T2 IRC Log: 2004-08-23

This is the log as captured by an IRC bot in the channel. The statements are those of the individual people and might not neccessarily reflect the policy and legal rules as set forth by the T2 SDE Project.

« prev | next »

--- Log opened Mon Aug 23 00:00:18 2004
05:37 -!- Netsplit kornbluth.freenode.net <-> irc.freenode.net quits: mnemoc_
05:38 -!- Netsplit over, joins: mnemoc_
12:31 -!- jsaw_ [~jsaw@volans.mpimf-heidelberg.mpg.de] has joined #t2
12:36 < rxr> hi jsaw_
12:39 < jsaw_> hi rxr
14:46 -!- rxr_ [~rene@p213.54.242.138.tisdip.tiscali.de] has joined #t2
14:46 -!- Topic for #t2: t2 - the upcomming rl successor ...
14:46 -!- Topic set by rxr_ [] [Thu Aug 12 02:15:41 2004]
14:46 [Users #t2]
14:46 [ jsaw_] [ mnemoc_] [ rxr] [ rxr_] [ valentin]
14:46 -!- Irssi: #t2: Total of 5 nicks [0 ops, 0 halfops, 0 voices, 5 normal]
14:46 -!- Channel #t2 created Sun Aug 8 21:15:33 2004
14:46 -!- [freenode-info] help freenode weed out clonebots, please register your IRC nick and auto-identify: http://freenode.net/faq.shtml#nicksetup
14:47 -!- Irssi: Join to #t2 was synced in 10 secs
14:55 -!- rxr [~rene@p213.54.195.188.tisdip.tiscali.de] has quit [Read error: 60 (Operation timed out)]
15:51 -!- You're now known as rxr
16:50 -!- mnemoc [~amery@200.75.27.56] has joined #t2
16:54 < rxr> hi mnemoc
17:04 -!- mnemoc_ [~amery@200.75.27.52] has quit [Read error: 110 (Connection timed out)]
18:11 < mnemoc> hi rxr
19:09 < valentin> hi
19:10 < mnemoc> hi valentin
20:19 < rxr> re
20:20 < mnemoc> re rxr
20:20 < mnemoc> jimmy did quit rock too
20:20 < mnemoc> OT: where does libao came from? (it's required by mpg321)
20:21 < rxr> ogg-vorbis
20:21 < mnemoc> thanks :)
20:21 < rxr> I heared that jimmy left rock from valentin ...
20:22 < rxr> yeah - I predicted more people leaving ...
20:22 < rxr> I consider mailing jimmy - but I guess he won't like the C++ thing of t2 I guess - and he already claimed to have choosen slackware ...
20:23 < rxr> however I find the build scripts of slackware really disgusting ...
20:23 < mnemoc> which distro has nice building scripts?
20:23 < rxr> really too simple without any real feature - just "build them damn thing and do nothing inteligent on the way" bruth force hacks ...
20:23 < rxr> former rock versions have not been too bad ...
20:24 < mnemoc> i'm very happy with 2.0
20:25 < rxr> I'm not happy how 2.0 shifts - they apply far too much stuff - and do not release
20:25 < rxr> I'm very courious when the next 2.0 release will be out - when left the 2.0-stable branch was in code freeze in order to update the linux kernels and do the next release ... - and now ...
20:26 < rxr> well - I have t2 on all production systems for update now - so I do only partially care ...
20:26 < mnemoc> clifford sent a mail about his idea.. he think he has hundrend slaves :)
20:27 < mnemoc> that's the good part of forks
20:27 < rxr> this the many rock trees thing?
20:27 < mnemoc> yes
20:27 < rxr> yeah - read it on the lurker web thing ...
20:27 < rxr> (after someone else pointed me to it ... ;-)
20:27 < mnemoc> =)
20:28 < rxr> I also wonder why clifford freaked up about the irc logs so much ...
20:28 < mnemoc> you have only one indirection ro RL-ML :)
20:28 < mnemoc> i guess that was just to troll
20:28 < rxr> they are neither that important - and if I would have liked to delete stuff - I would have choosen typo3 or my stable releases in the peoples directory ...
20:29 < rxr> why he thought I would wilful delete the holy irc logs is a puzzle for me ...
20:30 < mnemoc> i guess, in your case he would do that
20:30 < rxr> hehe - I could have even delete the svn database before pushing it to them ... *g*
20:30 < rxr> well - we'll see how this continous ...
20:30 < rxr> mnemoc: ah and btw. greetings back from susan ...
20:31 < mnemoc> oh... great :)
20:31 < mnemoc> grep apm.h /var/adm/flists/* please
20:31 < rxr> apmd I guess
20:32 < rxr> let me ssh to a x86 box - only ppc handy ;-)
20:32 < rxr> /var/adm/flists/apmd:apmd: usr/include/apm.h
20:32 < rxr> well and:
20:32 < rxr> /var/adm/flists/linux26-src:linux26-src: usr/src/linux-2.6.4-rock/include/asm-arm/apm.h
20:33 < mnemoc> battfink-acpi.c:34:17: apm.h: No such file or directory
20:33 < mnemoc> == 08/24/04 14:34:15 =[9]=> Finished building package battfink.
20:42 < rxr> do you need the apm stuff?
20:43 < mnemoc> i have acpi only on my kernel
20:43 < rxr> there should be a patch in the package that is only applied on powerpc that disables the built of it ...
20:43 < rxr> maybe I should (in t2) add to the .conf that the patch is conditionally applied when no apm.h is found ;-)
20:43 < mnemoc> yes
20:47 < mnemoc> :)
20:50 < rxr> hm - HEAD does miss the patch anyway
20:50 < rxr> and I fear I deleted it myself
20:51 < mnemoc> oh
20:58 < valentin> re
20:58 < rxr> hi valentin
20:59 < rxr> hm - why is retrieving the log so slow ...
20:59 * rxr going d' a new cappuccino ...
21:00 < rxr> wow - cliffords submaster mess left some filesize 0 files in the tree
21:01 * rxr finding for them
21:01 < valentin> rxr: i want to have a program/script for testing updates like that:
21:02 < valentin> ./bin/TestUpdate pkgname some-other-[D]-tags
21:03 < valentin> this program behaves like normal emerge except that it uses the alternate download tag and gives feedback wether this worked or not
21:03 < rxr> what do you need the some-other-D-tag thing for ?
21:04 * mnemoc ask the same
21:04 < valentin> for our smart update manager - alternatively i could mess around in the desc files, but i do not want that
21:05 < mnemoc> to update one specific [D] of a package?
21:05 < valentin> but wo altering the package description
21:05 < valentin> i could do a
21:06 < valentin> cp x.desc x.old ; sed s,old_tags,new_tags -i x.desc ..... cp x.old x.desc
21:06 < valentin> as an alternative
21:06 < valentin> just to test if posible updates work before aplining them to the desc file
21:07 < rxr> you do not yet need to apply them to the .desc file immediately
21:07 < rxr> I still do not get what your alt-D would do ...
21:07 < valentin> yes and i do not want to apply them !
21:08 < valentin> this is what i need my alt-D thing for
21:09 * valentin feeling missunderstood ,)
21:10 < rxr> yeah - please eplain in full detail - I miss you concept ...
21:10 < valentin> so i am on this like-update-patch-but-smarter thing
21:11 < valentin> at first there is a program that scanns all desc files for dl urls to find newer versions of the downloaded files
21:12 < rxr> yes, ack
21:12 < valentin> theese go into a list - btw: i propose to put those lists into the package directories themselfes, for reasons that will become clear
21:13 < valentin> now some voluntary people can run another programm (or the same whatever) to browse these list and select/reject possible updates
21:14 < valentin> the selected updates should be test-compiled in some kind of regression testing environment and that is what i need this alt-d thing for
21:15 < rxr> you want to commit them?
21:15 < rxr> so other people can run those ....
21:15 < valentin> commit what ?
21:15 < rxr> the alt-desc ...
21:16 < valentin> what exactly do you mean by alt-desc - the list of possible updates ?
21:16 < rxr> well do not make it that complicated and do nto seperate it too much
21:16 < rxr> yes - you said those alt-d's will contain the updated to be tried downloads ...
21:17 < valentin> yes - i want to store them, so i do not need to throw out packages each time i run my get-update-list programm
21:18 < valentin> but this is an proposal, of course we can go without this lists
21:18 < rxr> "need to throw out packages"? You talk in questions ...
21:18 < valentin> no
21:18 < valentin> i think you still did not get my point :(
21:19 < rxr> wanna call me *grin*
21:19 < rxr> I understood:
21:19 < valentin> so, lets start at the last point you said ack
21:19 < rxr> you test for updates and store the the result in package/*/*/*.desc-updated
21:19 < rxr> and then you want to test them later?
21:20 < valentin> i do not want to test them, i want them to be tested by t2 :)
21:20 < valentin> ok
21:20 < valentin> the reason why i need .desc-updated is
21:20 < rxr> sure that you want them to be tested automatically
21:21 < valentin> to store my decision if i want to update this package or not
21:21 < valentin> we have 1000+ packages
21:22 < valentin> now imagine how many "newer files" will be found each time you run Test4Updates
21:22 < rxr> yeah - and yet 4 devels ... :-)
21:22 < valentin> the thing is, i do not want to browse through a list of possible updates that i have rejected the week before
21:23 < rxr> we do not want to have so many lingering files - this would get conflicts when the tested runs svn up and packages got moved or updated ...
21:23 < rxr> what about this:
21:23 < rxr> test for update, run the build test as user update-tester (e.g. not root)
21:23 < valentin> as i said - we could store theese files elsewhere
21:23 < rxr> if (update build)
21:24 < rxr> mail (diff);
21:24 < rxr> else
21:24 < rxr> mail(some nice message what was done, how it failed);
21:24 < rxr> mark the test as done in tmp/updates-tested.lst
21:24 < rxr> or so ...
21:25 < valentin> the marking is what i want the .updated-desc files for
21:25 < rxr> but -desc is misleading, and in fact for svn up reasons it should not be in the package directories ...
21:26 < valentin> -desc was _your_ proposal
21:26 < rxr> you wrote alt desc in the first place ...
21:26 < valentin> where ?
21:26 < rxr> well - but when you do not want to parse some extra list
21:26 < valentin> i talked about something completely different in the begining of that discussion
21:26 < rxr> you can also store the .desc in tmp/tried-updates/$pkg.desc
21:27 < rxr> and get a code path like
21:27 < rxr> if file exists tmp/tried-update/$pkg.desc
21:27 < rxr> // use that one
21:27 < rxr> else use package/*/$pkg/$pkg.desc
21:27 < rxr> ---
21:28 < rxr> 21:02 < valentin> ./bin/TestUpdate pkgname some-other-[D]-tags
21:28 < valentin> this is code path for update build ?
21:28 < rxr> ^- sounded like that ...
21:28 < valentin> yes
21:28 < valentin> 21:02 < valentin> ./bin/TestUpdate pkgname some-other-[D]-tags
21:28 < valentin> was for
21:28 < valentin> if (update build)
21:28 < rxr> well - implement what you feel is right - I'm convinced you bing something useable into trunk :-)
21:29 < valentin> if it turns out to be unusable anybody here can revert it *g*
21:30 < rxr> I trust you coding and design abilities
21:30 < rxr> I'm sure I do not need to revert it - maybe fixing a single typo or so *g*
21:33 < valentin> thx for your trust :)
21:38 < rxr> oh my windows admins:
21:38 < rxr> My next question is: What's better to use fsfs or bdb. The server is windows 2000 box with apache. I have read the pro and cons, but I'm not able to choose one.
21:39 < rxr> they are not even able to choose s.th. according their needs when they even have a list of pros and conds :-(
21:39 < rxr> conts even
21:40 < jsaw_> re
21:41 < rxr> http://www.newscientist.com/opinion/opinterview.jsp?id=ns24611
21:41 < rxr> ^- Chernobyl engeneer interview
21:52 < rxr> hm - not that many new things in it ... no idea why /. runs it ...
21:52 < rxr> maybe they nearly never heared about "it" over there ...
21:52 < jsaw_> probably...
21:54 < rxr> the only new thing for me was: "Wow - someone standing in the reactor nearby overlived this up to today ... :-("
21:54 * rxr back to happier things ...
21:58 < jsaw_> read the comment by funkdid (Re: Unpatriotic)?
21:58 < jsaw_> http://science.slashdot.org/comments.pl?sid=119161&cid=10059286
21:59 < rxr> oh - this is uk - I thought it is us ...
22:00 < rxr> hm - I do not get discovery channel here in berlin via dvb-t ... :-(
22:12 < rxr> hm - xprint scheint das druckergebnis von Mozilla signifikant zu verbessern
22:12 < rxr> erhm - xpring seems to enhance the printing output of mozilla significantly
22:13 < rxr> just viewed some generated .jpg's from some sites and the output looked quite perfect
22:13 < rxr> not like the junk mozilla throws into lpr natively ....
22:13 < rxr> maybe one day I should try it ...
22:17 * rxr off for evening meal
22:20 < jsaw_> cu, /me off for experiments
22:36 < rxr> re
--- Log closed Tue Aug 24 00:00:32 2004