=== debianist [~pooh@80.179.93.85.forward.012.net.il] has joined #ubuntu [12:02] am gonna open a bug - /dev/dsp never works under rythmbox and there is no sound perior to to running alsamixer. i don't expect the layman to open alsamixer under a terminal [12:02] mdz: you maintain the zaptel and asterisk packaging? [12:02] after first install reboot [12:02] bdale: I used to [12:02] bdale: zaptel anyway [12:03] bdale: I maintained an asterisk/zaptel-based PBX at my previous job; I no longer have the relevant hardware and the packages need a new maintainer [12:04] mdz : do you have any lead on this? [12:04] mdz: I just picked up a digium developer's kit to play with (pci card with one each fxo and fxs daughter cards), and was a bit underwhelmed by what's in /usr/share/doc/, but will defer being annoying about it for now. [12:04] bdale: wnpp #251938 [12:04] debianist: I don't understand the problem. is this the issue where the volume is muted and/or set to a low value initially? [12:05] mdz: just delete the who categories line (to make menu entries disappear) [12:05] mdz: thanks for the pointer, I'll prod Mark [12:06] bdale: there's a #asterisk or similar somewhere [12:06] the lead upstream developer is also Mark [12:09] mdz: yes, that's the alsamixer issue [12:11] bdale: Heh. I'm getting complaints from people with Compaqs that won't work with their wireless cards now [12:11] jdub: morning [12:14] mdz: right. email sent. [12:14] mdz: fwiw, the kit I bought came with a floppy containing a set of config files for playing with this board set as delivered, and a script with associated README for doing an install from CVS that includes Debian-specific info in the README about build dependencies, but the script's automatic checker still only does RH/Mandrake. Still, it felt less Debian-hostile out of the box than a lot of things do. ;-) [12:14] mjg59: these are folks putting different minipci cards into notebooks? [12:16] ha ha xfs === spiv_ [~andrew@fuchsia.puzzling.org] has joined #ubuntu === Mithrandir [~tfheen@vawad.raw.no] has joined #ubuntu [12:18] bdale: Yeah === fabbione [fabbione@just.fugedabout.it] has joined #ubuntu [12:20] bdale: upstream is very cvs-update-make-make-install oriented; for a long time they didn't even really *do* releases [12:20] bdale: not a very fun position for the Debian maintainers :-) === haggai [~halls@pD9EA5F5B.dip.t-dialin.net] has joined #ubuntu [12:25] jdub: I'm a strange XFS user ... I've never had a problem with it [12:25] never seen a kernel panic, never had my data trashed, etc. [12:25] ext3 and reiser on the other hand ... lost drives galore to them! :p [12:26] that is indeed unusual, Keybuk [12:26] mdz : rythmbox says /dev/dsp doesn't exist [12:28] debianist, missing sound driver module? [12:30] mdz: I've sent email, maybe I'll be able to make an impression on them. actually, packaging an installer based on the script included in this development kit might not be an unreasoanble thing to do. worth pondering. [12:30] mjg59: my suspicion is that it's going to be hard to get the notebook folks to care a lot about this, but I could be wrong. [12:31] Yeah, I can imagine [12:36] bdale: may I whine a bit at you? About HP and Linux non-support for scanners? [12:37] Mithrandir: sure. it's a topic I like to whine about too. the fundamental problem is that HP has OEM'ed a bunch of scanner stuff that we don't own all the IP for, which means some scanners are well supported and some aren't. [12:38] seems like the current cheap-ish range of them aren't supported. [12:38] Mithrandir: I had a chance to poke at management in the right part of the company about a month ago. They "understand the issue" but I don't have a completely clear picture about if/when the situation might change. [12:38] that _really_ sucks, especially given that you've had _excellent_ driver support for a wide range of OS-es in the future -- I wouldn't be surprised if I could find a word perfect 5.1 driver for a laserjeg 4100TN [12:38] laserjet, even. [12:39] I really like your scanner and printer stuff, but if it doesn't work with linux, well, at least for scanners, we have canon. ;) [12:39] so please continue poking :) [12:40] Mithrandir: I will. meanwhile, I strongly suggest checking the SANE web site's list of supported devices before making purchases, regardless of the vendor in question. [12:40] yup, that's what I've been looking at. [12:40] I received a scanjet 3500C as a gift that I need to unload because of a lack of driver support === bdale suggests eBay === jdub suggests cousins [12:41] bdale: it must be sad to give such advice. :( [12:41] I keep my 5p for exactly this reason... it's a SCSI attached scanner that has excellant driver support. === kiko suggests bonfire === HcE_ [~hc@66.80-202-212.nextgentel.com] has joined #ubuntu [12:42] Mithrandir: well, yes. on the other hand, some Windows user will probably really enjoy it. [12:43] I actually haven't tried using my PSC-950 as a scanner while it's attached to a Linux system, but supposedly the hpoj stuff supports scanning with it. I should try that sometime. The printer/scanner/copier/fax widgets are pretty cool overall. [12:43] bdale: hah, I have the same one [12:43] I was hoping to replace it with a USB model because SCSI was getting terribly inconvenient [12:44] at least some of the PSC models are reported to work quite well under Linux [12:44] mdz: truly. my notebook supports USB 2.0, so that's become highly desirable as an interface for me personally. [12:44] mdz: yeah, printing on the PSC-950 from CUPS just works, and it hotplugs just fine === HcE_ is now known as HcE [12:45] bdale: I keep my computers in the closet and extend peripherals to my desktop via USB; it works quite nicely except that I still have a parallel printer (which requires another cable) and a SCSI scanner (which collects dust) [12:46] mdz: heh, my scanner mostly collects dust too. I use it maybe 3 times a year. [12:46] I would use it much more if it weren't such a pain to hook up; I prefer to scan documents rather than trying to maintain a physical filing system [12:47] some of the pins got bent on the cable I was using [12:47] however, I was reading up on how to get from paper to pdf the other night (sane scan pages in lineart 300dpi mode, then just give the set to convert seems to be the recipe) since I have some old machinery manuals in bad shape that I want to preserve the content from before they completely fall apart [12:47] and a new SCSI cable of the appropriate configuration would cost almost as much as a USB scanner [12:47] mdz: ick [12:48] bdale: but of course, I won't buy one of those, because I already have one I can't use and need to get rid of :-) [12:48] I've never sold anything on ebay, but I suppose I need to learn about that [12:49] scanjet 3500c seems to be going for all of $20, yuck [12:50] debianist: re #1071, dude, I do not own a time machine and therefore cannot retrospectively fix bugs in software I'd already released [12:50] debianist: Sounder 8 was released before I fixed that bug [12:50] Kamion: wimp! ;-) [12:51] is there a dvdauthor-equivalent in warty? === bdale recalls a facetious employee exam form with an option "e) allows facts to interfere with project progress" or some such... [12:53] Kamion : ok, was already close by mdz i think [12:53] Uclintu: what does dvdauthor do? [12:54] mdz: it seems to allow me to slap a bunch of mpegs together in a format suitable for writing with growisofs [12:55] I thought dvd+rw-tools had bits to do that [12:55] o [12:55] i'll check [12:55] I guess not === debianist beats himself unconsciously as a punishment === debianist [~pooh@80.179.93.85.forward.012.net.il] has joined #ubuntu === Keybuk pokes chinstrap playfully [01:33] c'mon little computer, serve a web page, you know you can do it .. [01:39] hehe, don't get too hopeful. [01:40] that explains the shit-for-beans bandwidth I have to little [01:40] rather, latency [01:41] (why do we have all this k-rad hardware and we're still all sitting around twiddling our thumbs? i blame arch.) [01:42] can I blame arch for not having k-rad hardware? [01:59] mdz: Live (not subset of) Supported is somewhat insanely weird to implement in Germinate [02:00] mdz: doable, I guess, but definitely not for 1am contemplation ... [02:00] or even 10am === StoneTable [~stone@c-67-167-111-186.client.comcast.net] has joined #ubuntu [02:02] Kamion: I don't mind if Live is required to be a subset of Supported for now [02:02] Kamion: it's just an idea that might change at some point [02:12] it would change? [02:13] oh [02:13] i see your mail [02:13] okay [02:13] i object to that idea entirely [02:13] ;-) [02:13] just for the record [02:15] odd [02:15] most of Germinate was written drunkenly at 3am ... I don't see how 1am could possibly be detrimental to it [02:18] heh [02:19] I'm going to have to invent some different seed promotion rules [02:19] yow; did somebody just flush little's outgoing mail queue? [02:20] I've started getting mails about old task changes. I'd wondered where they'd gone. :-) [02:20] I just unfucked canonical.com from within the LAN [02:21] ah yes, I'd changed it to mail cjwatson@canonical.com [02:22] elmo, nice ;) [02:23] elmo: did you find the entropy monster yet? [02:23] keybuk: no, I strongly suspect it's the kernel tho :( [02:25] what does the kernel need entropy for? [02:25] and why is that slowing things like tla down, which shouldn't be using it? [02:25] it's all very odd [02:26] I dunno exactly but it does call get_random_bytes() in a number of places, and I can't find anything using /dev/random or any other way entropy could be disappearing [02:26] we should invade some random country [02:26] this entropy crisis cannot continue [02:26] we must have resolve and courage [02:26] and shit like that [02:27] They may take our lives, but they'll never take our entropy!!! [02:27] fighting the war for more entropy! [02:39] ROCK! [02:39] laptop just arrived! [02:39] cool [02:39] jdub: rad === jamesh [~james@203-59-250-175.dyn.iinet.net.au] has joined #ubuntu [02:40] warty install coming riiiiight up :) [02:41] oh, funny: pia got a letter the other day, addressed to "P. Sith" [02:41] jdub: be careful with it, it'll break! [02:43] now i have my very own copy of windows xp [02:44] yeah, I have one of those ... [02:44] ah yes, there it is [02:44] *puts his cup back on the coaster* [02:46] cool, there's an extra battery charger in the mediabase thingy [02:46] haha === lamont goes to class. bbl [02:52] and then it dawns on me [02:52] after days of waiting [02:52] i did not think to come up with a computer name [02:52] I still haven't named mine [02:52] catsby [02:52] i'll donate you my name === jdub types 'UBUNTU' into the windows xp box [02:53] fortunately I got a suggestion from a friend for the amd64 box, otherwise I'd've had to think [02:53] and there's one spare for future use [02:54] jdub: what's your naming scheme? [02:54] i don't really have one [02:54] you're so doomed :-) [02:54] my desktop is lazarus [02:54] haha [02:54] my gateway is katia [02:54] think of a naming scheme, makes life so much easier [02:54] took me four years to come up with a naming scheme [02:54] jdub: leper? [02:54] my ibook is willow [02:55] lucifer [02:55] my qubes are quick and quark [02:55] to the untrained eye, my naming scheme is mostly "hard to spell" [02:55] Kamion: I thought that was Mithrandir's [02:55] riva, arborlon, cairhien, dhiammara, crydee, perimadeia, cittagazze [02:55] mine is alice in wonderland [02:56] Keybuk: there's a subtle difference between 'hard to spell' and 'norwegian' [02:56] jdub: he uses pwgen to invent them [02:56] they're really not words in any language [02:56] cool :-) [02:56] hfsnw [02:56] jdub: cockfosters [02:56] Kamion: characters from bad sifi? [02:56] eleven fucking system tray icons [02:57] and it's NEW [02:57] hey spiv_ [02:57] Keybuk: cities from mostly-bad fantasy [02:57] but close [02:57] i like cockfosters [02:57] that is so much cock [02:57] jdub: Hey. I'm about to sleep, but I thought you might like that suggestion before I zonk. [02:57] Kamion: am I okay to clean out the old kernel stuff now? both in terms of not built stuff and migrating to universe? [02:57] elmo: fine by me [02:58] mine is nanasawa, tycho, gabe (fd.o), twisp, catsby, brenna/kara (in reserve), kimiko, rentazilla, piro, largo, shirtguydom, ed, junpei, seraphim, boo, et al [02:58] if anything on cdimage is still using it I want to know about it anyway [02:58] the penny-arcade line is now the active one [02:58] I like "rentazilla" [02:58] Megatokyo fan, eh? [02:58] Kamion: used to be :) [02:58] daniels, where's the fruitfucker? :) [02:59] don't read it myself; several of my friends do, though [02:59] Hrdwr_BoB: the iBook I had during lca was fruitfucker2000, actually [02:59] Kamion: penny-arcade is great [02:59] heh, I amusingly keep a list of future names for machines [02:59] and I *never* use it [02:59] thom came up with syndicate [02:59] Keybuk: buy more machines, then :P [02:59] bluetooth, network, network, sound, 'safely remove hardware', dell quickset, intel wireless, on ac power, synaptics, control suite, starting with windows [03:00] daniels: isn't that, it's that they're just boring and not new :p [03:00] Hrdwr_BoB: so, I was standing in some homewares shop in Brunswick St, and I saw a Fruit Reamer, that was blurbing about how good it could ream the pulp [03:00] i burst out laughing and wondered if the creator read p-a :) [03:00] jdub: what's that, questions asked? [03:00] Kamion: systray icons [03:00] jdub: arse [03:00] ah [03:00] absolutely amazing [03:00] 'safely remove hardware', aka unmount [03:01] daniels, ehehehe [03:03] daniels: dude, you just tangoed my monitor [03:04] elmo: NO I DIDN'T! [03:05] elmo: (what happened? i assume 'tangoed' is some form of slang for 'completely rooted') [03:05] daniels: "sprayed drink over" I suspect [03:05] or, at least, caused elmo to do so [03:05] oh, cool [03:05] what keybuk said. tango's an orange fizzy drink, like fanta but nice [03:05] it's, um, "creamier" [03:06] heh. well, that's a little better than what I was thinking -- I was thinking that X had made the magic smoke escape or something in your sick quest to drive your monitor to eleventy billion x 2390 at 4Hz or whatever it was you were shooting for [03:07] lol 4hz [03:08] I like Belgian Fanta. You should get it in the UK. === Kamion imagines a wrap-around-entire-room monitor [03:08] mezzo mix is nice [03:08] Kamion: nice... [03:09] (coke + orange mineral water) [03:09] Kamion: don't forget the data glove! [03:09] Do iiyama do IMAX ? [03:09] Kamion: (although when anyone started talking about arch or any of the other projects, they'd end up screwing up their system totally) [03:10] hey, at least you could code by hand-waving! [03:10] in windows xp, you "repair" a connection [03:10] (welcome to my punchline.) [03:10] (dhcp release/renew) [03:10] jdub: does that automacially rebuild the links some clown punctured with an anchor? [03:10] jdub: it does kinda capture how people think though [03:11] "MY NETWORK IS BROKEN ... what happens if I click 'Repair'?" [03:12] i think one of my network cables is broken [03:12] then it should be "Repair Internet" [03:13] "Are you sure you wish to move The Internet into the Recycle Bin? Once done, you cannot get it back" [03:13] inn-err-net [03:17] actually [03:17] does Mac OS X hang the modem up if you drag it onto the Trash? [03:18] or scrub the firmware? [03:18] and if you drag "My Mac" onto the Trash, does it power off? [03:18] or didn't they think through the metaphor very well, after all [03:22] o acpi-modules-2.6.8.1-1-amd64-generic-di [03:22] o libdebian-installer-extra4-udeb [03:22] o ufs-modules-2.6.8.1-1-386-di [03:22] o ufs-modules-2.6.8.1-1-amd64-generic-di [03:22] kamion: all those are candidates for demotion for universe ? [03:26] elmo: acpi-modules isn't, let me fix the seed === haggai_ [~halls@pD9EA620E.dip.t-dialin.net] has joined #ubuntu [03:28] elmo: libdebian-installer-extra4-udeb is only used by cdebootstrap which we don't use, so I zapped that [03:29] ok, zapped in the archive too [03:29] elmo: ufs-modules can go on all architectures, I just updated the seed to remove it from powerpc too [03:30] it was there for potential use by os-prober to detect BSD systems, but that support's never actually been committed === jdub_ [~jdub@home.waugh.id.au] has joined #ubuntu [03:31] Kamion: hmm [03:31] Kamion: interesting installation wos [03:32] woes [03:32] i'll start again and step through [03:32] jdub_: hm, I'm going to bed now, sorry [03:32] kamion: cool, thanks [03:32] sucky hardware, most likely [03:32] Kamion: ok, i'll send mail === tvon [tvon@dsl093-119-225.blt1.dsl.speakeasy.net] has joined #ubuntu === kfish [~conrad@balgowlah-ep.tip.csiro.au] has joined #ubuntu [03:59] hrrrrmmm, interesting [04:09] daniels: ? [04:10] elmo: ! [04:10] daniels: xfree86 has a bogus b-d on libglide3-dev for amd64 [04:11] i think glide needs to be synced back, then [04:11] iirc glide in sid does amd64 love [04:11] yeah, I'm just seeing that [04:11] WTF [04:11] there's a libglide3 for amd64 but no libglide3-dev? [04:12] rad!! [04:12] <3 glide [04:14] mdz: hrm, must be related to scsi issues [04:14] Package: libglide3-dev [04:14] Section: libdevel [04:14] Architecture: i386 alpha ia64 amd64 [04:15] shouldn't dpkg-somethingorother blow up if it doesn't actually produce the binaries it's supposed to? [04:15] maybe it should, but it doesn't [04:15] e.g. dpkg itslef has had some binary it doesn't build in it's control file for eons [04:16] my howl package doesn't include a bunch of stupid shit [04:17] hrm === jdub_ ponders how to do the install [04:17] anyone done a netboot for a while? [04:19] we netbooted the DL140's before the conference [04:24] does d-i store any vars on the disk? === jdub_ can't find the debconf db === stub [~zen@dialup-131.80.220.203.acc05-dryb-mel.comindico.com.au] has joined #ubuntu [04:30] jdub_: there's a dir called cdebconf somewhere under /var [04:30] questions.dat and templates.dat [04:31] hrm, questions.dat maybe === jdub_ nastyhacks [04:32] i had to do that for partman (along with nano'ing the binaries it was calling) on the 4am-on-the-last-day cd :) === jame1 [~james@support.anchor.net.au] has joined #ubuntu [04:38] elmo: you going to file that glide thing in bugzilla for Mithrandir? [04:42] yo jame1 === stubish [~zen@dialup-30.83.220.203.acc06-dryb-mel.comindico.com.au] has joined #ubuntu [04:57] mdz: dude, your bong radar is on the money [04:58] we should so rip out that hdparm stuff [05:00] was that one of the problems with the real scsi stuff too? [05:03] jdub_: hey hey. Sorry, at work and only sort of here. [05:03] jdub_: seems plausible, dunno [05:03] hmm [05:04] so it's booting up for the second stage [05:04] it just loaded the ipw2200 (before tg3, grr) [05:04] then it hit loading usb modules [05:04] ieee194 [05:04] and now it's stopped [05:04] ctrl-c isn't loving me [05:05] ooh, there was an oops earlier [05:05] hci_usb_probe: can't set isoc interface settings [05:05] unable to handle kernel null pointer dereference === jdub_ boots again [05:06] hrm, there we go === Gman [~Glynn@amfea-proxy-2.sun.com] has joined #ubuntu [05:12] hans sure is out of his head === mdz reads LWN [05:14] new crack? === jorge_ [~jorge@arslinux.com] has joined #ubuntu === jorge_ is now known as whiprush [05:16] yp whiprush [05:16] yo, rather ;) [05:16] heya jdub, just the man I wanted to talk to... [05:41] whiprush: hey dude [05:43] hrm [05:44] no /dev/cdrom or /dev/scd0 [05:48] morning guys [05:54] hey daniels, jdub_. [05:55] ber [05:55] daniels: you missed the whole slashdot thing for your intervie while you were gone. heh. [05:56] whiprush: heh, I caught up on that a couple of days later [05:57] your fdo thing for the kde conference was sweet, btw, from what I saw from the fluendo coverage. [05:57] that second dude asking questions was gunning for you, heh. [06:01] heh :) [06:01] yeah, the small dude was aseigo [06:01] martin konold was being a dick, though [06:02] oh, I couldn't hear the name, was that the guy? the one whiner dude. [06:03] mdz: any weirdarse tricks to get ipw2200 going? [06:03] jdub_: nope, Just Works for me [06:04] hrm [06:04] as of sounder 8? [06:07] before [06:07] apt-get install linux-image-2.6.8.1-blah and it goes [06:07] is the driver not getting loaded? [06:07] mdz: can you approve the sync for freenet6? [06:08] driver's getting loaded [06:08] it's using firmware_class [06:08] but it's not getting any linkage [06:09] output from iwconfig is basically bare beyond the essid [06:09] no ap [06:12] fabbione: didn't I? [06:12] mdz: i don't see any answer in my inbox [06:13] you approved lvm2 and it is done [06:13] I must have confused the two [06:13] sent now [06:13] thanks :-) [06:15] ok [06:15] never mind [06:16] it's not grokking my ap [06:16] hrm [06:16] will there be a sounder 9 prior to the public release? Or should I test 8? [06:16] I missed it by a few days [06:19] grab 8, or a daily [06:19] k [06:19] jdub_: does explicitly calling 'key off' help? [06:20] one sec [06:20] upgrading via some random network [06:20] not my own! gar! [06:20] I'm wondering how people are going to handle the sudo thing. It certainly demolished all my bad habits. [06:22] mdz: approx how long does it take you to do i386 and ppc installs? [06:22] jdub_: not long, why? [06:22] whiprush: I think the current daily is no good [06:22] sounder 8 is probably your best bet [06:22] ok. [06:23] it has a couple of known issues, but nothing an apt-get upgrade won't fix [06:23] k [06:23] Colin's announcement to the sounder mailing list describes them [06:23] the sources.list still point to the cd, or do I need to update it manually? [06:24] mdz: i need times :) [06:24] jdub_: I'll time it the next time I go through [06:24] remind me [06:24] ok [06:24] ok, I'll redo the laptop install, I doubt there will be any major problems, but I'll report to the list anyway. [06:24] whiprush: you'll need to uncomment the sources.list entries for ftp.n-n-y.com [06:24] to get the latest packages [06:25] k [06:25] 118 upgraded, 0 newly installed, 0 to remove and 0 not upgraded. [06:25] man - I just upgraded yesterday or so, it seems... === justdave_ [~dave@24.247.63.44.gha.mi.chartermi.net] has joined #ubuntu [06:32] whoa [06:32] this is fast [06:33] your laptop? [06:33] yeah [06:33] is this the dell-crack300? [06:33] yeah [06:33] well, it's gotta be faster then the toilet seat, hasn't it? [06:34] it's only a 1.4GHz centrino [06:34] but blammy fucking faster than willow [06:34] with 1MB of fucking cache [06:34] let's just cover that again... [06:34] 1MB of fucking cache! [06:34] of course it's fast [06:35] how much cache do I have? [06:35] [davyd@pingu ~] $ cat /proc/cpuinfo | grep "cache size" [06:35] cache size : 512 KB [06:35] the new 2gig centrinos have 2MB as well [06:35] so not only is that faster then this CPU, it has 4 times the fucking cache [06:35] the dothans do, they go as low as 1.2 ghz, with the 2mb cache. === Riff continues to be bitter [06:35] whiprush: I didn't know they went that low [06:35] I've only seen 2gig versions [06:35] your machine is called pingu? that's so rad. [06:35] yeah, it's sick. [06:36] Riff: you just made my day :) [06:36] daniels: stephanie's is called pinga [06:37] Riff: the wattage difference from the banias centrino's is significant also. I've seen 20%+ improvement in battery life with the new chips. [06:37] whiprush: now if only someone would manufacture a new laptop I want to buy [06:37] the P-M is a sweet processor as well [06:37] it massively outperforms the P-3 and P-4 [06:38] Keybuk: could this be anything to do with it's 1MB OF CACHE!!?!?!?! [06:38] Riff: it's also because it wasn't designed by Intel Marketing [06:38] Riff: have you heard elmo's favourite song? [06:38] and the fact it's pipeline isn't 4 quadzillion instructions long [06:38] Riff: yeah I'm waiting for fujitsu to put them in their P series. [06:38] the P-4 was basically a " we want you to double the clock speed" ... " ok, we'll cheat, and double the pipeline length" [06:38] whiprush: I haven't looked at Fujitsu laptops, are they any good? [06:39] P-M uses most of its cache for instruction window and ROB stuff [06:39] daniels: undoubtedly, as Stephanie is a muppets fanatic [06:39] the P series is fantastic. Their battert life easily tops 10 hours. [06:39] however, you may have to remind me [06:39] 1MB is kinda useful for that :) especially with it's branch prediction magic [06:39] I was looking at an IBM X40, except that it doesn't have a synaptic touchpad [06:39] Riff: yeah, that's why I didn't buy one [06:39] which I found offensive to my delicate nature [06:39] they're also a bit under-specced for the price [06:40] My current dell laptop has a P4-M, which is an atrocious chip, thing heats up like the sun and lasts 3 hours with dual batteries. Weight is about ... 10 pounds. :( [06:40] that's just IBM [06:40] 1.2Ghz instead of 1.6 or 1.7 [06:40] whiprush: you mean a P-M ... there isn't a P4-M [06:40] or a P-4 ? [06:40] Keybuk: sure there is [06:40] I have one in here [06:40] there was a Pentium 4-M. [06:40] jdub: so what's the debugging artwork picture of? [06:40] lamont: the background? [06:40] ugh, that sounds like a marketing name [06:40] it was a horrible chip. [06:40] the sticker clearly says Pentium-4M [06:40] you tell me a street and I'm gonna scream [06:40] lamont: a road and the sky [06:40] Keybuk: I think it has extra speedstep crack [06:41] ah, that'd figure [06:41] was wondering if that was your house or something... :-) [06:41] It's a P4 with half the cache and speedstep. Horrible. [06:41] lamont: it's actually a garrett le sage photo, from the USA somewhere :) [06:41] ah, ok [06:41] finally had few enough windows on the screen to notice it. [06:41] quite why you'd put speedstep in a Netburst processor is just silly [06:42] "we'll scale down the speed of a processor designed specifically to fake faster clock speeds" [06:42] You're telling me ... it's a menace. Can barely tolerate it on the lap with the heat. [06:43] The newer dell's with the P-M's however, are sweet. The 2Ghz one will hang with any desktop chip. [06:43] HP's new blade servers use P-M's, totally sweet. [06:43] well, the fact a 2Ghz P-M performs like a 3+Ghz P-4 helps [06:44] thom: DUDE! I HAVE THE BEST BUG FOR YOU! [06:44] run away [06:44] the 1.0 thru 1.4 dothan's are ultra-low-voltage too, perfect laptop stuff, run great, and the fan doesn't even kick on. [06:45] so you're telling me I can get a 1.4 gig P-M with 2MB of cache [06:45] I can generally keep my laptop on my lap, only big compiles warm it up ... and then the extra fan comes on and my legs are happy again [06:45] and it will suck no power, and generally rock my world? [06:45] doko about? [06:45] well, the chip is out, wether the mfg's are shipping them yet is another thing. [06:45] dell for sure is shipping them [06:45] I don't much care of Dell [06:46] And certain IBM's also. But it depends on how you pick stuff on their website. [06:46] so [06:46] i have the battery module loaded [06:46] and there's nothing in /proc/acpi/battery/ [06:46] jdub_: apparently you don't have a battery [06:46] jdub_: bust DSDT ? [06:46] your laptop runs off a small nuclear generator [06:46] i have two :) [06:47] well [06:47] i'm plugged into AC [06:47] jdub_: you don't still have that nasty ass green toilet seat ibook do you? [06:47] with one battery in the docking station [06:47] and one battery in the laptop [06:47] Riff: HP are only shipping the nc4010 with 1.7 1MB cache at the moment [06:47] whiprush: that's what i'm typing on now, but i just got a new lappy today [06:47] oh, nice. [06:49] I'm looking for a new laptop myself, I really want a 15" powerbook, but I hear the ppc support for non-free codecs is kind of crap. [06:49] 10.6" screen" [06:50] ratio of 5:3 [06:50] I assume 10.6" is measured diagonally across the LCD surface [06:51] chris dibona works at google now [06:53] neat [06:53] cool === edd waves [06:53] yo edd [06:54] edd: dude! [06:54] hmmm... [06:54] i have bluetooth ready hardware :) [06:54] roughly in your timezone for once. well, PST. [06:54] jdub_: cool. i hope it just works :) [06:54] apparently the dimensions of this screen are in negative space [06:54] and going to get a bluetooth phone too :) [06:54] mdz around? === Riff thinks for a while about that === edd fears if more people use his software he'll actually have to finish iit [06:55] edd: yo :) [06:55] jdub_: the k700i is a very nice phone [06:56] edd: i use g-p-m and g-o-s, as well as the other gnome-bluetooth stuff, and it's mad phat [06:56] respek' [06:56] it looks like cock [06:56] the T630 looks nice though [06:56] eh? the k700i looks like love [06:56] nice screen, too [06:58] wait for the T700, it's rumoured to be sex itself [07:00] hey edd I dug your ifolder blog entry. One could almost get away with packinging it in experimental. *cough* [07:01] whiprush: heh. i think it's a little too buggy right now. but it wouldn't harm to play with the packaging [07:02] http://www.whiprush.org/2004/08/edd_dumbill_loo.html <--- I did my own thing but I don't think your blog does trackbacks. [07:02] it works great expect for the "sync when I feel like it" part. heh. [07:03] Keybuk: probably $bling when it comes out here, though [07:03] Keybuk: the k700i is a mere $au700 outright (not simlocked), and that's the exact same as the t630 [07:04] ouch [07:05] it's cheapish here [07:05] but it's aimed at the teen market anyway (the "K" is for "Kids") === cef_work [~cef@fw-01.amc.com.au] has joined #ubuntu [07:06] there's an S700i coming out RSN (Swivel phone) [07:06] Keybuk: tjat [07:06] Keybuk: that's daniels' market [07:06] good point === jdub [~jdub@dsl-202-52-33-118.nsw.veridas.net] has joined #ubuntu [07:06] ahr [07:06] veridas [07:07] that must be next door [07:08] we know there's a T700i coming soon because [07:08] a) it would be logical [07:08] b) there's been previews of it [07:08] c) they've started listing it under "Supported" for things like the power adapter [07:08] i tried to buy a cellphone here in the US today. no go with no SSN or US credit card [07:08] heh :) [07:08] edd: ah, suck :\ [07:08] i tried buying one in the uk, but the fascists simlock everything [07:09] http://www.mobile-review.com/forum2/attachment.php?s=8eab7f12f6488b6f22da7ee25e789835&attachmentid=1465 [07:09] daniels: cheap to unlock though [07:09] daniels: walk to the nearest market and pay a few quid for them to unlock it [07:09] ber [07:10] was someone's guess at what it would look like [07:11] the k700i was effectively free, anywya [07:11] old plan expired, $0 to get the same terms with a k700i [07:12] woo, people are moving out of the house [07:12] and it also comes with my telco's firmware, not vodafone's :P [07:13] this means I'm going to need to transplant the router into a new case [07:13] because he wants his case back [07:13] and I'm going to have to by a new printer [07:13] *buy === Riff wonders how much a colour laser printer costs === pitti [~pitti@195.227.105.180] has joined #ubuntu [07:15] Morning guys! [07:16] afternoon! [07:18] jamesh: you around? [07:18] the other good thing about a k700i is that two sim cards fit quite nicely between the battery and the cover [07:19] lamont: yeah. === Gman [~Glynn@amfea-proxy-2.sun.com] has joined #ubuntu === lamont declares victory, heads to bed. [07:27] lamont: night dude [07:46] hrm [07:46] Is nothing done with removable media at this point? [08:00] tvon: what do you mean in particular? [08:03] pitti: ah, not very clear was I? I'm talking about desktop icons and automounting. I popped a CF card into my laptop and there were some log messages but no mounting or icons. [08:04] I know there is a thread on the list about this stuff but I wasn't sure what was "supposed" to be happening with the current setup [08:11] hah! found someone with 18 active uids in his gpg key; i'm not the worst. [08:11] so i don't know what this means, but supposedly i need a new dsdt [08:11] which i have [08:11] but then i need some compiler [08:11] from intel, which i can't build [08:11] and then i need to rebuild my kernel [08:12] which i swore i would not do when using ubuntu [08:12] dsdt? [08:12] some acpi thing [08:15] mdz: so what is the consensum about apt lines? [08:16] fabbione: my feeling is that security and warty-via-network should be treated the same [08:16] jdub_: yeah, the DSDT is the map of ACPI thingoes you get from the bios [08:17] so the kernel knows where to look for things [08:17] jdub_: why do you need a compiler for that? [08:17] you should just be able to patch your kernel and compile with gcc [08:17] there's some intel iasl special acpi compiler thingy [08:17] hmm [08:17] jdub_: they don't just give you a binary blob? [08:17] jdub_: grahame bowland just did some stuff with this [08:17] for his laptop [08:17] tvon: actually it should automatically mount the device (see Computer window), and (unless you disabled it) a Nautilus window with the Card's contents should pop up [08:17] mail grahame@angrygoats.net [08:17] most of the ones on acpi.sf.net are not blobs [08:18] he should know what to do [08:18] hrm [08:18] tvon: I would be interested in debugging this. [08:18] tvon: does the device appear in the Device Manager? [08:18] tvon: (i. e. in hal-device-manager) [08:18] one min [08:18] mdz: ok.. i will see how complex it is to do something cool.. [08:19] mdz: but i would like a general consensum before implementing the changes [08:19] brb [08:20] fabbione: Mark and I agree that supported should be made available if there is a network, even if we are doing a CD install [08:21] mdz: ok... fine by me.. i will do the changes today [08:21] hal keeps crappping out it seems [08:21] thanks [08:22] tvon: so it does not show your CF reader? [08:22] mdz: do we agree to keep universe out? [08:22] mdz: or better.. commented out [08:22] tvon: actually you should see the device itself and one or more volumes [08:23] fabbione: I think so, yes [08:23] at least until we have some barrier in place [08:24] pitti: I'm getting some sorta loop here. [08:24] tvon: a hal loop? THAT sounds interesting [08:24] tvon: what do you mean [08:24] tvon: ? [08:25] pitti: http://paste.plone.org/1376 [08:25] tvon: BTW, yesterday I tested an USB CF reader myself, worked perfectly... [08:25] tvon: I'l look [08:25] I get a "Lexar ATA Flash" showing up in h-d-v..and it pops back and forth between having a volume under it and not [08:26] tvon: ugh! why does udev create and remove the device? [08:26] tvon: did you change anything in udev's rules? [08:26] pitti: nope [08:27] the line in question (29) is here: [08:27] BUS="ide", KERNEL="hd[a-z] *", PROGRAM="/etc/udev/removable.sh %k", RESULT="1", NAME="%k", MODE="0660", GROUP="plugdev" [08:27] tvon: ah, is this a PCMCIA card reader? [08:28] I believe so. Its the onboard CF reader on an X31 [08:28] s/onboard/builtin/ [08:28] tvon: ah, since these are neither USB nor FireWire, we currently don't support automounting them [08:28] dang [08:29] tvon: this will hopefully be sorted out in Hoary, but not for Warty any more [08:29] tvon: the problem is that we don't want to automount fixed hard disks [08:29] tvon: but it is very hard to decide which device node is a fixed one and which is a PCMCIA [08:29] ah [08:29] tvon: in fact, apart from the fact that the PCMCIA node may appear later, there is nothing in the kernel nor in sysfs that tells us that the PCMCIA thingy is removable [08:30] tvon: but still, this udev loop looks scary [08:31] yeah [08:31] tvon: is the device node actually removed and created? Can you see this by polling ls -l /dev/hde* ? [08:32] mhm, all the usual hde's are there, how can I tell? [08:32] (the card is out now and they are there) [08:32] ew....wth [08:32] pitti: card is out, but I'm getting this every few seconds in daemon.log: [08:33] Sep 9 02:32:43 localhost udev[4492] : removing device node '/dev/hde1' [08:33] more loopy goodness [08:33] tvon: udev keeps adding hde1 although the card is out? [08:33] nah, its removing it now it seems [08:34] woah [08:35] re-added the card while watching the log, it tries to make all the /dev/hde's but they already exist [08:35] http://paste.plone.org/1377 [08:35] Should they already exist? [08:36] nm, they must have been created last on an earlier load [08:36] tvon: actually not, udev should only create devices that actually exist, not all of hde1 to hde24234234 in advance [08:40] and they should be removed when the device goes away [08:40] tvon: right. but the cardmgr is the one that seems to create these devices [08:41] tvon: they probably still exist from an earlier invocation of cardmgr [08:41] cardmgr tries to create device nodes? eek [08:41] tvon,mdz: I've got an idea: udev wants to create these device nodes as root:plugdev, so it probably deletes the existing ones and creates them again [08:42] tvon,mdz: maybe cardmgr does the opposite... [08:42] mdz: but you are right, cardmgr should not care about creating devices (but it says so in the log) [08:42] mdz: a lot of failed mknod messages [08:43] tvon: BTW, I just got an idea how to get your card reader working :-) === Gman [~Glynn@amfea-proxy-1.sun.com] has joined #ubuntu [08:44] tvon: if "line 29" applied, this must mean that /sys/block/hde/removable is '1' on your system. Can you please verify that? [08:44] tvon: if so, then the change in pmount is trivial [08:45] pitti: yup [08:45] mdz: what do you think about the following pmount policy change: "device is removable" -> "device is removable || device node is writable by the user"? [08:46] pitti: if cardmgr tries to create devices, we should just disable that [08:46] because udev will handle it [08:46] pcmcia generates hotplug events [08:46] mdz: since udev assigns the plugdev group to removable nodes, the user can write the device anyway [08:46] mdz: agreed, I will look at cardmgr [08:47] tvon: Great! This did not work on another PCMCIA card, but then at least we can support some (including yours :-) ) [08:47] pitti: lovely :) [08:47] tvon: if mdz agrees, that is [08:47] pitti: I am not so sure about that policy change [08:47] cardmgr seems to be making them all root:root [08:48] tvon: eek [08:48] tvon: I will beat the hell^W^W device creation out of cardmgr [08:48] heh [08:48] mdz: but if the user has full write rights on the device anyway it should not be too critical to allow him to mount it, dont? [08:49] pitti: I am not so sure [08:49] mdz: okay, I will postpone that [08:49] mdz: we just should think over that. But before I'll deal with cardmgr [08:50] mounting lets you command something to happen in the kernel [08:50] whereas being able to write to the device is purely a user concept [08:50] mdz: right. OTOH we allow it on other removeable devices, too... [08:50] mdz: udev will only put removable devices in plugdev [08:50] yes, and I want to avoid being more liberal than that [08:51] there could be block devices that the user can write to which are not meant to be mounted [08:51] and we should not let them attempt it [08:51] mdz: so pmount should check that it is an IDE/SCSI volume [08:52] mdz: (which would be a good idea anyway) [08:52] that is not trivial to check, considering that Linus is always threatening to randomize the major device numbers :-) [08:52] mdz: I actually thought of looking into sysfs [08:52] I think that is probably the correct approach [08:52] mdz: I will think over it. I report back this evening. Before I will fix cardmgr and start to go over the security review [08:53] mdz: BTW, I have to change pumount to support lazy unmount. Do you see any problem with this? [08:53] mdz: we need the support to better handle the case if the user just rips off his devices [08:53] mdz: thom will modify hal to lazily umount removed (but still mounted) devices [08:54] mdz: he wants to call pumount -l for that [08:54] pitti: what needs to change in order to support it? [08:54] pitti: yes, I agree, we need to try to handle that case as well as possible [08:55] mdz: i have to add an option '-l', pass it to umount [08:55] mdz: and I have to replace the check "device node exists" by "device node exists or -l is given" [08:55] mdz: because if the user removes the device, /dev/sdxx will not be there any more for unmounting [08:55] pitti: in that case, you may as well just remove the check for whether the device node exists [08:55] mdz: but we still have the check that the device node must be present in mtab [08:56] mdz: so I guess that's okay [08:56] mdz: I will leave it in pmount and throw it out of pumount, I guess [08:56] Aight, bedtime [08:56] mdz: Okay, I can throw it out completely, but it does not hurt; maybe it saves from erroneous calls (not from malicious ones) [08:57] pitti: ok, if it protects against some error cases [08:57] night folks [08:57] as long as we are not relying on it for security [08:57] tvon: night [08:57] tvon: good night! Can I bother you when you got up with a new cardmgr? [08:57] pitti: sure, I'll give it a whirl [08:57] mdz: no, we don't. [08:57] pitti: just /msg me or something when you want me to try something [08:58] mdz: if a device is mounted and the user mounted it, he should be able to umount it even if the node disappeared [08:58] yes [08:58] mdz: that's reasonable, I think [08:58] tvon: Good night! [08:59] pitti: I just noticed somethin which does not look right [08:59] if( !realpath( argv[1] , device ) ) { [08:59] perror( "Error: could not determine real path of the device" ); [08:59] return -1; [08:59] } [08:59] /* does the device start with DEVDIR? */ [08:59] if( strncmp( argv[1] , DEVDIR, sizeof( DEVDIR )-1 ) ) { [08:59] I think this ^^^^ should look at device, not argv[1] [08:59] mdz: argh! [08:59] mdz: right, this was correct before I checked with realpath. [08:59] bedtime for me [08:59] mdz: thanks! [08:59] good night [09:00] mdz: good night! === Gman is now known as GmanAFK === silbs [~jane@dumbledore.hbd.com] has joined #ubuntu [09:01] silbs: Good morning! [09:06] silbs: morning [09:15] good morning! [09:32] anybody else out there think warty is starting to look really slick? [09:33] sabdfl: I talked to sb about our current security policy and its implementation, he was really impressed :-) [09:34] sabdfl: yesterday I showed the handling of removable devices to a Windows user, he was impressed that it "just worked" without installing any driver and stuff [09:34] sabdfl: so, yes! [09:35] so we have the full remove thing happening now? [09:35] daniels: see the mailing list. Right now nothing changed yet [09:36] daniels: but thom and I currently handle the "just rip it off" case [09:36] awesome :) [09:36] daniels: which should actually work not too badly [09:36] congrats, that's sensational [09:36] daniels: unless, of course, there are still processes that want to write to the device. These are lost, but we cannot help that [09:37] yeah === ploum [~ploum@191-221.242.81.adsl.skynet.be] has joined #ubuntu [09:38] Kamion: i am keeping a lock on base-config === sabdfl [~mark@wblv-228-55.telkomadsl.co.za] has joined #ubuntu === thom yawns === warthylog [~warthylog@port1845.ds1-khk.adsl.cybercity.dk] has joined #ubuntu === Topic for #ubuntu: SSDS | http://sounders:oink@wiki.no-name-yet.com/ | Bugs: https://bugzilla.no-name-yet.com/ === Topic (#ubuntu): set by Kamion at Tue Sep 7 21:03:40 2004 === #ubuntu [freenode-info] please register your nickname...don't forget to auto-identify! http://freenode.net/faq.shtml#nicksetup === warthylog [~warthylog@port1845.ds1-khk.adsl.cybercity.dk] has joined #ubuntu === Topic for #ubuntu: SSDS | http://sounders:oink@wiki.no-name-yet.com/ | Bugs: https://bugzilla.no-name-yet.com/ === Topic (#ubuntu): set by Kamion at Tue Sep 7 21:03:40 2004 [09:57] thom: morning! [09:58] hi pitti [09:58] morning thom [09:58] thom: I'm just at preparing an updated pumount which provides option -l for lazy unmount. Should help you with your hal changes [09:58] fabbione: Morning! [10:01] I've got the trash applet using the -accept icon when you hover over it now [10:03] jamesh: awesome === lucas_ [~lucas@cab-192023.calixo.net] has joined #ubuntu [10:05] hi [10:05] hey sabdfl [10:06] how is life on the other side of the world? [10:06] sabdfl: I don't know whether it is feasible to change the drag icon in time for Warty though. [10:07] which trash applet is this? [10:07] the one on your default ubuntu desktop panel, bottom right corner :-) [10:08] assuming I currently had one ;) [10:08] gcc -m32 -nostdlib -r spaceorb.o -o spaceorb_drv.o [10:08] spaceorb.o: file not recognized: File truncated [10:08] collect2: ld returned 1 exit status [10:08] make[7] : *** [spaceorb_drv.o] Error 1 [10:08] fabbione: good day today so far, lots of catching up. how's dannemark? [10:08] jesus, my system is cursed [10:08] I'm wondering if it's the one I was intending to push upstream [10:09] are there several? [10:09] no idea [10:10] this is Trash Applet 0.2 (c) 2004 Michael Stikkes [10:10] Riddell: http://luon.net/~michiels/trashapplet/ [10:10] what's the expected outcome of plugging a usb-audio device into an ubuntu machine? [10:10] ah, Michiel Sikkes [10:10] s/riddell/riff/ [10:10] aieeee, a gcc bug [10:11] jamesh: ok, I think this is the one I was mailed about [10:13] What would look cool would be to use Xrender cursors for dnd [10:13] Kamion: base-config ubuntu15 is up, but i will need a cd (even a daily) to test it completlely [10:13] so you'd be dragging a semi-transparent icon for the file [10:13] Kamion: since there are some differences in installing from cdrom and simulating an installation [10:17] you could drag a semi-transparent icon for the file easily, but it would look like complete crap [10:17] a) doing stuff like that is massively prone to tearing, doubly so if you don't do vsync (we don't), b) updates that often wouldn't be the best on the cpu === seb128 [~seb128@ANancy-111-1-23-50.w81-50.abo.wanadoo.fr] has joined #ubuntu [10:19] morning [10:20] hi seb128 [10:20] hey jamesh :) [10:20] got your mail, not replied yet [10:21] seb128: cool. Will have more patches soon :) [10:22] daniels: I remember jg and keithp talking about using Xrender cursors to handle semi-transparent window moves in the window manager ... [10:23] convert window contents to picture, adjust alpha, create cursor [10:24] daniels, mdz: X drivers update... as far as i can tell we need to update the nv driver and the wacom driver. daniels also had the ati stuff pending and there is a patch for r200 in the BTS to fix a segfault (or something similar) [10:24] yeah [10:24] it's doable [10:24] fabbione: i think mdz is asleep, but i believe the r4xx verdict is warty point release [10:24] should we take the chance? or we will leave it for hoary? [10:24] fabbione: do you mean the r200 dri stuff? [10:24] yes [10:25] fabbione: if so, i'm torn -- it's a pretty huge semantic change, but it does come from upstream cvs, and r200 dri seems quite solid there [10:25] fabbione: i don't see the harm if it's out there on a plate for us [10:25] daniels: yeah it's a 1600 lines patch :( [10:25] fabbione: i'll take the wacom stuff [10:25] daniels: what about the nv? [10:26] fabbione: i think we should just take the entire nv driver from 6.8, tbh [10:26] daniels: i don't mind if you want to do the updates at all [10:26] fabbione: cool [10:26] daniels: it was just to take a decision and coordinate [10:26] daniels: just send me the usual ubuntu15 -> ubuntu16 patch [10:26] fabbione: well, if i take wacom and i'll cop the ati stuff as well, do you want nv? [10:26] fabbione: yeah [10:26] fabbione: i'm chasing up some far more serious issues atm, however [10:27] but i'll be able to look at it tomorrow [10:27] daniels: i would rather prefer if you update stuff from upstream [10:27] daniels: since you know where to dig and what to replace [10:27] daniels: i can test the nv driver and the ati, if that's what you want [10:31] fabbione: ok, that'd be good thanks [10:31] i can test ati with the 8500 and 9000 i have here, but yeah, i have no nv or wacom hardware [10:31] so i'll get jaq to test wacom, and if you could test nv, that'd be great [10:36] yes i can test nv and one ati (mobility something) [10:36] 3 nv actually :-) === ddaa [~david@nemesis.xlii.org] has joined #ubuntu === Kinnison [~dsilvers@extgw-uk.mips.com] has joined #ubuntu [10:45] Morning [10:48] hi Kinnison [10:50] hey Kinnison === Kinnison hugs fabbione and waves at sabdfl [10:53] thom: are you already working at the lazy unmount hal? [10:53] Kinnison: hey dude [10:53] heyhi daniels === lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has joined #ubuntu [11:08] hiya, Kinnison === Kinnison bounces Mithrandir [11:14] hey Mithrandir [11:14] hiya fabio === fabbione tests crack of the day with new base-config === rburton [~ross@82-44-126-41.cable.ubr03.croy.blueyonder.co.uk] has joined #ubuntu === savs [~savs@jicbioinc1.ext.uea.ac.uk] has joined #ubuntu [11:21] hmm [11:21] what is the workaround for the "there was an error loading the theme Human"? [11:22] Mithrandir: just upgrade to the newest ubuntu-artwork package. This worked for me === ctd [~ctd@ctd.user] has joined #ubuntu [11:25] pitti: thanks, that helped [11:27] Mithrandir: it doesn't look very human yet, though :-( [11:27] pitti: yes [11:28] thom: I want to add lazy unmounting to pumount, but that got more difficult than I thought [11:28] oh? [11:28] thom: because if the device node is not present any more, I cannot stat /dev/whatever to find out the major device number of the device to be removed [11:28] thom: this means that I cannot check if the device is removable [11:28] pitti: caching them somewhere at mount time? [11:29] thom: and I cannot just use the name to lookup /sys/block/ because devfs may be used [11:29] keep a state file, yes. [11:29] fabbione, Mithrandir: I thought about this, but this involves many upstream changes [11:30] fabbione, Mithrandir: but I guess I cannot do without one. [11:30] pitti: you can still push the changes back to upstream ;) [11:30] pitti: on mount, echo devicename into /var/run/pmount/$username, on umount request, check if it exists in /v/r/p/$username and if not, refuse. [11:30] thom: I just wanted to say that this "little" item might last a bit longer because it might need to be discussed with mdz [11:30] Mithrandir: something along these lines, yet [11:31] pitti: no worries, i shall catch up on my email first then :-) [11:31] pitti: it sounds fairly simple and non-intrusive to me, but I haven't looked at the source code [11:31] Mithrandir: it is easy, but mdz abhors code changes for new features at this stage, so we must get this absolutely correct [11:32] pitti: ok [11:32] Mithrandir: I don't really like the fact to make pumount depend on external status information... [11:33] fabbione: yes, of course I will push the changes upstream. upstream is said to be a cooperative guy, up to now he always adopted my patches :-) [11:34] BTW, does anybody have PCMCIA drives? [11:34] I have PCMCIA smartmedia cards and such [11:34] (but not on my ubuntu system, though my gf has it on her) [11:34] Mithrandir: I just uploaded a new pcmcia-cs which does not create and remove device nodes on its own when devfs/udev is used [11:35] Mithrandir: and I need somebody to test it. tvon will do it, but he is asleep [11:35] Mithrandir: it's version 3.2.5-7ubuntu6, should appear in about 15 minutes [11:35] ok, I'll walk over to her in about that time, then. [11:36] Mithrandir: thanks! [11:36] has the upgrade-doesn't-add-the-initial-user-to-plugdev been solved yet, btw? [11:36] Mithrandir: no [11:36] Mithrandir: this is unlikely to be solved within Warty [11:37] Mithrandir: somebody proposed an external upgrade-to-warty package/script [11:37] Mithrandir: we cannot just put random users into this group [11:37] true [11:39] Mithrandir: I just got an idea about the state file: what if an USB drive is ripped off (/dev/sda), then I hotplug a real SCSI disk (/dev/sda as well)? The user should not be allowed to pumount it then. [11:39] Mithrandir: okay, this is a pathological case, but still... [11:39] pitti: if a removable drive is removed; doesn't the kernel automatically unmount it? [11:39] Kinnison: unfortunately not [11:39] pitti: you mean hotplug scsi drives aren't considered removable? ;P [11:39] pitti: you can still try to detect the device & co and use the state as last resource [11:40] pitti: can you do it by filesystem uuid? [11:40] pitti: and, won't pmount just mount it under /media/$UUID or something? [11:40] Mithrandir: well, in the current policy we only regard USB and firewire devices as removable [11:40] pitti: I was kidding. [11:40] Mithrandir: that's right, the mount point must be below /media/. That could help us. [11:41] it's not perfect, but it's a good start. [11:41] Mithrandir: I just want to avoid the case that root hotplugs a new SCSI device as /usr and the user can umount it [11:41] Mithrandir: but you're right, I still check that the mountpoint is /media/*, so this will not work [11:41] pitti: and that the mount point is mounted with user=$USER [11:41] (in mtab) [11:43] Mithrandir: right, uid=... [11:43] Mithrandir: but I don't know if this is the case with every file system [11:44] I think mount uses it to decide whether you can umount or not? [11:44] AFAIK mount looks in /etc/fstab, whether it has the 'user' option [11:45] yup, it does. [11:45] Mithrandir: but actually, isn't it enough to compare uid=xxx with getuid() in pmount? [11:45] I would think so. [11:45] Mithrandir: then we could circumvent the state file completely [11:45] thom: did you follow the discussion? [11:45] of course, if user mounts a SCSI drive over where a thumbdrive was, you lose. [11:46] (as in, the user will be able to umount the hotplug SCSI drive) [11:46] Mithrandir: right, but if the admin mounts a SCSI drive with uid=1000 (or so), he loses anyway [11:46] yeah [11:46] :) [11:46] pitti: yeah [11:46] looks reasonable from where i'm sitting [11:46] thom, Mithrandir: I will prepare an updated pmount package, but won't upload it yet. [11:46] I will upload it to my chinstrap homepage, discuss it with mdz, and when I got his blessing, upload it [11:47] in the meantime thom can use it for testing the new hal [11:47] yeah [11:47] agree? [11:47] okay [11:47] sounds good [11:48] ffs [11:48] warty will only see my ipod if its in when i boot [11:49] rburton: does it work correctly in debian? do you not get any events at all when you plug it in post boot? [11:49] just ieee1394: Node resumed: ID:BUS[0-00:1023] GUID[000a2700026ede8c] [11:49] ieee1394: Node changed: 0-00:1023 -> 0-01:1023 [11:50] sbp2 and sd_mod are loaded [11:50] i don't think i need anything else [11:50] rburton: I checked this with carlos' ipod, worked fine [11:50] rburton: does the device node appear? [11:50] nope [11:50] then we lose === rburton wonders if i need more modules [11:51] surely hotplug should be loading what is needed? [11:51] $ cat /proc/scsi/scsi [11:51] Attached devices: [11:51] silly thing [11:51] rburton: Can you ask Carlos about the modules? His' is working fine [11:54] email? [11:55] rburton: Carlos Perell Marn [11:55] ta [11:56] goody [11:57] security.no-name-yet.com is up and rocking [11:57] with base-config ubuntu15 you will get the apt lines automatically if network is available at install time [11:57] otherwise you will get them commented out [11:57] nice :) [11:58] yup [11:58] same goes for other stuff while installing from cdrom [11:58] but not universe === fabbione goes and cook some decent lunch === pitti becomes envious if he hears about decent lunch [12:05] bloody people in forwards timezones [12:05] i only just had breakfast [12:05] thom, you are lagging behind. :-) [12:07] heh [12:09] hrm, our dragon appears to have taken up climbing glass windows as a new sport [12:10] thom: ain't my fault if you live in the wrong country ;) [12:10] daniels: i was just thinking that perhaps the drivers updates should all come as ubuntu patches on top what's in the tree already. It will make it much simpler to revert if something is wrong. [12:11] fabbione: heh [12:14] jdub_: ? [12:14] thom: dude, its 11 and you've just had breakfast. get up earlier you slacker [12:15] rburton: i've been up since 8 :-) [12:16] (and if you believe that, there's a bridge over here I could maybe interest you in ;) [12:18] fabbione: what's with apache? [12:19] it's not got an ubuntu version num yet was an ubuntu upload? [12:20] elmo: nothing wrong.. it's the same as sid, i just upload it myself, without bitching you for a sync [12:20] elmo: did wireless tools get synced recently? [12:20] thom: not that I recal [12:20] +l [12:20] hrm [12:21] fabbione: mm, I thought the policy was that we should sync where possible... [12:21] thom: 1107 | wireless-tools | 26+27pre22-1 | 122 | 6427 | 2004-06-09 00:00:00 | 2 [12:22] (the date's the install/upload/sync date) [12:22] elmo: they are exactly the same package. [12:22] elmo: only s/unstable/warty to make katie happy [12:22] fabbione: yes, so what? we could just do uploads for everything and not sync anything [12:22] elmo: yeah [12:22] hrm [12:23] elmo: ok, i didn't do it with bad intention. it was just to avoid extra work for you. [12:35] fabbione: hmmm === doko [doko@dsl-082-082-140-133.arcor-ip.net] has joined #ubuntu [12:35] sweet, the only uninstallables in the archive now is amd64 oo.o [12:36] thom: crazney's just in the kitchen cooking dinner (i'm at his place stealing bandwidth) [12:36] get a real timezone [12:39] elmo: and that I'm building now. It'll need a bit of NEW love, though. === Mithrandir hugs his 100Mbit for that === reformed [~ben@ppp2C13.dsl.pacific.net.au] has joined #Ubuntu === thom wonders if spiv has gotten up yet so he can reclaim his computer [12:50] fabbione: so, do you need a new CD image then? [12:52] Kamion: yes please.. [12:52] i think daily didn't build [01:03] fabbione: it seems to have built fine [01:03] cjwatson@little:~/cdimage/www/daily$ ls 20040909 [01:03] MD5SUMS report.html warty-amd64-1.iso warty-amd64-1.list warty-i386-1.iso warty-i386-1.list warty-powerpc-1.iso warty-powerpc-1.list [01:05] thom: https://chinstrap.warthogs.hbd.com/~pitti/pmount/ [01:05] weird [01:05] 13:03 uh... "xscreensaver: 13:01:59: UNTHROTTLE ClientMessage received, but [01:05] not throttled." ? [01:05] my gf's getting that about every 30s when xscreensaver kicks in. [01:05] thom: this contains a proposed new package you can test for the hal change [01:06] thom: I will discuss that with mdz before, so please don't upload hal yet [01:06] Kamion: let me check.. sometimes i rsync and i don't notice stuff [01:06] Mithrandir: that's supressed in the latest default [01:06] thom: I tested the new pmount quite thoroughly now, so it should work [01:06] pitti: ok [01:06] Kamion: yeah .. it's there.. can you check if it has base-config ubunut15? [01:06] gar, bugzilla's on crack [01:07] Kamion: and if not.. mind to build an iso for me to test it? [01:07] pitti: so just 'pumount -l /dev/foo', right? [01:07] thom: ook [01:08] thom: yep [01:08] thom: however, this does only work as the same user who mounted the device [01:09] thom: this was not checked before [01:09] cjwatson@little:~$ grep base-config cdimage/www/daily/current/*.list [01:09] cdimage/www/daily/current/warty-amd64-1.list:/pool/main/b/base-config/base-config_2.44ubuntu14_all.deb [01:09] cdimage/www/daily/current/warty-i386-1.list:/pool/main/b/base-config/base-config_2.44ubuntu14_all.deb [01:09] pitti: uh, that's not gonna work for hal then [01:09] cdimage/www/daily/current/warty-powerpc-1.list:/pool/main/b/base-config/base-config_2.44ubuntu14_all.deb [01:09] fabbione: building a new one now [01:09] thom: I just got the same thought [01:09] since it'll be thom mounting, and hal unmounting [01:09] thom: actually this should be done by gvm [01:09] Kamion: thanks [01:09] pitti: it can't [01:09] thom: but we must check this, otherwise users can unmount root-mounted devices which are not in fstab [01:10] not with out rewriting huge amounts of hal [01:10] thom: hal should send a "ripped out" message to gvm, gvm should call pumount [01:10] fabbione: (the .list files are useful - you can do it with isoinfo -l too but it's more tedious) [01:10] thom: hmmm. Problem. [01:10] pitti: sure, but hal only sends the udi, not the device info [01:10] Kamion: yes .. i mirror only the .iso [01:11] isoinfo.. hmmm /me checks [01:11] ooh, so my ipod just got mounted with pmount [01:11] thom: gvm is capable of figuring out the device from an udi [01:11] rburton: great! What was missing? [01:11] pitti: not after the device is removed, it isn't [01:11] pitti: :) you have to unload all usb and firewire kernel modules and restart hotplug [01:11] thom: BTW, this was crap anyway. gvm should not do it, but gnome-vfs2 [01:11] fabbione: the idiom is 'isoinfo -lR -i whatever.iso' [01:12] rburton: sounds quite straightforward. [01:12] pitti: yeah, easy. carlos said he'll file a bug when he remembers with the details [01:12] pitti: shrug [01:13] thom: if I tell mdz that I modified pumount to allow hal to unmount any device, he will kill me [01:13] Kamion: i see.. i was RTFM ;) [01:13] pitti: ok, i'll look at fixing hal to expose the device that was removed [01:13] fabbione: hm, might need to do a 20040909.2 anyway, just noticed kernels on non-i386 are screwed [01:13] ok [01:13] that seems more reasonable, and more in line with how hal should work [01:13] just ping me when it's ready [01:13] (like, they don't have any) [01:13] thom: can't gnome-vfs figure out the device from the udi? [01:14] thom: right. hal should not have any policies, including (u)mounts === Mithrandir watches OOo build. Almost like watching paint dry. [01:14] thom: I think gnome-vfs is the right place to do that, and it runs as the normal user [01:14] pitti: the problem is that by the time the removed signal is emitted by hal and recived by g-v-m, the device is gone and hal knows nothing about it, so you can't resolve from udi to device [01:15] thom: by the time of the removal, isn't it still contained in hal's database? [01:15] pitti: nope [01:16] thom: this sounds strange. Just before hal purges it from its db (after getting the hotplug message) it still has the device entry [01:16] thom: can't the message be sent out in between? [01:17] thom: what does upstream do with this? [01:18] pitti: certainly this is what i'm seeing. when the hal_device_removed callback gets run, hal_device_get_property_string can't get block.device [01:18] thom: yes, the callback is certainly too late. [01:18] so we can change the callback to run earlier [01:18] thom: if that's possible? [01:18] which i will check now [01:18] thom: sth like hal_device_about_to_be_removed :-) [01:19] pitti: upstream have fixed it by umounting in hal :/ [01:19] pitti: *nod* [01:19] thom: providing a new callback sounds cleaner than just saving the device and using it later [01:19] yeah, agreed [01:20] thom: this even sounds general enough to get this upstream [01:20] hm, I forgot to unexclude openoffice.org on powerpc once it got fixed === cef waits for sounder8 to download [01:23] thom: I have to go for another round of fighting with the fiscal authorities, will return later (around 1400 UTC, I hope) [01:23] do we want to add a new one, or just change when Device Removed happens? i think the latter might be better - it means we don't change the api [01:23] k [01:24] thom: hmm. you mean just call deviceRemoved earlier? [01:24] thom: sounds as if it would make sense, too, then the clients could still query the db [01:24] thom: I don't know the exact implications of this, however [01:24] yeah [01:24] thom: so currently hal removes the db entry and calls the callback? [01:24] yep [01:24] thom: and you want to swap that? [01:24] wasn't it skanky goth chick? [01:24] er, nevermind, wrong window [01:25] thom: sounds reasonable [01:25] pitti: yeah [01:25] ok, i'll give it a go and see [01:25] thom: thanks! I will look at the interdiff when I'm back [01:25] erk, why did libfribidi0 just disappear out of Desktop? [01:25] thom: (or, it just will become a patch in debian/patches, I suppose) [01:26] thom: CU! [01:27] kamion: it's used by abiword-gnome which is in supported? === alextreme [~alex@am.xs4all.nl] has joined #ubuntu [01:32] elmo: yeah, just mailed mdz/jdub, it needs to be in Base really [01:32] (so that base-config can do right-to-left text) [01:32] whoa; after the kernel fixes and adding openoffice.org, powerpc just ballooned to 578MB [01:33] why's it so much larger than x86? [01:33] it's that damned spammy assembler language PPC uses === Riff smirks [01:33] sabdfl: three kernel variants [01:33] there's different MMU handling code for power3, power4, and everything else [01:34] nice [01:34] (there's an argument that power3 is a little pointless for us to support right now, since we probably can't boot on any such machine anyway due to only really supporting powermac-type machines ...) [01:35] power4 is what's used in G5 systems, though [01:35] hrm, I think my upstream is shit, or no-name-yet.com has a slow link [01:35] I'm betting more on my upstream.. only getting 30KB/s *sigh* [01:35] cef: dude, other side of the world [01:36] cef: I tend to get ~600KB/sec from nny.c === mako_ [mako@micha.hampshire.edu] has joined #ubuntu [01:36] I think there's also something a little bit strange with powerpc ISO generation; it's got 16/17MB extra in the .iso over what du says it should have [01:36] daniels: optus is apparently having upstream issues.. [01:37] Kamion: how does that compare with the x86 overhead? [01:37] Kinnison: that's the number compared against x86 overhead [01:37] powerpc CDs are ISO9660/HFS hybrids, which might well account for it, though [01:38] Kamion: merp; 16 megs more? [01:38] Kamion: I guess HFS hybridisation might cover that [01:38] about that [01:38] how much is the iso9660 overhead for x86? [01:38] du says 513MB, the ISO is 509MB [01:38] just got of the phone with one happy beta tester [01:38] so hard to calculate [01:39] only complaint was its failure to detect his ATI rage3d graphics card [01:39] Kamion: try du -B 2048 [01:39] for powerpc, du says 565MB while the ISO is 578MB [01:39] I've asked him to mail the list. [01:39] Kinnison: same answer [01:40] Kamion: okay; so it's not du using the wrong block size [01:40] daniels: From the Optus website: 'Due to a heavy traffic demand on the international web link, OptusNet customer's may be experiencing slow browsing speeds when attempting to access international sites after 6pm. We are aware of this problem and it is being investigated.' [01:40] Kinnison: oh, no, sorry; I had already been using -m [01:40] Kamion: is the x86 iso being built with compression? [01:40] cef: you'd be routing optus->telstra->usa->uk? [01:40] cef: ah, suck [01:40] -B 2048 says 262264 for i386, 288884 for powerpc [01:41] lifeless: tell him to mail the list with lspci and lspci -n [01:41] Kinnison: what, you mean the weird non-standard Rock Ridge extension thing? -z? [01:41] lifeless: i don't have an authoriative list for r128 [01:41] Kinnison: no, it's not [01:41] Kamion: right === mako__ [mako@micha.hampshire.edu] has joined #ubuntu [01:41] hm, actually, might have some ati devrel docs somewhere for r128 [01:41] i'll check it out when i get back home [01:41] Kamion: so the files on the x86 ISO should be taking up 537116672 bytes [01:42] Kamion: how big did you say the ISO itself was? [01:42] -rw-r--r-- 1 cjwatson cdimage 473008128 Sep 9 12:26 warty-amd64-1.iso [01:42] -rw-r--r-- 1 cjwatson cdimage 533405696 Sep 9 12:28 warty-i386-1.iso [01:42] -rw-r--r-- 1 cjwatson cdimage 605808640 Sep 9 12:30 warty-powerpc-1.iso [01:42] okay; that's really confusing === Kinnison ponders [01:43] daniels: I've emailed him [01:43] life ta [01:44] daniels: optus(melb-syd)-singtel(syd-sanjose)-level3(sanjose-newyork-london)-mnet(london-) and then the traceroute dies [01:44] ooow [01:44] well yeah, i doubt we'd allow traceroute traffic through, but auckland lives in level3 in london, so that sounds about right [01:45] yeah [01:45] 500ms pings on average.. *sigh*.. might try upping the tcp window scaling size if I can [01:45] might help [01:46] oh, I should do that server side. duh. [01:46] don't set it much higher than 2, or else broken routers REALLY bite you [01:47] hrm my side is already at 2.. hrm.. wonder if I've got a broken router upstream.. hrm [01:47] I mean the tcp_{r,w}mem stuff.. [01:48] elmo: that would be very nice, yes. [01:48] ok great, I've just got a 15-20kb speed improvementy by going back down to a scale of 0 [01:52] cef: if you want, i'll have a full warty mirror soon :) [01:52] cef: i'm over at crazney's house, rsyncing now [01:52] i'd imagine my connection isn't too great either [01:52] ooo took a *long* time (thanks Mithrandir ;) [01:53] daniels: might be faster than currently [01:54] though then again, I still need an iso to perform an install off [01:56] hence why I as asking about jigdo images === haggai_ is now known as haggai [02:06] ping? [02:07] cef: ah, fair enough [02:07] i don't think we quite had jigdo done, did we? === sabdfl [~mark@wblv-228-55.telkomadsl.co.za] has joined #ubuntu === simira [~simira@vpn-22318.vpn-s.ntnu.no] has joined #ubuntu [02:08] no.. but once we get a stable mirror of files, it's not that hard to generate jigdo files (apparently) [02:08] generating jigdo files can be easy or time-consuming [02:08] cef: well, if you want to come around and grab the i386 mirror, you're welcome to it [02:08] Steve McIntyre's JTE stuff makes it fast === fabbione kicks his ISP big times [02:10] daniels: the problem is that we need to generate a snapshot of the archive at the point when the jigdo is generated [02:10] elmo: what's the stayofexecution set to currently? [02:10] Kamion: IIRC 2 real days [02:10] Kamion: elmo upped it when thom complained [02:11] unless I'm very broken / remember incorrectly [02:11] Kamion: how hard would it be to do by hand from today's warty? [02:11] hi simira [02:11] so telling the archive to make a snapshot at the point when the jigdo has finished generating would be OK, even if the archive has changed slightly from the start [02:11] daniels: the world doesn't have access to little, and I don't have shell access to auckland [02:11] daniels: this makes it difficult [02:12] need to prepare something which makes a hardlink snapshot of the archive and get elmo to let me ssh-trigger it from little, I think [02:12] Kamion: use pdumpfs or glastree, perhaps? [02:12] or just cp -l [02:13] Mithrandir: oh, there are plenty of ways to do it, it isn't a fundamental problem [02:14] SOE is 24hours iirc [02:14] Kamion: oh right, I meant from a local machine [02:14] we're in Australia, remember ;) [02:14] daniels: oh, probably, but haven't done any of the debian-cd integration yet so I don't quite know [02:15] Kamion: righto, not mjcyh of a muchness [02:15] i'll probably just abuse uni's bandwidth for a bit :) [02:15] daniels: rsync works fairly well with CD images === Kinnison thought daniels was in ADSL wonderland now [02:16] bah amazing [02:17] i closed my adsl line like 6 months ago. [02:17] now my ISP claims that it didn't receive the router [02:17] and that they want the money or the shipping receipt [02:18] Kinnison: yeah, but i'm on a 4gb bandwidth limit, and 2 of that disappeared within 27h [02:18] how can tehy pretend a shipping receipt after 6 months... [02:19] so i can get bandwidth fine now, it just involves stealing it until nexttep get off their butts and process our plan change request (to unlimited) [02:19] daniels: merp === thom kicks hal in the nose === simira_ [~simira@vpn-22389.vpn-s.ntnu.no] has joined #ubuntu [02:21] sending the device_removed cllback after you remove the device from the store is so useless it's not funny [02:21] (stealing -> using a friend's connection, as in tonight, or going into uni) [02:21] thom: sounds sensible to me [02:21] Mithrandir: is amd64-libs yours or doko's to fix?? (#1114) [02:21] daniels: i dare say. next you'll claim that xprint is sensible, too [02:22] lamont: either is fine, he uploaded it, though [02:22] thom: watch it [02:23] syslinux | 2.10-1 | unstable | source, i386 [02:24] syslinux | 2.04-2ubuntu1 | warty | source, i386 [02:24] d'oh, I didn't realize we were that far behind [02:24] Changes in 2.09: [02:24] # Handle video pages correctly when using the API functions. [02:24] # Handle compiling on an x86-64 platform correctly. [02:25] I bet that latter one is the ubuntu1 part, yes? [02:25] guessing so, yeah [02:25] I'm wondering if newer versions fix the display problems we're seeing [02:25] I've not been hearing similar complaints about the Debian installer recently [02:27] daniels: i laugh at your blog entry about mod_c++. just use ISAPI [02:27] doesn't help if you want them to use !iis [02:28] daniels: apache2 supports ISAPI cross platform [02:28] HTHHAND === mako [mako@micha.hampshire.edu] has joined #ubuntu [02:31] daniels: I'm almost tempted to drive into the city, sit across the road from maccas with the lappy and the wireless card and leech it form there [02:31] s/form/from/ [02:33] woah, 120KB/sec atm [02:34] if only I could get the 650-700KB/sec that is pretty much the max I can attain.. ho hum *grin* [02:35] thom: what, on Unix too now? didn't know that [02:36] IIRC when I was at Zeus, Zeus was the only Unix server doing ISAPI === jame1 [~james@c211-30-13-146.thorn1.nsw.optusnet.com.au] has joined #ubuntu [02:44] Kamion: did you change something in base-config priority? [02:44] i got all the questions about TZ [02:48] fabbione: no [02:49] hmm weird [02:49] AHH hold on [02:49] Kamion: yeah, it got fixed up relatively recently in apache2 [02:49] never mind [02:51] bah [02:51] i knew there was something wrong ;) [02:53] Kamion: do we have a simple way to build cd locally? [02:56] s/simple// [02:56] would be enough :) [02:57] thom: stop clouding this discussion with facts [02:57] thom: (in all seriousness, that's way cool) [02:57] *g* [02:59] guys, has anyone seen grub failing with 'error 21' before? any idea how to fix it? [02:59] is that a boot error; or a grub-install error? [02:59] Kinnison: that's a boot error. Sorry. [03:01] well, guess that means it's time for install 2 for the night. === debianist [~pooh@80.179.66.56.forward.012.net.il] has joined #ubuntu [03:06] g'afternoon all [03:07] hi debianist [03:09] hi fabbione [03:09] you have any idea why there isn't any program to handle pdf files? [03:09] i mean with the correct assoc. without having to fire it up and choose a doc. [03:10] because nobody reported the problem before and we have forgotten to modify xpdf to do this [03:10] fabbione: not really, sorry [03:10] shall I open a bug? [03:11] fabbione: I tend to pick apart an existing ISO [03:11] Kamion: i am going with another "blind" upload.. i will need you to rebuild the cd after [03:11] ok [03:11] well.. than you need to run something to regenerate the Packages files and so on... [03:12] fabbione: usual procedure is: mount ISO, copy tree, copy in package you want, update Packages file by hand, update Packages.gz, update /dists/warty/Release, run "mkisofs -r -V 'Ubuntu 4.10 i386 Bin-1' -o warty-i386-hacked-1.iso -cache-inodes -J -l -b isolinux/isolinux.bin -c isolinux/boot.cat -no-emul-boot -boot-load-size 4 -boot-info-table new-i386" [03:12] I've got used to it :) [03:12] oh yeah [03:12] hold on :-) [03:12] let me take notes :P === debianist is taking note of this also [03:13] I've been meaning to write iso-editor to avoid having to do that [03:14] Kamion: ok.. it looks relatively simple [03:14] it's just a lot of mechanical work [03:15] why not through it all inside a perlly? [03:15] because I can't be bothered [03:16] k [03:17] debianist: you could take the challenge to do it :-) [03:17] to write a script for that? [03:17] debianist: like getting in input a .deb or a list of debs [03:18] and create a custom iso from a defined source [03:18] yeah [03:19] hmmm, it sounds rather an interesting job. I'd give it a try, what did Kamion mean with "update Package file by hand = Packages.gz" ? [03:19] btw, if any warty folks are bored, it'd be good to fix glide (see #659) so X can be up2date on amd64 [03:19] debianist: well.. if i will tell you i will spoil the fun of it :-) [03:20] is there a known bug that the times in the "list of appointments" under the clock don't match the times in the calendar in evolution? [03:20] elmo: mdz did it this morning iirc [03:20] fabbione: no, didn't work [03:20] fabbione : you are more than right! spare me those, and i shall be back with mastery :-)) [03:21] elmo: i'll take a look now === SteveA wonders where to look up such bugs [03:21] debianist: they are on the CD... just find them :) [03:21] thom: oh [03:21] thom: cool, thanks === fabbione needs to test base-config before that [03:22] thom: danke [03:22] actually, i've been little disappointed to find out warty did most of usb input setup, automagically mounts etc..where did all the fun work go? [03:26] ok guys, this time I'm getting grub failing without me having ignored any errors in the ubuntu setup. [03:38] hrm, just installing sounder8.. F8 at the boot prompt points ppl at the Installation Manual or FAQ on the Debian website [03:39] actually a lot of those pages need changing [03:41] elmo: uploaded, built for me on amd64 [03:42] fabbione : k, i'll give it a look and take tries at that. [03:42] whichever idiot invented sinuses needs a good smacking with a large sword [03:44] sinuses? [03:44] thom: neato thanks, pushing through now [03:44] like in the human body? [03:44] debianist: yes [03:44] mine hurt hugely right now [03:44] thom : i know what u'r tlking about [03:44] thom : they are buggy all the way from startup to me === Gman [~Glynn@amfea-proxy-1.sun.com] has joined #ubuntu === fabbione tests his first custom cd [03:46] first ubuntu custom cd.. even [03:49] cef: yes, can't change them until we have our own equivalents though :) [03:49] fabbione: standard mistake I make is forgetting to change Filename: in Packages [03:50] Kamion: oh.. i just removed the old entry and replaced it with a dpkg-scanpackages output [03:50] only for the packages i need [03:50] and now.. of course [03:50] .. [03:51] bdale: around? [03:51] for a second i tought it was wrong :-)) [04:02] elmo: is X depwaiting on both glide2-dev and glide3-dev? [04:02] is the usb drives not shown in (computer || other_place) and (nautilus_popup_only_access) is taken care of? [04:06] thom: just 3, it's building now [04:07] rock [04:07] hmmmm [04:07] something isn't clear to me... [04:07] why lamont told me that X was building on amd64? [04:08] and now it was depwaiting? [04:08] Kamion: i will have to upload another base-config [04:08] Kamion: testing on real cd is way different than on simulated installation [04:08] Kamion: other little bugs are coming up [04:09] is there a way I can use the ubuntu CD as a rescue CD? I have now thoroughly mangled my bootloader. [04:11] jame1: boot, run through the UI up until the start of partitioning (it hasn't written anything to the disk at that point, but does have all the necessary modules etc.) [04:11] jame1: then switch to tty2, mount any partitions you need, chroot if necessary [04:12] Kamion: will that let me see my RAID partitions? [04:12] jame1: yes it should... [04:12] I would expect so, yes [04:12] in the worst case modprobe the raid level you need [04:12] and start the raid manually [04:12] if not, md-modules-*.udeb is on the CD; install it with udpkg -i [04:13] at that point you should have it available [04:13] the installer is basically a package-managed mini-Ubuntu system that installs pieces of itself at run-time [04:15] an automatic rescue target is, I think, possible, but will require some scripting [04:18] anyone know of publically available info regarding the open secure digital card stuff? [04:18] damn, forgot to add acpi-modules to the d-i initrds [04:24] elmo: can i close the ftbfs on glide? (ie, are you happy that it works right?) [04:24] i have problems with "Applications"->"Multimedia"->"CDPlayer". just won't work OOTB, has problems with /dev/cdrom. totem works np though. [04:25] thom: yeah [04:25] cool, gone [04:29] thom: http://www.advogato.org/person/alex/diary.html?start=10. per-user apache daemons... [04:30] ok this should be it.. Kamion: base-config should be done now [04:33] rburton: heh, nice [04:33] not sure i'd want to run apache in the session, (seems uneccesary, too) [04:34] there are claims apache only takes 2m with their configuration. i hope so, as our apache2 is taking ~60M [04:34] thom: I think the idea is to not require any root privileges. [04:34] rburton: preforking with no modules loaded, 2M is about right [04:35] would you need anything other than the webdav module? [04:35] mod_zeroconf or similar, presumably === pitti [~pitti@195.227.105.180] has joined #ubuntu === pitti does not like lengthy bureaucracy [04:36] thom: he might be doing the mDNS advertisements in gnome-vfs [04:36] you would definitely need dav/davfs and userdir [04:36] thom: Hi, I'm back. Any news? [04:36] why userdir? [04:36] sounds like he is just making ~/Public the docroot [04:36] oh, although i guess with a per user daemon you'd just have the docroot as ~/Public or whatever [04:36] yar [04:37] pitti: hal sucks. still looking === jame1 [~james@c211-30-13-146.thorn1.nsw.optusnet.com.au] has joined #ubuntu [04:38] thom: agreeing to the hal statement. === StoneTable [~stone@65.169.94.254] has joined #ubuntu [04:55] I've got an error (on a red background, no less) proclaiming "Executing 'grub-install /dev/hdg1' failed. This is a fatal error." [04:55] any way I can find out what happened? [04:56] vt3 or 4 should have the full output [04:57] thom: it's saying "Error 6: Mismatched or corrupt version of stage1/stage2" [04:57] thom: is this a recoverable error or do I get to do another install? [05:01] jame1: never seen that one before, but Kamion'd be the one to ask [05:02] Kamion: you around? any ideas? [05:06] Is there any way I can get it to skip the verification of my swap partitions? After the 5th or 6th time through this, I'm pretty sure that they're ok. [05:22] jame1: um, sounds like the grub installation's broken, but if you're at that point it should be a fresh install, so I have no idea [05:23] Kamion: ok. I'm currently most of the way through the next install. [05:23] looking at the source I don't see any way to skip the swap check, sorry [05:24] given a default install, are there any "easy dialup" tools installed? [05:24] Kamion: ok. I kinda figured that'd be it. [05:24] hypatia: pppconfig? dunno about others ... [05:25] hypatia: Computer -> System Configuration -> Networking seems to have a wizard-type thing for PPP [05:25] Kamion: Unfortunately, pppconfig isn't easy enough for me, I have no way to work out which /dev device is my modem, or even if there is a driver for it loading correctly :( [05:25] Kamion: It requires wvdial, which did not get installed :( [05:25] yes, I just noticed that myself [05:25] jdub_: bug? [05:26] unfortunately that is a pain to remedy when I only find it out somewhere with no other way to access the net :) === hypatia is currently stealing time on someone else's laptop [05:27] Kamion: another data point for you: the red error screen did not appear when I told it to use EXT3. All other settings were the same. [05:27] people in Marseille don't seem as wireless-crazy as they were in London :) [05:28] jame1: oh, XFS [05:28] jame1: that warning is there for a reason :P [05:28] Kamion: no. reiser. [05:28] no warnings. [05:28] ah, dunno then [05:28] wvdial isn't even in Supported it seems [05:29] weird [05:29] I sort of just expected some "point and click" dialup love, alas === Kamion tries installing it from universe to see what happens [05:31] I hate printers [05:31] wow, wvdial depends on libdb2 [05:31] didn't even miss the dialup setup stuff [05:31] but yeah, that's missing [05:31] anyway, off to bed.. it's 1:30am here.. nite ppl [05:32] and it requires interaction to install ... [05:32] unless synaptic doesn't set DEBIAN_FRONTEND=noninteractive, which is always possible [05:36] *sigh*, is gnome-cups-manager supposed to be the way to handle printers? [05:36] Mithrandir: yes, believe so [05:36] is it only for me it locks up all the time? [05:37] and doesn't know that I have installed some new ppds, which it may use? [05:42] http://www.redhat.com/advice/tips/ [05:42] this is something interesting to implement [05:42] the samba "trash" [05:45] 'night all. === jame1 [~james@c211-30-13-146.thorn1.nsw.optusnet.com.au] has left #ubuntu [] [05:47] Kamion: huh, found an interesting germinate gotcha - if you run it separately for all arches, g++-3.4 is pulled in as a b-d mozilla only on amd64.. which is fine, it pulls in amd64's Depends for g++-3.4.. unfortunately it doesn't pull in i386/powerpc's which are diffrerent... [05:48] hm, isn't that correct? [05:48] if/when g++-3.4 is seeded on i386/powerpc, then their dependencies will be pulled in [05:49] a package is either in a component or it's not - you can't have per-architecture component-ness I don't think [05:50] ugh [05:50] uh, you could perhaps stick g++-3.4 in germinate workarounds in Supported for now, not sure what the right fix to that is [05:51] yeah, don't worry, it's a fairly rare case and britney catches it, so it's not a big deal.. [05:52] who knows maybe, per-architecture componentness is the right answer.. that would seem surprising to me as a user though if g++-3.4 was available on box but not the other === fabbione would like to socialize with elmo's girls ;) [05:52] that's already the case when things don't build, of course [05:53] fabbione: tart :P [05:53] elmo: did i tell you that security dns was updated? [05:54] fabbione: yeah, I saw you say so [05:54] ok.. [05:54] i start to melt down after 12 hours almost non-stop === fabbione should so something for his insomnia [05:55] s/so/do === thom -> TMBG gig [05:55] hal has made my brain explode [05:56] i am imploding [05:56] and it's always a bad sign when you get to the point of using FASCISTS! as a debugging string === fabbione fires up AC/DC [06:01] thom: what did I miss with glide? [06:01] I test-built it on amd64 and it did the right thing [06:01] hey mdz [06:01] mdz: control had just i386 in architectures for libglide2-dev, but you'd enabled the build of that package on any arch [06:01] mdz: Good morning! [06:01] thom: ah [06:01] morning [06:02] mdz: ahm - do you have some time for another pmount discussion? [06:03] pitti: yes [06:03] anyhow, hopefully my brain'll hurt less after They Might be Giants, so i'll hopefully fixup hal then [06:03] mdz: I tried to add the -l option for lazy unmounting, but there's a problem: [06:04] mdz: if the device node is not present any more, I cannot really check directly whether the device node belongs to an USB/FireWire device [06:04] mdz: so currently I dropped the device_removable() check from pumount and replaced it by a check if the mounted uid == getuid() [06:05] pitti: you could check that the device is mounted under /media [06:05] mdz: this adds the constraint that only the user who mounted a device can unmount it (this should be actually a good thing) [06:05] mdz: this is already done [06:05] pitti: or worse, create a pmtab so you know what was mounted by pmount and what wasn't [06:05] mdz: we already discussed the pmtab thingy, but it has its flaws, too [06:05] mdz: the current solution should not be worse because: [06:06] mdz: pmount only mounts removable devices (USB/FireWire), and the devices have uid=getuid() [06:06] mdz: pumount only umounts uid=getuid() devices, so transitively it only unmounts removable devices, too [06:06] mdz: this is not perfect, but from my POV its almost equivalent [06:07] mdz: and we have to check for uid==getuid() anyway (at least we should) [06:07] pitti: it sounds like it is probably close enough [06:07] mdz: the pmtab has an exploit: [06:07] pitti: you could add a dummy flag to the mount [06:07] to mark it as being mounted by pmount [06:07] mdz: if that's possible... [06:07] it might produce warnings [06:08] mdz: having an extra table is not a good solution, an user might pmount /dev/sda, rip it off, then the admin comes and hotplugs an /dev/sda SCSI device (fixed) to /usr [06:08] mdz: the user should not be able to umount /dev/sda then [06:08] hmm [06:08] mdz: I did not upload the current version yet, its on https://chinstrap.warthogs.hbd.com/~pitti/pmount/ [06:08] mdz: I wanted to discuss it first [06:09] mdz: if the pmount mount flag does not hurt, I can add this, too [06:09] mdz: but writing an extra pmtab has too many risks of doing sth wrong, I think [06:09] moof [06:10] lamont: hi! [06:10] pitti: I agree; I don't recommend a pmtab [06:10] lamont: may I bother you with a Debian buildd issue? [06:10] pitti: if you want to mail me the diff from the previous version, I'll look over it [06:10] sure, why not. Everyone else does... :-) [06:10] mdz: I'll do [06:11] lamont: postgresql 7.4.5-3 was built on m68k some days ago, according to the build logs. but the PTS still says its out of date on m68k and it isn't in the archive [06:11] lamont: 7.4.5-3 has critical bugfixes that need to go into sarge ASAP [06:11] m68k@buildd.debian.org - ask them to please upload [06:11] lamont: thanks, will do [06:11] or, ask elmo to remove it from the archive for m68k... :-) === lamont ducks [06:11] lamont: ahem - [06:11] never mind [06:11] elmo: are we going to be changing the archive to /ubuntu/ rather than /no-name-yet/ before release? just wondering if I should be keeping a more organized list of things to change when that happens [06:12] if it got removed (for m68k), then as soon as they uploaded, it would snap into testing, unless another upload had superseded it in sid. [06:12] lamont: no, sid also has 7.4.5-3 [06:13] mdz: http://www.piware.de/pmount-0.0.9.diff [06:14] mdz: it's quite large, but more than 50% is documentation change === lamont has a differen debian issue... postfix-tls Depends: postfix, which Conflicts: mail-transport-agent. exim is installed by default (debian, remember). apt-get install postfix-tls fails, because postfix will not be installed... What's the best way to make postfix-tls cause exim to go away, but still allow postfix to install... [06:15] kamion: not sure, I think sabdfl's happy to do that, but I need to double check [06:15] aka #270653 === mjg59 [mjg59@cavan.codon.org.uk] has joined #ubuntu [06:17] fabbione: heh! ati driver considered harmful, on debian-x [06:17] fabbione: should we start moving to using r128/radeon? i'm happy to make the discover1-data change [06:17] today I will go to town and fetch sounder 8 so that bonnie can install it tonight [06:19] daniels: easy... what/where? [06:19] daniels: changes in discover1 will be useless [06:19] fabbione: ... why? [06:19] fabbione: we use discover1-data for all our installs [06:19] daniels: X will need to be changed too [06:20] ah, you have a map? [06:20] daniels: because before there was the difference between r128 and ati.. [06:20] than it went away [06:20] i'd like to do it, but the nv/ati/wacom changes take precedence [06:20] now you want to split it again [06:20] there is code in config that handles that situation [06:20] i would like to split r128/radeon. i've always wanted to split r128/radeon. [06:21] i don't believe in the ati driver, except for the unknown card vendor-default case [06:21] and reimplementing the split is not easy if you were using the ati driver [06:21] so i rather prefer to keep it as it is for warty [06:21] sure [06:21] in any case, it [06:21] and we will split it again in hoary [06:21] 's 0220, and i was hoping to get an early night tonight. ha! [06:21] 'nacht [06:21] since we will restart from scratch we can do it [06:22] now.. i am off.. i need to do some stuff and i am freaking tired [06:22] daniels: let's just work on updating and testing the drivers [06:22] we will consider "better" features for hoary [06:23] daniels: also.. we need to talk with mdz and sabdfl to see if we can organize a X-men session [06:23] before the next conference === Kamion ponders branding the "First Stage GNU/Linux Bootstrap" string in yaboot [06:29] doko,lamont,debianist: are you subscribed to the SecurityReview wiki page? Then I don't need so send extra email updates [06:29] pitti: will be shortly if I'm not already [06:31] lamont: I just did an update and so far only mdz was CC'ed [06:31] doko, debianist: you are not yet subscribed, can you please do that? [06:33] pitti : aye aye sir [06:34] debianist: thanks [06:36] ew. [06:37] gnome-cups-manager -> print test page gives a huge "XIMIAN(R) DESKTOP" with logo and stuff. [06:38] Mithrandir: what's wrong with that? [06:38] glyph: it probably should print out a huge Ubuntu logo :-) [06:38] glyph: it should say Ubuntu [06:38] and an ubuntu logo [06:38] Mithrandir: yes! we want to waste people's color ink [06:38] Mithrandir: :-) [06:39] pitti: subscribed (to everything Warty :-) [06:40] lamont: thanks! (I confined it to *security* BTW) [06:46] pitti: was a matter of triviality to make it all warty... Used to be that before the wiki move. :-() === lamont grumbles at doko [06:47] _26_ flipping hours on hppa. [06:57] fabbione: you around? === kagou [~kagou@AMontpellier-251-1-37-160.w81-251.abo.wanadoo.fr] has joined #ubuntu [06:59] hi [07:00] kagou: hi === sabdfl [~mark@wblv-228-55.telkomadsl.co.za] has joined #ubuntu === sabdfl [~mark@wblv-228-55.telkomadsl.co.za] has joined #ubuntu [07:28] lamont: i am busy on the phone... === lulu [~lu@host217-37-231-28.in-addr.btopenworld.com] has left #ubuntu [] [07:49] who is the list moderator ?? === tvon [tvon@dsl093-119-225.blt1.dsl.speakeasy.net] has joined #ubuntu [07:51] re [07:52] kagou: Jeff Waugh, I believe === fabbione takes off for dinner after 1 hour of talk with the electrician [07:56] Kamion, you speak about jdub_ ? [07:57] yes [07:57] thanks [07:57] daniels: ping? [07:59] jdub_: so, where do we want to put the preview release ISOs, URL-wise? [07:59] jdub_: something like http://ftp.no-name-yet.com/cdimage/warty-preview/warty-i386-1.iso? [07:59] jdub_: I was also thinking of dropping the -1, it's a legacy from Debian's CD sets [08:00] jdub_: (and it confused one sounder who was thinking "are they only on Sounder CD 1?") [08:05] growisofs_mmc.cpp:static void *ioctl_handle=(void *)-1; [08:05] growisofs_mmc.cpp:#define ioctl_fd ((long)ioctl_handle) === mdz pummels whoever wrote this crap [08:10] * I leak some memory here, but I don't care... [08:31] Kamion: should today's be ok, amd64-wise? [08:39] Mithrandir: do you have the bandwidth to do amd64-libs easily? Or should I just deal with it? [08:39] I'm on a 2Mbit now, so I can do it without any big problems. [08:40] the problem was removal, right? [08:40] Mithrandir: thanks. just need a || true on each of the grep -v's in postrm, iirc. [08:41] pb is that when you remove the last line from ld.so.conf, grep exit(1)'s. [08:41] yup [08:49] mdz: can you please mail me if you took a look at the pmount changes? I would sort this out tomorrow, I still have an hour of RL things to do before I go to sleep === tvon [~tvon@h-66-167-234-52.mclnva23.dynamic.covad.net] has joined #ubuntu [08:52] tvon: hi! I uploaded a new pcmcia-cs package which should sort out this insane device creation - removal - creation - ... [08:53] tvon: can you please test them with your pcmcia device? [08:53] pitti: cool [08:53] pitti: sure [08:53] tvon: thx! [08:53] I'm at a coffee shop at th emoment [08:53] oh, but I have my camera [08:53] pitti: I'll take a look; next time, please email the diff, that way it automatically goes on my todo list === tvon snags the CF card [08:53] mdz: I can still mail it if you want [08:54] pitti: hmm [08:54] pitti: which uid does it use when doing the uid check for unmount? [08:54] if it is the invoking user, won't that break for hal lazy unmounts? [08:55] pitti: seems good in the logs [08:55] mdz: that's why we don't allow hal to pumount [08:55] mdz: first, because of the user, second because hal should stay policy free [08:55] hald keeps crapping out dammit [08:55] pitti: but hal already does this, with umount rather than pumount [08:55] pitti: how do we do lazy unmounts if not with hal? [08:56] mdz: thom and I decided that it would be the best if hald just send a message to gnome-vfs2 (gvm) and gvm does it [08:56] hmm [08:56] pitti: what happens if the user logs out of GNOME? [08:56] mdz: then it is not umounted, I guess [08:56] hrm, mayhaps I need a reboot after some updates [08:56] mdz: this is the same: a device is not mounted if the user is not logged in [08:57] volume-manager is set to browse the new stuffs, but it's not happening [08:57] mdz: after all, it's just a helper, not a final solution [08:57] can udev be restarted without things exploding? [08:57] tvon: no, the device won't be mounted automatically [08:57] tvon: yes, should be [08:57] pitti: ah [08:57] tvon: pcmcia-cs should just behave sanely now, without creating device nodes on its own [08:58] pitti: gotcha [08:58] tvon: why? does it still behave bad? [08:58] mdz: maybe gvm should unmount all pmounted devices on logout [08:59] pitti: nah, the logs look fine but I was hoping for some auto-browse magic on the CF [08:59] pitti: maybe [08:59] this gets tricky [08:59] tvon: I'm still negotiating this with mdz [08:59] pitti, what do you want to test with the last pcmcia-cs ? Because i'v an external DD SATA on pcmcia [09:00] kagou: the previous version created device nodes on its own, which badly interfered with udev. Nodes were created, removed, and created again, removed again, and so on [09:00] kagou: thus the device appeared as steadily unplugged and replugged [09:00] pitti: I don't see any problems with the patch [09:01] kagou: essentially the pcmcia packages should just behave as before, without stopping to work in general [09:01] pitti: did you see the sounder report where pmounting ext3 doesn't work? (due to uid= option) [09:01] nice pitti [09:01] kagou: so if you don't notice any difference, it's fine :-) [09:01] pitti: fixing that will probably require some adjustments to this new policy as well [09:01] mdz: no, I still have to catch up my mail [09:01] i test [09:01] mdz: if we are using grub only, should we turn off do_symlinks in /etc/kernel-img.conf? [09:01] mdz: I was busy with the security review until now [09:02] sabdfl: update-grub still uses the symlinks to provide the default and "(old)" options [09:02] no problems with my pcmcia :) [09:02] sabdfl: I use link_in_boot myself === tvon is now known as tvon-brb [09:02] kagou: fine, thanks! [09:02] your welcome [09:02] pitti: so that bug can be closed now, yes? [09:02] mdz: I think so [09:02] where's the doc on those? [09:02] mdz: I will close it now [09:02] sabdfl: kernel-img.conf(5) [09:03] ah, ok [09:03] no manual entry for that [09:03] damn i made a big report installation and i'm waiting for opprobation from jdub_ :) [09:03] mail is too big :p [09:03] sabdfl: apt-get install kernel-package [09:06] mdz: I think we should attempt the mount with option -s to fix that ext3 problem [09:06] mdz: I will reformat my USB stick and try that [09:06] pitti: oh, I did not know about that option. that sounds good [09:07] mdz: Will try that and put that into the 0.0.9 package before upload [09:07] mdz: so the mods are fine with you? [09:07] mdz: s/with/for/ [09:07] pitti: from what I saw, yes. I do not have time for a line-by-line audit [09:07] but I looked over them [09:08] mdz: of course not, I just want the new policy to be blessed by you :-) [09:08] mdz: the general idea of umounting with checking the uid= [09:08] no problem with the policy [09:08] mdz: I discussed that with Mithrandir, fabbione, andthom, they agreed as well [09:08] mdz: okay, then I'll upload this [09:09] mdz: thanks for reviewing === pitti severily needs something to eat now. CU later === tvon [~tvon@h-66-167-234-52.mclnva23.dynamic.covad.net] has joined #ubuntu === Mithrandir grumbles, geda in universe is broken. [09:20] what do you guys think of the latest panel menus? think we should move "Help" and "About Ubuntu" to the Applications menu? [09:20] that would make the menus about even in size [09:20] lamont: what's the status of buildd support for warty-security? [09:21] mdz: that's such a 2-days-ago question.. :-) [09:21] mdz: warty-security only kicks in after gold release, right? [09:21] mdz: you upload it, it'll build [09:21] lamont: ok, great [09:21] ditto for warty-updates [09:21] sabdfl: either gold release or FinalFreeze [09:21] but we'll want to test it before then [09:21] yes ok [09:21] testing is for losers [09:21] real men just Wing It(tm) [09:22] elmo: i will never forget your face when you Winged It (TM) with my mail... [09:22] sabdfl: pitti is beginning the Warty security review, and I think it makes sense to push at least some of those updates through warty-security to make sure it works [09:22] elmo: do we have a warty-amber or something? [09:22] mdz: err, nope [09:22] elmo: how should that part work? [09:23] i thought we'd want "warty" to be the same as the CD [09:23] mdz: well atm, it works more like p-u in Debian.. I didn't realise you'd want to do it amber-style [09:23] elmo: good point [09:23] amber? p-u? [09:23] elmo: how does stuff get from p-u to stable in debian? [09:23] sabdfl: p-u = proposed-updates, where things live before they go into a point release [09:23] sabdfl: I'm never going to be allowed to forget that, am I? :-) [09:23] nope :-) [09:24] >:-) [09:24] sabdfl: amber = tool used to install packages into the security.debian.org archive for an advisory [09:24] ok [09:24] elmo: I suppose soyuz should obsolete amber [09:24] so back to the panel menus. about ubuntu -> back to applications? [09:24] mdz: yeah, you'd have a web/mail interface as a replacement I guess [09:24] hopefully by the time we need a place to stage updates from random people before letting them into the security archive, we'll have enough of soyuz to oprovide that [09:25] sabdfl: I'd suggest that we have a Help menu, but I think jeff would set my house afire [09:25] oh, are we still going to do that opening a browser by default - I think that'd be cool for new users [09:25] (with a page pointing to docs, I mean) [09:25] it's looking pretty good at the mo [09:26] in fact, if we moved both help and about ubuntu over, then the menu's are dead even [09:26] elmo: local docs, or network docs? [09:27] Help is an application [09:27] ish [09:27] but About ubuntu isn't [09:27] wellll.... === sabdfl waves hands, hoping jdub doesn't notice [09:27] this application/computer distinction is so artificial anyway; lots of things don't fit. we should just put it all in one menu. yeah! we'll call it "start"! === mdz runs away [09:28] mdz: we could call it "Ubuntu", rather? [09:28] mdz: either/both.. just a general, welcome to ubuntu page with pointers to getting started docs, rather than being dumped into the plain desktop [09:28] at least, that's what I understood some people were suggesting [09:28] it'd be nice to be able to send them right to the website docs, if we can reliably guess that they can actually get there [09:29] elmo: there used to be a "Start here" icon on the desktop...:-) [09:29] what's our firefox default home page going to be? [09:29] mdz: note that when you test warty-security/warty-updates, uploads _will_ happen for them... [09:29] I think it should be something like search.ubuntulinux.org [09:29] sabdfl: to-be-determined.com is available [09:29] mdz: ping -c www.n-n-y.c && sensible-browser http://www.n-n-y.c/Warty/Start ? [09:29] would be a nice companion to no-name-yet [09:29] what about boss-never-decides-till-the-last-minute-then-changes-his-mind-again.com? [09:29] Mithrandir: do I even need to point out the failure conditions? :-) [09:30] sabdfl: a bit too long, I think. 63 chars max. [09:30] ;) [09:30] sabdfl: how about s/sabdfl/sbdfl/ where the s stands for 'schizo'? ;) [09:30] mizar:[~] echo -n boss-never-decides-till-the-last-minute-then-changes-his-mind-again |wc -c [09:30] 67 [09:30] dude, you should register sabdfl.org :) [09:30] just keep the b, just keep the b... [09:31] before our first ex-employee does ;-) [09:32] we have ubuntu-customized versions of 394 of the 923 binary packages in desktop [09:32] minus one false positive for ubuntu-artwork due to my laziness [09:32] mdz: just need to drop the '-again' then... :-) [09:33] elmo: thoughts on that ugly binNMU/powerpc thing? [09:33] BINNMU BAD [09:33] mdz: worse than that. [09:34] -6.0.1 > -6ubuntu1 [09:34] lamont: BINNMU BAD [09:34] lamont: i replied? [09:34] so we need to upload -6.1ubuntu1 === lamont checks his emails [09:34] I think we should just bite the bullet and upload 6.0.1ubuntu1 [09:34] it's hideous, but what choice do we have.. 6.1ubuntu1 is kind of worse, since debian's still at -6.. maybe [09:34] how about -6ubuntu2? [09:35] elmo: 6.0.1ubuntu1 sourceful? [09:35] mdz: gotta have a . [09:35] that looks crackful [09:35] lamont: why? [09:35] it's not a bin-nmu [09:35] because -6.0.1 > -6ubuntu* [09:35] bin-nmus are BAD [09:35] yeah, well done is done. [09:35] was a rebuild to get correct deps [09:35] mdz: dude, it's there, we can't ... [09:35] oh, we have a 6.0.1 in warty you mean? [09:35] oh, well, that's another option of course, we haven't released yet [09:35] yes [09:35] how about 6.ubuntu1 ? [09:35] we could kill the 6.0.1 out of warty :) [09:36] I wasn't in on the email conspiracy [09:36] and it's preventing 6ubuntu1 from getting there for ppc [09:36] elmo: it only breaks our ppc sounders upgrading to the preview instead of cold-installing... :=( [09:36] lamont: so did, e.g. the flush/rebuilds :P === lamont favors -6.1ubuntu1 [09:36] very true.. [09:37] and nuking -6.0.1 could be a purely elmo-activity. :-) [09:37] blaming elmo always works. ;P [09:39] Mithrandir: no. blaming elmo doesn't always work. [09:39] But it _is_ always fun. [09:40] well, true. [09:40] elmo: done [09:41] elmo: so you wanna just make it go away? [09:41] no-source-change upload is also no big deal... === lamont lets mdz/elmo decide. :-) === npmccallum [~npmccallu@69-162-252-7.ironoh.adelphia.net] has joined #ubuntu [09:42] of course, if we uploaded -6.1ubuntu1, we should do a no-source-change NMU to debian, since -6 is the latest in debian.. [09:42] maybe that's a vote for 'nuke', eh? === lamont hugs ccache [09:48] lamont: can you check why geda-gschem failed to build on amd64? [09:48] (universe) [09:49] Mithrandir: why, for the same reason it failed on i386, of course.. [09:49] which is? [09:49] +-lpango-1.0 -lgobject-2.0 -lgmodule-2.0 -ldl -lglib-2.0 -lX11 -lm [09:49] /usr/bin/ld: cannot find -lX11 [09:49] collect2: ld returned 1 exit status [09:49] missing xlibs build-dep? [09:49] looks like maybe a missing build-dep to me [09:50] whiptail x-dev xfree86-common xlibs-data xlibs-static-dev zlib1g-dev [09:50] 0 upgraded, 85 newly installed, 0 to remove and 0 not upgraded. [09:50] that's the tail end of apt-get install output [09:50] yup, since it built fine on my system here. [09:50] the real question is what happened on ppc... [09:50] it build-deps on libgtk2.0-dev, though.. [09:51] d-w libgeda-dev [09:54] ../../rpmdb/dbconfig.c:11:19: debug.h: No such file or directory [09:54] hrm.. wonder where that's supposed to come from... === fabbione yawns [10:02] elmo,lamont: whatever you guys think is best for the version numbering, as long as it gets resolved === tvon [tvon@dsl093-119-225.blt1.dsl.speakeasy.net] has joined #ubuntu [10:02] I've nuked 6.0.1 [10:02] what package is this anyway? [10:03] libgphoto2 [10:03] it's one of the two last remaining blemishes in the archive (out-of-date or broken deps) [10:03] crap [10:03] something people actually have installed [10:03] oh.. well.. we can still do a source upload if you want [10:04] it's fine with me if we're newer than a hypothetical Debian NMU [10:04] mdz: yeah, but they also have i386 and/or amd64 binaries from before the great flushing.. [10:04] mdz: so you want a source upload? will do [10:04] lamont: yeah, but then when we go to do a security update for it, they wouldn't get it [10:04] whereas that isn't a problem for the flushed binaries [10:04] yeah [10:05] yes, please do [10:05] sabdfl: rpm has issues... [10:05] mdz: momentarily [10:05] libelf is GPL'ed, but we also need libdw (dwarf), which isn't... [10:06] 2.1.4-6.1ubuntu0 [10:06] I like the sound of that version number... [10:08] blah, pbuilder doesn't have a warty.buildd [10:08] I have one for sbuild [10:08] could you send it to me? [10:08] or put it online? [10:09] Kamion: are you still around? [10:09] http://chinstrap/~lamont/warty.buildd [10:10] sabdfl: if you declared renaming time 2 days ago i would have love you more :) [10:10] argh [10:11] lamont: thanks [10:11] mdz: do apt-preferences apply to deb-src lines, or how do I do that? [10:11] lamont: no it doesn't [10:12] you have to do it manually iirc [10:12] fehg === lamont just removes the deb-src lines for sid. [10:15] libgphoto2 uploaded [10:19] mdz: should we recompile all the packages for v10 or go for v12? === SleepBoB [hrdwrbob@220-253-59-197.VIC.netspace.net.au] has joined #ubuntu [10:21] lamont: apt-get source = [10:21] fabbione: as I said in the bug, I think many packages can't use v12 because of the licensing [10:22] so we need to roll back to v10... [10:22] mdz: yeah, but I want to say apt-get source=warty [10:22] mdz: err, apt-get source =warty [10:22] lamont: you want apt-get source /warty :-) [10:23] does that work? [10:23] fabbione: that is what sarge is doing [10:23] lamont: no === lamont whaps mdz. :-) [10:23] mdz: yeah i read that... [10:23] there's a wishlist bug, and a possible patch, in Debian [10:23] ABI changes, etc. not going to happen for a while [10:23] mdz: looking at main, I'm not sure we care that much about v10 vs v12. Universe is much more populated [10:23] fabbione: so if we go for v10 we can perhaps sync packages from Debian [10:23] mdz: if the list of packages is that one in the bug i will take care of them tomorrow [10:24] fabbione: great [10:24] mdz: hmmmm i am not sure that's a good idea... [10:24] hopefully they all build and work with v10 [10:24] fabbione: you mean my list? === lamont can do the conversions today [10:24] mdz: i think it's easier to just change dependency and rebuild [10:24] lamont: yes sorry.... [10:24] fabbione: np. [10:25] mdz: you want them rebuilt with the v10 depends? [10:25] or do you want to rebuild the others with v12? [10:25] lamont: universe? no, not universe [10:25] lamont: with v10 [10:25] mdz: not universe. [10:25] lamont: see #1115 === lamont will rebuild his package list from #1115 to all b-d libmysqlclient10-dev [10:27] lamont: ok... [10:27] so i won't care about it tomorrow [10:27] hopefully i will get Xu packages [10:27] redland, mysql-dfsg,libdbd-mysql-perl [10:28] anyone know which end of australia Xu is on? [10:28] umm... mysql-dfsg... I think I'm not gonna do that one, since it delivers libmysqlclient12-dev..:-) [10:29] why my gf has to be late EVERY time [10:30] "i will be back at home max at 10pm"... [10:30] "wait for me..." [10:30] and it's 10:30 already === fabbione sighs [10:31] I asked her to be late so that you would come on IRC and talk to us instead :-) [10:31] mdz: i hate you :P [10:31] I like how volume control has 4 tabs on it, one of which is my modem [10:32] i am gonna crash in front of the TV === Mithrandir whacks geda [10:32] i start dreaming about our release plan... that's not a good symptom [10:33] good night guys [10:33] night fabbione [10:33] night === seb128_ [~seb128@ANancy-111-1-3-146.w81-250.abo.wanadoo.fr] has joined #ubuntu [10:35] elmo: what was that magic tool you mentioned which would try harder to apply patches? [10:36] mdz: iirc, he was really bitching about that tool later, since it horribly misapplied a fewl.. [10:36] well, never mind then :-) [10:36] yeah [10:36] it was late, but we feared elmo would get violent...:-) [10:38] mdz: wiggle [10:38] lamont: that was the gcc patch of doom - even patch fucked that one up [10:39] I could write a fricking thesis on how evil that patch was [10:39] s/patch/debian/ :-) [10:39] lamont: I don't get build failures automatically, can you investigate? [10:39] Mithrandir: oh. let me go check [10:39] lamont: no, as in patch(1) [10:39] lamont: else, I'll continue to prod you about random packages. ;) [10:40] Mithrandir: sigh.. What email? [10:40] tfheen@raw.no [10:40] address, that is [10:41] Mithrandir: fixed [10:41] thanks. === Mithrandir crosses fingers for gEDA [10:43] Mithrandir: did you ever get automated errors? [10:43] hrm... you want everything, or do you not want to see depwaits? [10:44] I'd prefer not to see depwaits. [10:44] I didn't get any no, just the ones you bounced to me [10:44] ok. You'll get failures that are beyond the automated handling, as well as successes, iirc. [10:44] thanks. [10:44] :) === lamont is not sure about successes.. :-0 [10:45] I don't really want successes, but if I get them, I get them. :P [10:45] they're, um, short. [10:46] ok [10:46] meh [10:49] Mithrandir: you get them. [10:49] yup, I see that [10:49] just got the libgphoto2 one [10:49] like I said, short... :-) [10:49] 25 lines [10:49] that's nice === lamont thinks that was what elmo's "meh" was about.. [10:51] no, the meh-ing is about random buildd log pimpage.. but I can't complain since it's my fault it's not fixed yet properly [10:51] (well and thom's.. can't forget to blame thom) [10:51] heh [10:51] good night guys! I'll try a sounder 8 installation and go to sleep then. [10:51] elmo: specific packages are actually my .procmailrc here. Mithrandir got added to the output side of the automated handler on all the amd64 boxes. [10:52] but it would be nice to get away from that... [11:07] Kamion: around? === ploum [~ploum@163-194.241.81.adsl.skynet.be] has joined #ubuntu === debianist [debianist@212.199.219.22.forward.012.net.il] has joined #ubuntu [11:34] mdz: just checking, is the default background now the one stating 'ubuntu debugging artwork'? [11:34] yes [11:35] kay, thanks [11:40] hmm, should it be used during bootsplash-phase too? === pitti [~pitti@195.227.105.180] has joined #ubuntu === tvon [tvon@dsl093-119-225.blt1.dsl.speakeasy.net] has joined #ubuntu [11:56] alextreme: there is no longer a bootsplash [11:56] for warty [11:56] what is being used then? [11:58] quietness [11:58] and pretty init scripts [11:58] hmm, i'll probably stick with the old one for now, would like to get this done and get back to work on the autobuilding :) [11:59] look in the ubuntu-artwork package and see if there is something in there that is intended for use as a boot splash [11:59] if there isn't now, there will be