[00:00:29] <jbk> evening [00:04:35] *** unixware has joined #opensolaris [00:04:40] <unixware> hi dudes [00:04:57] <unixware> is posible to use pf on opensolaris? [00:05:21] <boyd> unixware: Is there some reason that ipf is not working for you? [00:05:49] *** ChanServ sets mode: +o asyd [00:05:51] <unixware> boyd cause i know pf, and i have never try ipf before [00:06:53] <boyd> If you're new to solaris, surely there'll be a lot of that. Are you really going to spend all that effort making OpenSolaris feel like something else? [00:07:55] <unixware> no, i think i have to read about solaris from now [00:08:35] <boyd> :) Personally I don't find ipf that hard.. it's the concepts that take the learning, and they're the same in all packet filters. [00:09:59] <unixware> so, i will try and see what happend :P [00:10:48] *** sommerfeld has quit IRC [00:13:30] <boyd> So, there may be a way to get zfs into the Linux kernel after all... [00:13:34] <boyd> http://undeadly.org/cgi?action=article&sid=20070829001634 [00:15:35] *** Kitty has quit IRC [00:17:39] *** yippi has quit IRC [00:19:49] <nrubsig> boyd: you mean "remove CDDL and tack GPLv2 on the sources and claim it was always that way" ? [00:20:07] <boyd> Yah... seems to work for them. [00:20:28] *** Jondice_ver_2 has quit IRC [00:20:30] *** dunc has quit IRC [00:20:44] *** dunc has joined #opensolaris [00:20:57] *** dunc has joined #opensolaris [00:21:33] *** dunc has joined #opensolaris [00:21:49] *** dunc has joined #opensolaris [00:25:59] *** Lusitanian has quit IRC [00:26:37] *** cypromis has joined #opensolaris [00:33:18] *** nrubsig was kicked by nrubsig (Suicide!! Yippiieeee! <SMACK>) [00:33:42] *** bengtf_ has quit IRC [00:35:36] *** nostoi has quit IRC [00:35:45] *** estibi has quit IRC [00:38:31] *** sparc-kly has quit IRC [00:43:50] *** migi has quit IRC [00:45:24] *** maxpil has joined #opensolaris [00:49:52] <nachox> roland is damn crazy, cool guy though [00:51:14] <unixware> roland macdonal? [00:51:28] <flyingparchment> roland mainz, i guess [00:51:49] <unixware> i c [00:52:19] *** kFuQ has quit IRC [00:52:24] <kjetilho> aka nrubsig [00:52:33] *** kFuQ has joined #opensolaris [00:52:34] <unixware> ? [00:52:45] <kjetilho> his nick in this channel [00:52:53] *** apuc has joined #opensolaris [00:53:06] *** stevel has quit IRC [00:53:14] <nachox> yes, ksh's roland :) [00:53:49] *** pfa3rh has quit IRC [00:54:17] *** bengtf_ has joined #opensolaris [00:54:20] *** praseodym has joined #opensolaris [00:55:19] <praseodym> I'm doubting: should I install Solaris 10 or OpenSolaris as a development environment (subversion/ruby on rails/postgresql/apache) on vmware esx server 3? [00:55:29] <nachox> sun ray 2n :) i want that! [00:55:40] <flyingparchment> unless you want to develop opensolaris, install S10 [00:56:40] <praseodym> then the second question - Joyent is all over OpenSolaris. why is that? [00:56:51] <flyingparchment> because benr is an opensolaris fanboy [00:56:51] <Triskelios> flyingparchment: what is your reasoning there? SX is much more desktop-friendly [00:56:53] <flyingparchment> ;-) [00:56:57] <jmcp> praseodym: because it fits their needs [00:57:14] <flyingparchment> Triskelios: i don't want to mess with the OS when i'm doing development, and S10 is great for install-and-leave-it systems, with patching where needed [00:57:21] <flyingparchment> that's just my opinion, of course [00:58:08] <nachox> hmm, SXDE and CE are much better for laptops [00:58:12] <praseodym> and what about production? note we're running on vmware esx server, I'm not sure of opensolaris works all fine on that? [00:58:26] <flyingparchment> S10 also for production. no support for SX [00:58:35] <Triskelios> praseodym: S10 should be used for production [00:58:51] <Triskelios> praseodym: OpenSolaris is just the codebase, S10 and SX are derived from it [00:58:54] <nachox> i'd go with solaris 10 for production [00:59:00] <praseodym> jmcp: how does it fits their needs more than solaris 10? [00:59:02] * delewis uses Solaris 10 on his workstation [00:59:15] <nachox> and solaris 10 is not derived from opensolaris [00:59:18] <delewis> just because I can't afford to have breakage. [00:59:23] <nachox> i use solaris 10 in this laptop :) [00:59:33] *** maxpil has quit IRC [00:59:34] <Triskelios> nachox: a lot of SX stuff gets backported [00:59:35] <delewis> and there is significant breakage in SX, especially with JDS. [00:59:46] <praseodym> and when should I use SX over S10? [00:59:57] *** apuc has quit IRC [01:00:04] <Triskelios> praseodym: Joyent use S10 for most of their deployments [01:00:05] <jmcp> praseodym: you could ask them [01:00:17] <delewis> praseodym: anytime you value the functionality of your system are in such a situation that "upgrading" to the next SX release isn't an option whenever something is fixed. [01:00:20] <CIA-16> perrin: 6597381 BAD TRAP: type=e (#pf Page fault) rp=ffffff00043a48f0 addr=28 occurred in module "<unknown>" due to [01:00:30] <praseodym> I see :) [01:00:33] <nachox> delewis, the only problems i had with solaris 10 have been java related ironically, specially the latest sun connection patch [01:01:05] <delewis> nachox: I'm not saying Solaris 10 has breakage. I'm saying Solaris Express has breakage, but that's to be expected from a development release of anything. [01:01:11] <praseodym> and last question for now, is it possible to get a direct download link for solaris 10 instead of having to go through registration, sun download manager, etc.? [01:01:15] <flyingparchment> nachox: update connection _is_ a breakage [01:01:29] <flyingparchment> praseodym: no, but once you have that link, you can download it with wget or whatever [01:01:33] <delewis> I don't want Nautilus core dumping everytime I try and do something with it (which was the case with a fairly recent SX release) [01:01:59] <nachox> delewis, i meant in contrast with SXDE and CE where even the gnome trash bin is broken [01:02:02] <Triskelios> praseodym: if you want support for a production system you'll want to get a sunsolve account anyway.. [01:02:04] <delewis> I also build quite a bit of my own software and its beneficial to build that software on a Solaris 10 system, rather than an SX system for backwards compatibility with my Solaris 10 systems. [01:02:08] <praseodym> flyingparchment: I can't seem to get around the download manager [01:02:10] <delewis> nachox: precisely! [01:02:27] <Triskelios> praseodym: just click the links on the download page [01:02:33] <flyingparchment> praseodym: why not? once you accept the license agreement on the download page, just right click->copy the URL and do 'wget -O <file> <url>' [01:02:42] <praseodym> flyingparchment: k :) [01:02:45] <flyingparchment> praseodym: you need -O otherwise wget will try to write an insanely long filename and fail with a mysterious error [01:02:47] <nachox> i do have to say that SXDE is excellent considering it is beta software [01:02:55] <delewis> nachox: you should report that. The JDS team at Sun refuses to acknowledge the existence of that bug because they don't test JDS on SPARC. [01:02:59] *** stevel has joined #opensolaris [01:03:00] *** ChanServ sets mode: +o stevel [01:03:07] <praseodym> Triskelios: it's more that someone else has to download and deploy the iso image for me, and he hasn't got an account [01:03:17] *** stevel has quit IRC [01:03:48] *** stevel has joined #opensolaris [01:03:48] *** ChanServ sets mode: +o stevel [01:03:49] <flyingparchment> Triskelios: clicking the link is not much good for remote deployments :) [01:03:56] <flyingparchment> (unless you have much faster adsl than me) [01:04:05] *** fedorared has joined #opensolaris [01:04:11] *** Lusitanian has joined #opensolaris [01:04:15] <elektronkind> it's hard to find a good accountant these days [01:04:18] <elektronkind> oh, you said account [01:04:32] <elektronkind> make him one? [01:04:50] <nachox> http://www.flickr.com/photos/jimgris/683205344/in/set-72157600585529837/ :) [01:04:54] <Lusitanian> howdy [01:05:29] <moazamraja> wierd...i'm starting up a zone and it can't seem to 'sysidconfig' [01:05:34] <moazamraja> i zlogin and just get a prompt [01:06:21] <praseodym> one mystery here.. sol-10-u3-ga-x86-dvd-full.exe is 1153.94 MB and the split zips are 5x 559.91MB [01:06:28] <flyingparchment> moazamraja: zlogin -C [01:06:39] <moazamraja> yeah...did that [01:06:57] <Triskelios> nachox: those are the new sun rays? [01:07:00] <flyingparchment> reboot it and login quicker, maybe it failed for some reason [01:07:06] *** cypromis has quit IRC [01:07:23] <nachox> Triskelios, i think those are not available in the us... yet [01:08:29] <nachox> it looks like a really pretty device, i wonder how good sun rays are over wireless [01:08:43] <flyingparchment> tadpole makes a wireless sunray, so it can't be that bad [01:08:46] *** NikolaVeber has quit IRC [01:09:20] <jpdrawneek> done sunray over g wifi [01:09:28] <Triskelios> they look like my friend's laptop [01:09:29] <delewis> the bottleneck on Sun Ray performane is *not* the network throughput. [01:09:30] <jpdrawneek> it was ok [01:09:38] <delewis> you can run a Sun Ray successfully over twisted-pair. [01:09:55] <delewis> (this is why the Sun Ray 2's still only have fast-ethernet and not GigE) [01:10:02] <jpdrawneek> which knowing our netowrk we probably have [01:10:12] <praseodym> all, thanks for your help! I've got to get some sleep now. [01:10:20] <nachox> err, wireless is twice as slow as fast ethernet [01:10:35] <delewis> nachox: I said it does fine over twisted-pair, as well. [01:10:57] * boyd always cringes at lines like "twice as slow" [01:11:00] <delewis> which is either roughly equal to 802.11b or 1/5th of the throughput an 802.11g device is capable of. [01:11:03] *** praseodym has quit IRC [01:11:09] <nachox> boyd, ops :) [01:11:41] <boyd> : [01:11:43] <boyd> :) [01:11:56] <flyingparchment> delewis: er, fast ethernet runs over twisted pair [01:12:01] <jpdrawneek> get the wan firmware and your good on 512k adsl :) [01:12:21] <alanc> don't know why they're only selling the Sun Ray notebooks in Japan and not the rest of the world, but you can buy Sun Ray notebooks from other companies in the US [01:12:39] <jpdrawneek> because the west does not like small laptops [01:13:03] <jpdrawneek> all laptops must be like american cars [01:13:05] <boyd> Because the west does not like small laptops that are unusable unless you're online. [01:13:15] <delewis> flyingparchment: ok, 10Mbps ethernet. [01:13:22] <delewis> Sun Rays do fine on that. [01:13:46] <jpdrawneek> the point of having a laptop thats not online? [01:13:53] <nachox> then the problem is todays fat DEs [01:14:52] * boyd thinks..... train, plane, customer site where no net access is provided, coffee show with extortionate net pricing [01:15:00] <nachox> E17 for sunrays [01:15:21] <jpdrawneek> they finished E17 yet? [01:15:32] <boyd> ha! [01:15:33] *** fedorared has left #opensolaris [01:15:37] <delewis> it'd be great in an .edu environment where the majority of users would be on-campus, and thus have access to the Sun Ray server. [01:15:40] <nachox> of course not [01:15:52] <delewis> of course, I doubt the price Sun is charging for the laptop is economical by any means. [01:16:09] <delewis> (Sun Ray 2's are significantly overpriced, IMO) [01:16:13] <boyd> If it's anything like the desktop sunray pricing it's laughable [01:16:16] <jpdrawneek> always the problem [01:16:22] <nachox> i bet it costs a more than a low end dell laptop [01:16:31] <alanc> the trains and buses around here are mostly wifi-enabled these days [01:16:40] <delewis> I'd bet its roughly the same price as the Sun Ray 270. [01:16:43] <boyd> We don't all live in the vally [01:16:46] <boyd> valley [01:16:57] <delewis> (270 has a larger LCD -- 17", but that should be offset by the laptop functionality) [01:16:59] <jpdrawneek> mobile 3G networks [01:17:27] <boyd> Hereabouts 3G is only starting to become worth considering for anyone but corporations. [01:17:42] <jpdrawneek> boyd: where be you? [01:17:49] <boyd> Australia [01:17:52] <jpdrawneek> k [01:18:02] <jpdrawneek> europe is much nice [01:18:04] <jpdrawneek> r [01:18:09] <boyd> I'm sure. [01:18:21] <jpdrawneek> had 3G for a while [01:18:36] <jpdrawneek> in uk it just about does what the ads say [01:19:04] <boyd> There's one provider here just started offering 3G net access for a price I may consider... but it still suck compared to europe from what I hear [01:19:24] <jpdrawneek> unlucky [01:19:44] <jpdrawneek> maybe you all be saved by wifimax (or what ever intel decides) [01:20:06] <nachox> 3g is fast? [01:20:38] <jpdrawneek> think its up to a couple of megs in urba areas [01:20:39] <nachox> it's MIA in argentina [01:20:48] <vmlemon> I wouldn't download huge files over it, but it's fine if you just need to check e-mail or view small HTML pages [01:21:20] <jpdrawneek> watched football over it quite nicely etc... [01:22:45] <vmlemon> I'm using a certain 3G operator named after a low number in the UK, and the Internet access is hopelessly crippled unless you get an expensive contract from them [01:23:02] <boyd> That's the same provider hwew [01:23:03] <boyd> here [01:23:52] <unixware> does any one have install sun ray thin clients before? [01:24:06] <jpdrawneek> what do you mean [01:24:11] <vmlemon> You can get web page access through a proxy provided by "Wider Web" and the pages are modified slightly for viewing on mobile phones, but you can't do anything else [01:24:25] <unixware> jpdrawneek the server and clients [01:24:34] [01:24:38] <jpdrawneek> vmlemon: you get the larger phones with a proper browser [01:24:50] <vmlemon> And the obligatory contract, of course [01:24:59] <jpdrawneek> vmlemon: they guy who had it had it in free minutes etc... [01:25:11] <vmlemon> The Symbian Motorolas looked nice that they had on 3 at one point [01:25:16] <jpdrawneek> unixware: setup several sun rays [01:25:38] <alanc> installed clients - that was easy, plug into power & network, done! [01:25:39] <boyd> Here they're marketing usb and expresscard modems for laptops.. [01:25:44] <jpdrawneek> vmlemon: only ever seem nokia and sony erricson [01:25:50] <alanc> been a few years since I installed Sun Ray server [01:25:54] <unixware> jpdrawneek is it hard to do that? [01:26:03] <boyd> Tempt has one IIRC [01:26:05] <vmlemon> I have an old A835 on 3 that I use tethered over Bluetooth from time to time, and it works well [01:26:06] <jpdrawneek> unixware: can you read [01:26:07] <delewis> unixware: installing the Sun Ray Server takes 10 minutes. [01:26:15] <unixware> ohh [01:26:38] <jpdrawneek> unixware: just read what it asks you and your fine [01:26:56] <unixware> i just ned to doanload the sun ray sopftware o install in to the server rogth? [01:26:57] <jpdrawneek> you could just hold down enter if you really want [01:27:04] <vmlemon> Did have an early 3G NEC phone that's pretty useless, now, but it was a nice handset [01:27:07] <delewis> of course, doing things like failover groups takes significantly longer. [01:27:23] <unixware> i see [01:27:33] <delewis> unixware: uh, sure, provided you actually have Sun Rays around. [01:28:10] <jpdrawneek> unixware: fail overnot that bad - just find a key is the hardest part [01:28:13] <unixware> cause i want to buy some from ebay , and im not sure if is good idea [01:28:19] <boyd> Yes, you need a license key [01:28:20] <jpdrawneek> yes [01:28:25] <jpdrawneek> you don't [01:28:29] <boyd> !? [01:28:37] <jpdrawneek> you don't need a license [01:28:46] <jpdrawneek> never asks you for one [01:28:46] * boyd would have bought some already if he thought that was the case [01:29:04] <delewis> SRSS does require licenses nowadays. [01:29:12] <jpdrawneek> have you ever seen a sun ray licence? [01:29:15] <delewis> 4.x, anyway. [01:29:16] <boyd> That's what I thought [01:29:20] <delewis> 3.x does not. [01:29:39] <jpdrawneek> i have installed srss 4 and it did not ask me for a licence [01:29:44] <delewis> jpdrawneek: the SRSS 4.x demo on sun.com comes with a 30-day license. Each DTU now requires an RTU license. [01:29:50] <delewis> s/30-day/90-day/ [01:30:10] <jpdrawneek> back to 3 we go! [01:30:14] <unixware> delewis and is posible to get the 3.x? from there too? [01:30:23] <delewis> jpdrawneek: I guess you'll find out in 90 days past your initial install. [01:30:30] <delewis> unixware: its still offered on sun.com [01:30:31] <jpdrawneek> will do [01:30:34] <delewis> jpdrawneek: that's what I do. [01:30:52] <jpdrawneek> why they change it? [01:31:03] <delewis> moronic product managers would by my guess. [01:31:07] <boyd> Seems to be the opposite of the "everything open source" mantra [01:31:15] <delewis> boyd: welcome to the "real world" [01:31:31] <delewis> Jonathan has re-nigged on his promise to "free" Sun's software stack of RTU licenses. [01:31:32] <unixware> delewis that means that i can use them with no license? the 3.x? [01:31:38] <delewis> unixware: correct. [01:31:43] <unixware> delewis :D [01:31:45] <delewis> 3.x is free, technically. 4.x is not. [01:31:52] <unixware> ok [01:32:02] <jpdrawneek> run to srss 3.X and have fun [01:32:10] <unixware> :) [01:32:11] <boyd> delewis: happened with SGD too. [01:32:20] <jpdrawneek> the srss 4 firmware is pretty tho [01:32:25] <delewis> boyd: and it's really not just SRSS. There's several products in Sun's software stack that were free at some point and now require RTU. [01:32:37] *** yippi has joined #opensolaris [01:32:39] <delewis> Communications Suite, STK Volume Replicator, etc. [01:32:57] <delewis> Sun basically took the AVS, called it STK Volume Replicator, and attached a $20,000/TB price tag. [01:33:01] <delewis> way to go Sun! [01:33:15] <delewis> that's the "free" spirit, right there. [01:33:21] <boyd> delewis: except that sndr and ii are in current SX releases [01:33:28] <delewis> boyd: exactly. [01:33:47] <boyd> qfs too eventually, allegedly [01:34:08] <delewis> unfortunately, for those of us using Solaris 10 and lower, if we want AVS, we supposedly have to purchase an RTU license. [01:34:33] <delewis> (I'm not, as I see AVS as a free product because its being offered elsewhere for free) [01:34:49] <boyd> It's a funny situation. I wish they'd clear it all up. [01:34:57] * boyd seems to be saying that a lot about sun lately [01:35:09] <delewis> I don't think there's anything to clear up. [01:35:21] <boyd> I mean clear up the free/non-free plans [01:35:48] <delewis> The Pony figured out his 'strategy' of freeing $PRODUCTOFTHEWEEK and laying off another set of employees wasn't working. [01:35:57] <delewis> that's my thoughts, anyway. [01:37:14] <boyd> Sure, but if you announce one thing then change your mind you should announce the change [01:37:16] <nachox> delewis, of course opensolaris was an excuse to replace more engineers with community work [01:37:21] <nachox> :P [01:38:01] <boyd> Anyway, it seems to be that a time-unlimited 5 seat free license for sunray wouldn't kill them [01:38:03] <delewis> boyd: and there really haven't been any announcements, which pisses me off. It's a very subtle change between releases. [01:38:12] <boyd> That's what I mean [01:38:23] <delewis> the same goes for SSGD, which was 'free for evaluation' for a month or so. [01:38:32] * delewis holds onto his 100-user SSGD license [01:38:53] <boyd> Yes, I didn't bother to get a license for that since I (foolishly) assumed it would stay around [01:39:00] <jpdrawneek> we got a 20 pack licence for edu for 60 - all we got was a peice of paper saying well done [01:39:10] <delewis> boyd: and I think that lends truth the fact that Sun has recognized this strategy of freeing products and generating revenue off of support only isn't working. [01:39:24] <delewis> announcing their fuck-up is a sure-way to attract attention that you indeed, did fuck-up. [01:40:15] *** stevel has quit IRC [01:40:18] <wesolows> you assume that our revenue comes from software license sales [01:40:23] <boyd> Hmm.. Don't forget hardware. What I'm hearing is that at least part of that is the support organisation hasn't been ramped up enough to back up that plan, so customers are seeing poorer support [01:40:31] <wesolows> go have another read through our 10K [01:41:00] <delewis> wesolows: I'd say that depends on the hardware we're talking about. Right now, the Sun Ray division doesn't appear to be very profitable from Sun Ray client sales, so what does Sun do? charge of SRSS RTU licenses. [01:41:10] <jpdrawneek> well find a tec company that does provide good support these days? [01:41:27] *** Jondice has joined #opensolaris [01:41:51] <delewis> as for something like Solaris, which Sun does benefit largely from the hardware sales, they can afford to free that. [01:41:52] <wesolows> which isn't really a good model, since the clients are useless without the software. If we're a Solutions/Systems company, we should sell you the hardware at a profitable price and offer you our software free to complete the solution. [01:42:00] <delewis> wesolows: exactly. [01:42:18] <delewis> unfortunately, that's not the case with SRSS any longer. [01:42:39] <wesolows> well, you don't see 'VP' on my business card [01:42:45] <delewis> (yet I've had futile conversations with employees in here that claim to have some knowledge of the Sun Ray division, and still claim its 'profitable') [01:43:09] <wesolows> since we don't break out our financials that way, they should not be commenting like that [01:43:10] <delewis> and that's contrary to what I'm seeing which is a large amount of Fortunte 500s abandoning Sun Rays for practical reasons (namely, JDS) [01:43:16] <delewis> and .edu's [01:43:18] <wesolows> and in any case whatever they tell you is unreliable [01:44:06] <delewis> wesolows: I haven't put much stock in it :-) [01:44:24] *** Fullmoon has quit IRC [01:46:41] *** sarahj has quit IRC [01:49:08] *** sahafeez has left #opensolaris [01:50:00] *** uebayasi has joined #opensolaris [01:55:03] <delewis> another problem is that Sun owns hardware and software products that compete with one another, which is why a software Sun Ray solution has yet to be provided (outside of Sun) -- probably because they'd see a huge decline in SSGD sales. [01:55:11] <delewis> provided there actually are any SSGD sales. [01:55:34] <delewis> and vice versa with the SSGD RTU licenses, which are in direct competition with Sun Ray/SRSS. [01:57:27] *** karrotx has joined #opensolaris [01:58:44] <nachox> and marketing is not happy unless they change the name of their products every 2 years, top [01:59:03] <nachox> i wonder when they will add "java" to those [01:59:33] <wesolows> Sun has always competed with itself [01:59:38] *** unixware has quit IRC [01:59:43] <wesolows> in a way, it's a byproduct of being an innovator [01:59:56] <delewis> wesolows: true, but in this case, its hurting nothing but the other product line. [02:00:35] <delewis> Sun Ray adoption is dwindling and won't return unless Sun completely changes its stance on several things and SSGD adoption is zilch, anyway. [02:00:49] <wesolows> I hear you [02:01:35] *** mlh has quit IRC [02:02:25] <delewis> right now there's basically two options -- (1) you can kill either of the product lines to make the other profitable, or you can make them both profitable by integrating Sun Ray support into SSGD. [02:02:43] <delewis> the second option would be extremely cool, IMO. [02:03:06] <gdamore> I feel stoopid... what's SSGD? [02:03:16] <nachox> sun secure global desktop [02:03:26] <delewis> gdamore: Sun Secure Global Desktop a.k.a Tarantella (which Sun got from the acquisition of Tarantella, of course) [02:03:27] <wesolows> more integration across our product lines is really the holy grail for a solutions company [02:03:34] <wesolows> it's also something we've never been good at doing [02:03:54] <gdamore> i believe sunray and the tarantella product are intended one day to be integrated. [02:04:11] <nachox> delewis, wasnt there a blueprint about something like that? [02:04:26] <delewis> that'd be awesome, as when someone actually buys a Sun Ray deployment that could move their session on the Sun Ray to a real workstation seemlessly. [02:04:26] <nachox> delewis, something that also included vmware i believe [02:04:31] <gdamore> Tadpole (now General Dynamics) was actually going to implement a remote desktop (M$) client in their Sun Ray implementation at one point [02:04:38] <delewis> and that's always been a barrier to Sun Rays, IMO. [02:05:24] <delewis> nachox: you can use Sun Ray sessions on a workstation, but it's extremely nasty at the moment (makes use of VNC) [02:05:40] <delewis> and it doesn't solve the problem that Sun Ray and SSGD are in direct competition of each other at the moment. [02:06:09] <gdamore> a VNC/Rdesktop client in the Sun Ray terminal firmware would be a nice solution. [02:06:40] <nachox> in reality yes, but i believe ssgd is something you would deploy when you have a workstation infrastructure available and run ray is something you use when you dont [02:06:41] <gdamore> but then I don't claim to "really" understand what the heck the Tarantella piece really does... :-) [02:06:51] <delewis> gdamore: right now the VNC functionality is provided server-side (for each X session for each Sun Ray there's a vnc listener -- woopty doo) [02:07:25] <delewis> gdamore: same thing Sun Ray does, only in software, basically. Takes a display protocol (X11, RDP, etc.) and encapsulates that in its own protocol to be delivered to a client. [02:08:23] <delewis> and you get session support in the process. [02:08:29] <gdamore> yech. ETOOMANYDISPLAYPROTOCOLS [02:08:48] *** chris_d has quit IRC [02:09:19] <delewis> nachox: sure, but I'm always asked "can I move these sessions to my workstation?" whenever I'm doing a Sun Ray deployment for a client. [02:09:20] *** nrubsig has joined #opensolaris [02:09:20] *** ChanServ sets mode: +o nrubsig [02:09:21] <gdamore> From a protocol design standpoint, Sun Ray's protocol is something of a hack. But in practice it works better than all the others. [02:09:34] <delewis> it'd be nice if SSGD had the Sun Ray integration, then Sun could sell two products simultaneously. [02:09:37] <nrubsig> gdamore: define "hack" ? [02:09:40] <gdamore> "softray" :-) [02:09:57] <delewis> "buy our Sun Rays and SSGD and you'll have session support across all of your systems" [02:10:07] <gdamore> hack == random bits put together without a lot of formalized design. as a result there is a lot of inconsistency, etc. [02:10:08] <Tempt> Mmm, SunRay and SSGD [02:10:12] <delewis> "and receive the benefits of thin clients where needed" [02:10:41] <nrubsig> gdamore: isn't that normal for Sun products, e.g. GNU tools in /usr/bin/ are a good example... [02:10:45] <delewis> but unfortunately, Sun would rather those two products duke it out and see which one is left standing (neither will be). [02:10:56] <Tempt> SSGD is like the Citrix access gateway and Terminal Services and X11 and screen all rolled into one. [02:10:58] <nachox> delewis, the problem is with the hybrid implementations which unfortunately is what is available, yes, however, even if you move the sessions you would not be able to use the workstation processing power, the real solution imho is lowering the prices of sunrays and offer a migration plan with the hardware [02:11:01] <gdamore> well, lets not talk about /usr/gnu, etc. :-) [02:11:21] <delewis> nachox: that's another problem -- DTU pricing. [02:11:25] <delewis> it's outrageous. [02:11:29] <nachox> that last sentence probably didnt fully make it to irc :P [02:11:36] <Tempt> Morning everyone, btw. [02:11:41] <nachox> ETOOLONG [02:11:56] <nrubsig> gdamore: I could really vomit about the damn detail that the GNU stuff now squatts /usr/bin/stat and /usr/bin/readlink and now blocks the use of the matching ksh93 builtin commands. [02:11:56] <gdamore> DTU pricing is outrageous? why? its like $300 a seat (less than that, I think.) [02:12:05] <delewis> I've had several experiences with .edu's where they're dumping Sun Rays completely rather than migrating to Sun Ray 2's from their current Sun Ray 1's and just going LTSP/old desktops laying around. [02:12:10] <delewis> or buying new low-end Dells. [02:12:22] <delewis> gdamore: newsflash -- that's the same price as a low-end Dell. [02:12:32] <richlowe> there's less for students to destroy, though. [02:12:35] <nachox> gdamore, 300 is what a new low end whitebox workstation costs [02:12:39] <richlowe> at least, I'd hope the sunray would stand up. [02:12:39] <delewis> and in most cases, .edu's don't require session functionality [02:12:40] <gdamore> yes. but its hard to push the cost any lower. [02:12:42] <delewis> so they can just use LTSP [02:12:43] <Tempt> gdamore: The thing is the cost is DTU + License. [02:13:04] <Tempt> gdamore: There is no license included in your hardware purchase. [02:13:05] <delewis> and the power requirements for Sun Rays aren't really anything to behold. [02:13:06] <gdamore> i can't answer the license fee costs... but the DTU costs are damn near the smallest they can get without a *lot* more volume. [02:13:09] <nrubsig> alanc: ping! [02:13:21] <jpdrawneek> linux termial does seem a very cheap option [02:13:28] <gdamore> huh? Sun Ray 2 is like ~4W. [02:13:35] <delewis> my experience with power consumption on a Sun Ray 1 is that it was no more power efficient that a low-power peecee was with the disk removed. [02:13:38] <alanc> nrubsig: pong [02:13:45] <gdamore> Sun Ray 1 is *very* different from a Sun Ray 2. [02:13:50] <nrubsig> alanc: any new entries in the triage queue ? [02:13:58] <nachox> jpdrawneek, linux terminals are a lie, they are not thinclients [02:14:09] <gdamore> Sun Ray 1 had a microsparc IIep processor, and that sucked a fair bit of power. [02:14:14] <delewis> sure, it is, but a lot of customers out there have Sun Ray 1s or rather, adopting Sun Ray 2s would be no more beneficial than buying low-end Dells and doing LTSP. [02:14:14] <alanc> nrubsig: CR 6599659 Created P3 opensolaris/triage-queue RFE: Fix version propagation in OS/Net Makefiles [02:14:23] <nrubsig> alanc: and ? [02:14:24] <jpdrawneek> but you can get p100 to do anything [02:14:43] <alanc> CR 6599668 Created P3 opensolaris/triage-queue RFE: Move consumers of alias.sh over to ksh93 [02:14:45] <delewis> I know of an .edu that's dropping their entire Sun Ray deployment and doing old desktops+LTSP. [02:14:52] <gdamore> LTSP takes more effort, doesn't have session mobility, lacks integrated remote device support, audio support, etc. [02:14:53] <jpdrawneek> just point it at the back end box [02:15:02] <delewis> gdamore: .edu's don't use session mobility [02:15:05] <delewis> and LTSP has audio support, nowadays. [02:15:06] <jpdrawneek> gdamore: edu does not need that [02:15:10] <gdamore> Besides which, LTSP *sucks* over low band width link. [02:15:13] <delewis> and remote device support [02:15:17] <delewis> gdamore: bullshit. [02:15:25] <jpdrawneek> again edu does not need that [02:15:30] <delewis> using a USB device on a Sun Ray *kills* it. [02:15:40] <gdamore> Ever try using it over a 56K link? I've used Sun Ray over low bandwidth WAN links.... [02:15:42] <jpdrawneek> i been ok with usb [02:15:52] <delewis> jpdrawneek: try doing IO to a flash drive. [02:15:54] <gdamore> I'm assuming LTSP is just X11, isn't it? [02:15:55] <delewis> it's horrific. [02:15:58] <delewis> gdamore: no. [02:16:10] <delewis> its a combination of things now so that you get audio, remote device support, and X11. [02:16:16] <jpdrawneek> delewis: like a usb pen drive? [02:16:18] <nachox> it boots a linux kernel over the net too [02:16:21] <delewis> the only thing that's missing is sessions which .edu's don't actually use. [02:16:32] <gdamore> oh yes, IO on sun ray is horribly slow. but for saving your 1MB data files, or printing your e-mail, its quite ok. [02:16:58] <jpdrawneek> only thing good about true thin is the lack of things to go wrong [02:17:23] <jpdrawneek> but most edu have a mountain of crap they can just keep replacing it [02:17:31] <gdamore> if a thin client dies, you just grab another one off the shelf... nice being truly stateless... [02:17:32] *** alanc is now known as alanc-away [02:17:34] <delewis> so right now, there's zero incentive for an .edu to pick Sun Rays over LTSP/low-end or old systems (perhaps hardware failure, which can still be mitigated on thick clients) [02:17:36] *** jmcp has left #opensolaris [02:17:39] <jpdrawneek> especially with the new rules on disposing IT kit [02:17:40] <nachox> the hardware of LSTP is actually a real computer, not as easy to replace as sunrays [02:17:45] <delewis> gdamore: same goes for a peecee. [02:17:53] <gdamore> no, not true for a peecee. [02:17:54] *** jmcp has joined #opensolaris [02:18:01] <delewis> all that requires is that you change the MAC in your netboot config. woopty-doo. [02:18:03] <gdamore> you have to configure a peecee... [02:18:09] <jpdrawneek> it depends on how many support guys they got [02:18:10] <delewis> gdamore: not with LTSP. [02:18:19] <gdamore> with Sun Ray you don't even touch the MAC. [02:18:24] <delewis> you change the MAC in the config. Throw the peecee up, and boom, you're using X11. [02:18:25] <nachox> gdamore, you only need the pc to netboot [02:18:40] <jpdrawneek> seen sun rays being looked after by one guy [02:18:43] <delewis> gdamore: I know [02:18:44] <gdamore> a lot of older peecees *don't* netboot... without special PXE boot firmware. [02:18:59] <nachox> that can be hacked easily [02:19:00] <jpdrawneek> seen PCs on a grunt to 15 PC ration [02:19:01] <Tempt> jesus [02:19:06] <gdamore> (or a PXE boot cdrom/floppy... which isn't really the same) [02:19:07] <Tempt> give all the users a serial terminal [02:19:10] *** charlesnw has left #opensolaris [02:19:14] <Tempt> and them to STFU if they don't like it. [02:19:17] <nachox> that can be hacked easily1 [02:19:24] <nachox> ouch, damn arrows [02:19:31] <nachox> Tempt, that's sooo 80's [02:19:32] <delewis> gdamore: in any case, Sun Ray deployments are dwindling because of the rationale I've outlined. [02:19:47] <gdamore> i'm disappointed to hear it, but I have another theory for you. [02:19:49] <delewis> so you might have your enlightened Sun Ray philosophy but it doesn't match most of your customers'. [02:20:07] <jpdrawneek> sun rays need a bit of creative thinking and marketing [02:20:15] <gdamore> the theory is that these days folks want streaming video, e.g. for training videos, CNN, etc. [02:20:24] <gdamore> for shit like that, Sun Ray and LTSP suck. [02:20:29] <jpdrawneek> also that you do not need a quad core sli to read emails etc... [02:20:41] <nachox> sun rays will also need a DE that is not a drag soon [02:21:01] <jpdrawneek> solaris needs a DE that does not drag [02:21:02] <delewis> gdamore: there ways to get around that with LTSP. [02:21:13] <gdamore> gnome is fine with Sun Rays... but if you use a low bandwidth link you want to change the default background. :-) [02:21:18] <delewis> you can run a second X11 server on the netbooted peecee and do video locally. [02:21:19] <Tempt> Hmmm [02:21:26] <Tempt> You might have a point [02:21:29] <jpdrawneek> LTSP can use the cpu if there a hdd in it i have seen [02:21:31] <Tempt> Video on sunray is a non-starter. [02:21:40] <delewis> I know of several LTSP deployments that have very customized solutions to address this. [02:21:43] <jpdrawneek> its on the cards [02:21:46] <nachox> gdamore, gnome needs lots of memory which means you need lots of servers in your backoffice [02:21:54] <richlowe> great, this again. [02:21:56] <gdamore> delewis: of course, but all this presumes you *want* to hack it. if you're hacking the PC, why bother with thin client at all? [02:22:19] <jpdrawneek> because teh PC can be crap but still run ok programs [02:22:34] <delewis> gdamore: same rationale for Sun Ray. required maintenance and administration is zilch on a per-system basis after the customization process. [02:22:39] <richlowe> and what kind of crazy ass customer would *prefer* a solution they have to largely engineer themselves? [02:22:42] <richlowe> wait, don't answer that. [02:22:45] <gdamore> Tempt: video on Sun Ray is achievable. It requires investment. And an mpeg offload chip on the Sun Ray. Trust me... I've been involved with this exploration in *DETAIL* [02:22:46] <jpdrawneek> its all about cost - its cheaper to hack an old PC than get a sun ray [02:22:48] <delewis> richlowe: .edu's [02:22:54] <Tempt> richlowe: The sort of customer who buys SAP? [02:23:16] <gdamore> jpdrawneek: that is only true for small numbers of PCs. [02:23:17] <Tempt> gdamore: Oh, I can see many ways it could be *possible* [02:23:21] <delewis> gdamore: is there anyway to get acceptable video performance with current Sun Ray DTUs? [02:23:22] <Tempt> gdamore: Trust me, I can understand that. [02:23:30] <gdamore> start looking at 1,000's, and the labor cost dominates. [02:23:31] <Tempt> gdamore: But it won't happen for a long time. [02:23:33] <jpdrawneek> edu do a lot of crazy things [02:23:38] <delewis> as in is there anything that can be done with just newer firmware and SRSS? [02:23:52] <gdamore> delewis: not really... at least not without custom software. [02:24:05] <gdamore> custom *firmware*. which only Sun can upload. [02:24:13] <delewis> custom software, ok, that can be done if there's some sort of offload support. [02:24:20] <gdamore> the alchemy cpu in the SR 2's is actually fairly capable. [02:24:27] <delewis> it'd be trivial to write a new vo for MPlayer, for example, to get that offload capability if its there. [02:24:28] <Tempt> The MicroSPARC II was not renowned for playing full screen video. [02:24:43] <delewis> what about Xvideo? [02:24:45] <gdamore> the uSPARC-IIep was terrible... 100MHz only. [02:24:46] <Tempt> I've had some video success on sunray with showmetv [02:24:47] <nachox> what is the cpu inside those sunrays? [02:24:53] <gdamore> Xvideo is supported... but it only does YUV decode right now. [02:24:55] <jpdrawneek> so thats why the sun rsy 2 does not use it [02:25:06] <delewis> Tempt: ShowMeTV uses several private interfaces to get some level of offload. [02:25:07] <jpdrawneek> sun ray 2 is amd genod thingy [02:25:19] <Tempt> delewis: ShowMeTV is awesome. [02:25:20] <delewis> this private interfaces are not documented *at all* [02:25:27] <delewis> s/this/these/ [02:25:50] <gdamore> jpdrawneek: no, Sun Ray 2 started because of RoHS. infineon, the foundary for the copernicus cpu in Sun Ray 1gs, declined to do a lead-free respin. [02:25:54] <delewis> gdamore: and that's needed is decent Xvideo support on the DTU side. [02:25:57] <Tempt> Instead of worrying about open-sourcing the next craipshait of craip, why don't they open ShowMeTV so people can rip the goodness out [02:26:12] <delewis> if there's someway SRSS can directly access the graphics hardware in the DTU, Xvideo would work nicely. [02:26:19] <gdamore> delewis: you need more than that. THe problem is that you have to transmit the video *stream* over the network. [02:26:24] <jpdrawneek> gdamore: teh sun ray 2 have amd inside [02:26:46] <gdamore> AMD Alchemy Au1550. 550 MHz. Want to know more? [02:26:49] <jpdrawneek> gdamore: thats how there doing the vpn client [02:26:56] <gdamore> And they aren't AMD anymore. AMD sold that division to RMI. [02:27:03] <Tempt> Man, you can purchase a $20 DVD player that play DivX as well [02:27:15] <Tempt> it's all on one crappy chip (which must cost around $1) [02:27:15] <nachox> gdamore, that's x86 right? [02:27:20] <gdamore> No, MIPS. [02:27:33] <jpdrawneek> still fast [02:27:34] <Tempt> It won't be hard to put a dedicated video chipset into the thing. [02:27:39] <gdamore> (32 bit only... its a nice little chip. ) [02:27:46] * gdamore did the NetBSD port for the Au1550. [02:27:47] <nachox> that cpu is not capable of doing at least some processing? [02:27:58] <Tempt> Heh, run NetBSD on your SunRay. [02:28:08] <gdamore> I ran a Sun Ray client on top of NetBSD. :-) [02:28:19] <nachox> it's like your toaster but cooler [02:28:50] <gdamore> Tempt: the problem isn't the video chipset... it isn't getting access to the hardware from the server... its *the bandwidth* [02:29:20] <gdamore> doing the decode on the server would suck. so you need to decode/decompress the MPEG natively on the DTU. *that* is a bit tricky. [02:29:41] <gdamore> especially since then you get into patent wars, codec confusion (which codecs do you support), etc. [02:29:46] <delewis> MPEG? there are other formats out there. [02:29:47] <nachox> you mean you cant stream video over the existing 100mbps network interface? [02:29:52] <delewis> the only option is server-side decoding. [02:30:03] <gdamore> you can stream it *decompressed*, but it will suck... [02:30:24] <delewis> Sun needs to throw decent hardware in the Sun Ray then :-) [02:30:25] <gdamore> server-side decode will eat up *way* to much bandwidth. [02:30:44] <gdamore> its *not* a hardware problem... but an MPEG decode assist in the DTU would help. [02:30:46] <richlowe> push it all through aalib on the server side. [02:30:48] <delewis> the nasty thing with client-side decode is that Sun will only support a small subset of formats. [02:30:58] <gdamore> delewis: right. [02:31:01] <Tempt> gdamore: 100Mbit ethernet is fast enough. [02:31:11] <nachox> not to mention it will limit the clients that can be served with one server [02:31:19] <delewis> with server-side decode, existing software would be customized for DTU-side playback. [02:31:31] <gdamore> Tempt: do the math. only for tiny windows, or images where the pictures don't change much. [02:31:38] <delewis> and the existing software supports many, many more formats than Sun could possibly offer support for. [02:31:40] <gdamore> trust me... this is the *state* of Sun Ray video *today* [02:31:54] <delewis> especially given Sun's reluctance not to use patent-encumbered formats [02:31:59] <delewis> s/not // [02:32:01] <nachox> what about something hybrid? some formats get decoded on the DTU and the unsupported at the server? [02:32:04] <Tempt> Sun Ray is dead! Long live Sun Ray! [02:32:08] <gdamore> delewis: yes, so we wound up designing (but not implementing) a code download mechanism in the Tadpole client.... [02:32:36] <gdamore> nachox: yes, that would work. its achievable... but it requires a lot of careful design. [02:32:55] <gdamore> then there are some nasty protocol issues... like what happens to the stream when you remove your smartcard? [02:33:03] *** kFuQ has quit IRC [02:33:36] *** kFuQ has joined #opensolaris [02:33:45] <gdamore> Sun may still be working with GD on a solution to this problem... I don't know... but there was serious investigation in this direction... mostly targetted for the US government. [02:33:53] <nachox> what is done now with the sesion image? [02:34:09] <delewis> TBH, video was a huge issue for a Sun Ray deployment I was going to do recently for an organization. [02:34:10] <gdamore> If the US gov't got really interested, they could buy enough DTUs to make up for any lost commercial sales *trivially* [02:34:17] <delewis> they ended up backing out of it and doing LTSP instead. [02:34:51] <delewis> they'd rather spend the next two months writing a very customized LTSP solution than giving users Sun Rays that can't do video. [02:35:09] <gdamore> video is a huge issue for *everyone* doing thin-clients right now. nobody has a good solution. [02:35:24] <nachox> delewis, and who exactly would support that? [02:35:42] <jpdrawneek> they don't ask questions like that [02:35:56] <nachox> they should, it's part of the cost [02:36:02] <jpdrawneek> if its edu it great - never think that far ahead [02:36:04] <delewis> nachox: themselves. non-profits do all sorts of crazy things and generally have people that don't do by-the-book, but still know how to get things done. [02:36:24] <delewis> and for the majority of things they're using, they don't have support, anyway. [02:36:25] <nachox> delewis, that almost never scales infortunately [02:36:37] <nachox> *unfortunately [02:36:37] <delewis> nachox: 50 users, roughly. [02:36:39] <jpdrawneek> does not have to [02:36:45] *** Lusitanian has quit IRC [02:37:05] <jpdrawneek> with non-profit scalingis a non-issue [02:39:34] <gdamore> Sun Ray doesn't make sense for small numbers of users. [02:39:40] <nachox> nod [02:39:42] <gdamore> But when you have 100's or 1000's of them, it does. [02:40:00] <jpdrawneek> it depends what there doing [02:40:04] <delewis> it sort of does in this particular organization. There's a high-level of per-system maintenance with a small IT department. [02:40:11] <nachox> it doesnt make sense when you dont care about sesions [02:40:23] <delewis> and most of that maintenance could be cut back if every system had its disk removed. [02:40:28] <richlowe> it does maintenance/power wise. [02:40:44] <jpdrawneek> factor in ebay [02:41:19] <delewis> you guys should get out of the Fortune 500s and see what much smaller customers are faced with. [02:41:26] <Tempt> urk [02:41:27] <Tempt> No way [02:41:35] <nachox> hehe [02:41:38] <jpdrawneek> yep - the buggers will pay for nothing [02:41:40] <gdamore> "smaller" customers usually just run Windows. [02:41:43] <Tempt> I've done small business and it is the land of cheapskate fucktards who will do everything to avoid paying their bill [02:41:55] <Tempt> I'm *NEVER* going to be camped in reception waiting for a cheque again [02:41:58] <delewis> I'm not suggesting working for a smaller customer is better, but you clearly don't understand the problems they're faced with. [02:42:05] <jpdrawneek> and they can always do better than you [02:42:16] <delewis> and if Sun doesn't understand small customers, Sun won't be able to market a product to them. [02:42:19] <gdamore> delewis: don't confuse a lack of understanding with a lack of caring. :-) [02:42:40] <jpdrawneek> Sun has never understood the small customer [02:42:52] <nachox> actually they have the same needs big companies do but with a very small fraction of the budget [02:42:56] <jpdrawneek> also the UK vars are useless [02:43:00] <gdamore> Sun doesn't care about small customers..... they don't represent a target market. [02:43:07] <delewis> nachox: and that makes their needs very, very different. [02:43:39] <delewis> gdamore: unfortunately, that's a problem as a company like Dell presents products that all markets have access to. [02:43:49] <nachox> delewis, most end up with hacked up solutions and cheap-o-cluster [02:43:50] <gdamore> Small customers don't/won't run a UNIX/Solaris. They can't justify a full time system admin. (The fact that they need one anymore than they do for Windows is a fallacy, but a commonly held one.) [02:44:02] <delewis> if Sun stays cornered in the land of huge IT budgets, they're missing out on a segment of the market. [02:44:19] <jpdrawneek> windoze need a min of 2 admins [02:44:22] <delewis> gdamore: bullshit. Most of those smaller customers are running Linux and do employ some sort of fulltime sysadmin to manage those systems. [02:44:41] <delewis> Windows isn't applicable to a small shop any longer for same reasons Sun Rays aren't. Money. [02:44:41] <jpdrawneek> delewis: where you from again? [02:45:02] <delewis> jpdrawneek: Jackson, Tennessee, USA. [02:45:15] <gdamore> delewis: how small are we talking about? [02:45:18] <jpdrawneek> k: land of europe is not quite like that [02:45:30] <jpdrawneek> uk is really really bad on that front [02:45:45] <nachox> delewis, i dont agree there, check the situation in developement countries and china where there is piracy [02:45:46] <delewis> gdamore: 2-5 person IT department with a 50-100 user organization. [02:46:03] <gdamore> that's not "small". That's midsized. [02:46:19] <gdamore> small = < 50 users, IMO. [02:46:26] <delewis> non-profits over here (especially .edu's) are extremely interested in cost-effective solutions meaning non-Windows, but rather Linux. [02:46:32] <delewis> gdamore: fair 'nough. [02:46:35] <jpdrawneek> ya - usa def of small is evry body else mediam [02:46:53] <jpdrawneek> ya non-profit - main stay of linux [02:47:17] <gdamore> non-profits that don't want to pay for anything aren't interesting customers from a business perspective. [02:47:34] <delewis> and the thing a lot of non-profits have is some level of expertise (perhaps in implementing half-ass solutions) and a lot of labor time that can be involved in implementing these cost-effective, half-ass solutions. [02:47:54] <richlowe> and these days, Sun is *great* at the half-ass solution. ;) [02:47:55] <jpdrawneek> step out of non-profit is windoze till you hit the larger entrpise [02:48:01] <nachox> gdamore, only non-profit it end up in fortune 500 companies eventually and end up implementing whatever they are used to using [02:48:16] <gdamore> its called "volunteers"... and you get a lot of people that "volunteer" to be a webmaster or some shit, who think they can do it just because they've read the HTML bock... [02:48:41] <jpdrawneek> ya - thats the whole reason i got a job :) [02:48:54] <richlowe> so you sell them a system that requires less in the way of maintenance, and win. [02:49:09] <gdamore> no, that's a fallacy. you won't win. [02:49:25] <gdamore> the customer won't appreciate the lower maintenance, especially if it costs even $1 more. [02:49:25] <jpdrawneek> but for some strange reason sun rays have not taken over the world ? [02:49:43] <delewis> richlowe: unfortunately, that doesn't work with non-profits that have some level of expertise internally. They'd rather shave off the initial cost hurt and take the maintenance hit because they have "expertise". [02:49:56] <gdamore> and if they _do_ pay for it, then they go away, and you never get any more out of them... because they don't need to make more hardware purchases. [02:49:59] <jpdrawneek> you get odd compaines which just look at up front costs [02:50:15] <delewis> jpdrawneek: a lot of non-profits are that way, unfortunately. [02:50:23] <jpdrawneek> tell me about it [02:50:25] <delewis> I'm not saying its a good business model for someone to have. [02:50:30] <gdamore> their labor is often very cheap or free. [02:50:39] *** edp has joined #opensolaris [02:50:50] <delewis> and I've implemented plenty of decent solutions for non-profits that minimized maintenance and they still wanted to do half-ass hacks to it. [02:50:57] <jpdrawneek> they would have every body in second hand bentles if they could get them cheap [02:51:07] <nachox> initial cost is an important factory simply because you cant spend money you dont have even it you might have save some eventually, you need things working yesterday [02:51:09] <jpdrawneek> who care about the huge cost to run them etc.... [02:51:12] <gdamore> the point is, why would a big company like Sun give a rat's fart about these non-profits... they can't tolerate margin, won't pay for better equipment, and don't want to buy support contracts... [02:51:37] <jpdrawneek> its where your trainee grunts come from [02:51:37] *** gaber has quit IRC [02:51:53] <jpdrawneek> its called mind share [02:51:56] <delewis> I, myself, am one of those grunts. [02:52:00] <jpdrawneek> its why linux is winning [02:52:03] <nachox> i agree [02:52:19] <delewis> and I see other people at these non-profits that refuse to acknowledge Sun because of that stance you just stated. [02:52:25] <jpdrawneek> you need a setup quick - linux is the answer etc... [02:52:44] <gdamore> linux is never the answer... but I digress.. [02:52:53] <jpdrawneek> proof of concept have a nasty habit of becoming the production box [02:53:35] <jpdrawneek> seem to many peices of tat running the buisness that were a 5 min hack [02:54:07] <nachox> cant you guess why sun has had so many solaris downloads ever since opensolaris was released? opensolaris was great marketing but they downloaded solaris 10 because it was free [02:54:26] <Jiko> linux is often the answer, and saying that its not is just burying your head in the sand [02:54:37] <Jiko> it may not be the best answer, but it's often what people choose [02:54:53] <jpdrawneek> because you pretty sure its going to work [02:55:04] <jpdrawneek> just - but i have not got time ...... [02:55:08] <jpdrawneek> etc.... [02:55:24] <nrubsig> dmarker: ping! [02:55:30] <gdamore> i've seen pieces of shit like that running at major Fortune 500 companies.... [02:55:40] * gdamore remembers a certain incident at Qualcomm.... [02:55:50] <jpdrawneek> ya - knock something up [02:56:02] <jpdrawneek> wait for the buget to be approved [02:56:28] <jpdrawneek> think about doing it properly ..... one day ..... maybe ..... [02:57:03] <nachox> tell me about it, yesterday i learnt that the production nis server of a big company here was one of computers in the test lab [02:58:11] <jpdrawneek> this is teh problem people wont pay for software, especial if you can get something for free [02:58:48] <jpdrawneek> and then the bugger don't want to pay for support, especial if you can get it for free [02:59:24] <jpdrawneek> software is a bad place to be ..... [03:00:00] <jpdrawneek> got a new BTEC national Diploma - the cost of the pay software for it is silly [03:00:31] <jpdrawneek> you can't recoup it in the 5 year life of the course [03:01:00] <jpdrawneek> guess how quickly they ran to free/FOSS software [03:01:05] *** Fish has quit IRC [03:03:20] <jpdrawneek> sleep time [03:03:43] *** jpdrawneek has quit IRC [03:03:50] <Tempt> LeftWing: PING [03:04:56] <Tempt> nachox: Hey, we had something like that too. Sea of HP workstations, one just happened to be the NIS master [03:05:42] *** nrubsig has quit IRC [03:06:16] * jmcp recaffeinates [03:06:54] <nachox> Tempt, the first mistake was actually still be using nis btw [03:08:13] <Tempt> hmm [03:08:17] <Tempt> LDAP wasn't around back then [03:09:01] <masta> nis is still nice [03:09:22] <nachox> it is unsupported, period [03:09:35] <richlowe> eh? [03:10:03] <nachox> well, eoled [03:10:21] <Tempt> So was the SPARCstation 1+ [03:10:46] <Tempt> That's the thing about history, you can't look back at a ten year old deployment as say "Geez, they should have used T2000s and LDAP" [03:11:32] *** bengtf_ has quit IRC [03:11:41] <nachox> this particular deployment is definitely not 10 years old [03:13:03] *** simford has joined #opensolaris [03:22:44] <Doc> btw, when are we going to rename this channel? [03:23:55] <nachox> anyone remembers what rolands web page is? [03:25:32] <richlowe> Doc: come on then, I'm waiting to see which punchline is coming. [03:25:49] <Doc> punchline? [03:26:13] <Doc> i'm just suggesting that we'd get far more ppl if we had Java somewhere in the channel name [03:28:10] *** lloy0076 has joined #opensolaris [03:28:22] <richlowe> ah, that one. [03:28:57] <lloy0076> Is docs.sun having issues? I can get to docs.sun and do searches but getting to the results of a search (eg http://onesearch.sun.com/search/docs/index.jsp?col=docs_en&locale=en&qt=solaris+10+tmpfs) seems to be taking aeons. [03:29:18] <Doc> use google to search it [03:29:42] <lloy0076> Doc: Yeah, but then I get to go to http://docs.sun...and get the same problem). [03:30:20] <lloy0076> I'll get one page in Google's cache but of course, its links point to docs.sun :P [03:30:22] <richlowe> no, then you hit the "Cached" link. [03:30:39] <richlowe> and then you try and dig out a cached copy of the link destinations. [03:30:42] <Doc> apparently docs.sun.com has worked well in the past [03:30:49] <richlowe> Doc: how long in the past? [03:30:51] <Doc> once in 1998, and twice in 2003 from what i've been told [03:30:58] <richlowe> certainly not in 2003 [03:31:07] <Doc> each time for about 5 minutes each [03:31:42] <lloy0076> heh [03:31:58] <lloy0076> I seem to have stumbled across an offsite mirror of it of some description. [03:32:24] <Doc> docsview.eng works much better, although you do need to work for sun to get at it [03:32:34] <nachox> i had relatively few problems with it, some pages take a little too long to load, only that [03:33:37] *** bengtf_ has joined #opensolaris [03:44:33] *** lloy0076 has left #opensolaris [04:00:18] <CIA-16> wyllys: 6467709 spnego_gss_inquire_cred() can recurse indefinitely causing crash [04:04:38] <delewis> just put in an 'Open Hardware Documentation' request for XVR-1000 and MAJC technical documentation. [04:05:03] <delewis> given that's one of the framebuffers Xorg currently does not have support for and the SPARC graphics team isn't planning on porting to Xorg. [04:07:23] <masta> nice [04:09:45] <dlg> delewis: do it [04:10:01] *** alobbs has quit IRC [04:10:02] <delewis> dlg: already done. [04:10:23] *** alobbs has joined #opensolaris [04:10:40] <delewis> the XVR-500 is already being 'considered' and AFAIK Sun designed the XVR-1000 (it was one of the last framebuffers Sun did R&D on) [04:10:41] <dlg> cool [04:10:46] <dlg> whats majc? [04:10:51] <dlg> the chip on the xvr? [04:10:58] <nachox> nod [04:11:01] <delewis> and even the bulk of the XVR-1000, which is the MAJC processor was created by Sun. [04:11:15] <delewis> so Sun *should* own most of the IP associated with the XVR-1000. [04:11:16] *** gaber has joined #opensolaris [04:11:52] <jbk> evening [04:12:01] <nachox> http://en.wikipedia.org/wiki/MAJC yes Java again!! :P [04:14:39] <delewis> http://24.176.104.6/~dlewis/xvr1000_whitepaper.pdf (XVR-1000 architectural whitepaper if anyone is interested) [04:14:44] <delewis> I have the XVR-4000 WP, too [04:15:30] <Auralis> the 4000 is a very slick board [04:15:45] <delewis> http://24.176.104.6/~dlewis/xvr4000_architecture.pdf [04:16:58] <Tempt> ooh, XVR4000 [04:16:59] <Tempt> gimmeh [04:16:59] *** ashner has joined #opensolaris [04:18:33] <nachox> and then nvidia came :) [04:19:35] <delewis> nachox: XVR-1000 and XVR-4000 both have excellent image quality that's comparable to any Nvidia professional-level card. [04:19:45] <delewis> (comparable to 3dlabs is another matter, though) [04:20:00] <nachox> delewis, what about the cost [04:20:13] <delewis> they're EOL'd. [04:20:40] <delewis> as for the 3dlabs cards, they still have a market, otherwise the Realizm line wouldn't exist. [04:21:01] <nachox> back in the day i mean, though i'm guessing nvidia had nothing to compete then [04:21:07] <delewis> nope. [04:21:10] <delewis> this was 2002-2003. [04:21:18] <delewis> Quadros were fairly primitive then. [04:22:06] <delewis> XVR-1000 was a fairly pricey card (roughly $3k, IIRC) [04:22:21] <Tempt> The XVR-4000 was a pricey card [04:22:27] <Tempt> Given you had to buy an 880 to use it. [04:22:46] <delewis> pricey is Infinite Reality. :-) [04:22:59] <dlg> is the 4000 the zulu thing? [04:22:59] <delewis> which even the XVR-4000 doesn't compete with. [04:23:05] <delewis> dlg: yes. [04:23:06] <richlowe> dlg: yes. [04:23:09] <dlg> hahaha [04:23:09] <dlg> nice [04:23:10] <Tempt> Now, for lunch. [04:23:31] <nachox> zulu is now the upgrade with zones project thing :) [04:23:45] <richlowe> it's still less overused than 'mars' though. [04:24:15] <nachox> mars? [04:24:29] <richlowe> it seems sun have called one of pretty much everything 'mars', at least once. [04:24:55] *** uebayasi has quit IRC [04:26:19] <nachox> they should keep using things like happymeal :) [04:26:32] <nachox> easy to remember [04:29:34] <FunkyELF> whats up guys...I can't install OpenSolaris on my machine because it doesn't have SSE and there is a bug right now. If I were to install Linux with ZFS-FUSE would I later be able to use OpenSolaris whenever the bug is fixed without having to start the filesystem over from scratch? [04:31:18] <Triskelios> FunkyELF: there's no guarantee that ZFS-FUSE will be compatible [04:31:38] <Triskelios> FunkyELF: it can work but you'd need to create the EFI labels by hand [04:31:53] <Triskelios> FunkyELF: b70 has the SSE problem fixed, anyway [04:34:27] <richlowe> I don't think it does. [04:34:41] <FunkyELF> Triskelios, definately does not [04:36:11] <FunkyELF> Triskelios, http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6332924 .... the "Fixed in" field is empty [04:36:52] *** sahafeez has joined #opensolaris [04:38:00] <sahafeez> could someone with zfs guru status post a reply to the lame ass zfs vs linux etc all on digg.com [04:40:21] <Triskelios> FunkyELF: oh, that's lame [04:40:30] *** theRealballchalk has joined #opensolaris [04:41:10] <richlowe> and checking the media, it wasn't built with the older as to work around it, either. [04:41:23] <FunkyELF> Triskelios, really lame. Its not like I boot it, install it, it doens't work but I can fix it.......no, it doesn't boot at all [04:42:06] <Triskelios> richlowe: yeah,http://menzonius.wordpress.com/2007/08/19/opensolaris-not-ready-yet/ says b70 still doesn't boot [04:42:07] *** theRealballchalk has joined #opensolaris [04:42:56] <nachox> it doesnt boot in what systems? via ones? [04:43:00] <Triskelios> FunkyELF: if you're planning to reinstall later you can use S10 now... or even b65, bugs and all [04:43:05] <richlowe> nachox: cpus without SSE [04:43:16] <richlowe> nachox: the assembler is rather two conservative with tagging objects based on the hardware features they ned.e [04:43:22] <richlowe> it doesn't pay attention to conditional execution. [04:43:30] <richlowe> so it thinks libc needs SSE, because it thinks fpcw does'. [04:43:35] <nachox> richlowe, SSE was introduced in pentium2 right? [04:43:44] <richlowe> III I think. [04:43:45] <Triskelios> nachox: no, P3 [04:43:47] <richlowe> but older athlon's don't have. [04:43:49] <richlowe> it [04:44:02] <dlg> is sse floating point only stuff? [04:44:28] <FunkyELF> yeah, I have an athlon thunderbird without it [04:44:49] <richlowe> up until b65, the bits were built with an older as, so this problem wouldn't happen. [04:44:58] <richlowe> for various (semi-accidental) reasons, that changed. [04:45:39] <FunkyELF> actually, I tried using adobe premeier or something and it requried it.....thats what turned me on to open source. The fact that if you didn't have a particular instruction set you could just re-compile. Now comes along an open source OS which won't even boot the live cd because I don't have SSE [04:45:52] <nachox> hmm, system older than pentium III might be a drag to use anyway i guess [04:46:24] <FunkyELF> nachox, just for a file server. I want to use ZFS with RAID-Z. don't need a powerfull CPU I hope [04:46:51] <FunkyELF> 1.33GHz Athlon T-Bird should be enough for a file server [04:46:55] <sahafeez> 64 bit and lots of ram [04:47:16] <richlowe> sahafeez: depends on how heavily used it would be [04:47:47] <nachox> the sweetspot used to be like 2gb a while ago iirc [04:47:57] <sahafeez> zfs on none 64 bit sucks ass [04:48:08] <sahafeez> and less then 1gb of ram sucks ass [04:52:00] *** bengtf_ has quit IRC [04:57:20] *** uebayasi has joined #opensolaris [04:58:39] *** laca has quit IRC [05:04:24] *** deedaw has quit IRC [05:04:29] <FunkyELF> this is a local file server [05:04:44] <FunkyELF> just one one desktop and one laptop accessing it [05:05:07] <delewis> FunkyELF: irrelevant. bit-ness is important here. [05:05:08] <FunkyELF> right now my fileserver is on a 1.4GHz Linux box [05:05:39] <FunkyELF> what if I don't use compression? Any better? [05:06:32] <delewis> no, the fact is ZFS requires more virtual memory than physical memory, and that implies you need a 64-bit CPU to use ZFS optimally. [05:06:49] <delewis> and you need a lot of physical memory, as well (>1GB is ideal) [05:07:19] <e^ipi> I don't personally understand the drive to scale down instead of up [05:07:37] <sahafeez> i want less stuff [05:07:43] <FunkyELF> didn't read that in the "ZFS is the last word in filesystems" propaganda......perhaps it isn't [05:07:45] <Triskelios> e^ipi: makes things more accessible [05:07:51] <e^ipi> solaris isn't made for piece of crap 300mhz pentium2's [05:08:38] <sahafeez> like all things it is in context. zfs has its place and is very much the best filesystem given the right hardware [05:08:47] <richlowe> I still think you're all crazy. [05:09:04] <richlowe> a bunch of my files are on a crappy machine, on zfs, and in the use I give it, it's just fine. [05:09:11] * sahafeez is quite nuts [05:09:14] <richlowe> thus, y'all are insane, because anecdotal evidence is totally valid. ;) [05:09:18] <Triskelios> FunkyELF: I don't think you'll have too many problems with just a few disks anyway [05:09:19] *** kFuQ has quit IRC [05:09:24] <e^ipi> sahafeez: less stuff implies one really fast computer [05:09:33] <e^ipi> not a mountain of crappy obsolete junk [05:09:38] <Triskelios> FunkyELF: and not many simultaneous users [05:09:41] <sahafeez> yes and no [05:10:16] <sahafeez> in my case i said fuck it and have an external hd for my latop and a server in a colo and a usb disk box that is in the apple base station - no turend on servers. [05:10:23] <sahafeez> rysnc dups stuff. [05:10:24] <FunkyELF> well, I have an A7V133 laying around. It has the 1.3GHz athlon t-bird with 4 IDE controllers (8 hard drives). I would buy and use 4 drives, 500Gb each [05:10:25] <sahafeez> and it is simple [05:10:46] <sahafeez> should work....toss ram at it [05:11:09] * FunkyELF thinks it has 512+256 [05:11:38] <delewis> should work, won't be pretty, though. [05:11:50] <FunkyELF> Triskelios, you said S10....whats that? [05:11:53] <nachox> does anyone know how old roland is? [05:12:04] <jbk> i'd like one of those via chips w/ a small internal drive for a firewall/print server, but dunno how well the nics would be supported [05:12:24] <Triskelios> FunkyELF: you'll want 1GB RAM at a bare minimum for that many disks, we ran into some kind of I/O bottleneck on our 32-bit machine, results may vary [05:12:47] <Triskelios> FunkyELF: I meant Solaris 10 [05:13:06] <FunkyELF> is that free? [05:13:29] <Triskelios> has always been [05:13:42] *** bengtf_ has joined #opensolaris [05:14:00] <delewis> nachox: http://www.flickr.com/photos/jimgris/408133591/ (he's the one in the middle) [05:14:08] *** vmlemon has quit IRC [05:14:09] *** sioraiocht has quit IRC [05:14:41] <Tempt> Mm, lunch. [05:14:42] <nachox> delewis, mid 30's? [05:14:59] <delewis> nachox: that'd be my guess. [05:16:35] *** sioraiocht has joined #opensolaris [05:16:35] <nachox> hmm, i need his mail, lets check his opensolaris profile [05:16:50] * FunkyELF goes to bed [05:18:14] <nachox> not there, crap [05:18:23] <nachox> lets check the mailing lists :) [05:19:59] <boyd> delewis: Funny how people seem never to look like you imagine... I was expecting something more.... komodo dragon-like :) [05:20:05] <jbk> haha [05:20:08] <Tempt> Heh [05:20:44] <jbk> i always wonder about that myself [05:21:11] <Tempt> ksh integration driven by giant lizard. [05:21:13] <Tempt> news at 11. [05:24:47] *** nrubsig has joined #opensolaris [05:24:47] *** ChanServ sets mode: +o nrubsig [05:24:56] <nrubsig> gdamore: ping! [05:25:09] <nrubsig> gdamore: You're the networking guru, right ? [05:26:10] * nrubsig pokes gdamore [05:28:53] <richlowe> Tempt: now see what you did. [05:29:58] <nrubsig> Anyone familar with ARC stuff ? [05:32:24] <nrubsig> wesolows: ping! [05:33:30] <jbk> anyone familar with a jmicron ide controller? [05:33:35] <jbk> seems to have issues under nevada [05:34:21] *** derchris has quit IRC [05:34:28] *** derchris has joined #opensolaris [05:36:55] * gdamore pokes nrubsig back. [05:37:11] <nrubsig> gdamore: How can I ARC a nightmare ? [05:37:35] <nrubsig> gdamore: e.g. an application which relies in an instable interface of a provider on the internet ? [05:37:40] <gdamore> ? first you have to choose which ARC to use. I don't know if there is a special ARC for dreams... [05:37:54] <jbk> NARC? :) [05:38:04] <nrubsig> gdamore: did you see the "shnote" demo ? [05:38:47] <gdamore> no. [05:38:51] <nrubsig> s/in an/on an/ [05:39:06] <gdamore> but e-mail me the details, and i'll see if I can help direct you. [05:39:29] <gdamore> generally speaking reliance upon "unstable" interfaces is not a great idea. [05:39:31] <nrubsig> gdamore: one sec [05:39:34] *** sahafeez has quit IRC [05:39:44] <gdamore> i actually need to step away from the computer for a while.... [05:40:03] <gdamore> just e-mail me the details, i'll reply, but maybe not til tomorrow. [05:40:22] <nrubsig> Example usage: [05:40:24] <nrubsig> $ ksh93 shnote.ksh put "foo bar" [05:40:25] <nrubsig> SUCCESS: http://pastebin.ca/657301 [05:40:27] <nrubsig> $ ksh93 shnote.ksh get 657301 [05:40:28] <nrubsig> foo bar [05:41:00] <nrubsig> gdamore: my idea would be $ cp shnote.ksh /usr/bin/shnote ; chmod a+rx /usr/bin/shnote # [05:41:01] <jbk> Aug 30 23:15:29 mundens genunix: ID 228648 [05:41:09] <jbk> kern.info ata_set_feature: (0x66,0x0) failed <-- anyone know what that means? [05:41:10] *** slowhog has joined #opensolaris [05:43:38] *** nachox has quit IRC [05:54:51] *** mw46 has joined #opensolaris [05:54:56] *** oninoshiko has joined #opensolaris [05:55:54] <mw46> hello everybody, I've a question about the loopback interface in solaris;-) [05:56:26] *** nrubsig is now known as everybody [05:56:37] <mw46> Linux routes everything 127.0.0.0 to the loopback. [05:56:51] <mw46> How could I get the same behaviour with Solaris? [05:56:53] * oninoshiko lets everybody answer this one [05:56:56] <everybody> mw46: we're all dead, walking corposes. Noone can answer your questions! [05:57:11] * everybody hunts mw46 to get his blood [05:57:22] <jbk> you could add a route [05:57:26] <oninoshiko> 127.0.0.1 is the loopback interface on every OS [05:57:28] *** everybody is now known as blood [05:57:37] <jbk> solaris by default only does it explicitly for 127.0.0.1 [05:57:37] *** blood is now known as nrubsig [05:57:53] <jbk> route add 127.0.0.0/8 127.0.0.1 [05:58:01] <jbk> though i do not know what effect it might have on zones [05:58:40] <jbk> if you're not planning on using them, shouldn't matter [05:58:40] <nrubsig> mw46: AFAIK you want to ask gdamore once he is away again tomorror (e.g. in 16h) [05:58:51] <jbk> though, why do you need such behavior? [05:58:52] <oninoshiko> why would you want more then just .1? [05:59:09] <mw46> jbk: I tried that but it did not seem work:-( [05:59:18] <mw46> my ping still hangs there. [05:59:28] <oninoshiko> then you have something more wrong [05:59:31] <nrubsig> oninoshiko: sometimes you need such stuff to run multiple instances of the same application [06:00:02] <oninoshiko> but, but, but thats what ports are for [06:00:03] <mw46> oninoshiko: For testing I need a lot of different host names for localhost. [06:00:05] *** sponix has joined #opensolaris [06:00:16] <CIA-16> xc211169: 6551956 aac: fibctx->ctx_idx initialization problem, 6551959 aac: close of first context by FSACTL_CLOSE_GETADAPTER_FIB closes all contexts, 6584506 "-- AIF queue overrun" is displayed inaccurately [06:00:21] <nrubsig> oninoshiko: only if the application allows to select different ports [06:00:22] <oninoshiko> thats what the hosts file is for [06:00:22] <jbk> mw46: then perhaps you just need to create aliases on lo0 [06:00:37] <oninoshiko> there are applications that dont 0.0 [06:01:06] <mw46> how long can the lines in the hosts file become. [06:01:07] <mw46> ? [06:01:29] <jbk> no i mean an aliased interface [06:01:42] <mw46> I need like thousands of host names:-( oninoshiko. [06:01:46] <jbk> ifconfig lo0 addif 127.0.0.X netmask + broadcast + up [06:01:54] <oninoshiko> i've never hit a limit... i've assigned 5 or so aliases before [06:02:03] <jbk> well, you had do up to 8192 (I think) aliases [06:02:07] <jbk> just script it :) [06:02:12] <jbk> err can do [06:02:37] <Tempt> One could ask why the fsck you need to do this, but I'd assume insanity. [06:02:57] <mw46> something like a script;-) [06:03:02] <oninoshiko> mw46: i dont see how using more IPs is any more manageable then a little perl [06:03:36] <mw46> Tempt: I need a lot of virtual hosts for apache (but only on the lo interface). [06:05:49] * Tempt shakes head [06:06:27] <mw46> ;-) [06:06:57] *** tg has quit IRC [06:07:04] *** tg_ has joined #opensolaris [06:09:38] <jbk> why not host based entries? [06:09:56] <jbk> i mean are people still using netscape 3.x? (or whatever was the last version that didn't support them) [06:11:59] <Tempt> netscape 1 [06:12:00] <Tempt> :) [06:12:16] *** tg_ has quit IRC [06:12:21] *** tg has joined #opensolaris [06:13:11] <jbk> that would probably be far easier [06:13:28] <oninoshiko> i suppose you could just run a DNS server which is only authoritative for your loopback device... and only listens for requests on it [06:14:41] *** Kernel86_ has joined #OpenSolaris [06:15:24] <Tempt> wow [06:15:32] <Tempt> my employer's antispam strategy includes [06:15:36] <Tempt> blocking all email from a .org domain [06:15:37] <Tempt> :) [06:15:39] <Tempt> Looove it. [06:15:42] <jbk> haha [06:15:45] *** Kernel86 has quit IRC [06:15:57] <mw46> jbk: that seems to work. Thank you;-) [06:16:02] <jbk> you should see the footer my employer puts on all emails leaving the company [06:16:59] *** eboutilier_ has quit IRC [06:17:07] <mw46> jbk: I want/need to use host-name based vhosts. [06:17:24] <richlowe> in which case, you don't need thousands of interfaces, or addresses. [06:17:54] <mw46> I only have to add all of them to one line in /etc/hosts or run DNS. [06:18:49] <mw46> And I need a automatic way to add to that list;-) [06:19:18] <oninoshiko> it shouldnt be a hard script to write [06:20:30] <mw46> as long as you assume /etc/hosts does not have a limit on the line length;-) [06:21:07] <oninoshiko> I've never heard mention of one... [06:22:26] <mw46> Maybe it's easier to just start using a very limited DNS setup;-) [06:23:18] <mw46> Just a wildcard CNAME record and I should be done with it;-) [06:24:52] <Triskelios> mw46: apache doesn't care if its virtual hosts actually resolve to anything, btw [06:27:01] <mw46> apache itself not. But I need different host names for my cgi suff:-( [06:27:08] <mw46> and I need names. [06:28:00] *** duri has quit IRC [06:28:01] *** mw46 has quit IRC [06:34:25] *** duri has joined #opensolaris [06:36:07] *** ste12 has joined #opensolaris [06:38:40] *** dlynes_home has quit IRC [06:39:29] *** ste12 has left #opensolaris [06:43:59] *** [R]itchie has quit IRC [06:44:25] *** [R]itchie has joined #opensolaris [06:46:45] *** mw46 has joined #opensolaris [06:50:43] *** gm152 has quit IRC [06:52:11] *** dlg has quit IRC [06:52:11] *** sponix has quit IRC [06:52:11] *** bengtf_ has quit IRC [06:52:12] *** theRealballchalk has quit IRC [06:52:12] *** karrotx has quit IRC [06:52:13] *** cmang has quit IRC [06:52:14] *** apersson has quit IRC [06:52:14] *** coraline has quit IRC [06:52:14] *** mbalmer has quit IRC [06:52:15] *** AtomicPunk has quit IRC [06:52:15] *** medar has quit IRC [06:52:16] *** SYS64738 has quit IRC [06:52:16] *** jmcp has quit IRC [06:52:16] *** pjlv has quit IRC [06:52:16] *** WickedWicky has quit IRC [06:52:16] *** bondolo has quit IRC [06:52:17] *** sniffy has quit IRC [06:52:17] *** djgregor has quit IRC [06:52:17] *** rachel has quit IRC [06:52:17] *** tsp has quit IRC [06:52:18] *** fungUsm has quit IRC [06:52:18] *** mikaeld has quit IRC [06:52:18] *** dennis has quit IRC [06:52:18] *** Fetch has quit IRC [06:52:19] *** Dar has quit IRC [06:52:19] *** xinkeT has quit IRC [06:52:19] *** trs81 has quit IRC [06:52:19] *** loke has quit IRC [06:52:20] *** rarruda__ has quit IRC [06:52:20] *** lasseoe has quit IRC [06:52:21] *** johnniez_ has quit IRC [06:52:22] *** the-decider has quit IRC [06:52:23] *** moazamraja has quit IRC [06:52:23] *** pde has quit IRC [06:52:23] *** Stric has quit IRC [06:52:23] *** e^ipi has quit IRC [06:52:23] *** kohju has quit IRC [06:52:23] *** Auralis has quit IRC [06:52:23] *** bda has quit IRC [06:52:23] *** pjd- has quit IRC [06:52:23] *** psiekl has quit IRC [06:52:23] *** sparkleytone has quit IRC [06:52:24] *** vipe has quit IRC [06:52:24] *** networkdump has quit IRC [06:52:24] *** Bartman007 has quit IRC [06:52:25] *** Atomic-Punk has quit IRC [06:52:25] *** coffman has quit IRC [06:52:25] *** trochej has quit IRC [06:56:09] <nrubsig> mass extinction ? [06:56:37] * oninoshiko dies [06:57:16] *** coraline has joined #opensolaris [06:57:16] *** sponix has joined #opensolaris [06:57:16] *** bengtf_ has joined #opensolaris [06:57:16] *** theRealballchalk has joined #opensolaris [06:57:16] *** jmcp has joined #opensolaris [06:57:16] *** karrotx has joined #opensolaris [06:57:16] *** cmang has joined #opensolaris [06:57:16] *** apersson has joined #opensolaris [06:57:16] *** mbalmer has joined #opensolaris [06:57:16] *** AtomicPunk has joined #opensolaris [06:57:16] *** medar has joined #opensolaris [06:57:16] *** SYS64738 has joined #opensolaris [06:57:16] *** pjlv has joined #opensolaris [06:57:16] *** WickedWicky has joined #opensolaris [06:57:16] *** bondolo has joined #opensolaris [06:57:17] *** dlg has joined #opensolaris [06:57:17] *** sniffy has joined #opensolaris [06:57:17] *** djgregor has joined #opensolaris [06:57:17] *** rachel has joined #opensolaris [06:57:17] *** tsp has joined #opensolaris [06:57:18] *** fungUsm has joined #opensolaris [06:57:18] *** mikaeld has joined #opensolaris [06:57:18] *** dennis has joined #opensolaris [06:57:18] *** Fetch has joined #opensolaris [06:57:18] *** Dar has joined #opensolaris [06:57:18] *** xinkeT has joined #opensolaris [06:57:18] *** trs81 has joined #opensolaris [06:57:18] *** loke has joined #opensolaris [06:57:18] *** rarruda__ has joined #opensolaris [06:57:18] *** johnniez_ has joined #opensolaris [06:57:18] *** lasseoe has joined #opensolaris [06:57:18] *** kohju has joined #opensolaris [06:57:19] *** psiekl has joined #opensolaris [06:57:19] *** pjd- has joined #opensolaris [06:57:19] *** coffman has joined #opensolaris [06:57:19] *** bda has joined #opensolaris [06:57:19] *** Auralis has joined #opensolaris [06:57:19] *** e^ipi has joined #opensolaris [06:57:19] *** Bartman007 has joined #opensolaris [06:57:19] *** Stric has joined #opensolaris [06:57:19] *** Atomic-Punk has joined #opensolaris [06:57:19] *** pde has joined #opensolaris [06:57:19] *** sparkleytone has joined #opensolaris [06:57:19] *** trochej has joined #opensolaris [06:57:19] *** moazamraja has joined #opensolaris [06:57:19] *** vipe has joined #opensolaris [06:57:19] *** the-decider has joined #opensolaris [06:57:19] *** networkdump has joined #opensolaris [06:57:46] <nrubsig> mhhh [06:57:50] *** ChanServ sets mode: -o alanc-away [06:58:06] <nrubsig> mass... uhm... anti-extinction ? [06:58:44] * oninoshiko lives, again? [07:00:10] * nrubsig raises a knife from a concealed holster and starts stabbing... [07:00:29] * oninoshiko dies [07:00:45] *** nrubsig has left #opensolaris [07:08:20] *** mw46 has quit IRC [07:15:59] *** duri has quit IRC [07:17:56] *** duri has joined #opensolaris [07:18:49] *** oninoshiko has quit IRC [07:19:38] *** duri has quit IRC [07:22:28] *** duri has joined #opensolaris [07:31:12] *** woofman has joined #opensolaris [07:31:29] <moazamraja> *yawn* [07:32:04] <woofman> Heya is anyone else having trouble accessing http://docs.sun.com ? [07:32:39] <Marv|LG> nope works here [07:33:18] <woofman> OMG it started working again.. Must be at my end [07:33:40] <Tempt> moazamraja: heya. New host is up! [07:34:38] <moazamraja> oh cool [07:34:56] <moazamraja> lemme get a new host up over here this weekend (today is thursday here) [07:37:31] <Tempt> moazamraja: No worries, just drop me an email with your preferred hostname and I'll put your zone online [07:38:02] *** estibi has joined #opensolaris [07:38:24] *** Jondice has quit IRC [07:40:58] *** woofman has left #opensolaris [07:43:45] *** Chihan has joined #OpenSolaris [07:49:33] *** sponix has quit IRC [07:52:07] *** sponix has joined #opensolaris [08:15:53] *** slowhog has quit IRC [08:20:34] *** estibi has quit IRC [08:25:37] *** theRealballchalk has left #opensolaris [08:27:22] *** tsoome has quit IRC [08:36:20] *** nostoi has joined #opensolaris [08:41:37] *** bowdengl has joined #opensolaris [08:41:52] *** tamr has joined #opensolaris [08:43:32] *** praseodym has joined #opensolaris [08:47:48] *** mlh has joined #opensolaris [08:49:39] *** bengtf has quit IRC [08:51:56] *** yippi has quit IRC [08:52:09] *** tamr has quit IRC [08:53:15] *** Cyrille has joined #opensolaris [08:55:03] *** Erwann has joined #opensolaris [08:56:30] *** tsoome has joined #opensolaris [08:59:35] <WickedWicky> morning all [09:00:49] <quasi> morning [09:00:55] <WickedWicky> hey :D [09:01:04] <asyd> \_o< [09:01:09] <WickedWicky> when I put this in /etc/ipf/ipnat.conf [09:01:11] <WickedWicky> rdr rtls0 192.168.1.2/32 port 80 -> 192.168.1.3 port 80 tcp [09:01:18] <WickedWicky> that should work, should it not? [09:09:17] *** mbalmer has quit IRC [09:11:56] *** cydork has joined #opensolaris [09:13:51] *** logic has joined #opensolaris [09:19:25] *** estibi has joined #opensolaris [09:28:21] *** Irene has quit IRC [09:29:45] *** rootard has joined #opensolaris [09:40:55] *** bunker has joined #opensolaris [09:48:51] *** MattMan has joined #opensolaris [09:51:55] *** noyb has joined #opensolaris [09:52:56] *** bengtf_ has quit IRC [10:00:12] *** dmarker has quit IRC [10:00:24] *** dmarker has joined #opensolaris [10:02:58] *** mega has joined #opensolaris [10:09:06] *** iMax has quit IRC [10:14:08] *** bengtf_ has joined #opensolaris [10:14:11] *** bengtf_ is now known as bengtf [10:14:33] *** iMax has joined #opensolaris [10:15:30] *** bunker has quit IRC [10:18:12] *** NikolaVeber has joined #opensolaris [10:20:39] *** ludovicp has joined #opensolaris [10:21:45] *** bunker has joined #opensolaris [10:21:46] <Triskelios> hey, can someone with b70 default flash plugin and a microphone test if http://www.kirupa.com/developer/actionscript/microphone.htm works? [10:22:28] <Triskelios> I think the frequency conversion might be wrong, I get totally distorted sound from the mic [10:23:20] <Triskelios> the flash plugin that ships with SXCE is labelled 9.0.47 but is a different binary from adobe's official 9.0.47 plugin [10:30:40] *** peteh has joined #opensolaris [10:31:05] *** KermitTheFragger has joined #opensolaris [10:32:25] *** Gekkko has joined #opensolaris [10:35:43] *** nostoi has quit IRC [10:36:22] *** iceq has quit IRC [10:36:58] <sniffy> Hrmpf [10:39:30] <sniffy> If I have a non-global zone sharing the NIC with the global zone, but both are on different subnets. It doesn't help putting the gateway of the non-global zone in /etc/defaultrouter [10:39:36] *** migi has joined #opensolaris [10:39:59] <sniffy> As the non-global zone's route to that different IP subnet doesn't make it into the routing table of the global zone :/ [10:40:13] <sniffy> Any hints on that? [10:41:17] *** Fish has joined #opensolaris [10:41:52] <Fish> hello [10:43:15] <boyd> Wow... note to self... gmake -j99 is ... harsh on a 1GB box [10:43:16] *** Irene has joined #opensolaris [10:44:35] <sniffy> Aaah add tempif on bge0, setting default, then taking it down works. [10:44:57] <sniffy> Tacky solution, but works. [10:52:08] *** timsf has quit IRC [10:54:11] *** timsf has joined #opensolaris [10:56:43] *** Fish- has joined #opensolaris [10:56:44] *** uebayasi has quit IRC [10:57:55] <Triskelios> sniffy: the global zone won't let you add the non-global one's default route to its routing table? [11:02:53] *** logic has quit IRC [11:04:28] <sniffy> triske, Nope, but adding a temp if in the global zone, then adding the default route, and then removing the temp if works though. [11:09:48] *** Snake007uk has joined #opensolaris [11:12:13] *** Fish has quit IRC [11:16:10] *** Pietro_S has joined #opensolaris [11:16:51] <rootard> sniffy: crossbow will fix this I think. [11:17:20] <sniffy> Well until then this is my "hack" I'll abuse :) [11:18:07] <rootard> Yeah, I have come across the same issue. We ended up just putting an ip on the second subnet. [11:18:29] <rootard> Something to try: from the non-global zone, try to ping the global zone ip. [11:21:40] <sniffy> That doesn't work [11:21:43] <sniffy> Or two sec [11:21:59] <sniffy> Rather I've got an IP on that non-global zone on the same subnet so. [11:22:35] <sniffy> I'm not into separating the zones from the global one that much. [11:22:44] <sniffy> I have a management IP subnet it's all on. [11:23:14] <sniffy> I just don't want to "expose" the global zone to the Internet as it has no real use for it. For patches etc it gets NAT'ed by the Cisco gear. [11:27:05] *** Fullmoon has joined #opensolaris [11:27:22] <rootard> Cool [11:27:31] *** tomww has quit IRC [11:28:06] <rootard> If you add the second subnet ip to the global zone then the non-global zones gets access to the global-zone networks. At least on Sol10. [11:29:35] *** tomww has joined #opensolaris [11:29:39] <rootard> We have different vlans for internal/dmz/storage networks. Our non-global zones are only on a single net but they seem to talk directly with other vlans. :( [12:06:06] *** simford has quit IRC [12:08:44] *** cypromis has joined #opensolaris [12:11:08] *** xOmega has joined #opensolaris [12:16:35] *** kaiwai has joined #opensolaris [12:17:14] *** umdstu has joined #opensolaris [12:31:04] *** kaiwai has quit IRC [12:34:22] *** iceq has joined #opensolaris [13:06:43] *** deather_ has joined #opensolaris [13:09:18] *** Jondice has joined #opensolaris [13:09:19] *** deather has quit IRC [13:09:21] *** deather_ is now known as deather [13:10:49] *** bunker has quit IRC [13:21:50] *** MattMan is now known as MattAFC [13:30:45] *** estibi_ has joined #opensolaris [13:32:40] <bda> http://www.laughbreak.com/stories/survival_of_the_fittest.html [13:35:50] * PerterB laughs [13:36:04] <umdstu> cute [13:36:37] <umdstu> i wouldn' thave that kind of restraint [13:36:41] <umdstu> there would be many accidental deaths [13:38:34] <bda> My love is science is not so much I wouldn't just eat both either. [13:38:39] <bda> s/is/of/ [13:41:22] <JWheeler> now this can't be a good sign, I was trying to compile a package, and gcc had an internal error, so I try again, segmentation fault, try again, dump core, try again, it works.... [13:41:40] <JWheeler> bad ram maybe? [13:41:51] *** Risky has joined #opensolaris [13:42:48] <trygvis> might be [13:45:27] <JWheeler> well, it might be a good excuse to upgrade this machine [13:45:35] <Cyrille> some resource it wanted to use (like a temp dir or something) was in an unexpected state maybe? [13:45:42] <JWheeler> it'll be annoying though, I've already had to replace one pair of dimms on this hist [13:45:45] <JWheeler> *host [13:46:12] <JWheeler> I wouldn't have expected it to be getting internal errors and crashes like that if that were the case though? [13:47:13] *** estibi has quit IRC [13:49:56] <Cyrille> me neither, but if it's really an unexpected state it might fall into a gap in the handling code. [14:00:38] *** lloy0076 has joined #opensolaris [14:01:09] * lloy0076 sigh [14:01:30] <lloy0076> Key combinations CRASH my comp. Still. [14:01:42] <lloy0076> Never happened before b70 [14:06:30] *** alanc_away has quit IRC [14:06:54] *** alanc_away has joined #opensolaris [14:10:17] *** Gekkko has quit IRC [14:12:22] *** master_o1_master has joined #opensolaris [14:13:57] *** sponix has quit IRC [14:17:49] *** praseodym_ has joined #opensolaris [14:23:56] *** umdstu has quit IRC [14:24:48] *** master_of_master has quit IRC [14:27:04] *** laca has joined #opensolaris [14:29:28] <WickedWicky> who can help me with a quick ipf under onnv72 question? [14:33:19] *** praseodym has quit IRC [14:33:59] <ofu> nv72? [14:35:23] <WickedWicky> oui [14:36:25] <WickedWicky> I created a file, ipnat.conf [14:36:29] <WickedWicky> put this in it: rdr rtls0 192.168.1.2/32 port 80 -> 192.168.1.3 port 80 tcp [14:36:39] <WickedWicky> then I ran ipf -f ipnat.conf [14:36:50] <WickedWicky> and got the error: rdr rtls0 192.168.1.2/32 port 80 -> 192.168.1.3 port 80 tcp [14:36:53] <WickedWicky> not that [14:36:58] <WickedWicky> # ipf -f ipnat.conf [14:36:59] <WickedWicky> syntax error error at "rtls0", line 1 [14:37:00] <WickedWicky> this [14:37:08] <WickedWicky> but I dont see what I am doing wrong to be honest [14:40:57] *** ashner has quit IRC [14:44:36] *** Jagoma has joined #opensolaris [14:45:59] *** lloy0076 has left #opensolaris [14:47:54] *** umdstu has joined #opensolaris [14:48:02] *** eboutilier_ has joined #opensolaris [14:48:50] *** eboutilier_ has left #opensolaris [14:49:13] *** jafari has quit IRC [14:49:16] <SYS64738> ipnat.conf ? it's 2000/2001 stuff [14:49:35] <Gekkko[PDA]> burn it1! [14:49:50] *** Risky has quit IRC [14:50:22] <WickedWicky> is it [14:50:50] <flyingparchment> to reload ipnat.conf you une ipnat(1m), not ipf [14:51:59] *** eboutilier has joined #opensolaris [14:52:19] <WickedWicky> hmm [14:53:03] <WickedWicky> that did the trick, thank you very much [14:53:08] <Gekkko[PDA]> . [14:53:22] <WickedWicky> googleing for ipfilter + opensolaris only brought me to ipf [14:55:51] <SYS64738> I should have somethings in c1t1d0s7 how can I do to see inside of it ? [14:57:25] *** MattAFC is now known as MattMan [14:58:51] *** Pietro_S has quit IRC [15:01:47] <iMax> SYS64738: mount it? [15:02:19] <SYS64738> i don't reach to, mount -F zfs /dev/dsk/c1t1d0s7 /zones doesnt worl [15:02:22] <SYS64738> work [15:02:46] <timsf> Does "zpool import" show the pool ? [15:04:28] <SYS64738> damn i can't reconnect [15:04:38] <SYS64738> I hate when it does like this [15:05:29] <timsf> Reconnect to what ? [15:06:06] <SYS64738> I am playing with openvpn client and the ipsec tunnel towards the solaris server was interrupted [15:06:13] <timsf> aah [15:06:36] <SYS64738> I don't know why, cause the ipsec is between routers [15:08:32] *** nachox has joined #opensolaris [15:09:13] <nachox> morning [15:10:02] <SYS64738> what a coincidence while playing with openvpn the internet connection went down [15:10:29] <timsf> hi nachox [15:10:41] <SYS64738> http://www.comune.arezzo.it/ the city hall network is vaporized [15:13:28] <SYS64738> good night [15:17:55] *** sioraiocht_ has joined #opensolaris [15:17:56] *** sioraiocht has quit IRC [15:30:40] <bda> http://research.microsoft.com/camsys/avalanche/ [15:43:40] *** pfa3rh has joined #opensolaris [15:46:18] *** yarihm has joined #OpenSolaris [15:47:12] <trygvis> anyone know anything about how to install opengrok? [15:47:23] <asyd> well, there is a documentation [15:47:36] <quasi> trygvis: follow the docs? [15:47:58] <asyd> I have a quick'n'dirty howto in french that you can follow easily i guess, [15:48:00] <asyd> http://solaris-fr.org/home/docs/opengrok [15:48:14] <trygvis> the docs is rather .. lacking [15:48:24] <Doc> i do my best... [15:48:45] <trygvis> I want to be able to configure the "projects" list that is on src.opensolaris.org [15:50:17] <quasi> asyd: seems fairly simple to understand - even if I don't understand any french [15:52:38] <oxygene> bda: "Any peer can generate new unique combinations from the combinations it already has." seems to be the only new bit in there [15:53:14] <bda> oxygene: And the typical MS vendor-locks. [15:53:15] <oxygene> though given that they didn't state that these new combinations are actually useful, I guess every other redundant coder can do that, too *shrug* [15:53:25] <bda> Did you see the Harvard team working on "BT2.0"? [15:53:43] *** estibi_ has quit IRC [15:54:05] <Gekkko[PDA]> bittorrent 2? [15:54:10] <bda> Ya. [15:54:16] <Gekkko[PDA]> no [15:54:18] <oxygene> hmm.. even freenet does "healing" (recreating redundant packets once it has just enough to rebuild the full file) [15:54:19] <Gekkko[PDA]> but cool [15:54:28] <bda> http://torrentfreak.com/harvard-develops-p2p-client-that-uses-bandwidth-as-currency/ [15:54:38] <Gekkko[PDA]> lol [15:54:43] <oxygene> and a "no-one (needs to) know where the other stuff is" setup [15:55:24] <bda> oxygene: There is a really good Google TechTalk by Van Jacobson on dissemination network/content-addressable network stuff. I think I mentioned in here the other day. [15:55:36] <oxygene> you did [15:55:49] <bda> My mind continues to flee as I age. :\ [15:57:15] <Gekkko[PDA]> bda: very interesting [15:57:55] <bda> http://fe22.news.re3.yahoo.com/s/infoworld/20070830/tc_infoworld/91459 [16:03:29] <tsp> before I go reinventing the wheel, is there a program out there that will act like mv but keep the directory structure intact? [16:04:25] *** tsoome has quit IRC [16:04:28] <bda> Not sure what you're asking. cp -R? rsync -a? [16:05:35] *** communicator has joined #opensolaris [16:07:08] <tsp> I have a big list of duplicate files, and I want to move them somewhere for safe keeping while I see what happens when i delete them [16:07:23] *** RaD|Tz has joined #opensolaris [16:08:11] <bda> cp -R dir1 dir2 or rsync -a dir1 dir2 [16:08:22] <bda> If I'm not misunderstanding your intent, anyway. [16:09:24] *** dunc has quit IRC [16:12:18] <PerterB> I don't think he wants to move the whole subtree, plus he wants to move not copy [16:12:49] <bda> I am confused by "will act like mv but keep the directory structure intact", I guess. [16:12:56] <Cyrille> I still don't understand what he wants because he wants to move for safekeeping but delete... [16:12:58] <PerterB> cpio -pmduv destdir < filelist && xargs -t rm < filelist [16:13:17] <Cyrille> mv will keep the structure intact. [16:13:28] <bda> Hence my confusion. [16:13:29] <PerterB> but won't create directories [16:13:30] <timsf> zfs snapshot <dataset>@justincase ; rm -f <files> [16:13:33] <bda> haha. [16:13:47] <bda> timsf: cpio still beats out zfs snapshot. ;P [16:14:12] <timsf> in whose universe ? :-) [16:14:24] <bda> For sheer nerdery, I mean. :) [16:14:40] * PerterB puts on his nerdie propellerhead beanie [16:14:57] <bda> A few years ago I sent a buddy a Topless Steve Jobs propeller beanie for his birthday. [16:15:00] <bda> He was suitably enraged. [16:17:26] <PerterB> I want! [16:17:30] <bda> o_o [16:17:39] <PerterB> or perhaps not, come to think of it [16:18:45] <bda> Hm, it's still listed but I can't find the image. [16:18:48] <bda> http://www.geekculture.com/geekculturestore/webstore/buttons.html [16:19:04] <bda> I'm just dumb: http://www.geekculture.com/geekculturestore/webstore/webstoreimages/buttons/jobs.jpg [16:19:22] <PerterB> oh my [16:19:45] <bda> Indeed. [16:20:35] *** bnitz has joined #opensolaris [16:20:45] *** sarah has joined #opensolaris [16:21:09] <bda> Wow. That was 2002. :\ [16:22:19] * holcomb grabs your butt [16:23:52] <bda> \o/ [16:34:20] *** xOmega has quit IRC [16:35:10] * tsp never heard of cpio, it seems (from reading the manpage) to be extremely old [16:35:39] <timeless> ancient archive formats, yippee, doesn't the solaris package format rely on cpio? [16:35:51] *** praseodym_ has quit IRC [16:36:07] <timeless> istr using cpio when i wanted to extract and setup solaris packages w/o root (for use in LD_LIBRARY_PATH) [16:37:07] <kjetilho> tsp: even RPM uses cpio internally [16:37:12] <Cyrille> couldn't you have used pkgtrans to get the package in directory format and then manually copy the reloc aarea somewhere? [16:37:13] *** xOmega has joined #opensolaris [16:37:24] <PerterB> yeah, SVR4 pkgstream format is basically a header plus a bunch of cpio archives [16:37:36] <PerterB> but cpio -p doesn't create archives, it copies or links files [16:39:11] <PerterB> and just coz it's old don't mean it's broken ;) [16:39:26] <Marv|LG> i am old and i feelbroken though [16:40:42] <tsp> I'm the one that's old :) turned 20 recently, and still learning [16:40:51] *** Fish- has quit IRC [16:41:09] *** Fish has joined #opensolaris [16:41:39] <trochej> tsp: 20? Youngling. :) [16:41:51] *** Fish has quit IRC [16:42:23] <trochej> Unless you're talking your retirement, thet you ARE old :) [16:43:05] *** ludovicp has left #opensolaris [16:48:25] *** dlynes_laptop has joined #opensolaris [16:50:45] <tsp> heh [16:51:50] <tsp> is there a way I can run zfs commands without su'ing to root all the time? most notably zfs snapshot [16:52:09] <bda> RBAC. [16:52:13] *** danny_j has joined #opensolaris [16:52:14] <trochej> rBAC [16:52:18] <trochej> Oh, answered [16:52:34] <elektronkind> RBAC! [16:53:32] <nachox> wasnt there a project to delegate that to users in order to let them snapshots their homes? [16:53:47] <timsf> zfs delegation [16:54:09] <trochej> Oh, true :) [16:54:18] <timsf> http://blogs.sun.com/marks/entry/zfs_delegated_administration [16:54:18] <trochej> Is it already in SXCE? [16:54:47] <timsf> it's in snv_69, possibly earlier [16:55:01] <trochej> Cool, forgot about it completely [16:55:10] <timsf> Yeah, it's pretty nice [16:57:15] *** Fish has joined #opensolaris [17:06:15] *** nachox has quit IRC [17:06:36] * trochej compiling qemu [17:09:35] *** vmlemon has joined #opensolaris [17:09:42] <coffman> sudo! [17:09:46] * coffman hides [17:09:53] <PerterB> geshundheit [17:10:17] *** Jagoma has quit IRC [17:10:35] <PerterB> weird, there was only one h when I thought it! [17:10:51] <Cyrille> EFINGERS then [17:11:04] <PerterB> aye [17:12:27] <PerterB> "We're sorry, but your fingers are too fat to dial this number. If you should need a dialing wand, mash the keypad now..." [17:13:10] <coffman> pfexec is the new sudo [17:16:23] <trochej> :) [17:17:33] *** nachox has joined #opensolaris [17:25:16] <coffman> hey, rbac is nice, i can give my developers the rights they "need" (only on feburary 29. and only when the moon is half full ofc) [17:26:32] *** frostcs_ has joined #opensolaris [17:30:31] *** FrostCS has quit IRC [17:30:47] *** mega has quit IRC [17:34:24] *** derchris has quit IRC [17:34:33] *** derchris has joined #opensolaris [17:37:19] *** bowdengl has quit IRC [17:38:58] *** Dar is now known as Dar_HOME [17:47:48] *** Loof has joined #opensolaris [17:51:58] *** bengtf has quit IRC [18:04:49] <Loof> Can you build ON without the sun compiler? I.e. GCC + binutils? [18:04:58] <Loof> (+make, etc) [18:06:15] <bda> O_o [18:06:25] <PerterB> why would you want to? ;) [18:07:32] <PerterB> http://www.opensolaris.org/os/community/tools/gcc/status/ [18:07:47] <wesolows> you can use gcc [18:07:50] <wesolows> it works just fine [18:08:08] <wesolows> GNU binutils, no; the OpenSolaris linker has required features [18:08:18] <wesolows> and GNU make, definitely not; the makefiles assume OpenSolaris make [18:08:44] *** peyka951 has joined #opensolaris [18:09:06] <peyka951> i [18:13:38] *** bengtf_ has joined #opensolaris [18:13:40] *** bengtf_ is now known as bengtf [18:13:41] *** peyka951 has left #opensolaris [18:13:55] *** MattMan has quit IRC [18:14:38] *** peteh has quit IRC [18:14:51] *** RaD|Tz has quit IRC [18:15:47] *** bigjohnto has joined #opensolaris [18:16:09] <bigjohnto> anyone know what to do after replacing a pcu for a storedge t3? [18:16:34] *** xOmega has left #opensolaris [18:18:42] *** tg has quit IRC [18:18:51] *** tg has joined #opensolaris [18:19:18] *** tg has quit IRC [18:19:24] *** tg has joined #opensolaris [18:20:32] <trygvis> if I managed to fuck up /etc/vfstab (tried to break a svm mirror), how can I make / rw after rebooting? [18:22:35] <WickedWicky> mount -o remount,rw /dev/dsk/cYtXdZs0 / [18:22:50] <trygvis> yeah, tried that: ount: /dev/dsk/c0t0d0s0 is not this fstype [18:23:01] <trygvis> but it is ufs, and I've fsck-ed it [18:23:58] <WickedWicky> that's weird [18:24:11] <trygvis> indeed [18:24:16] <PerterB> check you don't have any typos left on the / entry in vfstab [18:24:41] <trygvis> duh, the whole point is that vfstab is r/o [18:24:43] <WickedWicky> giving a device and a mountpoint will bypass vfstab, no? [18:25:32] <trygvis> it should. this is the command I'm giving: mount -o rw,remount /dev/dsk/c0t0d0s0 / [18:25:40] <WickedWicky> is c0t0d0s0 / though? [18:25:43] <trygvis> (I forgot the s0 part in vfstab) [18:26:08] <trygvis> yep, it is one of two mirrors in a svm setup [18:26:20] <WickedWicky> then you should mount the submirrir [18:26:25] <WickedWicky> not the actual slive [18:26:26] <WickedWicky> slice [18:26:42] <WickedWicky> mounting the slice and write to it will give specteculare behaviour [18:26:49] <WickedWicky> like UFS corruption and panicas [18:26:52] <WickedWicky> panics too [18:27:01] <PerterB> oops sorry, doing 3 things at once... the point is it looks for stuff not on the command line in vfstab, so try explicitly adding -F ufs [18:27:15] <trygvis> I tried that as well, no help [18:27:37] <WickedWicky> try mounting the submirror [18:27:40] <trygvis> WickedWicky: ok, cool. but I though that if I wanted to stop using svm I could just use one of the mirrors directly [18:27:49] <trygvis> didn't work, could only mount d0 [18:28:11] <WickedWicky> metadettach the submirror you want to get rid of [18:28:16] *** umdstu has quit IRC [18:28:23] <trygvis> aha [18:28:32] <WickedWicky> metadetach mirror submirror_to_get-ridoff [18:29:52] <trygvis> oki, brb. got to reboot the mac [18:30:51] *** sbahra has joined #opensolaris [18:32:54] *** frostcs_ is now known as FrostCS [18:35:41] *** RaD|Tz has joined #opensolaris [18:37:49] *** yarihm has quit IRC [18:39:34] *** KermitTheFragger has quit IRC [18:43:39] *** charlesnw has joined #opensolaris [18:44:29] *** Cyrille has quit IRC [18:46:16] *** karrotx has quit IRC [18:49:16] *** bigjohnto has quit IRC [18:55:47] *** bondolo has quit IRC [18:57:05] *** stevel has joined #opensolaris [18:57:05] *** ChanServ sets mode: +o stevel [18:58:36] *** dlynes_laptop has quit IRC [19:00:10] *** bondolo has joined #opensolaris [19:00:20] <CIA-16> hiremath: 6597119 panic in daplt driver while running MPI tests, 6599547 Solaris initiated IB connections fails with OFED 1.2 [19:01:50] *** Erwann has quit IRC [19:02:29] <sbahra> masta, I didn't know you were here. [19:02:37] * sbahra was wondering what that smell was ;] [19:07:13] *** sstallion has joined #opensolaris [19:07:54] *** g4lt-sb100 is now known as g4lt-superbat [19:10:08] *** tsoome has joined #opensolaris [19:10:43] *** Fullmoon has quit IRC [19:13:51] *** communicator has quit IRC [19:14:50] *** Mdx4 has joined #opensolaris [19:18:50] *** Snake007uk has quit IRC [19:19:36] *** Kernel86_ is now known as Knerle86 [19:19:50] *** Knerle86 is now known as Kernel86 [19:20:03] *** tombhadAC has joined #opensolaris [19:23:41] *** migi has quit IRC [19:27:07] *** FrostCS has quit IRC [19:29:34] *** thedonvaughn has joined #opensolaris [19:30:32] <WickedWicky> heyas [19:39:03] *** yarihm has joined #OpenSolaris [19:39:45] *** e^ipi_ has joined #opensolaris [19:40:26] *** e^ipi has quit IRC [19:40:35] *** WickedWi1ky has joined #opensolaris [19:40:36] *** WickedWicky has quit IRC [19:45:13] *** chris__ is now known as shoe [19:45:19] *** stefanha has joined #opensolaris [19:45:26] *** Mdx4 has quit IRC [19:46:41] *** e^ipi_ is now known as e^ipi [19:46:54] <WickedWi1ky> re's [19:48:11] *** sioraiocht_ has quit IRC [19:48:11] *** sioraiocht has joined #opensolaris [19:51:34] *** SYS64738 has quit IRC [19:56:21] *** onlinebacon has joined #opensolaris [19:56:55] * quasi too [19:57:18] <onlinebacon> hey, i'm having a bit of a problem, I just installed Solaris Community release 70 on my new dell e521 (which is said to be compatable on the HCL) but after a full install it hangs at the SunOS release 5.11 version snv_70 64-bit paragraph [19:57:39] <onlinebacon> doesnt get to configuring /dev, even if i leave it for like 10 minutes [19:57:41] <flyingparchment> onlinebacon: edit the grub 'kernel' line and add this at the end: -kv [19:58:25] <onlinebacon> okey doke, it's just installing again atm, my dad insisted i must have messed up during install :) [19:58:30] <axisys> how do I add a default gw to a zone? [19:58:54] <axisys> netstat -nr from w/in one of the zone does not show default [19:59:01] <axisys> the other zones looks fine [19:59:08] <madhatter> re [19:59:50] <axisys> i know i cannot add a default gw from w/in the zone env [19:59:53] <madhatter> I am installing opensolaris b70 right now. I have been wondering why I haven't been able to merge my two harddisks to one pool [20:00:02] <axisys> is there a way to do it from global zone? [20:00:07] <madhatter> I thoght something like that would be possible [20:00:19] <CIA-16> jarrett: 6560317 TX NFS server needs to support NFSv3 clients, 6596123 TX NFSv4 should not allow a read-up mount via WebNFS [20:03:04] *** sfire||mouse has joined #opensolaris [20:04:10] *** cydork has quit IRC [20:04:33] *** yippi has joined #opensolaris [20:04:57] *** xOmega has joined #opensolaris [20:06:41] *** eboutilier has quit IRC [20:06:46] *** yippi has quit IRC [20:10:21] <onlinebacon> flyingparchment, that will show boot messages right? [20:10:26] <flyingparchment> yes [20:10:56] <richlowe> and it will run under the debugger, so you should be able to figure out where it is when it hangs. [20:11:14] <onlinebacon> ah great, thanks guys [20:11:30] <onlinebacon> i just got it today, and a nvidia card for it so im pretty excited :) [20:11:35] <sbahra> IS SOLARIS BETTER THAN LINUX [20:11:45] <flyingparchment> sbahra: no [20:11:48] <sbahra> OK [20:11:51] <flyingparchment> solaris is shit, it doesnt run any games [20:12:02] <sbahra> i knewi t [20:12:19] <wesolows> oh no, the trolls are stuck in a loop [20:12:30] * sbahra isn't a troll ;-[ [20:12:34] <onlinebacon> it runs stuff in wine though, and some games in linux emulation [20:12:36] <sbahra> this is my first troll here [20:12:44] <wesolows> no trolling [20:12:49] <wesolows> troll = kick [20:12:57] <madhatter> Hm, no answer is still an answer [20:13:02] <sbahra> wesolows, so kick me ;[ [20:13:20] <wesolows> madhatter: what are you trying to do? [20:13:29] <flyingparchment> wesolows: if you kick me, i'll write an awesome program that only works on linux so sun goes out of business and you lose your job [20:13:46] <wesolows> madhatter: merging zpools is not possible, but you can create a single one containing 2 disks in either a mirror or striped configuration [20:13:51] <wesolows> flyingparchment: go for it [20:14:22] <madhatter> wesolows: That's what I wanted to do in the installation process [20:15:01] *** comay has joined #opensolaris [20:15:01] <richlowe> there's no pleasant way to do it during install, yet. [20:15:12] <madhatter> wesolows: I thought when I can choose the installation destination I can choose both hard disks and create a pool [20:15:14] <wesolows> madhatter: is this for the boot pool, or just a data pool? boot/root zfs needs installer support; I don't know the current state of that work but it definitely isn't there before 70. If it's just a data pool, you can do it after installation at any time. [20:15:40] <richlowe> wesolows: lori has magical bits that zap pre-int installer bits onto the install image. [20:15:46] <richlowe> I have no idea if they worked more than once, however. [20:15:53] <wesolows> dunno [20:15:56] <richlowe> the last N times I've seen the mentioned was in the context of "These won't work on build $X" [20:16:00] <wesolows> it's Not There Yet. [20:16:10] <madhatter> wesolows: I wanted it for everything. So I don't have to know how much space I need for / or for /export/storage [20:16:31] <madhatter> wesolows: Not for swap of course [20:16:33] <wesolows> that's unwise [20:16:42] <madhatter> wesolows: Why? [20:16:46] <wesolows> your OS belongs on a pair of mirrored boot/root disks [20:16:53] <wesolows> everything else belongs on 1 or more pools [20:17:00] <ofu> has anybody tried to attach a sata-device to the onboard-controller of a T2000? [20:17:02] <wesolows> zfs boot can't support raidz2 for example [20:17:11] <wesolows> which is what you most likely want for data pools [20:17:50] <madhatter> wesolows: I had no raid for storage here yet. So i did no plan to do that [20:17:51] <ofu> mirrored zfs disks are better than sds-mirrored bootdisks in a thumper [20:18:10] <madhatter> But maybe I should think about that. [20:18:34] <wesolows> if you have only 2 disks total, you might as well just make them a single pool in a mirror config and use them for both the OS and your data [20:18:43] <madhatter> Right now I have 40G for / swap and $HOME, so the other disk can be storage alone then [20:18:53] <wesolows> but if you have 4 or more disks, I would do a 2-way mirrored boot/root and put the rest in a separate pool [20:18:54] *** micken has joined #opensolaris [20:19:21] <madhatter> wesolows: Just the two disks at the moment. [20:19:31] <wesolows> I guess it doesn't matter then [20:19:35] <wesolows> one pool it is [20:19:39] <wesolows> and wait for zfs install [20:19:55] <madhatter> But how can I create that _after_ installation? [20:19:56] <ofu> how long do i have to wait for zfs install? [20:20:02] <wesolows> until it's done [20:20:22] <wesolows> madhatter: well, I don't think you can, really [20:20:38] <wesolows> madhatter: Maybe there are instructions...let's see...it's not gonna be pretty though [20:20:51] <madhatter> wesolows: So, is there a better work around? [20:21:04] <madhatter> wesolows: Boot from DVD, create the pool and install then into it? [20:21:26] <wesolows> http://www.opensolaris.org/os/community/zfs/boot/ [20:21:57] <wesolows> yeah, it uses slices [20:21:58] <wesolows> yuck [20:23:37] <madhatter> wesolows: What do you mean by that? [20:23:42] *** shoe has left #opensolaris [20:23:51] *** slowhog has joined #opensolaris [20:23:54] <madhatter> wesolows: I thought slices are nothing unusual on Solaris [20:24:02] <bda> ZFS likes having full disks. [20:24:05] <wesolows> because you can only install into ufs, you have to use slices for that, which means you have to have another set of slices for zfs [20:24:18] <wesolows> which sucks because zfs works best with the whole disk [20:24:30] <wesolows> it will work correctly, of course, with slices [20:24:47] <wesolows> but performance and serviceability are better on whole disks [20:24:53] <ofu> 4GB cf cards or usb sticks are affordable these days [20:24:58] <bda> Install onto a USB drive, then install onto the system disks? :D [20:25:01] <bda> Or that. ha. [20:25:08] <wesolows> the other option is to write your own installer [20:25:15] <wesolows> which is highly recommended [20:25:20] <bda> haha [20:25:52] <madhatter> wesolows: Hm, so you normally use zfs only for storage and ufs for the boot partition? [20:25:59] <wesolows> no [20:26:03] <wesolows> I use only zfs [20:26:08] <madhatter> Then I don't understand it [20:26:10] <wesolows> but what I do is not a good example [20:26:15] <wesolows> it's complicated [20:26:28] <wesolows> if you use the Solaris installer, ufs boot + zfs for data is the current model [20:26:38] *** sahafeez has joined #opensolaris [20:26:45] *** timsf has quit IRC [20:26:58] <wesolows> I'm sorry that the installer forces you to use ufs [20:27:08] <madhatter> Okay, so you install a different way [20:27:23] <wesolows> I hate ufs and want it to die. If it were up to me, UFS would be marked Obsolete and mkfs.ufs would no longer be shipped in S11. [20:27:32] <madhatter> I probably will play around with zfs and stuff with my spare storage disk then [20:27:38] <wesolows> good plan. [20:27:51] <bda> ZFS++ [20:28:18] <madhatter> But thanks for the link. Maybe I will try this when familiar with zfs a bit [20:29:13] *** karrotx has joined #opensolaris [20:30:00] *** logic_ has joined #opensolaris [20:30:14] <nachox> i'm off :) [20:30:17] *** nachox has left #opensolaris [20:30:32] * ofu likes the new freebsd-installer, finstall... it even has a button for "install to zfs" [20:30:56] <sbahra> ofu, ewww. I didn't look at it. [20:31:02] <sbahra> ofu, was it completely successfully for SoC? [20:31:05] *** xOmega has left #opensolaris [20:31:06] <richlowe> FreeBSD aren't working on 3 separate, closed, installers at the same time, though. [20:31:20] <richlowe> so I'm not surprised there's would work better, and faster. [20:31:40] <sbahra> richlowe, our installer (sysinstall) sucks. Generally known fact. [20:32:11] <ofu> http://wiki.freebsd.org/finstall/Amnesiac [20:33:27] <sbahra> ofu, very cool. [20:33:47] *** Mdx4 has joined #opensolaris [20:38:57] <gdamore> code review request: http://cr.opensolaris.org/~gdamore/mac-notify/ [20:39:20] <gdamore> This is a nemo change to use a simpler thread based solution rather than relying on (abusing) the system_taskq for simple MAC notifications [20:43:50] *** RaD|Tz has quit IRC [20:46:12] *** mikefut has joined #opensolaris [20:47:59] <trygvis> WickedWi1ky: can't I just destroy d0? will it kill the submirrors too? [20:49:33] <WickedWi1ky> you wanna lose data? [20:49:52] <trygvis> not really :) [20:49:55] <trygvis> it is / after all [20:50:05] <WickedWi1ky> so, why not do this: [20:50:12] <trygvis> the idea was to break the mirror and use one side as a LU environment [20:50:15] <WickedWi1ky> metadetach mirror submirror_to_get_rid_of [20:50:36] <WickedWi1ky> format /dev/rdsk/secondsubmirror d [20:50:39] <WickedWi1ky> create a slice [20:50:50] <WickedWi1ky> create a filesystem, copy / over to the new slice [20:51:00] <WickedWi1ky> then metaclear the mirror [20:51:26] <WickedWi1ky> and dont forget to make adjustments to /etc/system [20:52:21] <WickedWi1ky> oh for LU.. [20:52:38] <WickedWi1ky> there was something for that.. I think it was flyingparchment who used the breaking-mirror technique for that [20:52:41] <trygvis> I wonder if I should just reinstall the entire box [20:52:52] <trygvis> it seems doable: http://www.sun.com/bigadmin/features/articles/using_lu.html [20:53:04] <trygvis> but honestly I don't need SVM so I just might do a clean install [20:54:03] <tomww> LU could do the detach/attach stuff for you. only needed: a new name for the to metadevice for the new ABE's disks [20:57:53] <trygvis> right. but I'm going for the "fuck it" solution. I want a full install instead of the old hack I seem to have made [20:58:04] *** henrik_ has joined #opensolaris [20:58:30] <trygvis> is it not possible to lofs mount an iso image and then nfs export that? [21:02:35] *** Emmedics4 has joined #opensolaris [21:02:36] *** Mdx4 has quit IRC [21:04:50] *** stefanha has left #opensolaris [21:07:51] *** leal has joined #opensolaris [21:08:01] <leal> hello, somebody here using AVS [21:08:04] <leal> ? [21:11:30] <bda> No, though I want to be. :) [21:11:38] *** bnitz has left #opensolaris [21:13:00] *** pfa3rh has quit IRC [21:14:22] *** apersson` has joined #opensolaris [21:14:25] <leal> :) [21:25:34] *** luosm_ has quit IRC [21:26:00] *** alanc-away is now known as alanc [21:29:08] *** pfa3rh has joined #opensolaris [21:30:00] *** leal has quit IRC [21:31:15] *** charlesnw has quit IRC [21:31:24] <onlinebacon> bah [21:31:42] <onlinebacon> solaris gets to the line NOTICE: Kernel debugger present: disabling console power management. [21:31:43] <onlinebacon> and then stops [21:31:53] <onlinebacon> when i use -kv on the grub boot line [21:32:02] <richlowe> F1-A, see where it is. [21:32:15] *** sioraiocht has quit IRC [21:32:20] <richlowe> $C would be a fairly good first thing to type at kmdb. [21:32:41] *** sioraiocht has joined #opensolaris [21:33:39] *** yarihm has quit IRC [21:36:04] *** fbo has joined #opensolaris [21:37:24] <onlinebacon> whoah [21:37:28] <onlinebacon> i tried it with just -v [21:37:34] <onlinebacon> and it is booting but taking ages :s [21:38:00] <onlinebacon> is that because its the first boot and it's normal, or is it a problem? [21:41:17] <seanmcg> onlinebacon: do you have a multicard reader in that dell box ? [21:41:20] *** sommerfeld has joined #opensolaris [21:41:20] *** ChanServ sets mode: +o sommerfeld [21:41:25] <onlinebacon> seanmcg, i do [21:41:54] <seanmcg> Theres the slowdown. I'm typing from the same model dell box E521 :) [21:42:09] <onlinebacon> ahh ok, so i can disconnect it and it will work fine? [21:42:50] <seanmcg> Yup, but then no card reader. Good news is that in snv_71 it'll boot :) [21:43:15] <onlinebacon> oh cool :D [21:43:29] *** mikefut has quit IRC [21:44:02] *** henrik_ has quit IRC [21:44:16] <onlinebacon> ok im booting it up :D [21:45:40] <onlinebacon> it works :P! [21:45:45] <onlinebacon> oh man thankyou so much :) [21:46:14] <seanmcg> np [21:46:23] <onlinebacon> that was a weird bug [21:47:04] <onlinebacon> brb [21:47:07] *** onlinebacon has quit IRC [21:47:39] <trochej> Weird bug is when your system end up changing labels of disks between two minor kernel releases [21:47:58] <axisys> how come I can ping the def gw from global zone and zone1 .. but not from zone2 ? [21:48:04] <axisys> here is the config http://rafb.net/p/viTubx77.html [21:48:25] <axisys> they are all on same ip segment [21:51:13] *** bengtf has quit IRC [21:51:27] * quasi gets a sense of deja-vu ... has axisys gone into an endless loop? [21:51:28] *** migi has joined #opensolaris [21:52:23] <seanmcg> trochej: theres maybe another bug with his system which he may find more weird :) [21:54:02] <coffman> seanmcg: oh fun, tell us! [21:57:30] *** FrostCS has joined #opensolaris [21:59:39] *** thedonvaughn has quit IRC [22:00:20] <CIA-16> meem: 6375443 dead declarations in <inet/ip*.h>, 6587730 MATCH_IRE_WQ logic has no consumers, 6592926 NCE_F_PROXY handling is a dead-end, 6592983 IP ioctl argument extraction logic can be simplified, 6592998 SIOC*MSFILTER ioctls can induce panic, 6594323 ipif_get_name() could be simpler and used more widely, 6595071 ipif_ndp_up()'s second argument is needless, 6597721 {O_,}SIOCGLIFCONF have bogus ipi_cmd_type values, 6599009 bogus conn_ref assertion in [22:03:17] <e^ipi> at the video store there's a pile of DVD collections where for a couple bucks more than a regular DVD you get two movies [22:04:45] <e^ipi> and it'll be things like "the ultimate action collection" or "the ultimate comedy pair" that'll have one good movie and some also-ran DVD which isn't B-grade, but isn't very good either [22:05:15] <e^ipi> I figgure they should take that a step further and publish a bunch of "ultimate contrast collections" [22:05:34] <e^ipi> have like, Schindler's List and White Chicks [22:06:17] <WickedWi1ky> that's quite a contrast :P [22:06:40] <delewis> yes, White Chicks is a real downer by comparison. [22:06:41] <alanc> "The Ultimate OS Collection: Solaris and Windows 98" [22:06:41] <WickedWi1ky> and I'd opt for Shindler's List being the better movie [22:06:59] <e^ipi> or american beauty and harold & kumar go to white castle [22:07:22] <WickedWi1ky> Quest for the Holy grail & Ghost Rider [22:07:22] <richlowe> driving miss daisy, and Junior. [22:07:44] <WickedWi1ky> Ghost Rider being to hidious and sad for words [22:07:49] <WickedWi1ky> to/too [22:08:09] <stevel> Animal Farm & Babe [22:08:16] *** pfa3rh has quit IRC [22:08:33] <e^ipi> heh [22:08:36] <WickedWi1ky> Momento & The Beach [22:09:14] <WickedWi1ky> oh and Requiem for a dream [22:09:15] <WickedWi1ky> wtf [22:09:21] *** tek-ops has joined #opensolaris [22:09:32] <richlowe> requiem for a dream needs to be paired with Friday. [22:09:38] <richlowe> or a cheech and chong movie. [22:09:48] <richlowe> Requiem For a Dream, and Nice Dreams [22:09:49] <richlowe> that'd fit nicely. [22:09:53] <WickedWi1ky> lol [22:10:23] <WickedWi1ky> how to screw up your life in a short time frame and see how peachy and cheerfull life is, wee [22:10:30] <alanc> 2001 & Battlefield Earth [22:10:34] *** tek-ops has quit IRC [22:10:35] <e^ipi> or the matrix, plus the sequels to the matrix [22:10:41] *** splunk has joined #opensolaris [22:10:51] <WickedWi1ky> Matrix vs Doom (The Movie) [22:13:37] <splunk> any idea why when I issue "svcs rmvolmgr" nothing is found on my Solaris 10? [22:13:40] <richlowe> I'd say Contact and They Live, but I'm not sure which would be the bad one. [22:13:41] *** bengtf_ has joined #opensolaris [22:13:43] *** bengtf_ is now known as bengtf [22:14:47] <WickedWi1ky> a huge contrast would be: Reservoir Dogs - Titanic [22:15:07] *** Hunger- has quit IRC [22:15:59] <alanc> isn't rmvolmgr the Nevada-only replacement for vold? [22:16:05] <richlowe> Yes. [22:16:18] <alanc> so you'ld have to look for the vold services on Solaris 10 [22:16:25] <wesolows> which is volfs, no? [22:16:25] <richlowe> though I can't speak to "Nevada-only", given the fashion for ill-advised backports. [22:17:27] <splunk> haha [22:17:33] <splunk> Solaris 10 [22:17:45] <wesolows> poor S10 [22:17:56] <bda> :( [22:17:57] <wesolows> it was so awesome when it was new, lo these 3 long years ago [22:18:06] <splunk> i just create a UFS fs on my external seageat 700 gig drive but its not automounting [22:18:14] <splunk> created* [22:18:22] <alanc> I think the only S10 machines I use any more are our S10 patch build machines [22:18:24] <splunk> i can manually mount it but [22:18:35] <splunk> its so flaky :( [22:18:36] <jbk> afternoon [22:18:43] <axisys> I am looking for info where it says the sun 10g card supports 1g.. aparently my sun support engineer could not find any such article to support my claim [22:18:59] <alanc> oh, and I guess our file & web server, which I haven't gotten around to scheduling downtime to move to Nevada [22:19:10] <axisys> all i have is the pkginfo -l SUNWnxge which shows 10g/1g driver in desc [22:19:55] <richlowe> Neptune is used in both dual 10GE and quad GigE parts. [22:20:05] <richlowe> which maybe what the DESC means. [22:20:34] *** fbo has left #opensolaris [22:21:11] <axisys> richlowe: so in other words it is possible that the 10g exprs crd do not go down to 1g [22:21:32] <richlowe> that's not what I said, I just pointed out that there are GigE-only cards that use nxge. [22:22:33] *** linux_user400354 has joined #opensolaris [22:22:52] <splunk> how do I initate rmvolmgr ? [22:23:12] <axisys> richlowe: yep.. that is what i meant.. just because pkginfo -l SUNWnxge says 10g/1g .. does not mean the card I have will support 1G [22:23:44] <axisys> my assumption was wrong is what I am trying to say [22:25:52] <WickedWi1ky> ye gods, about movies.. Mummie returns on tv [22:26:07] *** karrotx has quit IRC [22:27:40] *** Hunger- has joined #opensolaris [22:29:12] *** sarah has quit IRC [22:29:15] <axisys> anyone can explain why i can ping the df gw from zone1 but not from zone2 when they both are on same ip segment.. [22:29:23] *** yarihm has joined #OpenSolaris [22:37:18] *** Nishaway has joined #opensolaris [22:38:58] <axisys> tcpdump on global zone shows when i ping .254(gw) from zone1 all traffic goes to .253 and it works .. but zone2's ping traffic goes to .254 and never returns [22:39:03] *** pjlv has quit IRC [22:39:53] <axisys> same physical interface and same IP segment for both zones.. wonder if i am hitting a bug realted to tcp/ip stack [22:40:51] *** yarihm has quit IRC [22:42:52] *** wisnios has joined #opensolaris [22:44:30] <elektronkind> every time I see a reference to the elfsign command, I think of Dune and these large Elfs diving through the desert sand dunes [22:46:08] <jbk> haha [22:46:21] <flyingparchment> so i called the sun reseller that didn't send me a quote, they said they'd send it, and didn't [22:46:25] <jbk> 'stilgar, do we have elfsign?' [22:46:59] <elektronkind> I guess that makes me a total dork [22:47:56] *** wisnios has quit IRC [22:52:03] <WickedWi1ky> Elfs? Freemen! [22:52:20] <sommerfeld> no. sandworms. [22:52:27] <WickedWi1ky> sandworms and wormsigns [22:52:28] <WickedWi1ky> yeah [22:52:37] <WickedWi1ky> and that thingy to call them, maker? [22:52:50] <WickedWi1ky> god I feel old, haha [22:53:25] <Auralis> thumper [22:53:26] <WickedWi1ky> I remember when taxes where 17%, now I hear how it will go from 19 to 20 [22:53:29] <WickedWi1ky> that! [22:53:32] <sommerfeld> saw the Dune movie in the theater shortly after it came out. boy was that a waste of an evening. [22:53:40] <WickedWi1ky> I prefer the book [22:54:16] <gdamore> the movie isn't bad *if you have read the book* [22:54:41] <sommerfeld> first book was good; forced my way through the second one and ran out of steam halfway through the third. [22:54:43] <gdamore> when the movie first came out, they handed out a reference card before you went in... not a good sign. [22:55:10] <sommerfeld> gdamore: well, unless you're expecting the movie to actually stick to stuff actually in the book.. [22:55:11] <gdamore> yeah, the Dune books got progressively worse... sort of the way I imagine Frank Herbert's mind went. [22:55:31] <gdamore> heh. you can't expect that too much. :-) [22:55:54] <gdamore> i'm still pissed off about how much of the books were missing in Jackson's LOTR movies. [22:56:37] *** SirFunk has quit IRC [22:56:38] <gdamore> but then I'm probably unusual, since I have probably read all of the LOTR books at least a dozen (maybe even 2 dozen) times each. [22:56:50] <sommerfeld> That doesn't bother me; there's only so much you can fit in a movie people will sit through. [22:57:13] <WickedWi1ky> and three hours is quite a sit [22:57:16] <gdamore> i understood that. but there was major, major stuff that got cut out. i would have liked to have it in the directors cut. [22:57:44] <gdamore> the whole tom bombadil and the barrow wights, the return home, etc. [22:58:32] <flyingparchment> the film cut tom bombadil? [22:58:35] <gdamore> yes. [22:58:50] <flyingparchment> heh. that's the part of the book i always remember most, for some reason. [22:58:52] <gdamore> in the movie aragorn just hands the hobbits a few swords and says "here you'll need these" [22:59:02] <gdamore> (or something to that effect) [22:59:17] *** splunk has quit IRC [22:59:31] <sommerfeld> but that's pretty minor compared to what Lynch did to Dune. [22:59:34] <gdamore> there was a whole substory, that explained *why* those particular swords were important, particularly when one of the hobbits stabs the witchking at the end [23:00:15] <gdamore> the details of the hobbits home coming was kind of important too... it sort of dropped one of the major points of the novels, I thought, which was the hobbits coming into their own [23:00:57] <Auralis> i'm glad the films cut bombadil [23:02:14] <gdamore> why? [23:03:01] <Auralis> because ultimatly he is a pointless character for the overall story and would have eaten up a lot of time [23:03:04] <gdamore> in general I felt like the movies were compelled to rush everything at this frenetic pace, which doesn't map to the books at all. [23:03:09] <sommerfeld> IMHO the "scouring of the shire" sequence is important [23:04:06] <gdamore> bombadil gave the back history for how the hobbits came by the swords, one of which ultimately contributed to the downfall of the witchking. an ordinary blade (if you follow the books) could not have done so. [23:05:00] <richlowe> you could show that and still skip bombadil. [23:05:05] *** Jondice has quit IRC [23:06:36] * comay agrees with sommerfeld [23:06:47] * gdamore too [23:10:56] *** Jondice has joined #opensolaris [23:12:55] *** gcj has joined #opensolaris [23:13:22] *** Murmuria has joined #opensolaris [23:13:44] <gcj> hi all, i'm having some problems getting the network card driver (ne2000) running in opensolaris dev edition on qemu emulator, can anyone help me to see if the driver (ni) works and can be loaded? [23:16:52] *** micken has quit IRC [23:17:38] *** SirFunk has joined #opensolaris [23:17:39] *** kFuQ has joined #opensolaris [23:18:03] *** SirFunk has quit IRC [23:19:51] <Fetch> if I want to replace a raidz member vdev with a mirrored vdev, what are my options? [23:21:41] *** sfire||mouse has quit IRC [23:22:21] <sommerfeld> Fetch: at this point the best approach for that sort of thing is to create a new pool and transfer the contents via zfs send | zfs receive [23:23:03] <sommerfeld> you can do this with a narrow outage window through using incremental send/receive to send the diffs between snapshots [23:23:13] <Fetch> sommerfeld: hmm, horrific ;) thanks, good to know (I didn't know if a mirror vdev could be added to a zpool's spare pool, and set to replace an existing disk vdev) [23:23:34] <gcj> devices/pci@0,0/pci10ec,8029@3:ni0 exists, but i can't ifconfig ni0 [23:24:12] <Fetch> (I only say horrific because it necessarily requires having about 2x your existing pool's space. Alas an issue for some of us) [23:24:18] <sommerfeld> Fetch: it's expected to get better when zpool remove works for main pool vdevs. [23:24:36] <elektronkind> gcj: you've tried 'ifconfig ni0 plumb' ? [23:24:54] <gdamore> where did you get the "ni" driver? [23:24:59] <gcj> elektronkind, that i hadn't, and it worked, thanks! [23:25:06] <gcj> how can i get that to run automatically at boot? [23:25:15] <elektronkind> gcj: yeah, you need to plumb a interface before configuring it [23:25:16] <gdamore> touch /etc/hostname.ni0 [23:25:27] <gcj> thanks! [23:25:32] <gcj> sorry i'm new to solaris [23:25:38] <elektronkind> gdamore: ni is a NE2000 drive from that japanese dude [23:25:49] <gdamore> murayama's driver? ok. [23:26:05] <elektronkind> jah [23:26:24] <gcj> it's nice that sol210 drivers run on sol211 :-) better than darwin anyway :-) [23:26:43] <elektronkind> gcj, you can put the IP address you want that interface to have in that hostname.ni0 file [23:26:46] <gcj> although i had to hack the install script to make it think i was running 210 [23:26:52] * boyd cheers at the DDI being appreciated [23:26:56] <gcj> elektronkind, if i want it to dhcp? [23:26:57] <Fetch> back when snv_70 was up for download, I noticed snv_71 and snv_72 were tagged in the mercurial tree. Is there any rule of thumb for when it's a good/bad idea to build tags past the current release, for BFU? [23:27:18] <elektronkind> gcj: yeah, welcome to the land of (mostly) stable driver interfaces ;) [23:28:01] <wesolows> if only the nexus ops were stable :-( [23:29:04] <gdamore> wesolows: something I've been asking for since 1998. [23:30:03] <elektronkind> gcj: http://www.rite-group.com/rich/solaris_dhcp.html [23:30:10] <sommerfeld> Fetch: depends on what you're doing. if you're aiming to get a change back into the main tree you should generally be building the tip. [23:31:44] <Fetch> sommerfeld: typically, looking to get bugfixes that are fixedin a version not yet available for download [23:31:50] <gcj> elektronkind, thanks again [23:31:51] *** hohum has joined #OpenSolaris [23:31:58] <boyd> So, I've finished my port of zfs to linux. It's pretty much sed '/CDDL HEADER START/,/CDDL HEADER END/d;r gpl.txt' [23:32:07] <gcj> lol [23:32:10] <gcj> is that legal? [23:32:20] <sommerfeld> Fetch: specific fixes, or just wanting to run the bleeding edge? [23:32:27] <boyd> Seems to be working for them in other areas... just thought I'd help [23:32:30] <sommerfeld> hahaha [23:32:42] <Fetch> sommerfeld: specific fixes, generally. I don't especially like running the bleeding edge [23:33:00] <elektronkind> well you're no fun then [23:33:11] *** hohum has left #OpenSolaris [23:33:12] <Fetch> elektronkind: no fun at all [23:33:12] <sommerfeld> Fetch: so then you'd want to build off of the tag matching the numbered build. [23:33:37] <boyd> Solaris "bleeding edge is a bit less bleedy than many other orojects [23:33:40] <Fetch> sommerfeld: that's fine, I just was curious if there were any rules of thumb for running tags that are currently ahead of the currently available snv build [23:33:56] *** hohum has joined #OpenSolaris [23:34:02] <Fetch> (as, in my example, I built snv_72 when snv_70 was what was available as an iso) [23:34:40] <sommerfeld> Fetch: Generally works. Occasionally doesn't due to cross-consolidation dependencies (which are generally publicized on ON gate heads ups and the like) [23:34:57] *** Murmuria has left #opensolaris [23:35:02] <boyd> They have to do something like that to make the ISO, of course.. [23:35:46] <Fetch> of course. Any idea under what circumstances commits would be made to tagged branches? [23:36:07] <sommerfeld> Right. Generally those of us developers running the bleeding edge build the tip of ON and put that on top of a system running a numbered build with bfu [23:36:08] <Fetch> getting a snv_Foo that was not yet even a little bit snv_Foo would be lame, of course [23:37:02] <sommerfeld> Fetch: critical fixes to bugs which would render a build DOA sometimes land in the trunk after the build is closed and are then backported [23:37:45] <sommerfeld> Generally the set of differences is very small [23:38:13] <Fetch> thanks, that answers me well [23:40:10] <madhatter> Can somebody give me a hint/link on howto change/add window managers on opensolaris? I only find articles that tell you that you can choose between gnome and cde by default, but I'd like to get openbox installed. [23:40:19] *** sahafeez has quit IRC [23:40:57] *** scarab1 has joined #opensolaris [23:41:02] <scarab1> hey [23:42:06] <scarab1> i'm running zfs in an iscsi configuration... file transfers (windows copies) are great, but when we process lots of small files we get errors [23:42:28] <gcj> is it safe to set root's shell to bash in solaris? [23:43:09] <scarab1> gcj: i set bash as root's shell and haven't had any problems [23:43:18] <gcj> cool, thanks [23:44:21] *** pfa3rh has joined #opensolaris [23:44:23] <gcj> hmm, doesn't have any effect on "su" though [23:44:28] <scarab1> no [23:47:16] <scarab1> has anyone had issues with ZFS and lots of small (50-200k) files? [23:49:36] <sommerfeld> what sort of errors? [23:50:22] <sommerfeld> i run a few systems with multi-terabyte pools containing mostly solaris build trees.. which are lots of small files. haven't seen problems. [23:51:12] <scarab1> sommerfeld: I don't know the exact issues/errors (trying to get it from an end-user) [23:51:51] <scarab1> but they say that they can copy files over w/o any problems at all, but then when we start to process it (using "cracker" software--software that extracts/indexes, etc. data/metadata from e-files, and such) [23:52:00] <scarab1> when we use the software to process it, that's when we have issues... [23:52:39] <elektronkind> sounds like a job for dtrace [23:53:28] <sommerfeld> again, would help to understand what the "issues" are. [23:53:46] <scarab1> yeah, i know... [23:53:59] <scarab1> sorry i can't give that right now... yes, i've started looking heavily to dtrace recently [23:54:14] <sommerfeld> there are some program-visible differences but all well within the range of expected variation between different filesystems [23:55:18] <sommerfeld> for instance, stat() on a UFS directory returns the number of bytes used to store the directory in the st_size field. stat() on a ZFS directory returns the number of directory entries in st_size [23:55:37] <scarab1> hopefully i'll get an answer soon... [23:56:00] <scarab1> right now i guess i'm looking/curious about generally known issues/bugs related to small files and processing them [23:56:11] <sommerfeld> shouldn't matter, except that there's some code out there from BSD scandir which assumes no more than st_size/24 entries in any directory [23:56:30] <gcj> == 0xbadc0ded [23:56:30] <sommerfeld> and it's been cut & pasted too many times to know for sure that we've gotten to all of them... [23:57:23] <scarab1> all the calls from the app are windows. so, only samba calls would be used against the solaris/zfs setup [23:57:34] *** Fullmoon has joined #opensolaris [23:57:38] <scarab1> i'm using the shipped samba package [23:57:41] *** Fullmoon has quit IRC [23:59:14] <scarab1> what's dCache? [23:59:23] *** apersson` has quit IRC