=== chx [n=chx@unaffiliated/chx] has joined #ubuntu-motu [12:12] hi. I am not sure this is the right channel, sorry if not. WHere can I find a 1.0-8xxx nvidia-glx for Breezy? [12:14] shouldn't https://launchpad.net/distros/ubuntu/+source/xine-ui/+bug/6099 use a patch to correct this instead of modifying the source? [12:14] Malone bug 6099: "xine has a poor entry in the Menu" Fix req. for: xine-ui (Ubuntu), Severity: Normal, Assigned to: MOTU Reviewers Team, Status: Fix Committed === chx [n=chx@unaffiliated/chx] has left #ubuntu-motu ["Konversation] [12:16] raphink: the xine-ui package doesn't use a patch system, so I think it is fine to not move it over to one when adding another change [12:16] oh ic [12:16] good point [12:16] first bug fix for me as a MOTU [12:17] if I want to commit this change, I guess I should get the source package for xine-ui, then apply the patch, run debuild with my key and upload the source package again [12:17] right? [12:17] and document the changes in the changelog === ealden [n=ealden@ipdial-165-176.tri-isys.com] has joined #ubuntu-motu [12:17] azeem: well the patch documents the changes already [12:17] so applying the patch will document changelog [12:18] ah, ok [12:18] :) === allee [n=ach@allee.exgal.mpe.mpg.de] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [12:29] hub: k9copy failed to build some time ago, because of the xlib issue. Since you uploaded it, maybe you could request a rebuild. [12:32] Kyral: did you get anything from Katie (if you're whitelisted) [12:35] I uploaded gtkedit two days ago but it's not in dapper-changes yet [12:35] well it's NEW so I guess it takes more time [12:39] "whitelisted"? [12:40] yah I got gtkedit_0.1-0ubuntu1_source.changes is NEW from Katie yesterday [12:42] ok good [12:42] :) === Kyral still doesn't understand "whitelisted" [12:42] whitelisted means Katie sends you message [12:42] ah lol [12:43] you are in the list of people who receive messages from Katie [12:43] yah [12:43] and then you ask, "who's katie?" ;-) [12:43] I know who Katie is ;P [12:43] or what as it may be [12:44] hehe === jinty [n=jinty@G9e5d.g.pppool.de] has joined #ubuntu-motu === xhaker [n=xhaker@213.201.220.218] has joined #ubuntu-motu [12:58] anyone could throw an eye on http://revu.tauware.de/details.py?upid=1591 and maybe upload it again? === Burgundavia [n=corey@S0106000000cc07fc.gv.shawcable.net] has joined #ubuntu-motu [12:58] yo Corey [12:58] wb Burgundavia === Burgundavia_ [n=corey@S0106000000cc07fc.gv.shawcable.net] has joined #ubuntu-motu [12:58] by.. lol [12:58] lol === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === xhaker [n=xhaker@213.201.220.218] has joined #ubuntu-motu [01:12] raphink: how do I request a rebuild? and where do I check the build state? [01:12] raphink: I haven't found anything in the wiki :-/ [01:12] you check the build state on http://people.ubuntu.com/~lamont/buildLogs/ [01:12] and you request a rebuild by pinging elmo imo [01:13] Hmm, anyone know about the status of Xen in Dapper? [01:13] no idea [01:13] Kyral, someone integrated into the old kernel for breezy, but that was out of band [01:14] Kyral, currently waiting on either developer resources or integration into the mainline kernel. Never going to happen for dapper [01:14] Yah the guy who did that graduated from Clarkson last year [01:14] we still have his devel machine in the COSI ;P [01:14] he was a google CoC student that got dropped by google but then showed up on the last day and produced his results [01:15] unknown if he got paid [01:15] Yah [01:15] Ed Despard === Burgundavia_ considers google coc to have failed for Ubuntu [01:15] ouch [01:15] I [01:15] am playing with the idea of a Yast like thing for Ubuntu (NOT Yast4Debian) [01:15] I think the Ubuntu devs are simply too busy to be able to effectively mentor people [01:16] Burgundavia_: you mean SoC? [01:16] Kyral, better work would be to work with the upstream gnome guys. See http://live.gnome.org/PreferencesRevisited [01:16] ajmitch, yes [01:16] s/coc/soc [01:16] Burgundavia_: I don't mean it limited to GNOME [01:16] I mean like a universal control panle [01:16] raphink: ok [01:16] Kyral, system tools backends in desktop neutral [01:16] one that will work just as well in Flux as it would in GNOME ;P [01:16] raphink: mail or bug? [01:16] Kyral, you might also want to look at guidance [01:17] Link? [01:17] hub: mail I suppose [01:17] http://www.simonzone.com/software/guidance/ [01:17] the thing with s-t-b is that they are written in perl, which is not a preferring language for gnome or ubuntu [01:19] Burgundavia_: its based on KDE it said [01:19] Kyral, guidance is, but it written in python and I understand it has a fairly good seperation between backend and frontend [01:20] yah..maybe I should have clearified my intent [01:20] we have a good Sysadmin controls right now [01:20] but like extras [01:20] like setting up and monitoring HTTP servers, Xen, etc [01:21] then I would look at system-config stuff [01:21] the problem with all these preferences tools is that they are all written in different languages and for different projects without sufficent backend/frontend seperation [01:21] Like I was playing with SuSE in a VM yesterday [01:22] so there is no single set of tools to do the job [01:22] and I was quite impressed with all the Server options there are there === macgyver2 [n=eric@pdpc/supporter/student/macgyver2] has joined #ubuntu-motu [01:22] like setting up a LAMP only took like 3 minutes === freeflying_ [n=freeflyi@61.190.65.8] has joined #ubuntu-motu [01:26] raphink: mailed elmo [01:26] I dunno [01:26] hub: cool :) [01:26] its just a project idea right now [01:26] but it would be nice... [01:26] Written in PyGTK of course [01:27] why off course? [01:27] okay... [01:27] PyQt? [01:27] Py? [01:27] Python [01:28] yeah [01:28] yah? [01:29] Kyral, The KDE project's policy of leaving operating system specific software such as administration and configuration tools up to distributors to handle has not worked. The distributions have failed to produce a good working set of tools. Instead we've seen proprietry tools, non-KDE tools and often just plain poorly designed and implemented tools. <-- substitute GNOME in there and you see what I feel [01:29] Burgundavia_: thats what I mean [01:29] I mean this tool would rely on Apt... [01:30] but If I do it right it wouldn't be hard to swap in Yum for Apt [01:30] Kyral, make it as desktop agnostic as possible [01:30] "desktop adnostic"? [01:31] seperate out the backend and frontend === Kyral blinks [01:31] and make the backend package management agnostic [01:31] much like Ubuntu express [01:31] Which I still need to play with lol [01:32] not ready yet [01:32] post dev sprint at earliest [01:32] Yah I mean during install I could have it look at the system to determine which type its running on [01:32] I'd prolly write a module to handle it so the rest of the program doesn't care [01:34] Burgundavia_: proprietary tools? I know only one distro that do that in KDE [01:34] Burgundavia_: the rest is Free Software [01:34] Like if Yum is detected, load the Yum tools into the program, if Apt-Get etc etc [01:38] This would be a big project lol [01:40] If I pulled it off right...would it be big? === macgyver2 [n=eric@pdpc/supporter/student/macgyver2] has left #ubuntu-motu ["Leaving"] === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [02:05] hub, the fragmentation is as bad as it being non-free [02:06] GRCC [02:06] Burgundavia_: you are preaching the choir [02:06] Grand Unified Control Center :D [02:07] hub, yes === Kyral starts work on the project... [02:07] by learning Python lol [02:07] Burgundavia_: for the proprietary thing, my employer does it [02:07] Burgundavia_: and at one point that annoys me === teroedni [n=teroedni@ti411310a080-3401.bb.online.no] has joined #ubuntu-motu [02:10] hub, indeed, so does mine [02:10] hub: isn't your employer Xandros? [02:10] lol [02:11] raphink: yep. and they write network configuration tools for their desktop env which is KDE [02:11] and that are proprietary [02:13] a mine writes tools in GTK for setting up multiseat sytesms [02:14] who do you work for Burgundavia_ ? [02:14] Userful [02:14] ... [02:15] raphink: yep [02:15] anyway [02:15] time to go to bed really [02:15] I was just waiting for a build and it's done now :) [02:16] so later guys ;) [02:17] "hardware comes found and shaped in automatic rifle" <-- I love google translate [02:17] lol [02:18] ciao [02:18] it is translated italian talking about automatic hardware detection in Ubuntu/Ufficio Zero === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === psusi [n=phreak@54.161.205.68.cfl.res.rr.com] has joined #ubuntu-motu === punkrockguy318 [n=lukas@pcp02403042pcs.brdgtn01.nj.comcast.net] has joined #ubuntu-motu === Kyral wonders how hard it would be to get Xen in Dapper... [02:37] Kyral: need kernel support [02:38] hub: Not afraid to patch my own kernel ;P === mhz [n=mhz_chil@pc-252-84-215-201.cm.vtr.net] has joined #ubuntu-motu [02:41] Should be interesting if I can get it working :D === marcin` [n=user@194.114.146.58] has joined #ubuntu-motu === _maydayjay_ [n=maydayja@ip101109.101.nas.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [03:01] tseng:ping [03:01] oh Lathiat you pinged me a couple days ago? [03:02] ~> [03:02] i did ? [03:02] if i did i dont know what about :) === psusi wishes he could ping someone and get an answer back so far in the future that it comes before the ping === slomo_ [n=slomo@ubuntu/member/slomo] has joined #ubuntu-motu === thierry_ [n=thierry@modemcable051.65-131-66.mc.videotron.ca] has joined #ubuntu-motu [03:07] any MOTU that could review my package? [03:07] slomo , slomo_ : JohnnyMast told me you could review my package [03:08] url is http://revu.tauware.de/details.py?upid=1589 === ealden [n=ealden@ipdial-167-47.tri-isys.com] has joined #ubuntu-motu [03:14] Whee! Xen kernel compiling === _maydayjay_ [n=maydayja@ip101109.101.nas.net] has joined #ubuntu-motu === crimsun [n=crimsun@h-67-101-244-6.mclnva23.covad.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [03:38] wb LJ === bddebian [n=bddebian@71.224.172.103] has joined #ubuntu-motu [03:39] Heya gang [03:40] hi [03:40] hey bddebian [03:40] Whassup gents? === ajmitch is listening to debian vs ubuntu panel talk/debate at LCA ;) [03:40] Heh === Kyral is compiling the Xen Kernel === crimsun is setting his alarm clock === psusi is trying to figure out WTF hal-set-property is being a jackass when called from a hal callout script and saying "error: libhal_ctx_init: (null): (null) [03:41] mmm...Xen === bddebian is still being a LOSER [03:42] lol [03:42] I will be quite pleased with myself if I can get Xen running on Dapper [03:44] even moreso if I can get Debian running on Domain 1 === LaserJock [n=mantha@ubuntu/member/laserjock] has joined #ubuntu-motu [04:02] what component pops up the force quit dialog box? === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === bddebian [n=bddebian@71.224.172.103] has joined #ubuntu-motu [04:22] bddebian: hi! [04:22] Heya LaserJock [04:23] so does UVF mean no new packages can go in? [04:24] no, it means no new upstream versions of existing packages in the repo [04:24] ok [04:24] new packages can go in up til FF [04:25] so mod_dnssd could still go in... === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === Kyral [n=kyral@ubuntu/member/kyral] has joined #ubuntu-motu === freeflying [n=freeflyi@61.190.65.8] has joined #ubuntu-motu === SEJeff [n=SEJeff@12-203-76-59.client.insightBB.com] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === dcode [n=dcode@dpc671431125.direcpc.com] has joined #ubuntu-motu === Yagisan [n=jamie@60-240-12-18-nsw-pppoe.tpgi.com.au] has joined #ubuntu-motu [04:58] G'day all [04:59] dumb question. As there is no Breaks: field in debian/control can I emulate it with a conflicts ? [05:01] a question better for #ubuntu-devel or #debian-devel, I am afraid === Kyral [n=kyral@ubuntu/member/kyral] has joined #ubuntu-motu [05:03] Got it working... [05:04] Yagisan: Conflicts [05:04] Kyral: Contrats [05:04] too bad that Xen requires the 2.6.12 kernel [05:06] ok - thanks guys. [05:06] how does apt work out which package is more up to date from two different repositories? [05:06] is it just the version number? [05:07] the version number is taken primarily, but pinning has an effect, too [05:07] mr-russ: TTBOMK yes, the version number - unless you do apt-pinning [05:08] hmm.. apt-cache show, shows the wrong version, but apt-get installs the updated one ? === mr-russ will have to do a little more work. [05:08] so much to learn so fast. [05:08] use apt-cache policy [05:08] then you'll see which ones are available to apt === crimsun heads to bed, 'night all [05:09] Later crimsun [05:09] night bddebian :) === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [05:11] can somebody check their grub menu.lst for me and tell me something? === psusi [n=phreak@54.161.205.68.cfl.res.rr.com] has joined #ubuntu-motu [05:12] wrong channel [05:12] sorry] [05:12] :-( [05:12] so.... anyone know what happened to the perty new logout/shutdown dialog that was in dapper? it went away.. [05:16] psusi, unfortunately, I think it was something upstream === Kyral [n=kyral@ubuntu/member/kyral] has joined #ubuntu-motu [05:17] hrm... I want the other one back ;) [05:17] psusi: ditto :) [05:20] I like the current one better [05:21] the one with the timeout and the silly switch, or the one with the nice icons? [05:21] the icons on the previous one just make me wince [05:21] really? I thought they were perty [05:21] I like the one with "silly switch" [05:21] although I can't say I like the switch icon much [05:22] it felt more intuitive... was quicker/easier to shutdown/suspend/hibernate/reboot [05:23] I like the design, I just hate the icons [05:26] ajmitch: did your zope syncs go through? [05:26] minghua, well theming is fairly trivial to change normally [05:27] SEJeff: I am not against the old dialog, I am jest expressing my personal preference [05:27] SEJeff: although if dapper releases with those icons I probably will try to figure out how to change the theme === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === dcode [n=dcode@dpc671431125.direcpc.com] has left #ubuntu-motu ["Leaving"] [05:29] minghua, to each his own [05:30] minghua, I agree with you that the icons don't match the current style of the rest of the desktop [05:30] SEJeff: yes indeed [05:30] Burgundavia_: glad to see I am not alone here :-) [05:38] I suppose I like the layout more than the icons themselves [05:39] and yea, I can see how they don't mesh quite so well with the rest of the desktop... but they still look nice ;) === freeflying_ [n=freeflyi@61.190.65.8] has joined #ubuntu-motu [05:41] I'll take anything that lets me log out. ;-) === psusi wishes he could find decent info on reiser4... the mailing list is nothing but spam === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [05:47] psusi: why reiser4 ? [05:47] because I really like what I read on their web page [05:47] both about the on disk filesystem design, and the semantics... everything is a file [05:48] wondering logs.... all very interesting stuff [05:49] <\sh> if reiserfs 4 is at least so slow like reiserfs3 then forget about it...xfs is much faster and handling of the journals is much nicer... [05:49] plus... I just might be able to get it to organize files needed for boot so they are sequentially stored at the start of the disk so readahead-list can suck them up fast ;) [05:49] based on the reiser 4 page, reiser 3 is shit in comparison [05:49] \me prefers jfs === Yagisan prefeers jfs === psusi has only ever used ext2/3 and reiser3 === Yagisan can't type to save his life today [05:50] \sh: I think ReiserFS4 is quite a bit faster than ReiserFS3 but the fs choice is like gnome vs. kde, or emacs vs. vim [05:50] I saw benchmarks that showed reserfs4 was a regression in speed - let me see if I can find it [05:50] not really... speed is not subjective ;) [05:50] <\sh> LaserJock: I have reiserfs3 on partition and xfs on another (same disk connected via usb) [05:51] <\sh> LaserJock: mounting the reiser parition (80GB) let me wait at least 4-5 seconds, while mounting xfs partion let me wait less then 1 second [05:51] <\sh> on both partitions I have more then 10000 files [05:52] <\sh> partition even (I hate monday mornings) [05:52] \sh, that's because the current kernel reiser3 code reads in the whole free block bitmap on mount... there's a patch pending on the lkml right now to fix that so it only loads it on demand [05:52] psusi: speed is subjective to some degree because people want speed in different areas and it depends on the benchmark === hub is conservative === hub use ext3 only [05:52] LaserJock, yea... but when it's faster on all the benchmarks, it's hard to argue ;) [05:53] psusi: but it's not [05:53] psusi: i've seen other benchmarks that say it's not [05:53] imho, the design of ext2/3 isn't much better than fat [05:53] psusi: for me resier4 is like attaching a boat anchor to a car [05:53] <\sh> psusi: right..but another (in my eyes) advantage is, that I don't need to handle fscks anymore with xfs :) it just works :) [05:53] I figure there is a good that reiser4 is not in the mainline kernel and why it is a not the default fs [05:53] \sh, you don't fsck with ext3 or reiser3 either [05:54] my point is that I can (and have in the past) found benchmarks/data that convince me that Reiser4, ext3, or xfs are the best [05:54] Burgundavia_, the reason is poletics.... Hanz Reiser seems to not get along very well with the other kernel people [05:54] but I'm not really that into it so I just use ext3 [05:54] <\sh> psusi: i have to manually fix sometimes the reiser3 partition when I just turn off the computer because of some foo with crashing kernels :) [05:55] <\sh> psusi: while it was reading or writing something on this partition... [05:55] \sh, you shouldn't have to... my machine crashes all the time and I never fsck ;) [05:55] <\sh> psusi: well, if I don't do, I see many read errors and other funny messages from reiserfs [05:55] strange.... [05:55] <\sh> psusi: after fscking it everything is back to normal [05:55] again, I've seen where people have had lots of data corruption with reiser but not ext3 and vice versa [05:55] http://linuxgazette.net/102/piszcz.html reiser4 is pretty crap according to these benchmarks [05:55] that shouldn't happen since it's journaled [05:56] oh no [05:56] psusi, thats bs. We used bonnie++ to stress test reiserfs (on SLES9) and ext3. We consistently saw problems with the reiser volumes and nothing much more than a hiccup with the ext3 one [05:56] <\sh> psusi: but the best thing happend to an old customer of mine...he had everything running on reiserfs (it was a suse distro) and suddenly his important db drive was crashing [05:56] today's dapper updates broken xchat's url opening ability [05:56] SEJeff: thanks - that's the link I was looking for [05:57] SEJeff: Reiser 3 or 4? [05:57] <\sh> psusi: he tried to get the data from this drive (because he was a hard man and didn't have any backups) via ontrack rescue services [05:57] Yagisan, I found that very interesting when it first came out. I even used it + the results of our internal benchmarking to see that reiser is pretty much crap for our environment [05:57] <\sh> psusi: ontrack told him, they don't know how to rescue reiserfs data, because it's not a standard :) [05:57] SLES is reiser3 [05:57] SEJeff, it _shouldn't_ have problems because it is journaled... also, I've read that bonnie++ is a completely unrealistic benchmark [05:57] <\sh> psusi: that was 2002 [05:58] SEJeff: I think Reiser4 fares much better in most benchmarks than Reiser3 [05:58] SEJeff: yeah, the only system I use that has reiser(3) on it is my www server, as it was the only system I benchmarked that performed better with it [05:58] \sh, supposedly that long ago it did have a few bugs that ended up getting fixed... === psusi shrugs [05:58] LaserJock, actually I think that might be the wrong link. I saw one earlier that had reiser4 and showed reiser4 was horrible [05:58] btw, "it's not standard" is a cop out bullshit excuse of an answer === SEJeff looks around more [05:59] SEJeff: it was a tossup between XFS or JFS for everything else, and I chose JFS as it was just a bit quicker [05:59] psusi, bonnie++ is made to stress the harddrive and nothing else. That is an EXCELLENT benchmark for a database server [05:59] it's like ISPs who tell you they don't support Linux and refuse to do anything else when you call and tell them their DNS servers are fucked up [05:59] heh [05:59] Yagisan, yes [05:59] http://linuxgazette.net/122/TWDT.html#piszcz [05:59] ^^^ right link [05:59] Yagisan, yes, that is the correct one. I just didn't look at the date === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [06:00] I've seen benchmarks that showed the ext3 (or xfs) was best and some that say Resier4 is better but I don't think I've seen any that said Reiser3 was best [06:00] SEJeff, I believe I read somewhere that the usage patterns it tests aren't even very close to how a normal db behaves [06:00] psusi, Well I was in charge of running tests. And this included oracle stress testing. Reiser sucks, I am saying so from my personal experience [06:01] SEJeff, 3 or 4? [06:01] psusi, 3. reiser4 is not in SLES [06:01] and reiser4 is not stable [06:01] ++ [06:01] btw... I don't trust anything with the word "oracle" in it any further than I can throw it.... I got to mess with oracle for a while in a rdbms class in college and man it was a pile of shit [06:02] psusi: the best filesystem for you, depends on what you do with it [06:02] psusi, take a look at the link Yagisan gave and notice that the benchmarks show reiser4 getting beat by ext3 in the large majority of them [06:02] had such utterly retarded bugs in it... like if you fed it a text file to import into a table that did not end with a newline, it crashed.... pffft === cyberix [n=cyberix@hoas-fe36dd00-251.dhcp.inet.fi] has joined #ubuntu-motu [06:02] SEJeff, aye... for some things reiser3 can be slower than ext3... reiser4 apparently is a lot different [06:03] psusi, well weblogic and oracle are our standard at work. I dont have a say in that I'm just a lowly admin [06:03] psusi, look at that link and you will eat your words :) reiser4 is hype from Hans Reiser [06:03] SEJeff: I agree that Reiser3 is probably not going win any contests but I think that Reiser4 will be really good. I personally use ext3 since it is standard and I don't do anything where it would really matter [06:04] SEJeff, the design sounds very promising [06:05] psusi: yes, but the implementation isn't as good as the hype - can you say "longhorn" - because that's what it reminds me of [06:05] psusi, I'm *really* not trying to flame you... the design of minix is very promising. A microkernel is far superior to a monolithic one in theory. In implementation, linux wins and it is not microkernel based [06:05] certainly much better than ext2/3... which don't even use extents for allocations... still uses direct and indirect block pointers... that's nearly as bad as fat === SEJeff agrees with Yagisan again [06:05] microkernels never were superior in theory for performance ;) [06:05] I've heard that a properly tuned ext3 is the best FS but I haven't bothered to try it [06:05] psusi, There is 1 and ONLY 1 case that reiser is better than ext* [06:05] psusi: if it is a win for you, go for it [06:05] LaserJock, with dir_index enabled, it is very fast [06:06] psusi: but make your own benchmarks first, based on your usuage [06:06] reiser is very nice in that you can resize it online as in without unmounting it. If you use lvm, that will be great for you [06:06] Yagisan, aye.... [06:07] Actually I am again wrong. reiser allocates free space much better than any other filesystem out there so you can squeeze more out of your storage if you have many small files [06:07] SEJeff: which is why it is on my www server [06:08] very good for Maildirs ;) [06:09] anyway, that's why I say it is like gnome vs. kde or emacs vs. vim . You really have to find out which is best for you [06:09] agreed [06:10] in any case, ext3 is the worst ;) [06:10] not necessarily [06:10] in my experience ext3 and jfs both work well for the filesystem pbuilder does it's work (I actually have a seperate partition for that) [06:10] wow [06:10] I tried xfs but it was a but slower in unpacking the tarballs on a multiple pbuilder tun [06:11] s/but/bit [06:11] for that kind of thing I would use ext2 instead of ext3... don't really need a journal just for pbuilder and it slows things down [06:12] btw Yagisan, you ever play with that defrag package? ;) [06:12] psusi, well if you ever want to try out ext3, enable the dir_index option and it is much faster. tune2fs -O dir_index /dev/hdax === Gloubiboulga [n=gauvain@84.5.53.182] has joined #ubuntu-motu [06:12] psusi: yes - it doesn't work on jfs ;) [06:13] SEJeff, much faster if you are creating 1000+ files in a directory... not otherwise, right? [06:13] Yagisan, on your pbuilder partition silly ;) [06:13] psusi: my pbuilder partition is on a raid5 array, a journal won't slow it down much [06:13] psusi: I ended up with jfs for pbuilder [06:13] ahh [06:13] thought you said it was ext3 [06:14] and damnit this guy's graphs are retardedly unreadable [06:14] psusi, Well it makes it faster to access dirs with lots of files. think /usr/* and /var/* [06:14] psusi: I benched both ext3, jfs, and xfs for it. jfs won [06:14] s/both/each [06:14] SEJeff, I don't think anything in either of those has more than 100 or so? [06:15] psusi, I actually enabled dir_index today on my parents computer. I did hdparm -T before and after. Before 107.9MB/s after 174.8MB/s [06:16] psusi, I read it on ubuntuforums a few days ago [06:16] SEJeff, Ummm... hdparm doesn't go through the filesystem [06:16] it accesses the block device directly [06:16] psusi, well why would there be that big of a difference before and after? [06:17] caching is my guess... turn it back off and see if it goes down [06:17] SEJeff: -T is cache [06:17] thought so... [06:17] SEJeff: -t is disk [06:17] use -t [06:18] SEJeff: my system -T -> Timing cached reads: 1676 MB in 2.00 seconds = 836.45 MB/sec [06:18] that was on my parents old old system [06:18] SEJeff: and -t -> Timing buffered disk reads: 174 MB in 3.02 seconds = 57.61 MB/sec [06:18] ok thanks. [06:19] Yagisan: far faster than mine [06:19] I modified dd to use aio with the O_DIRECT option... gets like 20-30% better throughput with less than half the cpu of hdparm -t [06:20] btw, has anybody got some good resources on using the emacs modes for debian stuff? I'm not very familiar with emacs so I'm not sure how to use the stuff in devscript-el and debian-el for instance :( [06:20] ajmitch: I'm not satisfied I need more speed [06:20] Yagisan: I could probably get more speed by using the ide controller card on the shelf [06:20] ATA133 [06:20] ajmitch: that is 1x ATA 100 + 2x SATA 150 === _maydayjay_ [n=maydayja@ip101109.101.nas.net] has joined #ubuntu-motu [06:21] ajmitch: I have 2 x ATA133 ports empty, but need dappers kernel to use them [06:21] hdparm -t only gets 63 MB/s on my raid-0.... my modified dd gets 98 [06:21] <\sh> hey ajmitch [06:22] hey \sh, what's up? [06:22] 2x 10,000 rpm sata [06:22] <\sh> ajmitch: how's life [06:22] it's alright, LCA has started [06:22] psusi: 3 x 7200rpm seagates here [06:22] been at the debian miniconf today :) [06:23] met up wiht lathiat === Yagisan wished he could go [06:23] I've heard bad things about seagate's ide drives... I've only ever used their scsi products [06:23] I might not have much packaging time this week, and the network is a bit broken there still [06:23] <\sh> ajmitch: jdub should be attending as well, right? [06:23] \sh: yes, he was on the panel of debian & ubuntu discussion :) [06:24] and mjg59 is there, he took a look at my laptop :) [06:24] had one blow up... first generation 10,000 rpm drive... was a refurb though... the original one is still running today [06:24] <\sh> ajmitch: oh...no fights? [06:24] <\sh> ajmitch: no boxing no bloody noses? [06:24] almost [06:24] some strong words, anyway [06:24] really? that's sad [06:24] <\sh> .oO(hope nobody mentioned my name) [06:24] \sh: they did :) [06:24] I think I got that thing in 2000? damn... [06:24] <\sh> WOOT? [06:25] no... must have been 2001 [06:25] <\sh> wtf? [06:25] LaserJock: the strong words were more about time-based releases vs 'when its ready' === jsgotangco [n=jerome@ubuntu/member/jsgotangco] has joined #ubuntu-motu [06:25] psusi: I've always had good experience with seagate drives myself, but always bought mine new [06:25] ajmitch: oh, well that's not too bad then ;-) [06:25] \sh: your name came up when we were talking about motus & pushing stuff back to debian - it's understood that MOTUs will have differences of opinions, and can just be overworked [06:26] I had a bit of a chat with womble [06:26] <\sh> ajmitch: you are joking, right? [06:26] \sh: no, not really [06:26] <\sh> ajmitch: oh damn [06:26] Yagisan, yea.... I think I got ripped off when I got the second one... didn't say it was a refurb, but it turned out it was... that's the one that finally died... the original one is still going [06:27] blogging makes you famous [06:27] actually, it didn't even die... I stopped using it when my friends asked me where the jet engine was hiding [06:27] <\sh> I'm damned and doomed and should go away...now they are talking about me in a country far far away ;) [06:27] haha [06:27] then I realized it was just making so much damn noise because the bearings were shot... [06:27] \sh: don't leave us! [06:27] but it grew gradually and I slept with it next to me every night so I had been used to it [06:28] those first gen 10,000 rpms ran HOT though... oh boy [06:28] psusi: odd - I live seagates because they are quiet compared to other brands (longer warranty too, was 5 years when I checked last) [06:28] <\sh> ajmitch: please tell everyboody I'm just a fake [06:28] I think they drew like 20 watts each idle.... and you could cook an egg on them [06:28] s/live/like [06:28] <\sh> ajmitch: jdubs german alter ego or something like this...I'm not real. [06:28] Yagisan, yea... they have really high MTBF [06:29] their scsi ones anyhow [06:29] heard their ide lines aren't so reliable [06:29] \sh: lol [06:29] psusi: well, i've been psuuing mine almost 24/7 since I bought them - 2+ years ago [06:30] s/psuuing/pushing [06:31] yea.... the one first gen 10k scsi drive is still going in a machine at my dad's I think... been on nearly 24/7 since 2001 or so when I got it [06:32] also have a first gen 15,000 rpm cheetah still going at work... got that in 2003 [06:32] wait... is that right? [06:32] no... got that one in 2002 [06:32] feb 2002 [06:32] first batch that hit the market, I was there... had been waiting for 4 months trying to order them but they took forever to come out with production quantities [06:43] psusi: think they will actually get around to writing a repacker for resier4 ? I'm still waiting for a reiser3 repacker ? [06:43] bah - I can't type [06:43] beats me.... [06:44] all I know is I REALLY think the pseudo file stuff is cool === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === netjoined: irc.freenode.net -> niven.freenode.net === siretart [i=siretart@ubuntu/member/siretart] has joined #ubuntu-motu === irvin [n=irvin@ubuntu/member/irvin] has joined #ubuntu-motu === freeflying_ [n=freeflyi@61.190.65.8] has joined #ubuntu-motu === jinty [n=jinty@Gb5c5.g.pppool.de] has joined #ubuntu-motu === viviersf [n=cain@196.44.1.98] has joined #ubuntu-motu === robitaille [n=robitail@d154-5-117-228.bchsia.telus.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === lbm [n=lbm@cpe.atm4-0-1301006.0x50a0824e.vgnxx6.customer.tele.dk] has joined #ubuntu-motu === zakame [n=zak@210.213.77.164] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === irvin [n=irvin@ubuntu/member/irvin] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu === minghua [n=minghua@69-153-139-23.dsl.hstntx.swbell.net] has joined #ubuntu-motu === lucas [n=lucas@ubuntu/member/lucas] has joined #ubuntu-motu [08:59] morning [09:00] heya siretart [09:01] hi all [09:02] <\sh> good morning [09:03] hey lucas === pappan [n=garbage@bgepxyout-02.asiapac.hp.net] has joined #ubuntu-motu [09:11] for a non pooled mirror, does a .dsc file belong in the binary folder, or the sources folder? [09:11] sources [09:11] how does the .deb know it's signed when you install it? [09:12] from the Releases signing? [09:12] good question [09:12] okay, I'll look further into it. Quite a new experience setting up my own apt mirror. [09:13] http://wiki.debian.org/SecureApt [09:13] answer is: releases file [09:13] generally the .debs are not signed as far as I know [09:14]