IRC logs for #baserock for Friday, 2016-01-22

*** gtristan has joined #baserock05:03
*** mwilliams_ct has joined #baserock08:13
*** bruce_ has joined #baserock08:44
*** edcragg has joined #baserock08:59
*** bashrc has joined #baserock09:06
*** CTtpollard has joined #baserock09:08
*** rdale has joined #baserock09:20
paulsherwoodgtristan: has aboriginal's mail list gone down again?09:25
* paulsherwood can't find Previously posts aimed to entertain and educate on software engineering09:26
paulsherwoodissues in a 'humorous' way but as Paul says ""09:26
paulsherwoodbah!!!!!09:26
paulsherwoodi'm an idiot09:26
paulsherwoodcan't find http://lists.landley.net/pipermail/aboriginal-landley.net/2016-January/001486.html09:26
rjekIt appears the archiver hasn't run this month09:28
*** jonathanmaw has joined #baserock09:30
paulsherwoodit was there yesterday09:31
gtristanoh crap :(09:32
gtristanthat sucks09:32
gtristanThe hosting had eaten up 11months of history in 2015 fwiw09:32
paulsherwoodgtristan: maybe ask rob if he needs an alternative provider? maybe pepperfish could help?09:32
rjekPossibly.09:33
gtristangeez09:34
gtristansigh, I'll let him know his archives are busted *again* in case he doesnt know09:35
tiagogomes__maybe you should lorry the email archives09:36
paulsherwood?09:37
tiagogomes__sorry, nevermind, just a bad joke09:37
mwilliams_cttiagogomes__: it made me smile, which means it probably was a bad joke09:38
*** toscalix has joined #baserock09:40
*** ssam2 has joined #baserock10:12
*** ChanServ sets mode: +v ssam210:12
*** ssam2 has quit IRC10:16
*** ssam2 has joined #baserock10:18
*** ChanServ sets mode: +v ssam210:18
*** Lachlan1975 has joined #baserock10:21
gtristantoscalix, I'm about to go eat supper... I'll be back in an hour if you want to chat/meet10:21
*** gtristan has quit IRC10:28
rdaleafter making a change to ybd i'm having trouble building the bison chunk - i get a syntax error for a makefile. i seem to remember having this problem before10:52
rdalei thought it needed bash, although adding that as a build depend doesn't fix my current problem10:53
franredrdale, what is the error? log?10:57
rdaleit ends like this: http://paste.baserock.org/aheyipicuq10:58
rdalei think the make file syntax error is a red herring10:59
pedroalvarezthere was one problem... which looks related, addresed by you in 7c2bfac98e93e42ffcb23e0d43aee8f3d2c01c6910:59
richard_mawhmm, version UNKNOWN would be a problem for bison10:59
richard_mawit could try to get the version from git if it were available, or we had some stuff inject the version number manually11:00
pedroalvarezright, looks like this commit c6b6225e4f5872ea749fde416bca8e6b89cb5424 is not complete11:00
pedroalvarezrdale: I think that removing --no-git would solve it11:00
rdaleok, i'll try that11:00
*** gtristan has joined #baserock11:01
richard_mawwe need an FAQ: Q: Bison doesn't build! A: Bison is a low-level dependency that wants to get its version from git, which is a problem for bootstrapping. Ensure it's either got the version number explicitly written in, or that git is available and bison isn't configured to not use git.11:02
pedroalvarezheh11:03
pedroalvarezfor that we have git history11:03
tiagogomes__he have this http://wiki.baserock.org/guides/build-failures/11:04
pedroalvarezyeah, but that's more related to build issues because of the build tool11:05
rdaleremoving '--no-git' doesn't fix it11:05
pedroalvarezif we have to fill that with info for every component..11:05
pedroalvarezrdale: following that FAQ, is git available in the staging area?11:05
pedroalvarezbut anyway........ this shouldn't be failing11:06
rdalegit-minimal is a build depenency11:06
pedroalvarez(if this is master of definitions)11:06
rdalequite a few chunks in core.morph are missing perl which is needed to run autoreconf, and so i'm not sure how they build with morph11:07
richard_mawrdale: some of them have configure checked-in so they don't need to run autoreconf11:08
rdaleright, but the test for autotools in ybd is nearly the same as in morph, it only tries autoreconf when it can't find anything else11:09
pedroalvarezhm.. aren't both using what it's in DEFAULTS?11:10
pedroalvarezif not -> :(11:10
rdalethe order of the tests is slightly different, which is probably a bug in either morph or ybd11:11
rdaleah - is DEFAULT in definitions?11:12
pedroalvarezthey should both use DEFAULS, that's why we have it11:12
pedroalvarezyes11:12
toscalixgtristan: I have been in meetings but I will be free the next 90 min. WHen you are back, let me know11:13
rdaleybd does use that if present11:13
gtristantoscalix, I am back... I'll be around11:16
ratmice______i wonder if bison would accept clean/smudge git attributes for that version stuff11:18
ratmice______so the thing gets set on checkout..11:18
richard_mawooh, interesting idea11:19
mwilliams_ctcan anyone help me setting up a subdomain on baserock.org to point to our concourse instance? It's at paulsherwood's request11:24
rjekYou need Baserock Ops!11:28
rjekWhich I believe is gary_perkins, pedroalvarez, and ssam2.11:28
rjekAnd franred, apparently.11:28
radiofreei think franred stepped down11:29
pedroalvareztananananananna </batman>11:30
pedroalvarezI can't help with that, gary_perkins is your man11:30
pedroalvarezIf he is not around, I can find someone else11:30
ssam2rjek: I believe this is still done by Codethink ops11:31
ssam2if it has changed, nobody told me11:31
* gary_perkins is here11:31
rjekFortunately, gary_perkins is both.11:32
gary_perkinsmwilliams_ct: what's the IP and desired subdomain?11:32
rdalei get a missing libcrypto error with git-minimal11:37
mwilliams_ctgary_perkins: concourse. and it's to point to an aws load balancer so I might have to tweak stuff on AWS as well11:38
mwilliams_ctgary_perkins: will msg you the loader address11:38
gary_perkinsmwilliams_ct: ok, that's been setup. It might take anywhere from a few minutes to a few hours to propagate11:44
mwilliams_ctgary_perkins: fantastic thanks!11:44
mwilliams_ctalready visible on my machine :D11:44
gary_perkinscool :)11:44
mwilliams_ctIt's also visible over ssl using a self-signed certificate. I should probably do that properly at some point11:45
pedroalvarezwhat will make some people sad, is that it isn't public11:48
mwilliams_ctpedroalvarez: quasi-intentional. quasi because i'm not actually sure how to make it public without exposing us to attacks, and intentional because regardless it isnt quite ready to show of yet11:49
mwilliams_cts/of/off11:49
pedroalvarezI understand the first point11:49
mwilliams_ctI've asked upstream about the first point, though they wont be awake for a few hours11:55
paulsherwoodpedroalvarez: the idea is to make it public as soon as it's doing something useful. in the meantime i propose not to draw attention to it :)11:58
pedroalvarezit's fair :)11:58
paulsherwood:-)11:59
paulsherwoodseen elsewhere...12:02
paulsherwood10:33 < rdale>     - The build-depends code in ybd is installing the build-depends of each build-depends recursively12:02
paulsherwood10:33 < rdale>         - I believe this is wrong12:02
pedroalvarezheh12:02
paulsherwoodi believed this too, originally, but then folks told me that depends are transitive and popular opininon was against me12:03
paulsherwoodrdale: ^^12:03
paulsherwoodrichard_maw: ^^12:03
rdalehmm12:03
tiagogomes__I recall that12:04
paulsherwoodanyway, definitions are built with that in mind, by both ybd and morph12:06
paulsherwoodit would be a significant change to definitions (implied) semantics12:06
rdaleok, there is still a bug where gcc is pulling in stage2-gcc, which pulls in stage1-gcc - the rules need to be different for chunks built in bootstrap mode12:09
paulsherwoodwell, it's been building functional systems that way for some time... what is the actual issue?12:09
rdalethe final system contains all the build-depends chunks12:10
rdaleand strata12:10
paulsherwoodare you sure?12:11
* paulsherwood believed that only contents end up in systems, not build-depends12:12
rdaleyes, my minimal-system-x86_64-generic ends up with the chunks from the core stratum in it12:12
rdalewhich is only there because nasm needs core to generate its docs12:12
paulsherwoodoh, wow. that's a bug then, yes12:13
paulsherwoodbut is that the case with master ybd?12:13
rdaleyes12:14
rdalei'm using my branch with artifact splitting .meta files, but i haven't changed the basic build process12:14
paulsherwooderk, then.12:15
rdalethe top level system is composed of two strata, and i would expect to see only those two strata installed into it. but in fact all sorts of stuff including individual chunks are being installed directly into the system12:17
* paulsherwood reproduces this issue, and is amazed12:25
rdalegood, so it's nothing i've done12:25
paulsherwoodno... but i was *sure* this used to work12:26
rdalei thought it might be something to do with the transitive build-depends change12:27
paulsherwoodi don't think so, but i'll investigae over the weekend if you don't find it first12:27
pedroalvarezahemtests...12:28
paulsherwoodpedroalvarez: ack. i was thinking the same12:29
pedroalvarez:)12:29
pedroalvareztbf, I don't know if this is covered by Morph's test suite12:30
paulsherwoodheh12:30
pedroalvarezbut we would add such a test if we find this issue12:31
rdalei think transitive build depends might concievably be a good idea if there was tooling to show what they were - without that it seems a bit of a nightmare to me12:33
*** tiagogomes has joined #baserock13:26
*** tiagogomes__ has quit IRC13:26
*** tiagogomes has quit IRC13:41
*** tiagogomes_ has joined #baserock13:45
franredrdale, I think there were a tool in morph which showed the build-dependency graph, I can't remember the name14:12
rdaleok, it sounds a good idea then. in the core.morph you can change a chunk in one place, and suddenly another chunk will break because of these implicit build dependencies14:13
*** ssam2 has quit IRC14:16
*** ssam2 has joined #baserock14:26
*** ChanServ sets mode: +v ssam214:26
*** bruce_ has quit IRC14:44
*** bashrc has quit IRC14:44
*** faybrocklebank has quit IRC14:44
*** tiagogomes_ has quit IRC14:44
*** franred has quit IRC14:44
*** ssam2 has quit IRC14:44
*** edcragg has quit IRC14:44
*** nowster has quit IRC14:44
*** Lachlan1975 has quit IRC14:44
*** jonathanmaw has quit IRC14:44
*** bashrc has joined #baserock14:44
*** edcragg has joined #baserock14:44
*** bruce_ has joined #baserock14:44
*** faybrocklebank has joined #baserock14:44
*** nowster has joined #baserock14:44
*** ssam2 has joined #baserock14:44
*** ChanServ sets mode: +v ssam214:44
*** tiagogomes__ has joined #baserock14:44
*** Lachlan1975 has joined #baserock14:44
*** jonathanmaw has joined #baserock14:44
*** franred has joined #baserock14:50
*** bruce_ has quit IRC14:58
*** ssam2 has quit IRC15:04
*** gtristan has quit IRC15:06
*** bruce_ has joined #baserock15:09
*** gtristan has joined #baserock15:25
*** ctbruce has joined #baserock15:52
*** bruce_ has quit IRC15:52
*** ctbruce has quit IRC15:54
*** bruce_ has joined #baserock15:56
paulsherwooddo baserock pastes ever expire?16:08
CTtpollardyes16:11
CTtpollardiirc if not viewed for ~30 days16:11
paulsherwoodok16:11
CTtpollardI had it happen to me recently16:11
paulsherwoodthanks!16:11
*** bruce_ has quit IRC16:14
*** ctbruce has joined #baserock16:14
*** ctbruce has quit IRC16:17
*** franred has quit IRC16:35
*** franred has joined #baserock16:52
*** CTtpollard has quit IRC17:21
*** franred has quit IRC17:25
*** franred has joined #baserock17:40
*** ssam2 has joined #baserock17:40
*** ChanServ sets mode: +v ssam217:40
*** jonathanmaw has quit IRC17:41
*** ssam2 has quit IRC17:57
*** franred has quit IRC18:00
*** bashrc has quit IRC18:00
*** tiagogomes__ has quit IRC18:02
*** toscalix has quit IRC18:13
*** edcragg has quit IRC18:49
*** Lachlan1975 has quit IRC19:01
*** ratmice______ has quit IRC21:41

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!