IRC logs for #baserock for Thursday, 2015-07-02

richard_mawI knocked together a prototype tempdir daemon in python:
richard_mawso, even if I get nowhere with live atomic update since the last FOSDEM, I can always do a lightning talk on why we need FD based tempdir management00:35
*** inara has quit IRC04:55
*** inara has joined #baserock05:01
*** zoli__ has joined #baserock06:00
paulsherwoodrichard_maw: c00l! :)06:01
*** zoli__ has quit IRC06:35
*** zoli__ has joined #baserock06:59
*** paulw has joined #baserock07:07
*** franred has joined #baserock07:27
*** tiagogomes has quit IRC07:35
*** tiagogomes has joined #baserock07:47
*** zoli__ has quit IRC07:49
*** radiofree has quit IRC07:51
*** radiofree has joined #baserock07:53
*** mariaderidder has joined #baserock08:02
*** bashrc has joined #baserock08:02
*** sambishop has joined #baserock08:21
*** gary_perkins has joined #baserock08:23
*** zoli__ has joined #baserock08:26
paulsherwoodfolks - would someone here be willing/able cause a log of #automotive to happen using baserock infrastructure?08:26
*** ssam2 has joined #baserock08:28
*** ChanServ sets mode: +v ssam208:28
pedroalvarezI can be persuaded to08:32
paulsherwoodcan we get any +1s, or any dissenters?08:32
pedroalvarezgiven that all the logs of #baserock are now around 30M, I really don't mind08:33
pedroalvarezis it related to the baserock project? may08:34
pedroalvarezmaybe not08:34
pedroalvarez(this is not a -1)08:34
paulsherwoodwell, we support a genivi baseline, and #automotive is discussing automotive software including genivi. i would say there's some overlap, but the discussion will mainly be non-baserock.08:37
paulsherwoodon the other hand i see that more than a handful of participants here are also participants there :)08:37
persiaI'd generally prefer if we could eventually move logs to more closely related projects, but it seems a reasonable stop-gap measure.08:38
*** jonathanmaw has joined #baserock08:53
radiofreeqt 5.5 released, time to upgrade it in baserock (and ditch gstreamer 0.10 forever!)08:55
*** brlogger has joined #baserock08:57
*** brlogger has joined #baserock08:59
*** zoli__ has joined #baserock09:03
*** zoli___ has quit IRC09:04
paulsherwoodpedroalvarez: you're a star, thanks. it might be we need to scrap it, though... i'll keep you posted09:09
*** zoli__ has quit IRC09:10
*** zoli__ has joined #baserock09:10
*** lachlanmackenzie has joined #baserock09:24
*** edcragg has joined #baserock09:25
*** sambishop has quit IRC09:43
*** sambishop has joined #baserock09:58
*** zoli__ has quit IRC10:16
*** zoli__ has joined #baserock10:17
paulsherwoodsignin to wiki seems broken for me?11:10
paulsherwood has no useful content11:10
ssam2try https://11:16
ssam2that link worked for me though (already signed in)11:16
ssam2if I sign out and follow that link, I get useful content11:17
ssam2have to scroll down for it though for some reason11:17
ssam2maybe something wrong with the CSS used to do the sidebar11:17
paulsherwoodi got there by clicking the edit link. same result with https afaict11:17
ssam2I can't reproduce that in either Firefox or Epiphany, sorry11:25
ssam2you have definitely scrolled to the bottom?11:25
tlsait's style.css:476 setting clear:both for #login_btns11:27
tlsa(that causes the login stuff to be below the sidebar)11:27
ssam2ta! is there an easy way to fix that?11:27
ssam2i have no idea about CSS11:28
tlsaI'll see if I still have access11:28
SotKclear:left seems to make it at least visible to begin with11:30
tlsathere's no reason for it to clear anything11:30
tlsathere's nothing there to clear11:30
tlsaremote: open2: exec of clamdscan - failed at /usr/share/perl5/IkiWiki/Plugin/ line 197.11:38
tlsaTo git://
tlsa ! [remote rejected] master -> master (pre-receive hook declined)11:38
tlsaerror: failed to push some refs to 'git://'11:38
tlsaI've modified that file before (the last edit was also from me)11:40
ssam2argh, I don't know why that would happen11:42
ssam2seems like something broken at Branchable's end11:42
*** franred has quit IRC11:58
*** zoli__ has quit IRC12:16
*** zoli__ has joined #baserock12:16
KinnisonSomeone needs to poke branchable's sysadmin (joey hess)12:28
* Kinnison has asked12:28
Kinnisontlsa: Can you try now?12:35
KinnisonIf it doesn't work then Joey has been emailed so hopefully he'll fix it when he awakes12:37
*** gary_perkins has quit IRC12:53
*** gary_perkins has joined #baserock12:53
*** zoli___ has joined #baserock12:55
*** zoli____ has joined #baserock12:57
*** zoli__ has quit IRC12:57
*** zoli___ has quit IRC13:00
*** franred has joined #baserock13:01
*** jonathanmaw has quit IRC13:02
tlsaKinnison: still same error13:12
KinnisonHopefuly joeyh will work it out when he wakes13:15
*** straycat has joined #baserock13:17
paulsherwoodstraycat: long time no see!? :)13:17
Kinnisonpersia: I have seen him around as early as 2h before now13:17
straycatI am attempting to upgrade perl.13:19
straycatI don't understand why has 2 bsps?13:19
ssam2'bsp-x86_both-tools' is just SYSLINUX, I think13:25
straycatsyslinux and nasm13:26
straycatso that system has duplicated syslinux and nasm chunks?13:27
ssam2I don't think so, but worth checking13:28
straycatI do not understand the bsp setup, but is bsp-x86_both-tools not for use with the chroots, I also thought definitions was a global namespace?13:33
paulsherwoodi thought so too13:34
KinnisonOOI who authored the changes which made this madness?13:45
*** jonathanmaw has joined #baserock13:47
straycatOkay since no one knows what function this serves I expect I'll find it's safe to remove it, I'll test the build systems when I do so14:09
straycatthanks for your help14:09
*** straycat has left #baserock14:09
*** sambishop has quit IRC14:18
*** sambishop has joined #baserock14:26
nowsterHave people noticed this:  ?14:27
nowsterThe lorry probably needs updating.14:27
pedroalvareznowster: I believe that was a svn lorry, so there isn't a way to update it to the new url14:30
pedroalvarezcreating a new one is what we should do14:30
nowsterthere is libogg.lorry and libvorbis.lorry14:30
nowstertarball type lorries14:30
richard_mawpedroalvarez: if their import was also with git-svn then the imports _may_ be compatible but probably aren't14:32
nowsterit looks like git-svn was used14:32
*** sambishop has quit IRC14:32
pedroalvarezsha1's don't match14:35
pedroalvareznot sure if that makes them already incompatible14:35
nowsterspeex is also in git now14:36
pedroalvarezthat is currently a tarball in g.b.o14:37
nowsteras is vorbis14:38
pedroalvareznowster: are you looking at upgrading these components in the current definitions?14:38
nowsterI'm seeing gstreamer get stuck decoding ogg/vorbis files.14:39
nowsterIt'll decode an online stream of vorbis for about a second, then get totally stuck14:39
nowsterflacs and wavs it's happy with14:40
nowsterI don't know if we'd be happy lorrying mpg123, though, even though many potential applications would want it.14:40
paulsherwoodwhy not?14:46
rjekflacs and waves are integer-only; perhaps it's FP-related?14:47
rjek(tremor for Vorbis, libmad for MP3)14:47
* paulsherwood can't edit still... and sulks14:48
persiapaulsherwood: not even with git?14:48
paulsherwoodactually, i don't know. last time i tried that was a couple of years ago, i think14:50
persianowster: I'd recommend submitting a patch for lorrying anything you want.  On the surface mpg123 looks fine, although I don't know the whole story.14:50
persiapaulsherwood: I usually use git to edit the wiki, because I like to use my favorite editor.  Mostly works fine.14:51
nowsterOK. mpg123 is LGPL, but may be patent-encumbered (eg. Fraunhofer Inst)14:51
persiaI don't see any issues with lorrying it (doesn't violate).  Putting it in a reference system is another question.14:51
rjeklibmad sounds better anyway >:D14:52
persiaOlder versions of mpg123 should be fine: it is worth checking, but I understand that most code from 1997 is now patent-free.14:57
persiaIf the newer code is new algorithms that are patented, that is a different issue, but if the newer code is just new wrappers and reimplementations of the old algorithms, then there is little to fear.14:58
* persia is not really qualified to have an opinion about patent terms, so the above should not be considered authoritative.14:58
*** zoli____ has quit IRC15:01
*** jonathanmaw has quit IRC15:03
Kinnisontlsa: In theory things should work (pushing that attachment) now15:25
tlsassam2: wiki login position fixed15:26
ssam2yay! thanks!!15:29
ssam2I still have the 'Enter your OpenID' bit at the bottom of the page tho15:30
ssam2the buttons are correct now15:30
tlsaI don't see an 'Enter your OpenID' bit15:34
ssam2maybe for the same reason paul doesn't ?15:35
CTtpollarddifferent browsers and resolutions?15:36
tlsassam2: oh, you have to click on the open ID button, then it appears15:36
tlsaI can fix it now I see it15:37
rjekOh, it's a button15:37
tlsarjek: I didn't know it was a button either15:37
tlsaI assumed it was a logo to indicate what was supported15:37
rjekIf only we had some idiom, drilled into user's brained over a period of 30 years, for how to display something that can be clicked on.15:37
*** petefoth has quit IRC15:38
KinnisonSeemed fairly obvious to me15:40
Kinnisonit was in a box and when I hovered my mouse over it, it hilighted like a clickable thing15:40
Kinnisonthe odd vertical spacing is annoying though15:40
*** tiagogomes has quit IRC15:43
*** jonathanmaw has joined #baserock15:45
*** gary_perkins_ has joined #baserock15:56
*** gary_perkins has quit IRC15:56
*** franred has quit IRC16:27
*** ssam2 has quit IRC16:28
paulsherwoodhow do folks feel about me adding ybd instructions to the wiki? i'm happy not to do if it's controversial16:40
KinnisonGiven that at this point ybd does not implement everything used in the definitions set, I'd kinda against it on any default pages or getting started pages, but wouldn't object to an "Alternative Tooling" type page16:45
*** mariaderidder has quit IRC16:46
paulsherwoodwhat do you mean by 'everything used in the definitions set'?16:50
Kinnisondefinitions includes artifact splitting for various bits and bobs fr.ex.16:51
Kinnisonif ybd supports all that now, then excellent16:51
nowsterI'm looking into artefact splitting (as a background task)16:51
paulsherwoodit doesn't, yet. it won't ever support some things, because i don't think all of the things are correct16:51
paulsherwoodbut that's a conversation for another day, with alcohol.16:52
KinnisonI think that until and unless definitions is properly locked down behaviour-wise then no tool other than morph will be "valid"16:52
paulsherwoodthat's a valid opinion :)16:52
Kinnisonit's also correct16:53
paulsherwoodit's opionin. correctness is something different, i think16:53
Kinnisonhmm, well, you can think that if you want :-D16:53
paulsherwoodthere is no 'spec'. morph has bugs, and incomplete behaviours. so does ybd, but the incompleteness and problems are different between both programs. at the moment i think 'correctness' or 'validity' is not arguable for either program.16:56
*** bashrc has quit IRC16:56
KinnisonIn the absence of a proper maintained document about what definitions means, morph is de-facto the correct implementation, bugs'n'all.16:58
*** gary_perkins_ has quit IRC17:03
*** jonathanmaw has quit IRC17:05
*** mdunford has quit IRC17:12
*** edcragg has quit IRC17:22
*** sambishop has joined #baserock18:11
*** lachlanmackenzie has quit IRC18:37
*** juergbi` is now known as juergbi19:13
persiaReading backscroll, I support Kinnison's assertion that the morph behaviour is "correct".  I also happen to think much of is is "wrong", but being both formally correct and wrong is a common state of affairs for real things (as opposed to platonic ideals), so none of us should fuss that much.19:39
*** zoli__ has joined #baserock22:13
*** sambishop has quit IRC22:59
*** dabukalam has quit IRC23:24
*** dabukalam has joined #baserock23:25

Generated by 2.15.3 by Marius Gedminas - find it at!