> you really need 256MB of memory or more
[23:18:06] <delewis> and with ZFS, you *really* need 512MB (much more if you're on a 64-bit system)
[23:18:17] <darkcmd> do you have to use ZFS?
[23:18:21] <twincest> no
[23:18:40] <sickness> dclarke: 401gb free
[23:18:56] <delewis> darkcmd: Solaris is really meant to be usable and worthwhile, not to fit on thin hardware.
[23:19:03] <dclarke> sickness : yeah man .. let's run it
[23:19:14] <sickness> ok =)
[23:19:17] <dclarke> sickness : just run ./cruciple /path/foo/test
[23:19:19] <delewis> there's other operating systems that are more adept to that sort of thing.
[23:19:50] <darkcmd> hmm, I was planning on installing it on my Ultra 2
[23:19:57] <delewis> oh, you shouldn't have any problem then.
[23:20:02] <delewis> provided you've got some memory.
[23:20:07] <darkcmd> 512mb
[23:20:12] <delewis> I've got 1.5GB of memory in my Ultra 2 and Solaris flies on it.
[23:20:19] <delewis> darkcmd: you shouldn't have a problem.
[23:20:42] <twincest> ah, seems this may be related to cpu frequency scaling
[23:21:10] <sickness> dclarke: I'm running it :) zpool iostat shows mostly writes, as of now...
[23:21:26] <twincest> apparently solaris is confused when it's booted with an 800MHz CPU and the host scales it up to 2GHz later
[23:21:27] <dclarke> first pass it creates a ton of files
[23:21:43] <sickness> dclarke: crucible is at 36% cpu
[23:21:51] *** ShadowHntr has quit IRC
[23:22:11] <dclarke> sickness : are you using prstat ?
[23:22:14] <sickness> omg zio_buf_131072 gone down instead of up! now it's just 2490368
[23:22:29] <dclarke> sickness : okay .. thats just weird
[23:22:33] <sickness> dclarke: yeah
[23:22:40] <sickness> 29% is still too much?
[23:22:45] <dclarke> sickness : whats zio_buf_65536 look like
[23:22:59] <dclarke> sickness : prstat lies
[23:23:08] <dclarke> sickness : prstat or at least does not tell you the truth
[23:23:17] <dclarke> sickness : run prstat -m
[23:23:48] <sickness> zio_buf_65536              65536    984    991  64946176   3611126     0
[23:24:10] <dclarke> what does vmstat 2  say after five lines ?
[23:24:27] <dclarke> specifiacll vmstat 2 5
[23:24:43] <dclarke> specifically
[23:26:39] <sickness> 0 0 0 4143904 49808  0   0  0  0  0  0  0  0 148 140 151 1003 2951 1131 15 20 65
[23:26:39] <sickness> I don't really know how to read the output of prstat -m and vmstat, never used them =)
[23:26:39] <sickness> ok test1 is ended, now on test2 :)
[23:26:39] <sickness> 238328 files  avg=0.001569 sec  total=373.992328 sec  io_avg=39.828357 MB/sec
[23:27:06] <dclarke> 39.8 MB/sec is quite good
[23:27:25] <dclarke> what type of hardware ?  ATA ? SCSI ?
[23:27:52] <sickness> dirtiest cheap ass sata controller with sata disks :)
[23:28:07] <sickness> bought all at a local computer fair for like less than 400euro (box+disks)
[23:28:14] <sickness> I posted pics here some weeks ago, remember? :)
[23:28:22] <dclarke> er .. no
[23:28:30] <dclarke> but .. its working just great
[23:28:42] <sickness> yeah :)
[23:29:07] <quasi> sickness: what kind of box?
[23:29:08] <dclarke> sorry .. I have to run a GNOME test here
[23:29:11] <dclarke> be right back
[23:29:12] <sickness> I wanted the cheapest option available, just for the sake of it, so here comes the 512mb thing, 256 was not available :P
[23:29:19] <sickness> http://www.sickness.it/zfshomeserver.jpg
[23:29:20] <sickness> http://www.sickness.it/zfshomeserver.txt
[23:29:24] <sickness> http://www.sickness.it/zfshomeserver.png
[23:29:36] <twincest> you were going to build a ZFS fileserver with 256MB RAM?
[23:29:39] <sickness> quasi: it's a 64bit sempron 2600 in an asus vintage barebone
[23:29:46] <sickness> twincest: nope, I built with 512 in the end =)
[23:29:54] <quasi> sickness: ah
[23:30:06] <twincest> yeah, but you just said that was because you couldn't find 256
[23:30:22] <sickness> quasi: with a sil3114a pci sata1.0 controller and 3 sata2.0 320gb maxtor disks (root on 80gb pata disk)
[23:30:37] <sickness> twincest: yeah, if the price was lower, why not? I could try ;)
[23:30:49] <sickness> twincest: it's not a production box, only a home toy :)
[23:32:12] *** jafari has quit IRC
[23:42:48] <Error_404> hey dclarke
[23:43:33] <Error_404> what's this about genesi killing off their PPC workstations?
[23:44:11] <dclarke> yep
[23:44:16] <dclarke> you heard right
[23:44:32] <dclarke> the EFIKA 5K2 is their new gig
[23:44:38] <dclarke> the microscopic thingy
[23:45:28] <Error_404> the tiny slow one?
[23:46:06] <dclarke> er .. yeah
[23:46:07] <dclarke> that thing
[23:46:14] <dclarke> the 400MHz unit
[23:46:31] <Error_404> unfortunate
[23:46:36] <dclarke> its a specific solution for industrial applications I think
[23:46:42] <Error_404> now polaris needs to scavenge old macs, essentially?
[23:46:43] <dclarke> its real real small
[23:46:54] <dclarke> no .. big IBM gear !
[23:48:47] <dclarke> personally I want to see it run on big IBM gear
[23:49:05] <Error_404> me too, but IBM seems not to be too keen on the idea
[23:49:32] <delewis> heh, IBM would have a fit. :-)
[23:49:38] <delewis> losing AIX customers. ouch.
[23:49:45] <dclarke> you guys don't know the people that I know
[23:49:50] <dclarke> and they are keen for th idea
[23:49:51] <delewis> well, Linux/AIX customers, that is.
[23:50:03] <dclarke> it continues the viability of their hardware
[23:50:09] <sickness> well, I'd personally *LOVE* to see that kernel boot on the 800mhz ibook g4 12" that I have =)
[23:50:12] <sickness> ghgh
[23:50:21] <dclarke> Sun sells hardware that runs Linux and Windows and Solaris
[23:50:30] <dclarke> so .. IBM can run AIX and Solaris
[23:50:35] <dclarke> its a good thing
[23:51:00] <delewis> ugh. ssh is beginning to annoy me.
[23:51:09] <delewis> Nov  5 16:47:07 laplace.network.lan sshd[8466]: Failed publickey for dlewis from 10.0.0.7 port 65411 ssh2
[23:51:21] <delewis> I wonder if my home directory being shared between the two systems has anything to do with it.
[23:51:30] <delewis> (and thus, the .ssh directory being shared)
[23:52:28] <PerterB> yup... if ssh finds a private key in ~/.ssh/identity|id_rsa|id_dsa it will try to use it for public key authentication
[23:53:09] <delewis> right
[23:53:16] <Error_404> dclarke: so, we should continue the focus on PPC32 despite the fact that any PPC32 machines that'll be around to run it will be obsolete?
[23:53:24] <Error_404> somewhat like the VAX port of netbsd?
[23:53:35] <delewis> and the fact that my public keys were made using this particular system and not the other one that my home directory is shared with causes problems, right?
[23:53:50] <delewis> any slick workaround or is GSS SSO my only hope? :-)
[23:54:35] *** nbkk6fo_ has joined #opensolaris
[23:54:35] *** nbkk6fo__ has quit IRC
[23:54:44] <PerterB> no, it's more the fact that the public half of the key isn't in .ssh/authorized_keys
[23:54:51] <dclarke> Error_404:  until we have a shell and native tools we need to stay in PPC32 land
[23:54:53] <delewis> PerterB: ah!
[23:54:58] <delewis> well, I can fix that.
[23:55:07] <PerterB> yes, trivially :)
[23:56:24] <Error_404> fair enough I s'pose
[23:56:55] <Error_404> Is the ppc 970 boot sequence/memory controller/etc that much different?
[23:57:56] <dclarke> all the hardware specific layers will need a rewrite
[23:58:32] *** delewis has quit IRC
[23:59:00] *** piwi has quit IRC
[23:59:01] *** dakrone has joined #opensolaris





top