October 26, 2006 
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | 22 | 23 | 24 | 25 | 26 | 27 | 28 | 29 | 30 | 31
October 26, 2006 [00:00:37] <gdamore> (it does scare me that you had that URL so ready at hand. :-) [00:00:45] <myrkraverk> ;) [00:01:12] <alanc> I had been looking at it earlier today, so it was in my browser history [00:01:13] <myrkraverk> delewis: no luck at all - I'm starting to wonder if there is anything working at the other end of the usb cable :P [00:01:33] <gdamore> alanc: thanks, your advice wrt to dtlogin seems to work. now i just have to wait for it blank the screen... [00:01:45] <myrkraverk> delewis: aha! that was it (sorry for the ... bullshit I coused) [00:02:13] <delewis> myrkraverk: the 'devfsadm' worked? [00:02:29] <gdamore> alanc: btw, doesn't masa tend to hang out here, as well? [00:02:35] <myrkraverk> delewis: well - *something* worked ;P [00:02:46] <gdamore> it was probably devfsadm. [00:03:04] <alanc> does he? I hadn't noticed [00:03:06] <delewis> myrkraverk: what happened (and it's actually a similar situation with hotswap storage in Solaris) is you attached the USB drive after booting, thus the /dev entries weren't there, which means vold couldn't mount it. [00:03:19] <myrkraverk> delewis: ah, k [00:03:28] <twincest> shouldn't inserting a USB device cause the device entries to appear? [00:03:55] <delewis> not a mass storage disk device that's actually mounted. [00:04:02] <delewis> *that actually can be mounted [00:04:13] <myrkraverk> delewis: btw, can solaris somehow keep track of both of my usb hdds and mount them (automagically) at their designated places? [00:04:23] <delewis> myrkraverk: it will now [00:04:44] <delewis> another fix would've been to touch /reconfigure or append '-r' to your kernel boot arguments [00:04:49] <delewis> (both of these just run devfsadm on boot) [00:05:18] <myrkraverk> delewis: what I mean is, mount drive n at /place1 and drive m at /place2 - without me having to think about it? [00:05:29] <myrkraverk> (since I have 2 of them) [00:05:31] <gdamore> delewis: is this feature (automounting) something with the new HAL code? [00:05:46] <delewis> gdamore: supposedly, vold can do it, which I didn't know. [00:05:59] <gdamore> heh. vold has always been a crapshoot [00:06:11] <delewis> that's what the /rmdisk thing is for that vold(1) mentions. [00:06:16] <delewis> gdamore: indeed, which is why I was skeptical :-) [00:07:42] <delewis> myrkraverk: did vold that automagically mount it? [00:07:55] <delewis> er s/that// [00:08:12] *** axisys has quit IRC [00:08:25] <myrkraverk> delewis: something krept up that I think I'm using to format the thingy [00:08:29] <myrkraverk> (with pcfs) [00:08:30] <alanc> wow - you got the SPARC graphics senior manager to respond on an opensolaris list - just to pass the buck, but that's still impressive [00:08:38] <gdamore> wtf!?! xset says my monitor is turned on. [00:08:43] <gdamore> but it is blank. [00:08:51] <delewis> myrkraverk: yes, that means vold detected it [00:08:59] <myrkraverk> delewis: ok ;) [00:09:04] <delewis> if it's formatted with pcfs it should be mounted under /rmdisk/rmdisk0 for disk0 [00:09:08] <delewis> and /rmdisk/rmdisk1 for disk1 [00:09:09] * gdamore goes and squirrels away that e-mail address.... [00:09:15] <alanc> so Xsun thinks it told the ddx/driver to turn it on, but the driver fails to respond [00:09:16] <delewis> gdamore: :-) [00:09:25] <alanc> weird [00:09:49] <gdamore> no, this is right after dtgreet screen blanks [00:10:08] <gdamore> so it was on, then dtgreet blanked it. i ran xset -q _without touching the mouse_ [00:10:11] <myrkraverk> hmm, zfs is not part of b50 it seems - can I get it somewhere? [00:10:20] <delewis> myrkraverk: it is [00:10:23] <delewis> what makes you think otherwise? [00:10:31] <delewis> (ZFS was integrated in b38, IIRC) [00:10:47] <gdamore> so whatever dtgreet uses to blank the screen is _not_ DPMS [00:10:47] <myrkraverk> delewis: it wasn't listed as an option to format my hdd thingy :P [00:10:54] <delewis> myrkraverk: the installer does not support ZFS, yet [00:11:03] <delewis> you create your ZFS volumes after installation [00:11:15] <myrkraverk> delewis: ah, k [00:11:44] <myrkraverk> (that means I may want to re-install the thing) [00:12:14] <delewis> myrkraverk: right now, my only ZFS volume (on this workstation) is /export, which contains my home directory, zones, etc. -- basically, data volumes. [00:12:28] <alanc> we've been having weird issues in S10 with dtgreet not blanking the screen when it's supposed to, but I don't remember it not waking up when it's supposed to [00:12:29] <delewis> I use UFS for system filesystem. [00:12:53] <delewis> after installation, I create the /export zfs pool, create my user, etc. [00:12:57] <gdamore> aha. /usr/openwin/bin/xset s reset unblanks the screen [00:13:14] <alanc> oh, so it's the screen blanker, not the power management? [00:13:25] <gdamore> sure appears that way. [00:14:30] <myrkraverk> delewis: ok - I'll keep that in mind - just, on installation; how should I partition my disk? [00:15:13] <myrkraverk> delewis: do I leave most of it unused, or do I just leave some space for slices in the ufs partition (or am I misunderstanding the whole thing?) [00:15:40] <delewis> myrkraverk: well, my setup is fairly atypical for a workstation -- say, on my 36GB disk, I've got a 10GB slice for root, a 1GB swap, a blank slice (for Live Upgrade) that's also 10GB, and the rest goes to /export (which is my ZFS volume) [00:16:09] <richlowe> Hm, who was it asking about memory caps? [00:16:11] <delewis> if you don't want Live Upgrade, devote that space to root, and leave the rest for ZFS for your data volumes, like /export [00:16:14] *** jmcp_ has joined #opensolaris [00:16:16] <richlowe> now I notice 2006/598 [00:18:01] <myrkraverk> delewis: so, on installation, I only partition a 10 gb of the disk, or do I partition all of it, and leave a lot of empty space for slices? [00:18:14] <myrkraverk> (I think I left some empty place for slices) [00:18:19] <delewis> myrkraverk: you'll need to leave whatever empty that you want to use for ZFS [00:18:58] <myrkraverk> delewis: so I don't need an unpartition space? I can use "empty" slices? [00:19:05] <nrubsig> Does anyone know why http://pastebin.ca/221846 does not outout '-a-' as expected ? [00:19:28] <delewis> myrkraverk: you'll need to adjust the sizes of those empty slices after installation before you create the zfs pool [00:19:35] <delewis> that can be done with 'format' [00:19:57] <delewis> right now, just dedicate whatever space you want for root and swap and leave the rest empty [00:20:08] <myrkraverk> delewis: oh, k [00:20:39] <myrkraverk> . o O (lots and lots of stuff to learn) [00:20:46] <myrkraverk> well, I'm off to sleep; night all [00:21:58] <Tpenta> stevel: you around? [00:22:12] <richlowe> Tpenta, alanc: do you know of the reasoning behind Cc'ing the project alias on ARC stuff? [00:22:27] <richlowe> I'm trying to figure out if it's an artifact of the current open arc stuff being suboptimal (to say the least), or a general thing. [00:22:43] <Tpenta> which arc stuff in particular? [00:22:54] <richlowe> the majority of open cases with a related project or community thus far. [00:23:13] <Gman> it'll cc anything in the interest list, if that's what you mean [00:23:14] <Tpenta> in the case of a case, we include the project team so they can comment on discussions and aware of what is going on [00:23:15] <alanc> historically we've done it because most members of most project teams aren't on the ARC mailing lists, but should be part of the discussions for their project [00:23:28] <richlowe> I was thinking about this when replying to the 2006/569 discusson. [00:23:47] <richlowe> I'm not entirely sure whether that makes sense anymore. [00:24:05] <richlowe> since project team != project alias, and (ideally) subscription wouldn't be required to post to the arc list anymore. [00:24:14] <alanc> for instance, I'm the only engineer on the X team who gets ARC mail normally, but when I file an X case I cc the team list so the other X engineers can participate in that discussion [00:24:34] <richlowe> My thinking behind all this is that the same risk in the arc aliases being disconnected exists in this system. [00:24:35] *** Mazon is now known as mazon [00:24:41] <richlowe> where one accidentally dropped Cc fractures the conversation. [00:25:03] <richlowe> whereas keeping everything in one place, with just a heads-up to the project alias would keep everything together, and generally avoid some of the mailman snafus. [00:25:35] <alanc> ah - yes, we normally cc the alias for the engineers working on a project, not the general interest alias, which is what the opensolaris *-discuss aliases are closer to [00:25:41] <Tpenta> I think the big difference is that in general when we include a project team alias, we are talking a handful of engineers [00:25:52] <Tpenta> now, when we mail a project alias, we could be talking hundreds [00:26:03] <richlowe> Tpenta: whereas with both Roland's case, and the zones cases, it's Cc'd to fairly large lists. [00:26:08] <alanc> or for projects without a dedicated e-mail alias, we cc the project team members [00:26:12] <Tpenta> oh yes [00:28:30] <richlowe> alanc: is 3 of us jumping on John for the ON thing enough, d'ya think? :) [00:28:52] <Tpenta> :) [00:29:03] <alanc> yeah - but I'm the only one who ran into him in the cafeteria at lunch today 8-) [00:29:21] <alanc> he told me to keep up keeping him honest [00:29:24] <Tpenta> i kinda agree with what darren is saying, the opensolaris prefix is redundant, and i would prefer them being called psarc-* rather than arc-* [00:29:39] <Tpenta> i really gotta get over there to meet you folks [00:29:54] <nrubsig> which 32bit value has 'c' in UCS-4 ? [00:29:58] <alanc> then handed me a handful of lanyards with the old marketing slogin "Take it to the nth degree" on that he was cleaning out of some old storage hole [00:30:19] <_william_> gn [00:30:20] *** _william_ has left #opensolaris [00:30:27] <alanc> (I think that was the one after "Sun is the dot in dot com") [00:31:57] * delewis has both t-shirts [00:31:58] <icon> i still have my old 'the network is the computer' mousepad from the late 90's [00:32:07] <icon> i think they still have those dont they? [00:32:11] <delewis> I've got one of those mugs :-) [00:32:22] * nrubsig hates being ignored [00:32:31] * nrubsig wants to throw stones [00:32:43] <alanc> I'm still using my "UltraComputing" mouse pad from one of the original Ultras [00:32:46] <g4lt-U60> as opposed to "microsoft is the dot in ./exploit" ;P [00:33:16] <jengelh> http://img184.imageshack.us/img184/3685/bootunloadza1.png !? [00:33:36] <alanc> we're not ignoring you nrubsig, we're just not answering because we don't know the answer, and flooding the channel with everyone who doesn't know saying "I don't know" isn't useful [00:34:03] <delewis> no, we can just all /msg nrubsig "I Don't know" :-) [00:34:12] <g4lt-U60> wait, we're not ignoring nrubsig? when did THAT start? [00:34:23] <delewis> why waste channel bandwidth, when you just waste nrubsig's :-) [00:34:44] <alanc> if we ignore him, then we can't tease him [00:34:52] <g4lt-U60> says who? [00:35:11] <alanc> so we just tell ksh93 jokes without watching for his response? [00:35:28] <icon> hurm [00:35:30] <dwc-> ucs4? utf-32? wow [00:35:39] *** solarisjon has quit IRC [00:35:43] <icon> jengelh: how does sol perform for you in vmware? [00:35:57] *** triplah has quit IRC [00:36:08] <dwc-> if I had to guess, I'd bet utf-8 and 7-bit ascii probably overlap [00:36:22] <jengelh> icon: Works. [00:36:29] <icon> jengelh: have you tried parallels? [00:36:43] <jengelh> No. From what I have heard, its license sucks. [00:36:56] <icon> tempted to pick up a new macbook pro and throw sol on there instead of shelling into my opteron for dev work [00:37:09] <icon> core 2 duo's are pretty tasty chips [00:37:20] <jengelh> However my problem is that something unloads rpldev without hesitation [00:38:46] * icon & [00:43:48] <Tpenta> non-debug stuff for b51 is now up [00:44:40] <stevel> any jds guys here? [00:46:56] <Gman> no. [00:47:27] <Gman> those fuckers need to get back to work to make gnome faster [00:47:41] <alanc> and hog less memory [00:48:10] <stevel> lol [00:48:21] <stevel> gman: when's the next version of JDS (2.16 based?) due to hit snv? [00:48:25] <Tpenta> !seen kupfer [00:48:27] <Drone> kupfer (kupfer!i=kupfer@nat/sun/x-3d845599e2560142) was last seen in #opensolaris on Mon 16 Oct 2006 20:31 GMT, saying 'well, if they're both using the same v4 domain, I don't know what the problem is.'. [00:48:39] <Gman> stevel, about b53 is the hope [00:48:47] <stevel> cool [00:49:12] <Gman> there are packages available to install the latest and greatest if you don't want to wait [00:49:18] <Gman> seems reasonably stable to me [00:49:31] <stevel> what's great about it? or rather, what's compelling? [00:50:16] <Gman> newer firefox [00:50:19] <alanc> it should also have FireFox 2.0 - though that's not really a big change [00:50:34] <Gman> less of a memory hog, supposedly [00:50:48] <delewis> awesome :-) [00:50:51] <Gman> nicer default theme [nimbus instead of blueprint] [00:51:08] <delewis> isn't Nimbus already used (since b48)? [00:51:24] <Gman> stevel, you might even get a working sound juicer ;) [00:51:30] <Gman> delewis, possibly [00:51:31] <stevel> woot [00:51:32] <alanc> shipping any of the mono components? [00:51:37] <Gman> nope [00:51:42] <Tpenta> have they fixed the segv in nimbus yet? [00:51:53] <Gman> yes [00:51:59] <Tpenta> in which build? [00:52:00] <alanc> oh yeah, HAL support in GNOME [00:52:07] <alanc> thanks to tamarack finally being in ON [00:52:11] <Gman> Tpenta, mm, not sure [00:52:18] <Gman> definitely the latest [00:52:26] <Tpenta> oh well, I gues sI'll try it after installing 51 on the workstation [00:52:50] <stevel> what does HAL get me besides sound-juicer (possibly) working? [00:53:06] <Tpenta> better removable device support [00:53:09] <Gman> working removable media support [00:53:24] <stevel> hrm. what's wrong with the current removable media support? [00:53:27] <Tpenta> i was going to say "some would say working" ;) [00:53:29] <delewis> right now, it's broken if you've upgraded to > b51 and are using an older JDS. [00:53:36] <stevel> (i don't use much besides my usb key - which i haven't used recently) [00:53:50] *** MattMan has quit IRC [00:54:02] <Tpenta> i'll be interested to see what it makes of my external usb disk; with a fat32 and a zfs on it [00:56:18] <jengelh> what's the ccs in /usr/ccs/bin standing for? [00:57:03] <delewis> c compilation system [00:57:09] <delewis> you may find filesystem(5) helpful [00:57:58] <jengelh> interestingly misses /usr/ccs/bin/cc ;-) [01:00:52] *** lloy0076 has joined #opensolaris [01:02:20] <gdamore> yeah, but maybe someday we'll get a _bundled_ version of studio 11, and then we will have /usr/ccs/bin/cc. [01:02:46] <gdamore> in fact, all i want is basic C, C++ stuff necessary to build the OS. [01:03:18] <gdamore> all the IDE, editors, Fortan, and parallelization stuff can and probably should stay unbundled [01:03:55] <sp0rq> snoretran [01:04:20] *** sp0rq is now known as spawrq [01:04:25] * gdamore had to write lots of fortran code as a chemical engineering undergrad. the prof complained that all his code looked like "C". :-) [01:04:46] <alanc> much more likely is /usr/bin/cc - there's been discussion about moving much of /usr/ccs/bin to /usr/bin already so newbies stop asking "Why the hell does Solaris not include make?" [01:05:29] <gdamore> that would be nice. i always thought /usr/ccs/bin/ was a bit, ahem, strange [01:05:36] <delewis> and in most cases, the make in /usr/ccs/bin is useless to them [01:05:47] <delewis> because they're building applications that have Makefiles that require gmake [01:05:56] <gdamore> hahaha [01:06:43] <gdamore> i consider the fact that we have 3 major variants of make (Solaris, BSD, and GNU make) to be one of the great perversities of UNIX [01:07:10] <delewis> probably because POSIX 'make' "sucks" [01:07:21] <delewis> and various makes existed before the POSIX spec. [01:07:44] <gdamore> still, the variants seem to diverge further and further apart over time, rather than converging [01:07:57] <delewis> don't forget dmake :-) [01:08:17] <gdamore> dmake has Solaris make syntax. perhaps I should have said "make syntaces" [01:08:32] <delewis> then there's /usr/xpg4/bin/make which is POSIX make :-) [01:10:02] <nrubsig> delewis: no [01:10:32] <delewis> nrubsig: yes [01:10:48] <boyd> Can I go next? "maybe" [01:10:51] <nrubsig> delewis: /usr/xpg4/bin/make appears to be a sun make hacked to be closer to POSIX, including replacing the default shell (/bin/sh) with the sun posix shell (/usr/xpg4/bin/sh) [01:10:59] *** wesw has joined #opensolaris [01:11:08] <delewis> nrubsig: I figured that was the situation, but I wasn't sure. [01:11:13] <delewis> a lot of the /usr/xpg4 tools appear to be that way [01:11:48] <boyd> nrubsig: Are you saying it's *not* compliant? [01:12:00] <boyd> 'cos that would be a bug, surely [01:12:13] <delewis> boyd: I think it's not strictly POSIX-compliant. [01:12:54] <nrubsig> delewis: it's near posix with all sun extensions [01:13:04] <delewis> /usr/xpg4/bin/sh is a modified ksh88 (yes, I know they're are some slight variations), but POSIX specifies quite a bit less than what ksh88 offers. :-) [01:13:09] <boyd> posix compliant does not preclude extensions [01:13:13] <nrubsig> delewis: like the current usr/xog4/bin/sh is near posix but ksh88 based [01:13:25] <delewis> nrubsig: right [01:14:47] <boyd> The thing is, any implementation is going to be based on something... there's nothing wrong with being based on ksh88 if it's compliant [01:15:44] <delewis> boyd: absolutely, but it's wrong when you encourage users to use your less portable extensions over POSIX specs (as GNU sed does) [01:16:06] <delewis> Sun does a good of distinguishing between Sun behaviour and POSIX behaviour. [01:16:13] <boyd> Indeed... [01:16:27] <boyd> so my question is, is the zpg4 make compliant or not? [01:16:34] <boyd> if so, I don't see a problem [01:16:40] *** kimc has joined #opensolaris [01:17:13] <nrubsig> boyd: reality check: ksh93 is currently closer to POSIX than sun's hacked ksh88 [01:17:40] <gdamore> oh, gawd, not another ksh93 discussion.... [01:17:45] <delewis> "closer?" it's either compliant or it isn't. [01:18:07] <gdamore> does anyone know how to get dacf to notice a change in /etc/dacf.conf? [01:18:19] <boyd> delewis: Thanks, that was what I was trying to phrase [01:18:58] <boyd> gdamore: I've never even heard of it [01:18:59] <delewis> ;-) [01:19:45] <boyd> I don't see how the provenance of the code has anything to do with its compliance [01:19:48] <gdamore> dacf "replaces" (sort of) the /etc/iu.ap [01:20:01] <boyd> hehe.... never looked at that either :) [01:20:23] <nrubsig> delewis: uhm... well, /usr/xpg4/bin/sh has it's limitations... :-) [01:20:24] <gdamore> streams auto-plumbing. it less you configure a module to automatically be "pushed" when another module is opened [01:20:43] <nrubsig> delewis: this is why ksh93 will become /usr/xpg4/bin/sh quite fast. [01:20:46] <boyd> ... looks to be only for the console devices? [01:20:57] <delewis> nrubsig: but none of those limitations violate the POSIX spec, right? [01:20:58] <gdamore> that's mostly what it is used for. [01:21:07] <boyd> gdamore: I see [01:21:29] <gdamore> nrubsig: if there are specific variances from POSIX, then you should file a bug. this stuff was tested by the open group, right? [01:21:40] <delewis> of course it was [01:21:49] <delewis> Solaris was tested for UNIX98 and UNIX2003 [01:21:52] <delewis> and SUSv3 [01:21:57] <delewis> which covers the POSIX specs [01:22:02] <boyd> nrubsig: Are you saying that the problem with the xpg4 shell in solaris is that it does not have the extensions you want? They seems a silly position to take. [01:22:19] <boyd> ...non-posix extensions [01:22:51] <delewis> regardless, why should Sun completely rewrite a shell when they've got one that's nearly-compliant, but just requires a few changes. [01:23:02] <delewis> what extensions that shell contains are irrelevant to POSIX certification [01:23:10] <delewis> as long as it complies with what is specified by POSIX [01:23:18] <boyd> Exactly [01:23:48] <gdamore> hmm... i wonder if that is true. for example, I believe ANSI C requires you restrict your namespace collisions somewhat. I wonder if POSIX has similiar requirements for the shell. [01:23:56] <alanc> delewis: that's what Sun did with ksh88 to produce the current xpg4 ksh [01:24:10] <delewis> alanc: yes, which is why I used the example of a shell :-) [01:24:24] <boyd> alanc: That's how we got tot his point :) [01:24:26] <delewis> gdamore: ANSI C makes the distinction between strict compliance and compliance [01:24:39] *** coffman has quit IRC [01:24:46] <alanc> okay, I'll go back to sleep then [01:24:49] <delewis> your code is "strictly compliant" if it makes use of no undefined or implementation dependent behaviours [01:25:10] <delewis> otherwise your code is 'compliant' if it at least complies with ANSI C but makes use of undefined or implementation dependent behaviours. [01:25:21] <delewis> POSIX does not make this distinction [01:25:31] <gdamore> ah, thanks. [01:25:32] <boyd> So that's a judgment on the user of the compiler, not the implementer, right [01:25:37] <boyd> ? [01:25:38] <delewis> boyd: yes [01:25:43] <delewis> that's what -Xs does in Sun Studio, etc. [01:25:46] <delewis> (checks for strict-ness) [01:25:51] <delewis> or -Xc or whatever. [01:26:12] <delewis> and yes, some of it does reside on the user [01:26:15] <delewis> but it's up to the compiler to catch it [01:26:20] <boyd> sure [01:26:22] * gdamore decides that rebooting is faster than figuring out how to kick dacf into recognizing his changes to /etc/dacf.conf [01:26:32] <delewis> (I forget whether or not an ANSI C compiler is required to determine code-strictness) [01:26:37] <delewis> I'm leaning on "it doesn't" [01:26:50] <boyd> maybe implementation defianed [01:26:52] <delewis> regardless, the ANSI C spec is difficult to read :-) [01:26:53] <boyd> defined [01:27:10] <delewis> my knowledge of it mostly comes from "Expert C Programming," which an ex-Sun Studio developer wrote. [01:27:41] <delewis> Peter van der Linden (which also oddly enough wrote Java programming texts *cringe*) [01:27:54] * nrubsig was just saying that the sun posix shell has it's own limitations. They may fall under the category "bugs" ... but this discussion is worthless now that the successor is in preparation. [01:29:05] <boyd> nrubsig: I was wondering what you mean by limitations. Do you mean non-compliance? [01:29:29] *** raph_ael1 has joined #opensolaris [01:30:00] *** raph_ael has quit IRC [01:30:01] <nrubsig> boyd: lets phrase it like that... if it's a bug it needs to be fixed and that would draw more power from the real fix. [01:30:26] <nrubsig> boyd: that way we just label it as limitation. [01:31:20] <delewis> n [01:31:34] * nrubsig wishes the widechar API authors would burn in a very special hell. [01:31:43] <nrubsig> s/widechar/unicde, iconv/ [01:32:00] <nrubsig> The unicode junk makes everything much more compliciated. [01:32:12] <boyd> The only reason I can see to replace a working POSIX shell is if the one it's built on top of is being removed [01:34:02] <boyd> ... or if it makes the maintenance of the system significantly easier. Replacing it to add non-posix features is not justified, IMHO [01:34:03] <gdamore> crap. how does one boot using kadb in grub? [01:34:19] <boyd> on the multiboot line IIRC [01:34:25] <gdamore> so boot -k ? [01:34:28] <nrubsig> boyd: it's not about features. [01:34:28] <stevel> or -kd if you want it to drop into kmdb before starting the boot sequence [01:34:32] <stevel> if you need to set variables or whatnot [01:35:09] <nrubsig> boyd: it's more about the maintaince. the old solaris ksh has evolved in something which is considered worse than a death punishment to work on it. [01:35:37] <gdamore> okay, i don't understand. grub gives me a menu, at which i can use "c" to get to a command prompt. it doesn't understand "boot". [01:36:04] * gdamore suddenly misses the old pre-grub bootstrap [01:36:11] <nrubsig> boyd: there've been bitter complaints within sun in the past about this and the #ifdef massacre [01:36:17] <stevel> edit one of the existing boot entries [01:36:29] <stevel> change the "boot..." line to be "boot .... -k" [01:36:30] <gdamore> does the -k go before after the file name? [01:36:45] <gdamore> it says "kernel ...." not "boot ...." [01:37:05] <boyd> that's the line [01:37:08] <stevel> sorry, kernel - not boot [01:37:09] <stevel> i'm spacing out [01:37:14] <gdamore> tks. [01:38:29] <gdamore> that did it. now i can figure out what the frelling consconfig did. [01:38:32] <jmcp_> gdamore: the syntax is "kernel" + "/path/to/kernel" + $arguments [01:38:50] <richlowe> hey jmcp [01:38:56] <jmcp_> hey richlowe [01:38:58] <jmcp_> one-n-all [01:39:00] * jmcp_ doffs cap [01:39:12] <boyd> nrubsig: Ok, so it's maintenance... it would have saved a lot of time if you said that in the first place. [01:39:16] <boyd> Hi jmcp [01:39:24] <jmcp_> boyd - howdy [01:39:27] <boyd> jmcp_ even :) [01:39:31] <jmcp_> :) [01:39:32] <stevel> 'evening jmcp [01:39:39] <jmcp_> boyd: I'm trailing stuff everywhere these days :) [01:39:43] <boyd> hehe [01:40:02] <nrubsig> boyd: I could also quote bugs, but this would backfire badly. [01:40:08] <jmcp_> hmm. Need to find something on disk io QoS and containers [01:40:11] <richlowe> stevel: did you get the notification I forwarded? :) [01:40:26] <stevel> richlowe: yeah - that was completely retarded [01:40:53] <richlowe> you'll not find a better example of useless though. :) [01:41:20] <boyd> jmcp_: I don't think there is anything there for that [01:41:25] *** Ireul has joined #opensolaris [01:41:32] <stevel> indeed [01:41:42] <jmcp_> boyd: I know. What I need to find is something which addresses the concerns [01:42:21] <jmcp_> allegedly, HP's lpars can do some sort of disk qos [01:42:25] <boyd> The only thing I can generally offer is physical separation across HBAs [01:42:30] <boyd> (or disks) [01:42:40] <delewis> as does IBM's LPARs [01:42:55] <jmcp_> then again, I've seen some data recently which indicates that HP's scsi and fc stack doesn't perform. at all [01:43:00] <jmcp_> boyd: that's what I was thinking too [01:43:00] <delewis> gotta love virtual SCSI :-) [01:43:05] <jmcp_> delewis: ha! [01:43:06] <boyd> I wonder whether there are plans for ZFS to prioritise based on zone id [01:43:15] <jmcp_> maybe .... could be interested [01:43:18] <jmcp_> interesting even [01:43:26] <richlowe> I wouldn't expect that to live in the FS. [01:43:34] <jmcp_> no, me neither really [01:43:42] <jmcp_> but it's an interesting gedankenexperiment [01:43:42] <boyd> jmcp_: I'll do you a deal... you ask the q on zfs-discuss... and I'll read it [01:43:53] <jmcp_> boyd: I'll make sure I cc zones-discuss too [01:44:23] <boyd> I wouldn't expect it in the filesystem for lower-level partitioning, but zones are pretty high up. [01:44:44] <boyd> ZFS would want to know about ordering data accesses to it's pool from different sources [01:45:09] <boyd> I'm sure I saw people say [01:45:28] <boyd> I'm sure I saw people say " wouldn't expect that to live in the FS" about mirroring too at some point :) [01:45:41] <richlowe> it doesn't live in the FS, exactly, all the same. [01:45:49] <richlowe> it lives below the FS, in something that happens to share the same name ;) [01:46:03] <boyd> Well, sure... and it's that name that I used above :) [01:46:11] <richlowe> oh, well, be that way then. [01:46:13] <richlowe> ;) [01:47:15] <boyd> Can you say "flagrant layering violations"? :) [01:47:39] <twincest> yes, but for some reason i want to pronounce it 'fragrant' [01:47:59] <boyd> Hehe.... as opposed to "stinkin'" [01:48:12] <boyd> ... or are you japanese? [01:48:19] <jmcp_> boyd: and whenever I go to type "debian" I mostly-type "deviant" [01:48:38] <boyd> hehe [01:49:17] <jmcp_> boyd: it was "rampant" not flagrant [01:49:27] <boyd> Oh... oops [01:49:31] <jmcp_> boyd: nowurries [01:50:13] <jmcp_> I think my comment to dclarke when he was creating the blastwave-listed "top 10 reasons to use ZFS" was along the lines of (quote) Because the only people who care about "rampant layering violations" don't live in the real world (endquote) [01:50:18] <jmcp_> or something like that [01:50:46] <boyd> Yeah, I remember that [01:51:12] <twincest> who said that? [01:51:23] <jmcp_> twincest: me :) [01:51:40] <twincest> ah, i thought you were responding to someone in particular [01:52:14] <boyd> jmcp_: but you didn't originate "rampant layering violations" did you? [01:52:23] <jmcp_> no, that was somebody else [01:52:26] <jmcp_> just a mo .... [01:52:48] * elektronkind get frustrated at mediawiki [01:53:06] <jmcp_> Andrew Morton [01:53:20] <boyd> Aaaahhhh [01:54:01] <twincest> elek: sup? [01:54:14] *** jsubl2 has joined #opensolaris [01:55:29] <gdamore> anyone know of a command to quickly open a device and list all STREAMS modules on it? (before I write one...) [01:56:02] <boyd> would strconf < /dev/blah work ? [01:56:19] <gdamore> it might. thanks. [01:57:26] *** piwi has joined #opensolaris [01:58:07] <gdamore> crap, of course i'm trying to see below a mux driver.... [01:58:14] <gdamore> (namely consms) [01:58:37] <boyd> Don't cross the streams!!! [01:59:14] *** uebayasi has joined #opensolaris [01:59:15] <wesw> anyone know the rough eta for zfs root [installer] w/ either opensolaris or "production"? [01:59:20] <boyd> Sorry about that... channelling venkman for a second there [01:59:57] <jmcp_> wesw: s10u5 is the last rough target that I heard of [02:00:14] <wesw> jmcp_: thks. [02:00:21] <gdamore> this is weird. hwc is above consms on a ps2 mouse stream on sparc, but not on x86. [02:00:48] <gdamore> oh, figures, i'm looking at an snv_49 system on sparc. [02:00:59] <boyd> You have a PS2 mouse on a sparc? [02:01:06] <gdamore> yes. [02:01:16] <boyd> Didn't know we did that... [02:01:19] <boyd> Oh... wait [02:01:23] <gdamore> this is true on snv_50 as well... [02:01:26] <delewis> boyd: SPARC portable.. :-) [02:01:41] <gdamore> right. there was an OEM board (Ultra AX? ) that did that too [02:02:32] <gdamore> so what is "hwc"? [02:03:18] *** Ireul has left #opensolaris [02:03:19] <boyd> delewis: Yeah, I realised.. hence the "oh, wait" :) [02:03:25] <gdamore> hardware cursor? [02:04:22] <gdamore> no wonder i can't find it in open source. it must be a closed module. damn you closed source! [02:07:56] *** lloy0076 has left #opensolaris [02:07:59] <richlowe> a quick check suggests it's in an entirely closed consolidation. [02:08:05] <richlowe> never mind just regular old 'closed' [02:08:16] <richlowe> imps SUNWxwmod [02:08:18] <richlowe> whatever that means. [02:08:27] <gdamore> lemme guess: sparc graphics group. [02:10:16] <gdamore> i'm trying to get my vuidsyn (synaptics tpad) working on x86, in case anyone was curious. it works beautifully on sparc, but on x86 it only "sort of works"... i can't issue any ioctls to it down the /dev/mouse stream [02:11:08] <twincest> hmm, synaptics support is one of the things i need to run solaris on my laptop :) [02:11:45] <gdamore> there is something a bit more primitive in frkit. but i have a much nicer module that works without hacking the xserver but by adding a vuidmodule. [02:12:16] * jmcp_ aways to readup everything about containers [02:12:38] <boyd> jmcp_: That shouldn't take long :) [02:13:31] <Tpenta> good morning jmcp [02:15:47] <boyd> Hi Alan [02:15:57] <alanc> hmm, wonder if we can assign 6486407 to nrubsig [02:16:19] <boyd> I can't see that bug [02:16:30] <stevel> do it :) [02:16:40] <alanc> it just came in to the triage queue [02:16:43] <stevel> boyd: it was just filed an hour ago or so [02:16:49] <alanc> CR 6486407 *HOT* Created P1 opensolaris/triage-queue Proposal: Migrating /bin/sh to ksh93 before /bin/ksh [02:16:52] <boyd> I guess that's why :) [02:16:59] <stevel> it hasn't been synced up to b.o.o. yet [02:17:10] <boyd> P1 !?! [02:17:21] <alanc> P1 RFE [02:17:28] <boyd> Ah [02:17:48] <Tpenta> p1???? [02:17:52] <Tpenta> you gotta be kidding [02:18:01] <stevel> it hasn't been triaged yet :) [02:18:02] <alanc> and really, it's just P1 because that's what the person who filed it on bugs.os.o clicked [02:18:11] <richlowe> Not really. [02:18:12] <boyd> Hey, stevel, no pressure or anything, but is there an RSS feed of putbacks to the hg repo? I ask since the native hgweb interface will do it [02:18:18] <richlowe> You have to justify anything above P3 [02:18:22] <richlowe> I'd love to see the justification :) [02:18:27] *** jbalint has quit IRC [02:18:43] <boyd> Justification: Hissy fit [02:18:44] <alanc> === *Justification* ========================================================== [02:18:45] <alanc> Priority changed from [] to [1-Very High] [02:18:46] <alanc> ? [02:18:48] <stevel> Priority changed from [] to [1-Very High] ? [02:18:51] <stevel> :-) [02:19:19] <stevel> boyd: not at the moment, we haven't looked at setting up hgweb yet though [02:19:24] <alanc> the form doesn't judge your justification, just makes you type something in [02:19:33] <boyd> stevel: Cool... just wondering [02:19:54] <stevel> boyd: i would like to setup an RSS feed of putbacks to the Hg repo though [02:19:57] <gdamore> did gisburn file that bug? [02:19:58] <twincest> i think the form should reject unsufficient justifications [02:20:09] <alanc> nope [02:20:09] <elektronkind> what would old man Bourne think of that [02:20:12] <boyd> Justification: "Daddy, *I* want an oopma-loompa. get me an oompa-loompa *now* [02:20:43] <gdamore> hahaha [02:20:45] <boyd> twincest: You're planning on making it pass the turing test [02:20:47] <boyd> ? [02:20:49] <Auralis> heh, my P2 XVR-1000 opengl crash is still in triag [02:21:08] <twincest> boyd: i'm sure the people who invented ZFS could solve a trivial AI-complete problem [02:21:17] <boyd> hehe [02:21:30] <alanc> btw, I just got to login to JDS running on a SB1500 with PGX-64 running Xorg with the opensource ATI mach64 driver on Nevada [02:21:50] *** jbalint has joined #opensolaris [02:21:52] <boyd> No I don't want them wasting time on that.. I want do_my_work --well [02:22:05] <boyd> & [02:22:13] <alanc> it still needs a lot of work, but it was cool just to be able to use the GNOME desktop res tool to change the resolution from 1024x768 to 640x480 on a SPARC [02:22:22] <stevel> alanc: now that's cool [02:22:45] <alanc> the XVR-100 on the other hand segfaulted in the open source radeon driver [02:24:31] <gdamore> heh. the radeon code is "scary" [02:24:52] *** kleppari has quit IRC [02:24:54] <gdamore> i'm not sure it has ever been run big-endian. although maybe on macs? [02:25:16] <gdamore> gdamore& [02:26:19] <richlowe> alanc: So, I've been trying to get you a clean-seaming Xorg core dump for a while, and am generally failing. [02:26:24] *** nwf has joined #opensolaris [02:26:31] <richlowe> alanc: would mildly smashed pstack output help at all? [02:26:39] <alanc> maybe [02:26:42] <richlowe> (Xorg radean driver crashig reminded me) [02:27:00] <richlowe> Ah, hm, maybe not. [02:27:12] <richlowe> _start -> main -> Dispatch -> ???????? -> SEGV [02:27:32] <alanc> someone else complained about that today [02:27:41] <alanc> I've never seen that on my machines [02:27:45] <richlowe> it's been killing me about once a day fore a couple of builds. [02:27:50] <richlowe> but I haven't managed to get a core that's generally clean. [02:28:00] <alanc> weird [02:28:07] <richlowe> (and this machine has hardware issues, so I'm wary about it's validity in general) [02:28:20] <richlowe> for all I know, it's bad memory and spite. [02:28:37] <richlowe> (is "malice" a valid Evaluation?) [02:32:16] *** Jiko_ is now known as Jiko [02:33:34] *** stevel has quit IRC [02:37:40] *** eboutilier has joined #opensolaris [03:00:11] <nrubsig> !summon cyril [03:02:33] *** jonkelle has quit IRC [03:03:14] <jamesd> !pkill -SIGHUP nrubsig [03:03:53] <nrubsig> rm -Rf /export/home/jamesd/* [03:04:02] <gdamore> fg garrett [03:04:08] <jamesd> permission denied [03:04:26] <nrubsig> /usr/bin/fuck /dev/satan # [03:04:34] *** bengtf_ has quit IRC [03:04:37] <gdamore> no such command [03:04:40] <jamesd> core dumped all over your shoes. [03:04:54] <spawrq> damnit [03:04:55] <nrubsig> gdamore: why ? bengtf_ left the channel... [03:05:18] <spawrq> any of you go through the break-fix sun classes? [03:05:27] <spawrq> or whatever they're called [03:08:09] <jmcp_> spawrq: I did one of them [03:09:21] *** pseudo_ has joined #opensolaris [03:09:47] <spawrq> still have the books? [03:10:36] <jmcp_> nope, sorry [03:11:16] <spawrq> fault analysis workshop [03:11:18] <spawrq> that's the one [03:14:09] <gdamore> hahaha.... I think Linda is going to be very, very sorry she mentioned ATI or 3DLabs in public. [03:14:28] <richlowe> gdamore: Eh? [03:14:47] <gdamore> she is saying the reason Sun can't release ati/3dlabs sources is because of ati/3dlabs nda. [03:15:10] <Auralis> i took that as a given [03:15:14] <gdamore> i have contacts at both companies, and am pretty sure that i can get permission on both accounts to release the sun source code -- at least as far as ati and 3dlabs is concerned [03:15:18] <richlowe> it's not as if each and every one of us doesn't know they're the sparc gfx OEMs. [03:15:33] <gdamore> and now, i've thrown the gauntlet. :-) [03:16:22] <gdamore> remember, i've gone thru this process to release an ati radeon driver with even more proprietary info than what is in pfb. [03:17:12] <richlowe> pgx32 is mfb? [03:17:21] <gdamore> dunno [03:17:29] <gdamore> pgx32 is weird. [03:18:27] <gdamore> anyway, one of her common excuses just got shot full of holes. [03:18:28] *** slowhog has left #opensolaris [03:18:51] <richlowe> Probably just means the *existing* agreement doesn't allow it, and nobody cares to ask again. [03:19:02] * richlowe highly doubts you'll see code out of sparc graphics too [03:19:13] <richlowe> that may just be the reputation alanc has talked up for them though ;) [03:19:19] <gdamore> i have pfb and m64 source. :-) [03:19:27] <gdamore> (under NDA) [03:19:55] <nrubsig> gdamore: send it to pastebin.gov, please [03:19:59] <gdamore> basically, i've publically offered to work it out with ATI and 3Dlabs, if Sun will agree to release their interest. [03:20:08] <gdamore> .gov? [03:20:56] <nrubsig> gdamore: why not ? They already hide the stargate program from he public... there is more than you know... =:-) [03:21:07] <nrubsig> s/ he/ the/ [03:21:15] <gdamore> you watch too much TV. :-) [03:22:11] <nrubsig> gdamore: no. [03:22:14] <twincest> the government has a secret document that explains why ksh93 can't be integrated, too [03:22:27] <gdamore> rotf. [03:22:28] <nrubsig> gdamore: it's all in illuminati-driven plot to drive out the Ori. [03:22:45] <nrubsig> twincest: that, too. [03:23:25] <gdamore> at this point, i'm surprised ksh93 _isn't_ integrated, so the rest of us can stop talking about it. :-) [03:23:52] <gdamore> (or stop listening to others talk about it. - duck - ) [03:24:04] <nrubsig> First Afganistan, Iraq, North-Korea and Iran and THEN Gemany (well, you know... we have enougth plutonium to build 17 nukes) [03:24:11] <nrubsig> 17 nukes! [03:24:34] *** nwf has quit IRC [03:24:45] <nrubsig> We're bad nazi terrorists here, in league with iran and the evil muhatschheddin.# [03:25:57] <nrubsig> And we make&&dring beer. [03:25:58] *** nwf has joined #opensolaris [03:25:59] <nrubsig> Evil. [03:26:12] <gdamore> the rest of us just drink it. :-) [03:26:29] <nrubsig> gdamore: terrorrrist! [03:27:00] <gdamore> no, only the people who _don't_ drink beer are terrorists. look at afghanistan, iraq, iran... :-) [03:27:12] <nrubsig> bah [03:27:48] <nrubsig> grumpf... [03:27:51] <gdamore> or even the dry counties in the US south... :-) [03:27:59] * nrubsig is waiting... now for cyril to upload B51 [03:28:28] <gdamore> i tell you, nothing good ever came of denying a man his inalienable right to get mind staggeringly drunk. [03:28:28] * nrubsig is unhappy when he can't do his work. [03:31:16] <elektronkind> http://dlc.sun.com/osol/on/downloads/b51/ [03:31:46] <alanc> gdamore: as of this morning, ati is no more though - have to convince amd now [03:31:47] <nrubsig> ((glenn_releases_new_ksh93_sources && gisburn_uploads_new_sources)&& (cyril_uploads_b51 && gisburn_moves_tree_to_b51) && code_review && rti_approval && os_net_commit && echo "project done." [03:32:38] <nrubsig> elektronkind: I need it in svn [03:33:00] <nrubsig> elektronkind: and I let cyril do the work since he nows a way to do it efficiently [03:33:06] <elektronkind> nod [03:33:43] <nrubsig> elektronkind: which means I have to wait. [03:33:50] <nrubsig> elektronkind: I have to wait anyway for glenn. [03:34:13] <alanc> though amazingly enough, we started getting answers from ati people today who'd been ignoring us for weeks [03:35:02] <nrubsig> alanc: maybe they were buy worrying about their jobs... [03:35:16] <nrubsig> s/buy/busy/ [03:35:21] <gdamore> i'm not surprised. but again, i have contacts at ati/former ati. i suspect that to a certain extent they are going to be cooperative [03:36:00] *** nwf has quit IRC [03:36:11] <nrubsig> gdamore: you have the m64 code... how feasible is it to turn the 8+24 code into 12+24 code ? [03:36:30] <gdamore> NFI. I've not stared at it much. why would you want that? [03:37:02] <nrubsig> gdamore: 12bit pseudocolor is better than 8bit pseudcolor... you have 4096 color cells instead of 256 [03:37:03] <alanc> how could you store 36 bits in a 32-bit number? [03:37:16] <gdamore> i've never heard of anyone wanting to use 12-bit graphics. [03:37:25] <nrubsig> gdamore: pseudocolor [03:37:36] <alanc> it's rare these days you hear of anyone who still wants pseudocolor [03:37:48] <nrubsig> gdamore: IRIX machines had since the beginning [03:37:55] <Auralis> amiga had a 12bit color mode [03:37:56] <gdamore> again, why not use truecolor? the only people i know who want 8bit color have no choice because they have ancient apps [03:38:02] *** piwi has quit IRC [03:38:20] <twincest> alanc: http://en.wikipedia.org/wiki/Hold_And_Modify ;-D [03:38:52] <nrubsig> gdamore: 1. legacy apps. 2. pseudocolor "allowes" the trick to change colors on the fly via changing the palelle instead of updating the whole screen. [03:39:22] <gdamore> i have no interest in this. i'll work to open source the driver, then you can try. [03:39:26] <nrubsig> gdamore: boing sells a flight control software which uses this feature, and they are urgendly seeking for a replacement [03:39:34] <nrubsig> s/boing/boeing/ [03:39:43] <nrubsig> since IRIX is going away [03:40:16] <gdamore> radeon even in true color can use a palette to change the colors [03:40:27] <alanc> they might consider moving up to a 1990's method of color display [03:40:40] <gdamore> alanc++ [03:41:00] * Auralis would want at least 30bit color or even better 32bits with HDRI [03:41:01] <nrubsig> XVR-1000/4000 appears to have the hardware-capability but sun only implemented 3*8bit palattes instead of 8*10bit palattes for pseudocolor [03:41:04] <gdamore> take a look at the m64 ati driver in Xorg to see how hard it would be [03:41:24] *** gm152 has joined #opensolaris [03:41:29] <nrubsig> Auralis: note that pseudocolor can have 16bit per palette entry, e.g. 48bit color resoution [03:42:05] <nrubsig> gdamore: the Xorg m64 driver had no overlay support, sun's driver has it. [03:42:06] <Auralis> yes, the XVR-1000/4000 have 30bit color [03:42:32] <gdamore> radeon (XVR-100) uses 30bit color internally [03:42:33] <nrubsig> Auralis: and XVR-1000/4000 has support for up to three 8bit/10bit hardware palettes [03:42:49] <Auralis> nrubsig: i know, i have a 1000 :) [03:42:51] <nrubsig> sun only implemented 8bit ones [03:42:52] <alanc> I think the XVR-1000/4000 had 30-bit color available via OpenGL but not X [03:42:58] <Auralis> super image quality, just getting slow [03:43:37] <Auralis> and with a sony gdm-fw900 i can see the difference [03:43:54] <jamesd> Apple Computer reported today that it has developed computer chips that can [03:43:54] <jamesd> store and play music inside women's breasts.This is considered to be a major [03:43:54] <jamesd> breakthrough because women are always complaining about men staring at [03:43:54] <jamesd> their breasts and not listening to them. [03:43:55] <nrubsig> Auralis: soon X11 will be 24bit only without a way to implement 30bit at all. [03:44:13] <nrubsig> Auralis: thanks to keith packard's "superiour" Xrender extension [03:44:31] <Auralis> bah, frell that [03:45:02] <Auralis> i want more, 32bit hdri at least [03:47:14] <gdamore> wouldn't it be a hoot if i got pfb/m64 opensourced after all? [03:47:24] <alanc> yes [03:47:26] <Auralis> yes [03:47:27] <g4lt-U60> iboob? [03:47:28] <nrubsig> yes [03:47:44] <alanc> though then I suppose I'd have to release the Xsun sources too [03:47:51] <nrubsig> g4lt-U60: iBoobs (with milk extension) [03:47:56] <alanc> or just wait for people to port them to Xorg [03:49:48] * jmcp_ lol @ jamesd [03:49:55] <jmcp_> actually, /me rotflmao @ jamesd [03:50:01] <jamesd> :-) [03:50:11] <spawrq> that's weird [03:50:26] <gdamore> yes, Xsun would be useful. :-) I have those too, btw. :-) [03:51:11] *** nwf has joined #opensolaris [03:51:47] *** piwi has joined #opensolaris [03:51:57] <jmcp_> jamesd: got a url for that, btw? [03:51:57] <alanc> but then I'd have to go dig through Xsun to see if there's anything other than libserverdps that isn't either open source or Sun owned, and that's a hell of a lot of source [03:52:00] <Auralis> anyway, limiting the display to 24bit is not good, especialy with hdtv and shit moving up to 30bit already [03:52:19] <jamesd> jmcp no it was pasted into another channel. [03:52:28] * jmcp_ nods [03:52:39] * jmcp_ forwards to closely vetted friends .... [03:52:40] <jmcp_> :) [03:53:50] <gdamore> as i said, xvr-100 is already 30 bit. [03:54:09] <gdamore> (though the driver just uses a 24-bit map, and 30-bit color ramp) [03:54:24] *** noyb has quit IRC [03:54:33] <piwi> hi, could anybody confirm that blogs.sun.com is down? [03:55:03] <jmcp_> piwi: seems to be down from my end [03:55:15] <piwi> ok, thanks [03:55:17] *** noyb has joined #opensolaris [03:55:31] <alanc> works for me, though that may be coming from the proxy cache, since it shows the last post being at 6pm [03:55:50] <piwi> mmh, i get a "Follow these steps to finalize your Roller installation:" welcome message [03:56:10] <spawrq> this solaris early access cd doesn't seem to have the installer on it [03:57:20] * jmcp_ lunches [03:57:30] <spawrq> whats the solaris installer called? [03:57:39] <spawrq> for text based console [03:57:47] <jmcp_> spawrq: evil-installer-spawn-of-doom [03:57:50] <gdamore> suninstall, i think [03:57:56] <spawrq> hahah [03:57:57] * jmcp_ really lunches [04:01:11] <spawrq> is there some weird problem with 50's installer routine? [04:01:33] <spawrq> the cd boots up, spits out some info about the NIC then dies saying it can't find the installer [04:02:33] *** zarathustra has quit IRC [04:02:41] *** zarathustra has joined #opensolaris [04:04:07] *** eboutilier has quit IRC [04:08:14] *** jafari has joined #opensolaris [04:08:26] <jafari> # [04:08:31] <jafari> sorry [04:08:40] *** gm152 has quit IRC [04:08:43] *** gm152 has joined #opensolaris [04:16:58] *** jonkelle has joined #opensolaris [04:19:43] *** deedaw has quit IRC [04:22:10] *** alanc is now known as alanc-away [04:22:30] *** yongsun has joined #opensolaris [04:26:50] <gdamore> i'm going nuts trying to figure out why ioctls don't seem to get from consms to my vuid module. [04:27:07] <gdamore> and why this works on SPARC but not on x86 is what is _truly_ annoying me. [04:28:05] *** aliquis has joined #opensolaris [04:28:20] <aliquis> Hello hey [04:28:37] <gdamore> oh wait, i think i've figured it out. [04:32:20] *** kleppari has joined #opensolaris [04:32:31] <aliquis> Off topic: Nerds as you are, which is the better option: 1) Keep my not-so-shitty-but-on-a-budget-PC with old CRT and crap gfx and run whatever free software on it or hackintosh 2) Buy 20" iMac 3) Buy 15" Macbook Pro I'm poor and unemployed but I can afford them atm. [04:34:02] <Auralis> with the macbook pro i would wait till the core2 duo models are aviable [04:34:05] <gdamore> what are you planning on doing with it? if you want to run MacOS X, the choice is clear. :-) But otherwise you have to figure out what you're needs are [04:34:10] <gdamore> s/you're/your [04:35:01] *** Dr_Jekyll has joined #opensolaris [04:35:50] <aliquis> Auralis: core 2 models was released two days ago [04:36:11] <aliquis> Auralis: low end config got 1GB ram default, faster cpu, more hdd, and all models got FW800 [04:36:26] <aliquis> mid model and max model got 2GB ram as default [04:37:23] <Auralis> ah cool, wasn't sure if they are out yet [04:37:40] *** piwi has quit IRC [04:37:58] <aliquis> gdamore: Well, usually I think KDE is ok, amarok > itunes, k3b is probably as good as whatever macos comes with, Opera > Safari, kmail/mutt/thunderbird are probably > mail.app, and so on. But it's hard to hide the fact that more stuff "works" with mac os, thought I always belive I can get everything to work anyway but that is rarely the situation ;), also you can't deny that there are some commercial apps for the mac [04:38:33] <aliquis> gdamore: But even if I would use mac os x i can install it on my regular pc, thought future updates might not work, and I should really buy a ati card so i get accelerated graphics in that case [04:39:07] <Error_404> aliquis: kmail/thunderbird aren't as good as mail.app [04:39:24] <delewis> pine > * [04:39:25] <delewis> :-) [04:39:38] <gdamore> you can't install macos on a regular pc. [04:39:43] <delewis> actually, I use Thunderbird, but I had to throw a bit of late-nite sarcasm into the mix. [04:39:46] <Error_404> gdamore: not legally [04:39:47] <aliquis> gdamore: For imac vs macbook I don't have any friends to visit longer and therefor not so huge usage for a notebook and high res would be nice. But with the macbook pro I get almost the same machine and I can always get a 23" TFT or whatever for it, so I can make it just as good as the desktop, harder to make the iMac into a laptop ;) [04:39:56] *** Dr_Jekyl1 has quit IRC [04:40:01] <delewis> what's Solaris support like on Apple hardware, nowadays? [04:40:12] <gdamore> i heard it mostly works. [04:40:17] <delewis> "mostly" :-) [04:40:18] <gdamore> i think audio was one shortcoming [04:40:19] * delewis smirks [04:40:39] <aliquis> Error_404: Ok, I have never used mail.app, just read someones opinion was that thunderbird was better (he was on mac and switched), so i trusted him. Many of the apple bundled software looks great and all but they seem rather limited compared to others. (Say iPhoto looks hot as hell, but doesn't Picasa give more options?) [04:41:05] <aliquis> gdamore: I have had mac os installed on my pc, so yes, i can [04:42:03] <aliquis> but I don't have SSE3 which is a problem for latest versions of quicktime and itunes, but that will probably be fixed (emulated function call), and since I have nvidia gfx there are no accelerated driver, but I could buy a X1600 card and get that aswell. That would give me a much cheaper mac, but not a real one, not a legal one, not a good looking one, not a silent one, not a huge screen one ;) [04:42:04] <gdamore> okay, well, then it is a non-supported hack. :-) [04:42:07] <Error_404> as far as safari vs. opera and k3b vs disk utility, i concur though [04:42:11] *** Gman is now known as GmanAFK [04:42:43] <aliquis> Error_404: Thought of course opera is available in mac os aswell, but the point "safari looks and is great" isn't valid if you will use another browser :) [04:42:53] <Error_404> safari is ugly and crap [04:42:59] <aliquis> :D [04:43:01] <Error_404> no adblock + filterset.g = useless [04:43:06] <AbeFroman> it's fine, but with no adblockafjd [04:43:17] <AbeFroman> good call [04:43:19] <gdamore> i didn't know that it had been hacked for official MacOS X. I though the hack was only for the ADC prerelease. but i don't follow mac too closely anymore. [04:43:24] <boyd> Error_404: firefox has no adblock either [04:43:26] <Error_404> it has pith helmet, but that's not free [04:43:31] <Error_404> boyd: doesn't it? [04:43:32] <aliquis> They just got the 10.4.8 kernel running on hackintosh =P [04:43:42] <Error_404> i wonder why i haven't seen an ad in 2 years then, that's wierd [04:43:48] <boyd> Error_404: It's an addon... but there are addon for safari too [04:43:58] *** jonkelle has quit IRC [04:44:05] <Error_404> and none of them can import filterset.g [04:44:19] <aliquis> http://static.flickr.com/101/265167820_9ae0f0f586_o.png mac os x on my pc ;) [04:44:32] <aliquis> http://static.flickr.com/112/265175677_97514418f6_o.png with cpu info ;) [04:45:08] <aliquis> http://img320.imageshack.us/img320/7467/dopusdt6.png Win xp shoot, but the theme broke the installation a lot :D [04:45:09] *** karrotx has joined #opensolaris [04:45:27] *** jonkelle has joined #opensolaris [04:46:10] <aliquis> Error_404: Also I suppose finder is better than windows explorer? [04:46:21] <aliquis> Error_404: But directory opus is better than both? [04:48:49] *** delewis has quit IRC [04:51:36] * aliquis killed the channel :/ [04:57:29] *** delewis has joined #opensolaris [05:03:07] *** nwf has quit IRC [05:03:40] <Tpenta> http://blogs.sun.com/tpenta/entry/sjvn_s_rant_on_oracle [05:07:44] *** sahafeez has quit IRC [05:12:28] *** paxc has joined #opensolaris [05:12:58] <paxc> ola! [05:14:24] *** jafari has quit IRC [05:19:22] <aliquis> ketall? [05:19:35] <aliquis> Que'tal maybe, who knows [05:19:36] <aliquis> :D [05:20:36] *** laca has quit IRC [05:24:14] *** laca has joined #opensolaris [05:31:37] *** jonkelle has quit IRC [05:32:52] <boyd> Hmm... now that there's an hg repo, it looks like the full hg bundles to download have gone, is that right? [05:33:34] *** jonkelle has joined #opensolaris [05:35:49] <aliquis> Error_404: your take on file managers? [05:37:30] *** qbit has joined #opensolaris [05:38:04] *** jmcp_ has quit IRC [05:38:19] *** Sieghard has joined #opensolaris [05:39:37] <Sieghard> is it normal for the system to panic every time i run 'zpool scrub' ? [05:39:59] <boyd> Umm... no. [05:44:14] *** jsubl2 has quit IRC [05:46:21] <Sieghard> boyd: it isn't? hmm, seems to happen on all my machines [05:47:55] *** qbit has left #opensolaris [05:49:22] <boyd> That's odd. [05:49:44] <boyd> I'd get a stack trace and search for a bug [05:50:01] <Sieghard> boyd: how do you do a stack trace? mdb? [05:50:13] <richlowe> What's the panic message? [05:50:21] <richlowe> and is the pool redundant? [05:50:25] * boyd teaches.... sorry [05:50:37] * richlowe is wondering if the scrub is just touching off an I/O error in a non-redundant config [05:51:29] *** qbit has joined #opensolaris [05:53:15] <Sieghard> richlowe: the pool is redundant [05:53:53] <Sieghard> richlowe: panic is page fault [05:54:09] <Sieghard> zfs:vdev_queue_io_to_issue+1ec () [05:55:28] <Sieghard> I do get one "Illegal Request" error on the drives (iostat -nE) when i start the scrub, but the drives themselves are fine [05:57:31] <richlowe> I'd ask zfs-discuss then. :) [05:58:02] <Sieghard> richlowe: already done, now just waiting for response :) [05:58:19] <Sieghard> hopefully i'll figure out what's wrong this time, these zfs problems are driving me insane [05:58:44] <Sieghard> last time i got no luck, had to destroy old pool and start from scratch [06:03:29] *** gnu2it2 has joined #opensolaris [06:05:30] <gnu2it2> my cdrom died on my solaris box, is there a way to setup the /jumstart directory structure on a linux box from the CDs? [06:11:12] <Sieghard> gnu2it2: nfs [06:18:20] *** laca has quit IRC [06:25:59] *** wesw has quit IRC [06:50:24] *** nwf has joined #opensolaris [06:57:18] *** drio has joined #opensolaris [07:00:27] *** drio has quit IRC [07:21:46] *** paxc has quit IRC [07:23:28] *** nwf has quit IRC [07:33:27] *** drio has joined #opensolaris [07:33:30] *** drio has left #opensolaris [07:34:12] *** drio has joined #opensolaris [07:38:54] <nrubsig> ahhgggrrllll [07:38:57] <nrubsig> Heeeelpppp [07:39:04] <nrubsig> Alien attack: http://maps.google.com/maps?hl=en&t=k&q=Germany&ie=UTF8&z=18&ll=48.857699,10.205451&spn=0.002404,0.006738&om=1&spn=0.002404,0.006738&om=1 [07:39:23] <nrubsig> Europe is dooooomed.... [07:40:36] *** drio has quit IRC [07:40:37] <dwc-> germany first [07:40:46] *** nrubsig was kicked by nrubsig (User was terminated by giant insects.) [07:40:57] <dwc-> better get your komodo dragons ready for combat [07:41:26] *** drio has joined #opensolaris [07:48:18] <aliquis> dwc-: no problem as long as it stays in germany only ;) [07:48:34] <aliquis> if it walks down to italy that is ok aswell [07:48:36] <aliquis> it would actually be prefered [07:51:25] *** drio has quit IRC [07:51:56] *** drio has joined #opensolaris [07:53:12] *** drio has quit IRC [08:04:23] *** EdLin has joined #opensolaris [08:06:22] *** EdLin has left #opensolaris [08:29:31] *** tsoome has joined #opensolaris [08:41:37] *** Ezechial_ has joined #opensolaris [08:43:06] *** bunker has quit IRC [08:46:16] *** cypromis_ has joined #opensolaris [08:49:16] *** sahafeez has joined #opensolaris [08:55:29] *** Cyrille has joined #opensolaris [08:57:45] *** cypromis has quit IRC [09:01:15] *** bougie has joined #opensolaris [09:12:13] *** Dar_HOME is now known as Dar [09:14:05] <bougie> hello [09:15:48] *** mazon is now known as Mazon [09:19:13] *** rachel has quit IRC [09:19:54] *** rachel has joined #opensolaris [09:24:35] *** tsoome has quit IRC [09:49:19] *** astinus has quit IRC [10:02:56] *** yongsun has quit IRC [10:07:19] *** dunc has joined #opensolaris [10:08:57] *** aijin has joined #opensolaris [10:10:06] *** alanc_away has quit IRC [10:10:06] *** GmanAFK has quit IRC [10:10:48] *** GmanAFK has joined #opensolaris [10:10:52] *** alanc_away has joined #opensolaris [10:13:57] *** axxl has joined #opensolaris [10:14:42] *** damienc has joined #opensolaris [10:16:43] *** aijin has quit IRC [10:17:27] *** aijin has joined #opensolaris [10:17:59] *** bunker_work has joined #opensolaris [10:31:43] *** raph_ael1 is now known as raph_ael [10:33:15] *** MattMan has joined #opensolaris [10:35:00] <raph_ael> hello [10:35:25] *** Triskelios has quit IRC [10:59:22] *** sahafeez has left #opensolaris [11:02:04] <PerterB> morning [11:05:00] <asyd> hello PerterB / raph_ael [11:09:28] <raph_ael> hello asyd :) [11:36:41] *** Cyrille has quit IRC [11:37:24] *** Cyrille has joined #opensolaris [11:51:00] *** GmanAFK is now known as Gman [11:52:10] *** Vanuatoo has left #opensolaris [11:56:32] *** Triskelios has joined #opensolaris [11:56:48] *** triplah has joined #opensolaris [12:00:50] *** r3boot has quit IRC [12:05:10] *** kloczek has joined #opensolaris [12:06:36] *** Triskelios has quit IRC [12:09:47] *** Vanuatoo has joined #opensolaris [12:11:50] *** Triskelios has joined #opensolaris [12:12:48] <asyd> ./s 2 [12:12:51] <asyd> oups [12:15:38] <Gman> fuck i hate lawyers. [12:19:44] <Cyrille> any specific reason? [12:19:53] <Gman> osr stuff [12:20:40] <Gman> there's not a single lawyer i've met that seems to understand software and how it all links together [12:20:49] <Gman> [other than, say, eben] [12:21:05] <clee> that's depressing [12:21:14] <Gman> indeed [12:21:15] <clee> you would think that of all companies, Sun would probably have lawyers that grokked open source [12:21:23] <clee> even RH's lawyers aren't that great at it [12:21:25] <Gman> they probably do [12:21:40] <Gman> they're probably in short supply and the good ones get better projects ;) [12:23:01] <clee> hehehe [12:25:36] * dunc wonders if u could use the word grok in court [12:25:48] *** LordKing has joined #opensolaris [12:29:02] *** aijin has quit IRC [12:36:45] *** salamanders has joined #opensolaris [12:41:41] *** LeftWing_ has quit IRC [12:41:41] *** LeftWing has joined #OpenSolaris [12:44:02] *** damienc has quit IRC [12:44:55] *** damienc has joined #opensolaris [12:45:24] *** jonkelle has quit IRC [12:46:01] *** jonkelle has joined #opensolaris [12:46:24] *** Triskelios has quit IRC [12:46:29] *** Triskelio- has joined #opensolaris [12:51:59] *** gm152 has quit IRC [12:57:52] *** calumb has joined #opensolaris [13:00:57] *** Odin-LAP has joined #opensolaris [13:01:14] *** MattMan is now known as MattAFC [13:02:20] *** calumb is now known as calAFK [13:03:55] *** uebayasi has quit IRC [13:04:54] *** jamesd has quit IRC [13:16:21] <jengelh> How do I copy the hardware clock time to the software clock? [13:17:11] <lplatypus> dunc: no, i think her majesty would be a bit old to recognise that word [13:18:30] <Cyrille> well he just asked whether it could be used, not understood ;-) [13:18:47] * boyd wonders what the legalities of including Gentium in Osol would be... http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&id=gentium [13:23:09] <oxygene> boyd: the license seems to be very similar to bitstream vera's license [13:23:20] <boyd> I guess that's good... [13:23:23] <oxygene> yes [13:23:34] <boyd> Hey, nice work on the latest pmpkg, BTW [13:23:39] <oxygene> boyd: the only nit I see is 3): No Modified Version of the Font Software may use the Reserved Font Name(s), in part or in whole - what's "a part" of those names? [13:23:54] * boyd nods [13:24:24] <boyd> You can't use a "G" :) [13:24:40] <oxygene> ie. if I make a font called abcde..wxyz01..89 under this license - is any of the characters "part of the name"? [13:24:52] *** coffman has joined #opensolaris [13:25:17] *** lplatypus has quit IRC [13:25:25] <boyd> I see the point. [13:25:28] <twincest> hmm, that's quite a nice font [13:25:48] <boyd> Yeah, it's pretty comprehensive in the glyphs too [13:25:54] <oxygene> boyd: but they write that a "1.1" version of that license is work in progress, with clarifications - maybe that fixes it [13:25:58] <oxygene> hmm.. [13:26:06] <jteo> zfs hangs systems during copying. hmm... [13:26:07] * oxygene packages that font ;) [13:26:34] <boyd> oxygene: Hehe... they hav some others too... http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=CharisSILFont http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=DoulosSILFont [13:27:06] <oxygene> hmm.. okay, that will be a "sil-fonts" bundle (similar to the non-free agfa-fonts I already have) [13:27:43] *** Odin-LAP has quit IRC [13:28:44] <boyd> There are a couple of others http://scripts.sil.org/cms/scripts/page.php?site_id=nrsi&item_id=OFL_fonts [13:30:07] <oxygene> boyd: thanks for the pointer [13:30:16] <boyd> Thanks for the packages :) [13:30:35] <oxygene> :) [13:30:41] <jteo> where? [13:32:45] <boyd> http://openbios.org/~oxygene/projects/Projects/pmpkg [13:34:22] *** Griff has joined #opensolaris [13:35:24] *** wesw has joined #opensolaris [13:37:53] <jteo> ah. [13:41:49] <jteo> does anyone here actually *want* a mirror of raidzs? [13:42:02] <boyd> Not me. [13:42:53] *** Griffous has quit IRC [13:44:09] <quasi> jteo: I contemplated it for a moment as pure fun [13:44:53] <jteo> so the use case is..esoteric. [13:46:07] <quasi> raidz/raidz2 in it self isn't quite flying on this thumper... [13:46:42] <spawrq> eh? [13:46:48] <spawrq> software raid is meh. [13:46:59] <spawrq> MEH I TELL YOU [13:47:28] <quasi> but then there's various odd hw raid versions as well - so saying nobody would is probably too optimistic ;) [13:47:40] <boyd> spawrq: you are unlikely to get a positive response here. Your comment sounds uninformed in the like of ZFS [13:47:55] <spawrq> boyd- i've been using zfs for a little while now. [13:49:56] <quasi> -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- [13:49:59] <quasi> Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP [13:50:02] <quasi> x4500 16G 113938 96 104538 19 80421 20 103010 96 392274 41 400.9 1 [13:50:11] <quasi> is raidz2 on 9 disks [13:50:36] <quasi> -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- [13:50:40] <quasi> Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP [13:50:43] <quasi> x4500 16G 118527 99 200102 46 100964 29 98140 95 248235 28 589.7 2 [13:50:49] <quasi> is raidz on 9 disks [13:51:21] <quasi> -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- [13:51:22] <spawrq> wonder what it would be on a 3320 using just zfs? [13:51:25] <quasi> Machine Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP [13:51:28] <quasi> x4500 16G 27939 23 26610 5 20642 5 89223 83 108337 8 622.8 1 [13:51:31] <quasi> is mirror of 2 disks [13:51:48] <trygvis> dude .. paste on rafb.net/paste please [13:52:40] <spawrq> maybe i'll run some more tests today [13:54:52] <quasi> spawrq: those numbers were out of bonnie++ [13:55:26] <lasseoe> quasi, tried filebench? [13:57:03] <quasi> lasseoe: yes, but that was totally broken - I've been planning to rebuild it myself and see if it works any better than the package off sf.net [13:57:23] *** axxl has quit IRC [13:57:37] <lasseoe> quasi: heh yeah I've had nothing but problems with the package as well, haven't tried rebuilding from source yet [13:57:47] *** tfb has joined #opensolaris [13:58:19] <quasi> IO Summary: 376839 ops 6238.1 ops/s, (960/960 r/w) 30.6mb/s, 157us cpu/op, 8.1ms latency [13:58:34] <quasi> IO Summary: 458921 ops 7593.1 ops/s, (1168/1168 r/w) 37.4mb/s, 148us cpu/op, 6.6ms latency [13:58:50] <quasi> IO Summary: 274736 ops 4547.1 ops/s, (700/700 r/w) 22.1mb/s, 133us cpu/op, 11.2ms latency 1957: 622.336: Shutting down processes [13:59:15] <jengelh> `svcadm enable application/x11/x11-server` says "Patten '..' does not match any instances", but it exists in `svccfg list` and /var/svc [13:59:17] <quasi> those are raidz2, raidz and mirror of varmail [14:01:28] <icon> hrm curious [14:01:39] <icon> well maybe not [14:01:49] <icon> quasi: using two parity disks on the second test? [14:02:21] <boyd> jengelh: there is no service called that... check the output of svcs. I suspect you want cde-login or gdm [14:02:54] <quasi> icon: no magic at all - just basic zpool of the chosen type [14:03:18] <icon> quasi: how many disks? [14:03:31] *** Tpent1 has joined #opensolaris [14:03:47] <quasi> icon: 9 in both the raidz2 and raidz, 2 in the mirror [14:04:13] <icon> same number of parity disks in both of the raidz pools? [14:05:19] <quasi> it doesn't say - they were created without specifying anything like that, so whatever it is, it is properties of raidz(2) [14:05:36] <icon> ah [14:05:50] <boyd> raidz2 has 2 raidz has one [14:06:59] <quasi> I suppose that explains part of the speed difference [14:07:27] *** B|nTaRa-village is now known as B|nTaRa [14:11:07] <oxygene> boyd: http://en.wikipedia.org/wiki/SIL_Open_Font_License [14:12:03] <boyd> Looks ok [14:12:04] <oxygene> boyd: the sil fonts are packaged - but it seems like gentium is really the best of them, the others look weird at some sizes (eg. regular becomes bold at >13pt) [14:13:06] <boyd> I suspect that's a limitation of the renderer in Solaris. They look fine on MacOS [14:13:36] <oxygene> ah, right.. in 10u2 it was still buggy (but I think it's fixed in nevada) [14:25:11] *** jbenedetto has joined #opensolaris [14:28:46] <Stric> quasi: and raidz2 has more complex maths than raidz or mirror [14:32:34] <coffman> raidz on a thumper.. i thought mirror+strip is way to go fo performance? [14:32:55] *** laca has joined #opensolaris [14:35:24] *** dunc has quit IRC [14:36:03] *** dunc has joined #opensolaris [14:48:29] *** LordKing has quit IRC [14:49:32] <quasi> coffman: perhaps - I'm just comparing the 3 on s10u3 beta [14:54:51] *** MattAFC is now known as MattMan [14:58:17] <icon> coffman: raidz isnt the same as raid 0+1 [14:58:34] <icon> it resembles a more stable version of raid 5 [14:58:39] <oxygene> icon: for a start, it's slower ;) [14:58:42] <icon> yup [14:58:58] <icon> parity is much better than mirroring stripes, not to mention you get more capacity for it [14:59:05] <oxygene> icon: I see no claim in coffman's statement that raidz would be similar to raid1+0 [14:59:06] <coffman> i know that [14:59:28] <icon> sorry, i need coffee :) [14:59:40] <oxygene> :) [15:00:26] <icon> the part that threw me was that the cpu usage was lower on the raidz2 pool [15:00:33] <icon> i would have expected more [15:01:08] <coffman> if i had 48 disks i would do mirror+strip on them :) [15:01:20] <icon> coffman: why? [15:01:45] <icon> that would basically give you 24 disks worth of capacity - which is a huge waste [15:02:04] <icon> plus, you can only recover from 1 disk failure [15:02:26] <Stric> if you do it wrong, yes [15:02:29] <quasi> icon: you should be able to recover from up to 24 disks [15:02:43] <icon> ah, mirror the stripes, not the otherway around ;) [15:02:52] *** Teknix has quit IRC [15:03:04] <Stric> stripe over 24 small mirrors [15:03:07] <icon> yeah [15:03:22] <icon> still a nasty affair when you have a failure [15:03:41] <Stric> but if you're unlucky and two disks within the same mirror dies, you're toast.. [15:03:47] <jengelh> so use raid10 [15:03:57] *** qbit has left #opensolaris [15:04:39] <Stric> jengelh: what type of "raid10" do you mean? [15:04:59] <Stric> people mix 0+1, 1+0 etc all the time, so it's kinda hard to know what someone mean.. [15:05:02] <jengelh> what, what type? [15:05:07] <jengelh> I said raid10 (1+0) not raid01 [15:05:28] <coffman> well, i would split, like half in raidz half in mirror strip [15:07:09] *** nachox has joined #opensolaris [15:07:19] *** calAFK is now known as calumb [15:08:13] *** mrdeviant has joined #opensolaris [15:15:59] *** Headphone has joined #opensolaris [15:23:13] *** coffman has quit IRC [15:31:02] <quasi> one odd thing with the raidz/raidz2 - running 5 iterations on bonnie++, raidz2 performace jumps up and down by ~20% - on raidz it is pretty much constant (and no, the machine isn't doing anything else) [15:32:05] <jteo> i need to start running benchmarks too.. [15:34:51] *** wesw has quit IRC [15:37:16] <Headphone> I had tried to install fedora core 6 on a machine which has soalris on it [15:37:23] <Headphone> Now : dmesg |grep sol [15:37:23] <Headphone> Console: colour VGA+ 80x25 [15:37:23] <Headphone> Xen virtual console successfully installed as ttyS0 [15:37:23] <Headphone> sda4: <solaris: [s0] sda5 [s1] sda6 [s2] sda7 [s3] sda8 [s5] sda9 [s7] sda10 > [15:37:28] <Headphone> on linux [15:37:54] <Headphone> But I cannot mount the solaris partition from linux ! [15:38:23] <Headphone> for example: mount -o ro /dev/sda5 /mnt/solaris gives an error [15:38:33] <Headphone> mount: wrong fs type, bad option, bad superblock on /dev/sda5, [15:38:36] <jteo> obviously you can't. Linux can't read UFS. [15:39:01] <Headphone> jteo, I don't think you are right [15:39:12] <Headphone> jteo, because I done it in the past [15:39:20] <Headphone> on a different machine [15:39:27] <jteo> Headphone: hmm. [15:39:30] <richlowe> linux will sometimes read UFS. [15:39:31] <Headphone> you hace ufs support in linux [15:39:42] <jteo> that UFS is..slightly different. [15:39:43] <Headphone> lsmod |grep ufs [15:39:46] <richlowe> how well it does depends on which UFS in specific. [15:39:58] <Headphone> returns ufs [15:39:59] <richlowe> but unless things changed since I last tried, I'd strongly suggest not trying to write to it. [15:40:13] <richlowe> the linux ufs module takes (or at least took) arguments indicating which kind of UFS. [15:40:17] <Headphone> I did try mount with all possibile ufs subtypes [15:40:24] *** calumb is now known as calAFK [15:40:48] <Headphone> richlowe, you cannot write to it; it is mounted ufs only [15:41:43] <Headphone> the problem is that it does not permit even to mount it as read only ! [15:41:54] <Headphone> but on a different machine it does. [15:42:24] <Headphone> the difference is , that in that different machine FIRST there was linux and than solaris [15:42:32] <Headphone> but here it is vice versa [15:43:10] <Headphone> However, the linux installer does NOT touch the solaris [15:43:25] <Headphone> so this is very very strange for me [15:43:40] <Headphone> what i see in the kernel log while trying to mount is : [15:43:46] *** rfds has joined #opensolaris [15:43:53] <Headphone> ufs_read_super: bad magic number [15:44:06] <rfds> hello [15:44:13] <rfds> solaris 2.8 correspond to solaris 8? [15:45:27] *** bank__ has joined #opensolaris [15:45:30] <asyd> yup [15:45:45] <Stric> not that 2.8 exists, but people say 2.8 when they mean 8, yes. [15:46:07] <rfds> asyd Stric thanks! [15:46:11] <rfds> :) [15:46:49] <Stric> (and the kernel in solaris 8 is sunos 5.8, just to add a bit more to the confusion) [15:47:41] *** karrotx has quit IRC [15:47:44] <rfds> okay :D [15:48:45] <bank__> I expect that after replace DS with DSlocalhost I can succesfully sendmail to my google account. [15:50:03] <bank__> but it doesn't. [15:50:18] *** rfds has quit IRC [15:54:45] *** calAFK is now known as calumb [15:55:44] <hile_> you should NOT be hacking up the sendmail.cf directly unless you know a hell of a lot more about what you're doing than you appear to know [15:56:08] <delewis> greetings, hile_ [15:58:25] <bank__> mailq then /etc/mail/sendmail.cf: line 201: unknown configuration line "0 ResolverOptions=-DNSRCH" [15:58:46] <bank__> but it doesn't need I think.. [15:59:28] <bank__> hoooleeyy [15:59:30] <bank__> :D [15:59:34] <bank__> i got mail [15:59:46] <Headphone> let me rephrase my mounting question: is it possible to install a dual boot system such that first we install solaris and than linux ? does anynody know about such possibility ? [16:00:41] <jengelh> as you said [16:04:46] *** axisys has joined #opensolaris [16:05:51] *** r3boot has joined #opensolaris [16:08:51] *** triplah has quit IRC [16:20:47] * quasi starts heading for http://nl.sun.com/nlosug/ [16:21:26] <jteo> quasi: have fun. [16:22:15] <trs81> Headphone: linux swap shares a partition number with solaris iirc [16:22:38] <quasi> jteo: thanks [16:22:51] <trs81> so you might need to specify -t ufs on the mount line. and possibly other options, check the man page [16:23:00] * quasi just wishes more of it would be in a language I understood [16:26:22] *** calumb is now known as calAFK [16:26:38] *** postwait has quit IRC [16:28:20] *** Dar is now known as Dar_HOME [16:31:03] <Headphone> trs81, right, but I mistakingly did not install swap partition here... [16:31:23] <Headphone> trs81, what seems to me is that the linux installation [16:31:38] <Headphone> damaged the soalris installation from before; and googling a bit [16:31:41] *** bunker_work has quit IRC [16:31:53] <Headphone> show that perhaps I should have isntalled linux [16:32:03] *** postwait has joined #opensolaris [16:32:10] *** Wez has quit IRC [16:32:32] <Headphone> with install Grub NOT on the MBR [16:32:51] <trs81> what linux? [16:32:57] <Headphone> installing Grub on the MBR is the default. [16:33:07] <Headphone> you mean which disto ? FC6 [16:33:20] <Headphone> or do you mean somethng else? [16:33:23] <trs81> yeah, distro [16:33:40] <Headphone> FC6 , released 2 days ago [16:33:47] <trs81> debian has pretty good control about what partitions are used for what things [16:33:54] <trs81> never used fedora myself [16:34:26] <Headphone> it seems to me that I will have to install solaris again [16:34:30] *** calAFK is now known as calumb [16:35:03] <Headphone> does anybody knwo about installing GRUB NOT on the MBR? this is the default [16:35:10] <trs81> yeah, if the partition has been mkswapped it's probably gone [16:36:12] <Headphone> trs81, I did not create a swap ! [16:36:35] *** woland_ has joined #opensolaris [16:36:40] <trs81> Headphone: but the installer probably saw the solaris partition and thought "ah, a swap device" and ran mkswap on it [16:37:11] <woland_> hi guys. is it possible to install opensolaris on a machine adaptec scsi/raid controllers? [16:37:18] <woland_> solaris 10 doesn't see the disks [16:37:23] *** loke has joined #opensolaris [16:37:24] <Headphone> If it did , then a swap partition shpuld have beeb creatd ! fdisk -l does not show any swap (It was an error) [16:37:51] <trs81> http://www.sun.com/bigadmin/features/articles/multiboot_laptop.html [16:37:57] *** bengtf_ has joined #opensolaris [16:38:01] *** bengtf_ is now known as bengtf [16:38:28] <trs81> what does file -s on the partition say? [16:38:35] *** richlowe has quit IRC [16:38:37] *** rachel_ has joined #opensolaris [16:39:31] <Headphone> trs81 , from linux you mean ? on which partion ? [16:39:38] <trs81> the solaris partition [16:39:57] <Headphone> file -s /dev/sda4 [16:39:58] <Headphone> /dev/sda4: x86 boot sector, GRand Unified Bootloader (0.94); partition 1: ID=0xc, starthead 1, startsector 63, 61432497 sectors; partition 2: ID=0x83, starthead 254, startsector 61432560, 208845 sectors; partition 3: ID=0x83, starthead 254, startsector 61641405, 102398310 sectors; partition 4: ID=0xbf, active, starthead 254, startsector 164039715, 81947565 sectors, code offset 0x48 [16:40:13] <trs81> interesting ... [16:40:14] <Headphone> ID is solaris [16:40:21] <Headphone> 0xc [16:40:47] <trs81> what about /dev/sda5? [16:41:19] <Headphone> interesing ... [16:41:20] <Headphone> file -s /dev/sda5 [16:41:21] <Headphone> /dev/sda5: Unix Fast File system (little-endian), last mounted on /, last written at Wed Oct 25 10:30:57 2006, clean flag 253, number of blocks 20482875, number of data blocks 20172604, number of cylinder groups 417, block size 8192, fragment size 1024, minimum percentage of free blocks 1, rotational delay 0ms, disk rotational speed 60rps, TIME optimization [16:41:26] <trs81> yep [16:41:30] <Headphone> UFS it seems to me [16:41:31] <trs81> /dev/sda4 is an extended parition [16:41:36] <Headphone> I know [16:41:46] <AbeFroman> email from our linux admin: [16:41:47] <AbeFroman> > If you want to have the best chance of surviving OS release [16:41:48] <AbeFroman> > boundaries, link executables statically. [16:41:57] <AbeFroman> this is why you fail [16:42:01] <hile_> oh geez [16:42:05] <trs81> are you familiar with the way dos parition tables work? [16:42:17] <Headphone> trs81, very little [16:42:22] <asyd> s18 [16:42:23] <trs81> Headphone: anyway, try mounting /dev/sda5 with -t ufs -o ufstype=sunx86 [16:42:24] <asyd> oups [16:42:33] <Headphone> I tried [16:42:40] <Headphone> the magic number error [16:42:44] <trs81> Headphone: the extended partition contains a pointer to another partition table [16:43:46] <Headphone> why fdisk -l does not show it ? [16:44:00] <Headphone> it should show it I think [16:44:12] <trs81> because it's an implementation detail [16:44:13] <Headphone> Device Boot Start End Blocks Id System [16:44:14] <Headphone> /dev/sda1 1 3824 30716248+ c W95 FAT32 (LBA) [16:44:14] <Headphone> /dev/sda2 3825 3837 104422+ 83 Linux [16:44:14] <Headphone> /dev/sda3 3838 10211 51199155 83 Linux [16:44:14] <Headphone> /dev/sda4 * 10212 15312 40973782+ bf Solaris [16:44:23] *** richlowe has joined #opensolaris [16:44:28] <trs81> or .. [16:44:31] <trs81> solaris uses disk slices [16:44:37] <Headphone> ok [16:44:38] <trs81> on x86. I remember. [16:44:44] <Headphone> x86_64, right [16:44:47] <trs81> which are subdivisions of a partition [16:45:16] <Headphone> this is reasonable, and fact is that dmesg | grep solaris [16:45:19] <Headphone> gives sda4: <solaris: [s0] sda5 [s1] sda6 [s2] sda7 [s3] sda8 [s5] sda9 [s7] sda10 > [16:45:34] <Headphone> so it reads the slices [16:45:40] <trs81> have you tried mounting sda6 to sda10? [16:46:06] <Headphone> yes [16:46:11] <Headphone> sda6 and sda10 [16:46:16] <Headphone> and the same error [16:46:35] <trs81> sda7? [16:46:44] <Headphone> is there a way to know if it is sunx86 or ufs2 or something elses? [16:47:00] <Headphone> maybne with the "file" command [16:47:05] <Headphone> sda7 in a moment [16:47:24] <Headphone> the same [16:47:25] *** alanc_away is now known as alanc [16:47:50] <Headphone> it takes time because I try also with old ufstype=ufs2 [16:48:02] <Headphone> and without ufstype at all , which is the odl [16:48:09] <myrkraverk> erm, can anyane guide me to some info on putting zfs on an usb disk? [16:48:10] <Headphone> the "old" I mean [16:48:10] <trs81> ok, do you have UFS in your kernel? [16:48:18] <Headphone> of course [16:48:31] <Headphone> lsmod |grep ufs [16:48:32] <Headphone> ufs 96841 0 [16:48:40] <elektronkind> solaris's ufs is a bit different from other ufs implementations [16:48:49] <elektronkind> the divergence happened quite a while ago [16:49:02] <Headphone> elektronkind, this works ok on a differnt macine [16:49:11] <trs81> Headphone: I'm out of ideas [16:49:12] <Headphone> with dual boot solaris/linux [16:49:17] <Headphone> trs81, thnks [16:49:31] *** bank__3 has joined #opensolaris [16:49:32] <Headphone> just one last question : is there a way to know [16:49:43] <Headphone> I am repeatin it : if the slice [16:50:09] *** Jiko_ has joined #opensolaris [16:50:09] <Headphone> is sunx86 or ufs2 or sun maybe or something else? [16:50:31] <Headphone> maybe with the file command ? [16:50:55] <elektronkind> are you talking about the logical partition on the disk itself? [16:50:57] *** pseudo__ has joined #opensolaris [16:51:02] <elektronkind> (sorry, coming into this convo late) [16:51:45] <Headphone> I am talking about the slices the I see when running [16:51:50] <Headphone> dmesg |grep sol [16:51:59] <Headphone> which correspond in this case to [16:52:07] <Headphone> /dev/sda5 - / dev/sda10 [16:52:17] <elektronkind> ah gotcha [16:52:37] <Headphone> beacuse I think that it is the old one [16:52:39] <trs81> is the other machine running fc6? [16:52:46] <Headphone> No, fc5 [16:52:55] <Headphone> and there it was ufstype=old [16:53:09] <Headphone> in fact I had mounted it without ufstype at all [16:53:26] <Headphone> and it takes the "old" as default [16:53:31] <Headphone> see man mount from linux [16:53:34] *** jonkelle has quit IRC [16:54:57] <Headphone> in fact the question is : is there a way to read a magic number of a file [16:55:03] *** jonkelle has joined #opensolaris [16:57:40] <trs81> there's an open bug in the redhat bugzilla [16:57:49] *** MattMan has quit IRC [16:57:56] <trs81> 209921 - fc6 prerelease [16:58:04] *** Jiko has quit IRC [16:58:13] <trs81> Headphone: are you using x86_64 or x86? [16:58:24] <jengelh> boyd : Something like cde-login or gdm is not listed svcs at all :( [16:58:43] *** pseudo_ has quit IRC [16:59:02] <jteo> jengelh: svcs -a |grep login [16:59:37] <jengelh> ah, -a did it, thanks! [16:59:48] <Headphone> x86_64 [17:00:03] <Headphone> trs81, on both machines [17:00:43] <trs81> so fc6 and solaris are both the x86_64 versions on both machines? odd. [17:01:01] <Headphone> NO [17:01:16] <jengelh> solaris 64-bit would be news [17:01:21] <Headphone> on the other one on which it works it is FC5 [17:01:22] <jteo> jengelh: solaris is 64bit. [17:01:40] <jengelh> well, the userland [17:01:51] <Headphone> trs81, he talks in that bug about ufs2 [17:02:32] *** bank__ has quit IRC [17:02:45] <jteo> i just realised my zpool is on a possibly faulty drive. and it's a single drive zpool. -sigh- [17:02:46] <trs81> Headphone: is there a fedora core 6 livecd? you could boot the 32bit version and try mounting it [17:03:13] <Headphone> I don't know; I have belenix on USB [17:03:26] <trs81> jengelh: sun cc defaults to 64bit on x86_64 iirc, the performance benefits of more registers outweighs the size increase [17:03:39] <Headphone> and suse live cd [17:03:48] <richlowe> trs81: I'm not certain that's the case. [17:04:01] <icon> hrmm i dont think it does [17:04:02] <richlowe> (or to be more explicit, I recall that not being the case, but I also think I maybe wrong :)) [17:04:12] <Headphone> richlowe, the question is : [17:04:22] <myrkraverk> is it possible to have more than one partition on an usb drive? [17:04:23] <Headphone> is it possible according to what is known [17:04:35] <Headphone> to first install soalris than linux and then mount [17:04:50] <Auralis> sure it is possible, usb is just the connecting technologue, like ide, scsi [17:05:26] <Headphone> a moment : the solaris from linux; this with installation of Grub on MBR [17:05:29] <Headphone> OR [17:05:37] <Headphone> must the GRUB be installed NOT [17:05:40] <Headphone> on the MBR? [17:05:56] <Headphone> or does it not make a differnece at all ? [17:06:16] <trs81> richlowe: ah, you're right [17:06:38] <myrkraverk> Auralis: so I just fdisk (or that's what I do in linux) the usb drive? (c3y0d0?) [17:06:52] *** bunker_work has joined #opensolaris [17:07:59] <myrkraverk> c3t0d0 I meant [17:10:14] *** Mazon is now known as mazon [17:11:25] <trs81> Headphone: I would download http://torrent.fedoraunity.org/unity/FC-6-i386-Server-LiveCD-Unity-20061024.torrent/view and see if you can mount the partitions [17:11:45] <trs81> if so, it's a 32bit/64bit problem introduced in fc6 and you should comment on that bug [17:13:36] <Headphone> trs81, ok , i will try it [17:13:55] <Headphone> any idea about the MBR question [17:14:34] *** drio has joined #opensolaris [17:14:58] <trs81> you don't want to install one grub over the other [17:15:49] <trs81> in your case it doesn't look like that's happened - solaris has installed its grub in the partition boot record, and is relying on the default mbr to see the boot flag on the partition and boot from there [17:16:00] <Headphone> trs81, do I understand right ? should I choose to install Grub NOT on the MBR? [17:16:20] <Headphone> when installing linux on a machine with solaris on it? [17:16:33] <trs81> the safest option is not on the mbr [17:16:40] <jengelh> grub in mbr is always good :> [17:16:59] <myrkraverk> erm, can I get a list of slices somehow? [17:16:59] <Headphone> trs81 , OK ! This is important lesson [17:17:18] <Headphone> trs81, so to install it let's say on / ? [17:17:25] <trs81> or /boot [17:17:30] <Headphone> ok [17:17:55] <Headphone> so this means that the older MBR is kept; so i understand [17:18:34] <Headphone> and the older MBR should recognize the new grub [17:19:35] *** rachel_ has quit IRC [17:21:17] <jengelh> hm the gdm2-login does not allow to change to console [17:23:02] *** jonkelle has quit IRC [17:25:05] *** dunc has quit IRC [17:25:18] <richlowe> is it just me, or is b.o.o responding real, real slow? [17:32:39] *** ProfMikey has joined #opensolaris [17:33:20] *** stevel has joined #opensolaris [17:33:20] *** ChanServ sets mode: +o stevel [17:35:07] *** dunc has joined #opensolaris [17:35:28] <elektronkind> richlowe: file a bugon oit [17:35:32] <elektronkind> s/oit/it [17:35:41] <myrkraverk> is it possible to rename a zfs pool? [17:36:01] <richlowe> elektronkind: hah. [17:36:19] <richlowe> stevel: is b.o.o choking somehow, or am I just real unlucky? [17:36:23] <richlowe> stevel: it's leading real, real slow. [17:36:25] <richlowe> 'loading' [17:36:55] <stevel> yeah - it's not loading for me either [17:36:57] <elektronkind> myrkraverk: yeah, you need to export the pool (zpool expost poolname) and then re-import it under a new name (zpool import poolname newpoolname) [17:37:04] <stevel> or it's loading really really slow rather [17:37:05] <elektronkind> s/expost/export [17:37:41] <myrkraverk> elektronkind: ok, thanks [17:38:34] <gdamore> hi ( [17:38:39] <gdamore> i meant hi * [17:38:52] <gdamore> is ffox any less a pig than 1.5? [17:39:02] <gdamore> s/ffox/& 2/ [17:39:07] <richlowe> stevel: want me to email website-discuss, or do you want to prod someone? [17:39:15] <elektronkind> const char *hi [17:39:15] * gdamore needs another cup of coffee. [17:39:16] <stevel> i prodded someone [17:39:44] <jteo> gdamore: it will never be as slim as what you want it to be. [17:40:08] <gdamore> yes, i know that. but is it _improved_? [17:40:46] <gdamore> (i.e. should i bother to install it? ;-) [17:40:56] <richlowe> isn't improved synonymous with scarily large? [17:41:00] <richlowe> see also vi v. vim ;) [17:42:35] <delewis> in the case of vim there's only 30,000 patches between each release. [17:42:58] * delewis wonders when the last time usr/src/cmd/vi was touched [17:43:19] * gdamore thinks ~never. [17:43:48] <delewis> I've seen a vi patch every now and then [17:43:55] <gdamore> solaris userland is widely cursed as being very stale and unmaintained/incompatible in circles outside of solaris. :-) [17:44:14] * twincest saw a twm fix go back into S10 a while ago [17:44:16] <richlowe> "incompatible" is a word with many meanings. [17:44:24] <richlowe> and so few people ever say the important bit. [17:44:26] <richlowe> ie, *with what* [17:44:34] <twincest> rich: posix *cough/bin/shcough* [17:44:39] <asyd> like "girs are incompatible with boys" [17:44:49] <delewis> don't forget grep :-) [17:44:54] <richlowe> twincest: sometimes they mean posix, sometimes they mean "What linux does", sometimes they mean "What I think should happen", sometimes they mean... [17:44:55] <Cyrille> really? I seem to recall a few adequate interfaces... [17:45:01] <delewis> Solaris grep does not support EREs in *any* case [17:45:08] <delewis> one must use egrep *or* /usr/xpg4/bin/grep with -E [17:45:15] <jteo> this sounds like a schily discussion. [17:45:20] <gdamore> one of the biggest complaints i've heard about is grep -r. [17:45:41] <twincest> i don't think i've ever used grep -E, i learnt it as egrep and fgrep [17:45:42] <richlowe> gdamore: there was a PSARC case moderately recently regarding that. [17:45:45] <richlowe> still no actual change though. [17:45:51] <richlowe> iirc, it was for grep -r/-A/-B [17:45:52] * delewis maks a motion to cp /usr/xpg4/bin/* /usr/bin [17:45:55] * delewis ducks [17:45:56] <gdamore> go hang out on the netbsd lists and say something about solaris. everyone agrees it has the best kernel around. but almost everyone there loathes its userland. and quite a few of them use it a lot [17:46:06] *** bunker_work has quit IRC [17:46:21] <gdamore> delewis++ [17:46:22] <delewis> gdamore: I loathe the userland, which is why I generally put /usr/xpg4/bin *before* /usr/bin [17:46:32] <gdamore> *precisely* [17:46:59] *** kloczek has quit IRC [17:47:17] <gdamore> my path is /usr/xpg6/bin:/usr/xpg4/bin:/usr/ccs/bin:/bin:/usr/bin:/usr/sbin:/sbin followed by X, dt, SUNWspro, and opt places [17:47:48] <gdamore> (the reason /bin is there with /usr/bin is that sometimes I log into NetBSD) [17:48:29] <richlowe> I had (and still to some degree have) intents on attempting to add the minor annoying missing bits to various userland things. [17:48:45] <richlowe> "attempting" because I can guess just how big an argument that will kick off. [17:48:48] <alanc> delewis: xpg4/bin is supposed to just be things incompatible with existing /usr/bin options - there's a bunch of bugs though for things that were added to xpg4 but not /usr/bin that aren't incompatible so should be in /usr/bin as well, like a bunch of grep flags [17:48:55] <gdamore> it would probably help solaris enormously if you did. a lot of people apparently hate those annoyances. [17:50:22] * gdamore thinks sun should EOF old POSIX-incompatible behaviors in /usr/bin, and move them to /usr/sun or /usr/old or somesuch. [17:50:44] <gdamore> which will not be in my path, just as /usr/ucb is not in my path [17:51:08] <alanc> easier said than fixing all the old scripts can be done [17:51:09] <richlowe> as alanc said, I think that merging the things that aren't incompatible would help a great deal. [17:51:19] <gdamore> it probably would. [17:51:40] <richlowe> after that, 50 people all fixing their favourite userland nit would leave you almost done :) [17:51:52] <alanc> so go out fix them and request sponsors... [17:52:01] <richlowe> alanc: scroll up a little. [17:52:03] <gdamore> *basename* is incompatible?!? bizarre [17:52:28] <alanc> adding missing XPG4 flags to /usr/bin commands should be automatic approved ARC cases for most of them [17:52:31] <gdamore> alanc: i have enough pending sponsor requests for now. :-) [17:53:51] <gdamore> id being a good example. [17:54:38] <alanc> ooh, next week's PSARC agenda should make some people happy...Inception: CIFS Client on Solaris (2005/695) [17:54:44] *** bougie has quit IRC [17:54:58] <alanc> and I guess I need to show up in a couple weeks for Inception: Virtual Console (2006/591) [17:55:11] *** stevel has quit IRC [17:55:45] <elektronkind> I suppose the CIFS client will see gnome-vfs integration? [17:56:00] <gdamore> maybe i should request-sponsor for the changes for /usr/xpg4/bin/id, just as an example. [17:57:06] <alanc> 4521640: *id* /usr/bin/id missing options from /usr/xpg4/bin/id [17:57:34] <gdamore> waiting for bugs.o.o [17:57:38] *** bougie has joined #opensolaris [17:57:49] <jteo> gdamore: everyone is. [17:57:59] <gdamore> is it down? [17:58:01] <alanc> there's a missing-std-feature keyword for bugs in which /usr/bin tools are missing compatible XPG4 options - don't remember if b.o.o lets you search by keyword though [17:59:01] <gdamore> after spending time with our synaptics code yesterday, i was somewhat disappointed to discover it is going to require some surgery to make it work with regular PCs. [17:59:28] *** jamesd has joined #opensolaris [17:59:31] <jteo> gdamore: that's a time constraint issue for you? [17:59:36] *** ChanServ sets mode: +o jamesd [17:59:43] <richlowe> gdamore: kinda. [17:59:46] <richlowe> gdamore: it responds eventually. [17:59:53] <richlowe> gdamore: go get coffee, it'll be done by the time you're back maybe. [17:59:58] <richlowe> ... unless they took it down to fix it [18:00:15] <gdamore> no, it came up now. [18:00:51] <gdamore> jteo: it might be. the problem is that our firmware "annotates" the protocol and participates in ps2 synaptics initialization in funny ways [18:01:00] <gdamore> mostly in an attempt to support both internal and external mice [18:01:29] <alanc> 4843344 *grep* should have -q, -e, -f, -x, -E, and -F options [18:04:16] <gdamore> i just submitted the sponsor request for the /usr/bin/id stuff. my good deed for the day. [18:05:20] <jteo> gdamore: you work fast. [18:05:42] <jteo> gdamore: are the docs for synaptics touchpads open? [18:05:44] <gdamore> jteo: heh. your turn. you can do 4843344. :-) [18:05:49] <gdamore> jteo: yes. [18:06:00] <jteo> ah. [18:06:17] <gdamore> synaptics are very developer friendly, from what i can see. [18:06:45] <gdamore> their docs include compatibility issue details, and details for all of the previous models as well as their latest ones [18:06:58] <gdamore> (and how to tell the difference between models) [18:10:37] <jteo> dumb qns: not all touchpads on them lappies are synaptic are they? [18:10:56] <gdamore> no, there is an alps model as well, but i am pretty sure those are mostly usb. [18:11:29] <alanc> 6231496 *tail* non-XPG4 tail could support -n and possibly -c [18:11:57] <richlowe> alanc: if you're searching for these, could you tag them oss-bite-size while there? [18:12:00] <alanc> 6415421 *sh* legacy ^ operator causes interoperability problems [18:12:02] <richlowe> (put the tag in Description as well as keywords) [18:12:08] <richlowe> since b.o.o doesn't search keywords. [18:12:15] <richlowe> because, well, that would make sense, and b.o.o doesn't. [18:12:45] <gdamore> i would love to see /usr/bin/sh become posix compatible. so many problems would just "go away" if that were true [18:13:00] <richlowe> talk with nrubsig about it, he shares that sentiment. [18:13:29] <gdamore> no, nrubsig wants ksh93 to be /usr/bin/sh. i don't care if we just copied /usr/xpg4/bin/sh to /usr/bin. [18:13:46] <richlowe> /usr/xpg4/bin/sh is ksh88 [18:13:51] <richlowe> I believe he intends to make it ksh93 [18:13:51] <alanc> richlowe: sure [18:13:58] <gdamore> yes, i know that. [18:13:59] <richlowe> ... hell, there was a period he wanted to make zonecfg ksh93, so... [18:14:29] <gdamore> frankly, i'm not entirely clear that ksh93 is so much vastly superior to ksh88. [18:14:36] *** dunc_ has joined #opensolaris [18:14:37] * gdamore hides from nrubsig [18:14:43] <hile_> bwhaha [18:15:13] <alanc> ksh93 is clearly 5 ksh'es better than ksh88 - it's basic math! [18:15:16] *** Darwin has joined #opensolaris [18:15:23] <Darwin> hi [18:15:42] *** bougie has quit IRC [18:17:05] <gdamore> haha. i mean, unless you want to write _programs_ in ksh, i doubt anyone would notice much difference. [18:17:33] <gdamore> the rest of us, who merely write _scripts_ in ksh, are happy with ksh88, i 'spect [18:17:44] *** Jiko has joined #opensolaris [18:18:36] *** nwf has joined #opensolaris [18:20:42] <gdamore> /bin/sh however is a PITA, particularly when dealing with configure scripts that hardcode /bin/sh, and assume basic POSIX capabilities like $(), [[]], and (( )) [18:21:03] <gdamore> cross-building NetBSD is one such case, btw. :-) [18:21:24] *** dunc has quit IRC [18:22:27] <alanc> 6398988 /bin/sh should support $(), just like POSIX [18:25:00] <alanc> heh - that should confuse people: -r-xr-xr-x 66 root bin 7628 Jul 13 10:14 /usr/X11/bin/Xorg* [18:25:09] <alanc> the 7kb Xserver! [18:25:56] <hile_> wrapper script? [18:26:04] <alanc> at least until they realize it's just another link to isaexec, and the real X servers are: [18:26:43] *** bank__3 has quit IRC [18:26:52] <richlowe> gdamore: did you get a response about the pgx32 stuff yet? :) [18:27:45] <gdamore> if you mean, did i get a fix for the dtgreet problem, no. [18:27:54] <gdamore> if you mean did i get permission to follow up with ATI/3Dlabs, no. [18:27:57] <richlowe> no, I meant the kernel fb driver ranting. :) [18:28:11] *** Jiko_ has quit IRC [18:30:05] <Stric> gdamore: I just run gnu configure scripts through bash instead, it's like 2-3 times faster than solaris /bin/sh.. [18:30:16] <Stric> ksh is way faster too [18:32:54] <Stric> mostly due to built-in [ I guess.. or something.. [18:33:24] <gdamore> exactly, builtins are a Good Thing. at least using them for numeric expressions and substring parsing. [18:33:44] *** mkhl has joined #opensolaris [18:34:21] <gdamore> way back when, i cut ~12 hours off of a 24 hour patch run (Solaris 2.5.1) by converting just _one_ pipeline to use a few builtins instead of a pipe thru sed | grep to do substring extraction. [18:34:31] <gdamore> (this was changes to patchadd) [18:34:47] *** loke has quit IRC [18:34:48] <gdamore> of course sun didn't integrate _that_ change until S8. [18:38:52] *** dunc has joined #opensolaris [18:43:10] *** Cyrille has quit IRC [18:45:19] *** salamanders has quit IRC [18:51:34] *** Headphone has quit IRC [18:53:07] *** alanc is now known as alanc_away [18:53:18] *** stevel has joined #opensolaris [18:53:19] *** ChanServ sets mode: +o stevel [18:54:00] *** calumb has quit IRC [18:57:42] <stevel> richlowe: b.o.o. is down due a storm [18:57:43] <stevel> :-P [18:58:34] <richlowe> eh? [18:59:12] *** rfds has joined #opensolaris [19:00:03] <stevel> storm in denver is apparently causing issues with power at our broomfield campus [19:00:07] <stevel> where b.o.o. is housed [19:00:12] <gdamore> ah. [19:00:25] <richlowe> Oh, you meant it literally. [19:00:29] <gdamore> wow. i'm having fun looking at the bullfrog power management/i8042 code. [19:00:31] <richlowe> the :-P threw me off. [19:00:51] <gdamore> apparently it doesn't have a real i8042, but only _emulates_ one with a microcontroller. weird. [19:00:58] <stevel> i thought it was a funny reason :) [19:01:12] <richlowe> stevel: You're just happy it wasn't you this time. [19:01:17] <stevel> damn straight [19:02:16] <postwait> Anyone know if adding 119 to the PARAM_REQUEST_LIST in dhcpagent will make Solaris set "search" correctly in /etc/resolv.conf? [19:05:19] <oxygene> gdamore: I don't think real i8042 are still deployed today [19:06:29] <gdamore> this is quite amusing. [19:07:04] <gdamore> reading code in C to emulate i8042. that code has bugs, and our vuidsyn module is bug-compatible. [19:12:47] *** hile_ has quit IRC [19:13:01] *** damienc has quit IRC [19:13:20] *** Kernel86 has quit IRC [19:13:35] *** bondolo has quit IRC [19:19:57] *** alanc-away is now known as alanc [19:22:51] *** hile_ has joined #opensolaris [19:23:41] *** mdj has joined #opensolaris [19:25:07] *** gnu2it2 has quit IRC [19:26:24] *** Wez has joined #opensolaris [19:30:31] *** Kernel86 has joined #OpenSolaris [19:38:47] *** slowhog has joined #opensolaris [19:42:25] *** paxc has joined #opensolaris [19:43:38] *** deather_ has joined #opensolaris [19:43:46] *** dunc_ has quit IRC [19:44:26] <jamesd> can some sun guys tell sun maketing guys that 501.x.x.x is not a valid ip address... [19:44:54] <paxc> its not and IP [19:44:55] *** stevel changes topic to "bugs.opensolaris.org currently experiencing technical difficulties | SXCR: 50, ON build: 51, ON nightly: 20061023" [19:44:58] <richlowe> I'd assume it's being used like a 555 phone number. [19:45:15] <richlowe> stevel: you may want to adjust that. [19:45:22] <richlowe> stevel: perhaps "more than normal..." added in somewhere. [19:45:41] <jamesd> http://www.sun.com/2006-1017/feature/index.jsp the chapter 2 of there sun virtualizion videos... use a bunch of bogus ip#'s [19:45:49] *** stevel changes topic to "bugs.opensolaris.org currently experiencing technical difficulties (more so than usual) | SXCR: 50, ON build: 51, ON nightly: 20061023" [19:46:35] <elektronkind> ok, who decided to run b.o.o off of sunsolve [19:46:38] <elektronkind> ;) [19:46:47] <stevel> i blame richlowe [19:46:55] <jteo> i blame stevel. [19:47:01] <stevel> he's been talking smack about b.o.o. for ages now [19:47:09] <richlowe> *I* have? [19:47:16] <elektronkind> <b.o.o> screw you all, I'm going home [19:47:27] <richlowe> I've talked no more smack than anyone else :) [19:47:42] <mrdeviant> what is b.o.o ? [19:47:49] <elektronkind> bugs.opensolaris.org [19:47:50] <mrdeviant> bugs.openoffice.org ? [19:47:51] <mrdeviant> ah [19:48:04] <stevel> jteo: while i (unfortunately) have the ability to take down almost all of opensolaris.org spectacularly - bugs.opensolaris.org is the one piece i can't touch :) [19:48:15] <dwc-> file a bug on the bugtracker [19:48:34] <richlowe> stevel: I'd rather you could. [19:48:37] <richlowe> stevel: you fix things. [19:48:45] <stevel> richlowe: only because i break them in the first place ;-) [19:48:49] <richlowe> crashing twice a week, but not sucking in between would still be improvement :) [19:48:53] <jteo> that's besides the point. [19:48:55] <richlowe> ... oh, you mean *that* smack talking. [19:49:00] <richlowe> Ok, then, yeah, I'll take the blame. [19:49:06] <stevel> :-D [19:51:44] *** dunc has quit IRC [19:56:25] *** mazon is now known as Mazon [19:57:37] <richlowe> Hm, crossbow thingy at SVOSUG [19:58:24] *** Teknix has joined #opensolaris [19:58:30] <stevel> didn't we just have crossbow last month? [19:58:33] <stevel> i thought tonight was IP instances [19:59:03] <richlowe> Oh, hah, yeah. It helps if my fingers and my brain line up when I'm typing, huh? [19:59:45] *** deather has quit IRC [20:08:33] <Teknix> so what does it mean when the pool and raidz is showing CKSUM errors, but none of the disks in the pool are? [20:10:09] *** nachox has quit IRC [20:24:28] *** jamesd has quit IRC [20:34:44] *** bengtf has quit IRC [20:38:12] *** bengtf has joined #opensolaris [20:38:29] *** nwf has quit IRC [20:40:02] *** nwf has joined #opensolaris [20:42:46] *** mdj has quit IRC [20:42:53] *** cub has joined #opensolaris [20:45:05] <cub> anybody know the terms for different FC cable connectors? Sun ships us a setup including SAN box and FC cables....however the FC card on the Sun box ....the holes are way apart that the regular FC cable can't plug in. I wanna call them that they ship me the wrong card but don't know the name [20:45:09] *** opSuse has joined #opensolaris [20:49:25] *** AbeFroman has quit IRC [20:49:28] *** AbeFroman has joined #opensolaris [20:50:12] <Stric> cub: sun shipped us a disk array with fibre fcal and a machine with copper fcal.. didn't work out too great either. [20:50:30] *** AbeFroman has quit IRC [20:51:15] *** AbeFroman has joined #opensolaris [20:51:33] <gdamore> anyone got any suggestions for low end VIA boards to run SNV on as a NAS server? [20:53:14] <mrdeviant> i'm running b50 on a en15000g. i'm not using it for NAS, though. [20:53:41] <lasseoe> cub, you have LC and SC connectors, LC is small and SC is bigger and bulkier [20:53:42] <mrdeviant> you have to use a 3rd-party driver for the on-board nic, though. [20:54:06] <gdamore> that's okay, i suppose. i want to run zfs and nfs on it. [20:54:14] <gdamore> does it support onboard sata? [20:54:37] *** nwf has quit IRC [20:54:38] <mrdeviant> yea the onboard sata works [20:54:54] <gdamore> cool. [20:54:54] <mrdeviant> well, hm. let me retract that. [20:54:58] <gdamore> uh oh. [20:55:01] <mrdeviant> it has both sata and pata connectors. i think this box is using pata [20:55:14] <mrdeviant> i'd have to pop the case to see how it's setup. so i can't say for sure. [20:55:24] <gdamore> ah. so i could use pata for the OS, and sata for the zfs store? [20:57:32] <mrdeviant> in theory, but like i said, i don't know if i've ever used the sata on this board. [21:01:03] *** Mazon is now known as mazon [21:01:09] <trs81> mmm, nothing like hitting internal compiler errors in sun studio's cc [21:04:11] *** Cybernd has joined #opensolaris [21:06:51] <myrkraverk> trs81: better than finding you wife in bed with another woman? [21:07:02] <myrkraverk> (oh, sorry, this isn't #solaris) [21:07:05] <mrdeviant> gdamore, i just popped the case. i'm using pata. [21:07:58] <trs81> myrkraverk: no, not really. [21:08:17] *** nwf has joined #opensolaris [21:09:12] <myrkraverk> trs81: ;) [21:13:45] *** tsoome has joined #opensolaris [21:15:21] <gdamore> toobad. [21:15:43] * myrkraverk tries to ask again: [21:16:02] <myrkraverk> e1000g: [ID 80193 kern.warning] WARNING: pci8086,1000 - e1000g[0] : Could not identify hardware <---- anyone here who knows what that means? [21:16:33] <gdamore> how well does ZFS deal with adding new storage? does it immediately start taking advantage of the new spindle? does it change redundancy? any one really know? [21:18:41] <gdamore> myrkraverk: the pci id isn't known to the driver? [21:19:19] <myrkraverk> gdamore: it's aliased in the driver_aliases file (that's all I know) [21:19:22] <twincest> if the id isn't known why did it try to attach the driver? [21:20:01] <gdamore> it might also need to be added to the .c. The aliases file tells the system to use that driver, but it does nothing to help the driver itself recognize it. [21:20:21] <myrkraverk> twincest: that isn't the question; the question is: why doesn't it identify the hw - it's aliased for it [21:20:32] *** kgrimm has joined #opensolaris [21:20:35] <myrkraverk> gdamore: hmm, k [21:20:42] <gdamore> Usually the driver needs to know about special hardware in its .c because it needs to be able to handle quirks, etc. [21:20:50] <twincest> myrk: uhm, maybe it's not your question.. [21:21:10] <myrkraverk> twincest: what do I know? I'm just a wannabe mathematician ;P [21:21:20] <myrkraverk> gdamore: hmm, ok [21:21:21] <gdamore> also will upgrading Solaris properly preserve zfs/raidz? [21:22:03] <opSuse> question about belenix live cd and partitions: if I boot a solaris machine with belenix (On a USB) will I see by prtvtoc (or somehow else) the soalris partitions ? can I mount them ? [21:22:30] <gdamore> i would think so. its the same kernel. [21:22:45] <gdamore> is prtvtoc the same program as well? or does it come from linuxland [21:23:09] <twincest> if linux has prtvtoc, they took it from solaris [21:23:15] <myrkraverk> gdamore: any chance for me to check the e1000g driver? do you know where I can find the source? [21:23:22] <opSuse> I think it is not from linux; linux has fdisk [21:23:52] <opSuse> myrkraverk, it is closed as far as I know [21:24:10] <myrkraverk> opSuse: damn [21:24:21] <opSuse> myrkraverk, I agree [21:24:38] <opSuse> myrkraverk, it is not the only nic driver which is closed [21:24:53] <opSuse> also rtl 8139 and many others [21:25:13] <myrkraverk> opSuse: ok, then I'll need to file a bug, it seems ;/ [21:25:15] <opSuse> i don't understand why; because intel linux drivers are open [21:25:16] <trs81> gdamore: zfs makes the new storage available immediately, but it doesn't transfer any data to it until something writes it [21:25:26] <gdamore> ah. [21:25:28] <gdamore> tks. [21:25:53] * gdamore is suddenly trying to spin up on ZFS. [21:26:12] <richlowe> opSuse: the why would probably be "Because they don't have permission to open it" [21:26:13] <trs81> gdamore: redundancy is dealt with per vdev (I think that's the name), which is a group of disks [21:26:25] <richlowe> why they don't have permission is a matter for them, and whoever would have to give permission. [21:26:53] <richlowe> I'm far more interested in why some of the (iirc sun designed) sparc nic drivers are closed, beyond them not being in ON. [21:26:56] * richlowe points at Cassini [21:26:56] <trs81> gdamore: so you add a group of disks in raidz or raid 1. mixing raid levels is not supported, but does work [21:26:59] <opSuse> richlowe, who should give permission to whom ????????? [21:27:15] <twincest> opsuse: the owner of the driver needs to give permission to sun, or open the driver themselves [21:27:26] <twincest> (or the owner of the proprietary IP used in the driver) [21:27:27] <mrdeviant> that, and the fact that cassini hasn't been ported to nemo [21:27:29] <richlowe> and "Linux's driver is open" does not necessarily imply anyone else has that permission. [21:27:40] <opSuse> twincest, intel opens their nic drivers [21:27:49] <opSuse> rtl opens their nic drivers [21:27:50] <twincest> their solaris drivers? [21:27:55] <twincest> realtek solaris driver was never open source [21:27:57] <opSuse> linu [21:27:58] <opSuse> linux [21:28:03] <twincest> not even realtek's own one [21:28:17] <tsoome> open for linux doesn't make them open for others.... [21:28:18] <opSuse> linux nic drivers [21:28:22] <trs81> gdamore: upgrades work fine, although zfs on root might need handholding, but zfs on root isn't recommended for the non-adventurous [21:28:26] <richlowe> and iirc, linux's realtek driver is not from realtek. [21:28:30] <richlowe> whereas Sun's might be. [21:28:37] <tsoome> the word "open" is most widely misused one [21:28:39] <gdamore> i'll keep zfs "not on root" for now. [21:28:56] <richlowe> (and having read the comments in wpaul's realtek driver, one could understand why they'd want to hide that pile as best as they can) :) [21:28:57] <opSuse> richlowe, now i remeber - you are right about realtek] [21:29:06] <gdamore> are we talking about the realtek gige driver, or the 10/100 driver? [21:29:15] <twincest> i'm talking about the 8139 & family [21:29:15] <richlowe> gdamore: rtls is the 8139 driver. [21:29:30] <gdamore> the 8139 is a PoS. you wouldn't want it in your machine if you choice. [21:29:37] <richlowe> rge is open. [21:29:39] <richlowe> rtls is not. [21:29:42] <twincest> so? they work fine for me, and lots of other home users [21:29:43] <gdamore> right. [21:29:58] <gdamore> have you looked at the code in the linux driver? it will explain a lot [21:30:04] <opSuse> who is the manufacture of rge , may I ask? [21:30:07] <twincest> explain what, why it works fine? [21:30:16] <twincest> i'm aware of the 8139's shortcomings [21:30:18] <richlowe> opSuse: realtek, it's the realtek GigE driver. [21:30:21] <gdamore> (i.e. there are some horrible performance-limiting bugs that have nasty software workarounds) [21:30:24] <opSuse> richlowe, thnks [21:30:48] <richlowe> well, it's realtek hardware, I don't know who did the driver. [21:30:57] <gdamore> right, the problem is the hardware, not the driver. [21:31:08] <sickness> uhm [21:31:21] * richlowe has been assuming that many of the NIC drivers are vendor-written [21:31:28] <gdamore> the _driver_ works, or at least as well as can be expected with drainbamaged 8139 hw [21:31:28] <richlowe> though of course, nobody is actually allowed to confirm nor deny that. [21:31:37] <opSuse> just little remark: I remember that I saw 8139 drivers for solaris on realtek web site ; closed [21:31:38] <gdamore> many aren't. [21:31:57] <gdamore> i _think_ rtls is realtek written though. [21:32:12] <twincest> ISTR a message from the rtls driver in S10 that was identical to realtek's driver [21:32:18] <twincest> so it seems likely that it is [21:32:19] <opSuse> gdamore, I also had that impression [21:32:24] <gdamore> what's funny, is that IIRC the entire 8139 design is open. so its not like they're hiding any proprietary info. [21:32:41] <richlowe> gdamore: I was specifically referring to closed nic drivers that are not locked up with the sparc hardware stuff. [21:32:42] <twincest> gda: well, most companies don't open source by default [21:32:45] * richlowe mutters about cassini again. [21:32:56] <mrdeviant> on a related note, it's ironic that the ncp driver (the crypto driver for niagara) is closed, but the niagara rtl is open [21:32:57] <richlowe> ce(7D)'s reputation for horribleness makes me really want to see just how bad it is :) [21:33:07] <gdamore> cassini is ugly. you don't want to even look at that driver. :-) [21:33:32] <gdamore> i've seen it. trust me. you don't want to see it. [21:34:07] *** laca has quit IRC [21:34:08] <gdamore> first read (and fully understand) all of bge. if you had _fun_, then come back and ask for ce source. [21:34:48] *** kgrimm has quit IRC [21:35:29] *** gnu2it2 has joined #opensolaris [21:37:24] * gdamore wishes there was a ZFS best practices guide for systems with 2, 3, or 4 drives. [21:38:02] *** jamesd has joined #opensolaris [21:38:03] *** ChanServ sets mode: +o jamesd [21:39:02] <richlowe> gdamore: The closed nic drivers listed as not having plans to release as source are e1000g, elxl, iprb, ixgb, pcelx, pcn and rtls (though that page only counts stuff in ON) [21:39:30] <richlowe> intel, 3com, intel, intel, 3com, AMD, realtek [21:40:06] <gdamore> what about nge? [21:40:21] <richlowe> oh, yeah, I missed that, one. [21:40:39] <gdamore> now i've read pcelx and pcn, and I'd practically be willing to work on a from-scratch implementation of those, if it would get them opensourced. [21:40:41] <richlowe> spwr, too, I didn't realize that even existed. [21:40:42] * myrkraverk has filed the bug [21:41:13] <gdamore> nge isn't part of ON, actually, IIRC. [21:41:15] <opSuse> which bug? [21:41:20] <opSuse> id? [21:41:21] <richlowe> I'm not sure how accurate or up to date that list is kept though (http://www.opensolaris.org/os/about/no_source/) [21:41:29] <twincest> rich: doesn't that cover nearly every common 10/100 card on intel? :) [21:41:45] <richlowe> twincest: nah, bge's are pretty damn common. [21:42:07] <richlowe> the cheaper end mostly seems to be rtls or afe. [21:42:14] <richlowe> either that or I just get unlucky when buying cheap nics :) [21:42:58] <twincest> bge's GE, isn't it? [21:43:06] <myrkraverk> opSuse: I'm going to post the id as soon as I find it, the bug database seems slow to me ;/ [21:43:06] <mrdeviant> yes [21:43:07] <twincest> or is there an FE model too? [21:43:12] <myrkraverk> opSuse: it's my e1000g bug [21:43:18] <opSuse> myrkraverk, ok... [21:43:41] <opSuse> myrkraverk, what you mean by it ? (?"my?) [21:43:44] <myrkraverk> opSuse: in the mean time, I'm going to give the em driver a try [21:43:51] <myrkraverk> opSuse: I found it ;P [21:44:09] <myrkraverk> opSuse: it's not so bad as if I put it there ;) [21:44:49] <opSuse> what is em ? is it open source ? [21:45:12] <myrkraverk> opSuse: yes, an "open source alternative to e1000g" [21:45:27] <opSuse> BTW, I got a panic in e1000 tx thread because of b_next=0 [21:45:35] <richlowe> twincest: yeah, it's GE, I'm just an idiot. [21:45:37] <opSuse> But it could be an error which I had made [21:45:46] <twincest> actually most of the ones you listed are GE too, so ignore me [21:45:57] <twincest> but still, most of those are quite major :) [21:46:15] <richlowe> petition gdamore to replace the ones that murayama hasn't :) [21:46:19] <richlowe> they seem to like NIC drivers. [21:47:06] <richlowe> and there's jteo's elxl replacement too, I wonder what happened to that. [21:47:21] *** jamesd has quit IRC [21:47:38] <myrkraverk> opSuse: if you find the one who maintains the bug database, ask why I can't just look up the bugs I've submitted (or if so, where I find the link) [21:48:03] <opSuse> it is a BUG!!!!!!!!!! [21:48:05] <richlowe> myrkraverk: b.o.o is having issues (see the topic) [21:48:15] <richlowe> you can't find bugs you filed however, because it sucks. [21:48:25] <richlowe> you won't get the ID until it the notification gets forwarded to you. [21:48:27] <richlowe> again, because it sucks. [21:48:28] <myrkraverk> richlowe: ah [21:48:44] <opSuse> Threre should be a bugzila for the bug db and management ! [21:48:51] <gdamore> i think the bug submission thing goes to a different database, where it gets triaged, before it gets entered into the "real" databse. [21:48:52] <richlowe> I think they keep it around to make jive look good ;) [21:48:58] <richlowe> gdamore: not quite. [21:49:01] <myrkraverk> richlowe: so, should I hire a dominatrix to whip the webdesigner? [21:49:09] <richlowe> gdamore: bugs-by-mail to bt2, but in the opensolaris/triage-queue cat/subcat [21:49:12] <gdamore> really, it seems like CR numbers are different than bug numbers [21:49:28] <richlowe> gdamore: then it gets filed 'properly', but it's still valid and visible if there. [21:49:33] <richlowe> but b.o.o only syncs every 24 hours-ish. [21:49:36] <richlowe> the whole system is moronic. [21:49:41] <gdamore> heh. [21:50:00] <alanc> opSuse: the problem with using bugzilla is the hundreds of thousands of existing bugs in Sun's existing bug db and all the tools built to interface with that db [21:50:15] * gdamore wonders if folks are still using my bug history reporting tools that query the sybase database directly [21:50:35] <richlowe> alanc: you're going to find ever increasing pressure for an alternative until someone gets off their ass and starts making changes for the better. [21:50:38] <myrkraverk> alanc: plus, bugzilla isn't the very best I can think of either [21:50:38] <alanc> doubtful, since the bug db isn't in sybase any more [21:51:05] <gdamore> is it really bugtraq2 now? it was still bugtraq+ when i was around. bt2 was just an idle threat in my time. [21:51:14] <alanc> we did do this exercise 2-3 years ago of switching the bug db - it was a painful, multiyear process we don't want to repeat [21:51:20] <myrkraverk> richlowe: I've volunteered to hire a dominatrix to help with that effort ;) [21:51:37] <alanc> yes - the project originally started to make bugtraq+ Y2k compliant finally got finished around 2004 [21:51:44] <gdamore> dominatrices are cheap. engineers, otoh, are expensive. [21:52:01] <richlowe> precisely, and the more time they have to spend dealing with b.o.o... [21:52:12] <gdamore> or did i get that backwards? hmmm..... [21:52:24] <richlowe> I still think a group of friendly Sun folk should do all their work via our tools, and add their voices to the complaints :) [21:52:36] <myrkraverk> gdamore: maybe - I don't know the rates in the us [21:52:41] <alanc> except that the engineers don't deal with b.o.o - we use the real tools so we can get our jobs done [21:52:58] <alanc> b.o.o is a horrible front end to the bug db [21:53:06] <richlowe> alanc: kinda my point, nobody who 'matters' is having to deal with the suck. [21:53:07] * gdamore wonders if the real tools are any better than bt+ was. [21:53:08] <richlowe> alanc: so nothing changes. [21:53:15] <richlowe> alanc: maybe someone who matters having to deal with it would change things. [21:53:48] <alanc> b.o.o won't let us edit existing bugs so we can't use it [21:53:58] <richlowe> more suck, more need for change :) [21:54:03] <gdamore> clearly this should be a pre-requisite to opening up write access to the svn/hg repos [21:54:15] <alanc> the client is at least not a sparc-only Motif binary any more [21:54:39] <richlowe> alanc: The reason you can't use it, as you said, is because it's unusable. [21:54:43] <gdamore> haha. the old client was a p-o-s. i wrote a bunch of perl scripts to do other things talking to the db directly. [21:54:44] <richlowe> ... so why should we again? :) [21:55:14] <alanc> it's java webstart stuff, so you can go to the URL from any machine on the intranet/vpn and get a local client running - SPARC, x86, Mac - even windows [21:55:41] <alanc> much nicer for work-at-home than remote X display over VPN [21:55:47] <gdamore> alanc: it seems like they need to make that work with an ssl-protected gateway. [21:55:56] <gdamore> then they could hand out access to external users. [21:56:12] <gdamore> (without letting them on SWAN) [21:56:21] <alanc> they'ld have to find a better way of separating the confidential vs. public information first [21:56:24] *** opSuse has quit IRC [21:56:25] <richlowe> gdamore: they can't let us see large amounts of the data. [21:56:46] <richlowe> don't know what the reason for not letting us edit our own data is. [21:56:50] *** _william_ has joined #opensolaris [21:56:58] <alanc> (which is one of the reasons b.o.o sucks, because they use such a bad way of doing that now that most information isn't shown since some of it may not be publishable for some bugs) [21:57:21] <gdamore> if access was authenticated, they could fix that. let you see only the data you're entitled to. [21:57:27] <_william_> hi all [21:57:32] <alanc> richlowe: first they'd have to figure out what data you provided [21:57:34] <richlowe> There's 3 interfaces to the system, depending on where you are. Sun engineers use bugster, tools use monaco, those of us outside use Sun engineers. There's no other way to go about stuff :) [21:58:09] *** sahafeez has joined #opensolaris [21:58:27] <gdamore> alanc: the database (if it is a real database) should know that. but data submitted via b.o.o probably has no real submitter id on it [21:58:48] <_william_> i have a compilation question under solaris with sunstudio... maybe some one can help me please ? i compile some sources (from mozilla) and have this warning cc: Warning: -xarch=native has been explicitly specified, or implicitly specified by a macro option, -xarch=native on this architecture implies -xarch=v8plusb which generates code that does not run on pre UltraSPARC III processors [21:58:50] <richlowe> gdamore: "Opensolaris" as the company, I think it's Linda Bernal as the human submitter. [21:58:51] <myrkraverk> erm, I shut off vold earlier - how do I mount a CD now? [21:58:58] <richlowe> the user filing the CR is in hook6 [21:59:01] <alanc> all data from b.o.o currently comes in under the userid of the PM who runs the interface [21:59:05] <richlowe> (email address from os.o account) [21:59:06] <_william_> anyone has some doc ref about this ? any clues of whats wrong ? [21:59:37] <gdamore> william: don't use -xnative if you want to run on older systems. its pretty much in the standard studio docs [21:59:44] *** nwf has quit IRC [22:00:00] <_william_> gdamore, yes :) but... i am not using -xarch=native [22:00:03] <gdamore> if you only want to run on the system you're building for, the warning is harmless and you can ignore it [22:00:10] <gdamore> did you specify -fast? [22:00:11] <_william_> this flag use the processor type of the build system [22:00:14] <alanc> _william_: you told it to generate the fastest possible code for the machine you're running the compiler on, so it's just warning you that it's following instructions and making code that won't run on older machines [22:00:15] *** jamesd has joined #opensolaris [22:00:16] <_william_> i did [22:00:17] <sahafeez> were are at the change logs keep for lets say if i wanted to see the difference between 49 and 50 and did not want to dl the iso [22:00:18] *** ChanServ sets mode: +o jamesd [22:00:24] <_william_> okayyyy :) [22:00:35] <_william_> thanks i understand the link between fast and native [22:00:39] <richlowe> sahafeez: dlc.sun.com/osol/on/downloads/bxx [22:00:40] <_william_> i'm stupid :) [22:00:45] <alanc> sahafeez: only the ON & X changelogs are published so far - see the onnv & X communities [22:00:47] <_william_> thanks guys [22:00:52] <gdamore> np [22:00:54] <richlowe> s/xx/50/ for the b50 putback log, say. [22:01:09] <richlowe> in the download directory you have the full htmlized logs, and a text file just listing CRs and ARCs. [22:01:12] <myrkraverk> anyone ? I need to mount this CD to test the em driver ;P [22:01:31] <richlowe> alanc: JDS is kinda published by virtue of their svn history. [22:01:39] <myrkraverk> (I don't even remember what I did to turn vold off) [22:03:32] <sahafeez> thanks! [22:04:23] <sahafeez> i use sxcr as my desktop i am trying to figure out if it is worth upgrading all the time [22:05:23] <sahafeez> as far as my upgrade choices, live is out as i have /usr /var, etc... [22:05:33] <richlowe> that's not a problem for liveupgrade. [22:05:39] <richlowe> as long as you have duplicates of each of them. [22:05:46] <richlowe> or your second BE's / slice is big enough to hold all 3. [22:06:16] <sahafeez> my second slice is not. [22:06:31] <sahafeez> and it is a silly thing. [22:06:42] <sahafeez> can i jumpstart and upgrade? [22:06:48] <sahafeez> s/and/an [22:07:11] <richlowe> Yes. [22:08:57] <sahafeez> cool. i have an openbsd box here i am going to try to setup as a jumpstart box. i know that it has been done on linux and free so... [22:09:00] *** yarihm has joined #OpenSolaris [22:09:58] <sahafeez> what should i use to burn an dvd under solairs? [22:10:11] <jamesd> cdrecord [22:10:30] <sahafeez> thanks. now i just need to find a scsi dvd burner on ebay [22:10:31] <elektronkind> or cdrw [22:10:35] *** mrdeviant has left #opensolaris [22:10:40] *** mrdeviant has joined #opensolaris [22:10:43] <sahafeez> growiosfs? [22:11:39] *** yarihm has quit IRC [22:12:17] <Gman> morning dudes [22:12:27] <elektronkind> howdy [22:13:15] *** yarihm has joined #OpenSolaris [22:13:31] *** woland_ has left #opensolaris [22:13:33] <Gman> fuck yeah, sun made money again! [22:13:52] <elektronkind> q3 report is out? [22:14:20] <Gman> yep [22:15:13] <elektronkind> congrats [22:15:32] *** tfeb has joined #opensolaris [22:15:51] <Gman> http://www.sun.com/aboutsun/investor/index.jsp [22:16:46] *** tfeb has quit IRC [22:18:11] <jamesd> Net loss for the first quarter of fiscal 2007 on a GAAP basis was $56 million or a net loss of ($0.02) per share, as compared with a net loss of $123 million, or net loss of ($0.04) per share, for the first quarter of fiscal 2006 [22:18:32] <jamesd> not really made money.. it was all eaten up by costs... but it did do better [22:19:00] <dwc-> well, it's always -made- money .... it just doesn't have a profit ... still. [22:20:16] <AbeFroman> most importantly it seems to still be >5 after the announcement [22:20:17] <sahafeez> sun needs to make it. i have a bet riding on it ;) [22:20:30] <dwc-> lol [22:21:07] <jmcp> the Wall St consensus was that SUNW would lose 0.04c on a GAAP basis for q1 [22:21:10] *** onlinebacon has joined #opensolaris [22:21:20] <onlinebacon> hey is schillix still alive? [22:21:22] *** doownek has quit IRC [22:21:43] <dwc-> beating estimates is good [22:21:59] *** dunc has joined #opensolaris [22:22:03] <jmcp> add making a profit to that and it's even better [22:22:10] <Gman> jamesd, nod, i guess i didn't read the fine print :( [22:22:11] <jmcp> hopefully this FY can see that happen [22:22:20] <onlinebacon> because it looks pretty dead, on their site, nothing changed since august [22:23:29] <Gman> growth is still up seemingly [22:23:32] *** tfb has quit IRC [22:23:48] *** onlinebacon has quit IRC [22:24:06] <jamesd> onlinebacon, what are you talking about... sun has released x4100 mc2 x4600 mc2 ... and also is hyping blackbox [22:24:23] <jmcp> the investor webpage doesn't get updated much [22:24:24] <richlowe> jamesd: ... schillix [22:24:25] <Gman> i'd love to understand our costs a little better [22:24:31] <richlowe> jamesd: Sun may have, I'll bet schilly didn't ;) [22:24:35] <jamesd> oh [22:24:44] <jmcp> Gman: it's the best 1st quarter for a long time [22:24:48] <Gman> yeah [22:25:04] <jmcp> and as good as q306 [22:25:05] <jamesd> yeah schiliy is too busy fighting ghosts/demons/debian supporters [22:26:08] *** delewis has quit IRC [22:26:11] <sahafeez> hey is there a tim j here? [22:26:45] <jmcp> Gman: more importantly, I reckon, SUNW has started fy07 better than it finished fy03, fy04 and fy05 [22:27:06] <Gman> jmcp, that's good [22:27:21] <Gman> jmcp, i'm struggling to grok why we're not actually profitable still [22:27:44] <jmcp> restructuring charges and accounting requirements :( [22:28:33] <jmcp> oh, goody - gross margin up over the last 3 quarters [22:30:26] <richlowe> you've got another boom coming, silly web stuff and all. [22:30:26] <richlowe> the interesting bit will be whether it kills everyone again. :) [22:30:27] <alanc> costs a lot of money up front to cut costs long term (layoffs need severance packages, closing campuses needs moving expenses, etc) - we're still getting hit with those short-term costs now [22:30:50] <alanc> a lot of that falls into the "one-time charges" that show the 4 cents-per-share loss [22:30:50] <sahafeez> sun got big and bloated and needs to cut down. that cost $$ so the profits will not be there until the diet is over - put simply [22:33:51] <jmcp> yay, I got paid [22:34:20] <Darwin> and I got laid [22:34:20] <Gman> richlowe, heh, dot ohhhh :) [22:34:51] <Gman> alanc, we always seem to have 1 time charges thoguh [22:34:51] <Gman> very frustrating [22:36:00] <sickness> oh really? so schillix will not be updated anymore? :/ [22:36:41] <hile_> bout time, eh, jmcp? [22:37:00] <jmcp> hile_: damned right [22:37:20] <elektronkind> I'm seriously wondering what Bonwick has in store for sun's storage product line [22:37:32] *** Tekni has joined #opensolaris [22:38:20] <sickness> we're all waiting for (cheap?) zfs jbods :') [22:38:29] <Gman> i'd actually buy a sun mp3 player ;) [22:38:50] <Gman> doubt sun would be bold enough to produce one [22:38:50] <sickness> iSun [22:38:50] <sickness> ghgh [22:38:53] <sickness> SunPod [22:39:19] <jmcp> I'd love to see Mac apps runnable on Solaris [22:39:20] <hile_> and you'd find a way to make 1GB of its capacity just for the JDS UI, eh gman? [22:39:27] <Gman> hile_, heh, that would be evil [22:39:57] <sahafeez> no, it would be perfect. solaris kernel and mac gui [22:40:22] <sahafeez> and NeXT runs on sparc so i am thinking it would not be that hard [22:40:22] <hile_> except that my mac's gui annoys the shit out of me sometimes [22:40:35] <jmcp> hile_: which is why I just want the apps :) [22:40:51] <Gman> i'm looking forward to the day my ipod dies so i can upgrade ;) [22:41:20] <richlowe> just wait for the gnome folks to get done cloning them all. :) [22:41:21] <alanc> you mean a Sun MP3 player more portable than http://www.dme.org/log/2004/06/16/project-icece ? [22:41:21] <sahafeez> bah. gnome is a toy vs. aqua [22:41:50] <richlowe> haha, them's fightin' words. [22:41:51] <sickness> gnome and aqua are both useless toys, but gnome is also unstable as hell ;P [22:42:20] <sahafeez> haha. i was just about to add an very unstable toy. [22:42:21] <gdamore> both are pigs. [22:42:21] <sahafeez> my mac never crashes [22:42:50] <sahafeez> and it gets used hard 8 hours a day [22:42:50] <sahafeez> JDS on the other hand.. [22:42:50] <Gman> ok, enough gnome bashing [22:43:21] <Gman> i dunno about you, but it's been a while since i've had a crashing gnome app [22:43:50] <sahafeez> start rhythbox [22:43:51] <sahafeez> or run snv_48 w/vermillion 49 and try to drag and drop [22:44:20] <richlowe> the last JDS app I had crash was gdm. [22:44:21] <gdamore> gnome has gotten pretty stable, i'll give it that. you just have to throw enough memory at at it. [22:44:22] <richlowe> and only then because it segv's when X crashes under it. [22:44:29] <richlowe> (as do a fairly large number of other things) [22:44:37] *** Teknix has quit IRC [22:44:50] <Gman> hrm, works for me [22:44:51] <alanc> hah! beat richlowe to a bugs.o.o bash on tools-discuss! [22:44:51] <richlowe> alanc: No, I ranted once in here, I'm done for a while. [22:44:52] <Gman> i think that bug got fixed recently though [22:45:15] <richlowe> wow, talking of fightin' words. [22:45:22] <richlowe> > Well, I'm not ON, and that's a stupid guideline. [22:45:25] <richlowe> Shannon should go into politics. [22:45:51] <alanc> even with such obvious flamebait as "don't put comments in SCCS, put them in the bug report"? [22:45:51] <Gman> bill would hate the way we do things in gnome ;) [22:45:56] <Gman> we keep a ChangeLog too ;) [22:46:21] <gdamore> i think we should ask the jds developers to share a sun ray server -- 10 jds developers to a single V280R with 4GB ram. i bet some of the problems I have with JDS would get fixed really fast. :-) [22:46:21] <richlowe> alanc: actually, your response was rather tasteful. [22:46:21] <Gman> gdamore, sunray is incredibly broken [22:46:50] <richlowe> gdamore: did you miss them giving Glynn a blade100 to debug the sparc stuff on? [22:46:51] <sahafeez> hum, can you do something like that for the xsun team to [22:46:52] <richlowe> karma. :) [22:46:52] <gdamore> no sunray isn't. the philosophy that everyone has unlimited ram and 3D acceleration is broken [22:47:21] <alanc> gdamore: they've been doing that in the beijing office lately - running the latest vermillion JDS builds on a Sun Ray server for the team there [22:47:21] <richlowe> alanc: but no, I'd have left it alone, totally not worth diverting that thread yet more. [22:47:22] <Gman> you're absolutely over the barrel with how much stuff you can throw down that network cable [22:47:22] <alanc> don't know if they're using it as their primary desktop or not [22:47:51] <gdamore> Gman: yes, but you shouldn't have to throw _every_ friggin' pixel down stream. [22:47:52] <gdamore> with CDE sun ray is quite snappy. [22:47:58] <alanc> sahafeez: there isn't really an "Xsun" team any more [22:48:21] <gdamore> even over 11mbps WLAN. [22:48:22] <Gman> that's because cde doesn't do shit. [22:48:51] <gdamore> what is so friggin' great about gnome? all that eye candy costs, man. [22:48:52] <alanc> I'm probably the last member of the X team who still uses Xsun on his primary desktop - most switched to Xorg on x64 boxes already [22:48:52] <gdamore> and frankly, it doesn't look that much better, as far as i'm concerned [22:49:21] <gdamore> i use both Xsun and Xorg. but that's because I run both nvidia/x86 and sparcs. [22:49:22] <tsoome> it's pathetic attempt to copy windows.... [22:49:22] <alanc> I'm on Xsun primarily because I use the x64 boxes in my office for testing and leave the SPARC as a glorified stable xterm [22:49:52] <gdamore> that's a power hungry xterm... :-) [22:49:52] <Gman> gnome is definitely not windows [22:49:52] <gdamore> alanc needs a sunray. :-) [22:49:53] <Gman> if it was, the button order would be different [22:50:22] <gdamore> no, windows runs pretty well over RDP. gnome sucks. [22:50:23] <sahafeez> alacn: will it support sparc with 2 elite3ds yet? [22:50:23] <hile_> I love my sunray [22:50:23] <alanc> I have Sun Rays, but I'd only use one as my primary desktop if I could set it up to not use a server managed by our IT group [22:50:23] <sahafeez> alanc even [22:50:47] <gdamore> you can, but you have to have some control over your network. :-) [22:50:48] <Gman> we accept bug reports [tm] [22:51:22] <stevel> alanc: you can - that's what danek does i believe [22:51:22] <alanc> sahafeez: not yet - we've only gotten it up on PGX64 so far, but it's only been a couple of days since that became working [22:51:51] <gdamore> i don't believe it. i'm _totally_ convinced that everyone in the gnome project is so hopped up on eye candy that they have no idea about how people with more limited systems will use their environment. [22:51:52] <gdamore> as i said, give the gnome/JDS people sun rays, and they will find that a lot of that eye candy sucks. [22:51:55] <tsoome> today I started some app from gnome menu. cd-ripper or like. it gave error like /dev/sp0 is missing (cant remember exact device name). [22:51:56] <alanc> Sun IT has this nasty habit of customizing the GNOME desktop in bizarre ways and breaks my settings every time I log in to one of their Sun Ray servers [22:52:21] <sahafeez> well my desktop will be thye u80 with those cards for the next year so.. [22:52:22] <Gman> there's really not that much eye candy in gnome [22:52:22] <Gman> some of the gtk-engines/themes are badly written [22:52:22] <tsoome> anyhow, there has been no device with such name in solaris [22:52:22] <stevel> gman: or rather, there's as much eye candy as you (the user) wants [22:52:22] <gdamore> the worst was JDS 2, that background was almost 100% incompressible over sun ray protocol. [22:52:26] <stevel> a properly written gtk theme engine can be very efficient [22:52:52] <dduvall> alanc: the problem that you run into is that your DTU will automatically connect to the itops fog, but your machine is a simple utswitch away, until either your DTU loses its IP address (never, if things are working okay) or power-cycles. [22:52:52] <sahafeez> alanc: is it on the road map - i mean does the install change to xorg on sparc if you have a supported card [22:52:52] <xinkeT> gdamore: at least some of the jds team have sunrays [22:52:53] <xinkeT> and are well aware of the problems [22:52:54] <Gman> gdamore, yeah, but you're expecting graphic artists to know that stuff [22:52:57] <sahafeez> E17 build on sparc? [22:52:59] <Gman> which well, isn't going to happen [22:53:22] <gdamore> Gman: well, I expect whoever drives the technology to teach the artists, if necessary [22:53:22] <Gman> the default login on sunray is jds internally [22:53:22] <alanc> gdamore: the gnome community is working on that now that they're being used in the One-Laptop-Per-Child project machines [22:53:29] <alanc> sahafeez: way too soon to tell [22:53:52] *** onlinebacon has joined #opensolaris [22:53:52] <Gman> we have a lot more people running jds on overpacked servers [22:53:52] <onlinebacon> hey [22:53:52] <gdamore> remember how the artists had to figure out how to reduce the palette size on web pages. they eventually figured it out [22:53:52] <elektronkind> would anyone find an infobot being a useful addition to the channel? something that auto-tinyurl-izes things, answers common questions, and other nifty stuff? [22:53:59] <gdamore> define overpacked? [22:54:22] <alanc> !infobot when will ksh93 integrate? [22:54:23] <alanc> !infobot and what's up with the komodo dragons? [22:54:24] <gdamore> infobot: 2093. [22:54:52] <elektronkind> <infobot> not bloodly likely [22:54:53] <richlowe> I agree with the point alanc is making. [22:54:53] <richlowe> at least, the point I think he's making. [22:54:53] <gdamore> infobot eats alanc. infobot is a komodo dragon. [22:54:54] <sahafeez> there was a point? [22:55:00] <onlinebacon> is Schillix still alive? [22:55:03] *** deedaw has joined #opensolaris [22:55:22] <alanc> though if it could automatically answering "What's the difference between OpenSolaris, Solaris Express [22:55:23] <richlowe> I've already had to ignore one bot, I really don't want to have to ignore however many more :) [22:55:23] <sahafeez> onlinebacon: no. stop asking ;) [22:55:23] <alanc> and SX:CR?" that would be useful [22:55:23] <richlowe> though the commits bot thingy could be cool, I guess. [22:55:31] <onlinebacon> sorry i didnt see the answer before cos i was lagging and got disconnected :( [22:55:52] <onlinebacon> thanks though :D [22:55:53] <alanc> onlinebacon: ask schily [22:55:53] <richlowe> onlinebacon: we don't know, you'd have to ask schily. [22:55:53] <onlinebacon> thanks :) [22:55:54] <sahafeez> !seen schily [22:55:55] <Drone> schily (schily!i=schily at p54BED88F dot dip.t-dialin.net) was last seen in #opensolaris on Wed 18 Oct 2006 23:12 GMT, saying 'dclarke: does star link now?'. [22:56:22] <gdamore> i never understood why shillix existed in the first place. [22:56:23] <sahafeez> onlinebacon: he pops in here - he would be the best guy to ask as it is his distro [22:56:30] <tsoome> why there are all those linux distros?;) [22:56:31] <Gman> gdamore, proof of concept [22:56:32] <onlinebacon> wicked, will do :) [22:56:51] <Gman> because he could? [22:56:53] <onlinebacon> it came out like a week after solaris 10 didnt it? [22:56:54] <Gman> i'd like more people playing around with that sort of thing [22:57:22] * Gman is rather intrigued to see gentoo/portage on the roadmap for belanix [22:57:23] <gdamore> okay, i guess. nexenta i understand. but i don't think we need a bunch of other distros. but thats just my opinion [22:57:25] <Gman> sure we do [22:57:26] * hile_ notes that eh's the bastard child when he uses JDS... JDS with dtterm since, well, gnome-terminal sucks [22:57:52] <Gman> solaris express isn't open [22:57:53] *** mkhl has quit IRC [22:57:54] <Gman> hile_, 2.16 is significantly improved [22:57:55] <tsoome> open source you mean [22:57:55] <onlinebacon> so there is nexenta and belenix - they are the open, opensolaris's? [22:57:57] <gdamore> open schmopen. i don't want to be Linux. [22:58:22] <Gman> tsoome, yeah, free software [22:58:22] <sahafeez> <=- is going back to WM! [22:58:23] <tsoome> it's definitely open, all interfaces are documented;) [22:58:30] <gdamore> haha [22:58:31] <mrdeviant> no they're not :P [22:58:32] <Gman> onlinebacon, even they include some non-opened components [22:58:53] <hile_> mrdeviant: all public supported interfaces perhaps [22:58:53] <onlinebacon> ah ok [22:59:00] <mrdeviant> right. which sure ain't everything [22:59:01] <richlowe> gdamore: distros are distros, they'll happen. [22:59:22] <onlinebacon> I saw a pretty good link, comparing CDDL, GPL and EULA in a table :): http://linux.wordpress.com/2006/10/04/at-a-glance-2-gpl-cddl-and-bsd-vs-eula/ [22:59:23] * gdamore has no time open source zealotry. wants source to be open in order to get work done, not as a political statement. [22:59:24] <richlowe> you don't have to run them, so far you don't even really have to care about them. [22:59:27] <richlowe> as long as nobody mandates I test on all of them, I don't care :) [22:59:52] <gdamore> richlowe: yeah, i was just expressing mild curiousity, is all. i don't care that they exist, as long as they don't fracture the community. [23:00:12] <myrkraverk> ok, since I'll have to wait for sun to fix e1000g - what are my alternatives? (the em driver didn't work either, afaik) - pcmcia cards? usb ethernet cards? [23:00:23] <gdamore> imo, Linux is too fractured by all the 18 gazillion different distros. [23:00:24] <Gman> gdamore, it's a good encouragement to creating a good base for them to work off [23:00:53] <gdamore> rge works with cardbus [23:00:59] <gdamore> that would be my first choice. [23:01:00] <richlowe> The only real problem will be when someone has a change that can't go back to ON. [23:01:12] <myrkraverk> gdamore: you mean the NIC? [23:01:23] <onlinebacon> I think it is ok to be semi open source like opensolaris, stops LOADS of different distributions popping up like in linux, but allows you to learn off of the code [23:01:23] <richlowe> and the only "Why not?" argument fires up. [23:01:52] <gdamore> yes, rge is available in cardbus. Netgear FA511. [23:01:59] <myrkraverk> onlinebacon: it's only a matter of time until ppl have replacements for the non-open parts [23:02:23] <myrkraverk> gdamore: what is cardbus? [23:02:30] <gdamore> pcmcia 32-bit. [23:02:39] <myrkraverk> gdamore: ok [23:02:54] <gdamore> you have to have a 32-bit pcmcia controller. not all controllers/slots support cardbus. [23:02:54] <onlinebacon> so is belenix like, the best opensolaris fork? [23:03:19] <myrkraverk> gdamore: do you know how I can check? [23:03:23] * Cybernd likes gnusolaris [23:03:25] * hile_ likes SXCR [23:03:30] * gdamore & [23:03:33] <onlinebacon> SXCR? [23:03:53] <onlinebacon> oh, sorry [23:03:53] <hile_> solaris express community release [23:04:01] <Cybernd> onlinebacon hmm i also got such a link in my bookmarks - http://blogs.sun.com/chandan/entry/copyrights_licenses_and_cddl_illustrated [23:04:28] * jmcp departs [23:04:53] <onlinebacon> heh Cybernd, mine was completely lifted :( [23:05:30] <Cybernd> onlinebacon always interesting to see the same pictures apearing in different resources [23:05:54] <onlinebacon> I just want something small, so I can build it up, with my own software, because I dont like tonnes of stuff I am not going to use [23:05:56] *** mrdeviant has quit IRC [23:07:37] *** silly_girl22 has joined #opensolaris [23:07:47] <richlowe> the last couple of zfs putbacks have had some of the most vague synopses. [23:09:17] <sahafeez> onlinebacon: gentoo stage 1. [23:10:18] <sahafeez> onlinebacon: openbsd. both very small UNIX like distros. solaris is far from small ;) but it is a great powerfull robust os [23:10:57] <onlinebacon> Used both before :D, openbsd is pretty slow, and gentoo is full of idiots who want to be the next linus torvalds :/ [23:11:17] <sahafeez> slow? how so? for a desktop? [23:11:17] <onlinebacon> I usually use FreeBSD, just looking for a change now [23:11:45] <sahafeez> gentoo will be the fastest unix distro you will find. gnome flies under it. [23:11:46] <onlinebacon> I found it a lot slower than free and net on my desktop [23:12:00] <sahafeez> it is slower then freebsd. [23:12:06] <onlinebacon> what do you use sahafeez? [23:12:07] <sahafeez> i use it for my servers [23:12:26] <sahafeez> bash-3.00$ uname -a [23:12:26] <sahafeez> SunOS carmenere 5.11 snv_49 sun4u sparc SUNW,Ultra-80 [23:12:42] <onlinebacon> cool :D [23:12:44] <sahafeez> # uname -a [23:12:45] <sahafeez> OpenBSD cork.beastproject.org 3.9 RAID39#1 i386 [23:12:52] <sahafeez> and a Mac. [23:13:11] <onlinebacon> I only have one machine :( although on my dad's laptop right now [23:13:29] <sahafeez> for a day to day desktop - apple rules [23:14:02] <sahafeez> for my working on geek stuff i like my sun for the keyboard and mouse for the most part (silly i know) [23:14:22] <onlinebacon> lol, thats fair enough, i like the type 5's, nice and clicky :D [23:14:30] <sahafeez> yes. [23:15:59] <sahafeez> things are reaching a point re: oss software were they are all the same for the most part - gnone on every desktop so it becomes more about what you like to do with the underlying stuff i think. [23:16:37] *** gnu2it2 has quit IRC [23:17:05] <sahafeez> and what you know. i am not a solaris expert at all - just ask anyone here ;) but i have had enough time building and running large networks to know that sun hardware is rock solid and i never had issues with the 1000+ solaris boxes i ran in production [23:17:11] <onlinebacon> I think free software is great, but so is software quality, and the linux world is full of awful distributions, and a kernel that isnt up to par with bsd at all [23:17:26] <sahafeez> nope. linux is crap vs. bsd [23:17:37] <sahafeez> but linux will run on hardware that bsd will not. [23:17:38] <onlinebacon> sahafeez: Nowhere near your knowledge here then, still 15 and learning as fast as I can :) [23:18:03] <myrkraverk> is google down? [23:18:03] <onlinebacon> sahafeez: But that is becoming irrelevant, people can afford new computers, and BSD has always worked on every computer I have tried it on [23:18:06] <sahafeez> haha. cool. just forget eveything you know about linux and you will be find on soliras [23:18:18] <onlinebacon> :p [23:18:21] <onlinebacon> myrkraverk: Nope [23:18:27] <sahafeez> that is like asking if the sun is off [23:18:33] <onlinebacon> exactly :) [23:18:39] <onlinebacon> google has like 300 server farms [23:18:40] <sahafeez> 100K servers, no it does not go down [23:18:41] <myrkraverk> onlinebacon: k, that means http promblem is local [23:19:35] *** silly_girl22 has quit IRC [23:19:35] <onlinebacon> oh well, guess I will download Solaris Express then :) [23:19:48] <sahafeez> read the install docs before you install. [23:19:53] <onlinebacon> take all the stuff I do not need, like java and stuff, and strip it down [23:20:03] <sahafeez> anyone? i know there is a nice link on big admin on how it works. [23:20:05] <jamesd> it would have to be a major router/backbone or maybe dns ( which has happened in the past) to brinkg down google. [23:20:06] <onlinebacon> sahafeez: I think i should be ok, ran solaris a while ago on my sparc :) [23:20:13] <sahafeez> ok [23:20:39] <onlinebacon> gaaah i need to borrow my brothers DVD drive :D [23:20:40] <sahafeez> solaris - stable as hell, if you can put up with the install client [23:20:55] <hile_> people install from media? [23:21:01] <onlinebacon> how else? [23:21:06] <sahafeez> jumpstart [23:21:09] *** stevel has quit IRC [23:21:12] <sahafeez> onlinebacon: just ignore him [23:21:26] <sahafeez> ;) [23:21:33] <onlinebacon> lmao, im intrigued [23:21:37] <onlinebacon> i like network installation :D [23:21:40] <onlinebacon> saves Cds [23:22:04] *** krozinov has joined #opensolaris [23:22:12] <tsoome> it's faster [23:22:13] <sahafeez> just but you need another box to run it from and you need to set it up just so etc. there are scripts to setup if you are running it on a solaris box. any other have fun [23:22:30] <onlinebacon> oh, i only have one box so i gues media it is :) [23:22:32] <onlinebacon> guess* [23:22:40] <hile_> I really should rewrite all my finish scripts again... [23:23:03] <sahafeez> why you are at it i need JS setup scripts for openbsd ;) [23:23:24] <xinkeT> i have nexenta setup for jumpstart now [23:23:30] *** krozinov has left #opensolaris [23:23:36] <onlinebacon> whats the fastest solaris? [23:23:43] <jengelh> Empty Set. [23:23:48] <onlinebacon> nexenta, belenix or community release? [23:23:55] <hile_> uh? [23:23:59] <hile_> that question doesn't parse [23:24:05] <asyd> "lol" [23:24:12] <hile_> at least not in anyway that would make sense [23:24:13] <jamesd> solaris express is the fastest because its always based on a non-debug build [23:24:21] *** mazon is now known as Mazon [23:24:27] <onlinebacon> wicked, thanks jamesd [23:25:02] <myrkraverk> what is SMBus ? [23:25:14] <sahafeez> onlinebacon: you do know solaris is called slowaris right [23:25:19] <onlinebacon> yeah [23:25:33] <sahafeez> myrkraverk: new replacent for the isa bus in mbs. has the montioring stuff on it. [23:26:05] <sahafeez> http://www.smbus.org/ [23:26:20] <sahafeez> they moved all the legacy stuff from the isa to it.. [23:27:50] *** cub has quit IRC [23:32:31] <onlinebacon> anyways, im out, thanks for all the help everyone :) [23:32:34] *** onlinebacon has left #opensolaris [23:35:21] <myrkraverk> sahafeez: ah [23:36:00] <myrkraverk> does the solaris cardbus driver drive ene cb 712/4 ? [23:36:06] <myrkraverk> any good way to check? [23:46:55] *** ffidelis has joined #opensolaris [23:49:39] *** deather_ is now known as deather [23:50:53] <axisys> what process does auto power / shutdown ? i see ^M character which looks to me as power loss on this one systems messages file [23:52:38] *** jonkelle has joined #opensolaris [23:52:40] <Tpent1> !seen kupfer [23:52:42] <Drone> kupfer (kupfer!i=kupfer@nat/sun/x-3d845599e2560142) was last seen in #opensolaris on Mon 16 Oct 2006 20:31 GMT, saying 'well, if they're both using the same v4 domain, I don't know what the problem is.'. [23:54:26] <axisys> this string `genunix: [ID 540533 kern.notice] ^MSunOS Release 5.8 Version Generic_117350-39' usually means the system lost power.. whats the internal process that can control autoshutdown? (soory for the repeat .. trying to rephrase the question) [23:55:07] <rodrickbrown> powerd [23:55:17] <rodrickbrown> man powerd [23:55:40] <rodrickbrown> man power.conf [23:56:27] <axisys> rodrickbrown: thats not running.. one less too wory.. doing POE .. thnx [23:56:40] *** ffidelis has quit IRC [23:57:36] *** alanc is now known as alanc-away [23:58:04] <axisys> *worry