Log for #openttdcoop.devzone on 25th July 2012:
Times are UTC Toggle Colours
05:47:39  *** Nat_aS has quit IRC
05:48:00  *** Nat_aS has joined #openttdcoop.devzone
06:31:18  *** Alberth has joined #openttdcoop.devzone
06:35:21  *** Alberth has quit IRC
06:53:47  *** Nat_aS has quit IRC
06:54:07  *** Nat_aS has joined #openttdcoop.devzone
06:59:27  *** Alberth has joined #openttdcoop.devzone
07:02:00  <Brot6> OpenGFX+ Trains renders - Revision 35:43563922bb96: Update: Altered livery and layout of passenger m... XXotic750X @
07:03:13  *** Nat_aS has quit IRC
07:03:14  <Brot6> OpenGFX+ Trains renders - Revision 36:6c55fc766661: Add: Blender models and render run for sh125 liv... XXotic750X @
07:03:14  <Brot6> OpenGFX+ Trains renders - Revision 37:cbd851818740: Add: Blender models and render run for early mon... XXotic750X @
07:03:26  *** Nat_aS has joined #openttdcoop.devzone
07:21:04  <Brot6> repository /home/hg/zbuild registered in Redmine with url /home/hg/zbuild
07:21:04  <Brot6> repository /home/hg/zbuild created
07:23:07  <Alberth> going for the shorter name? :)
07:23:12  <Brot6> OpenGFX+ Trains - Revision 738:b86b0f164921: Update: 32bpp Passenger sprites for normal use XXotic750X @
07:23:25  <planetmaker> well :-) I want to keep the zbasebuild repo
07:23:42  <planetmaker> this is just a very slim top-level repo which has zbase, zbasebuild and opengfx as sub repos
07:23:56  <Brot6> OpenGFX+ Trains - Revision 739:5e7b12ccb268: Add: Sprites for sh125 livery XXotic750X @
07:23:59  <planetmaker> thus the relative paths as used in zbasebuild will continue to work
07:24:12  <planetmaker> hopefully :-)
07:24:48  * Alberth keeps fingers X-ed
07:24:57  * planetmaker builds a test clone now
07:25:07  <Brot6> OpenGFX+ Trains - Revision 740:b55a7c030c10: Add: Sprites for early monorail passenger XXotic750X @
07:26:45  <planetmaker> hm, it fails on gimp (which it always does when I have no X11 running). So far looking good :-)
07:27:52  <Brot6> OpenGFX+ Trains - Revision 741:a750353d5ff4: Feature: Use 32bpp sh125 livery sprites XXotic750X @
07:29:56  <Brot6> OpenGFX+ Trains - Revision 742:e9b305cb6da9: Feature: Use 32bpp cargo refitted engine sprites for MU... XXotic750X @
07:32:22  <Brot6> OpenGFX+ Trains - Revision 743:3192a37b6db7: Cleanup: Changed spacing and EOLs for easier reading XXotic750X @
07:40:20  <planetmaker> Ammler: it seems that the CF did not run. Or did not build nmlc at least...
07:43:44  <planetmaker> btw, Alberth, maybe you should indeed gray-scale ajust your monitor if you don't see the shades of gray in the church roof which you posted yesterday :-)
07:44:18  <Alberth> perhaps :)
07:45:02  <planetmaker> check for instance
07:45:03  <Webster> Title: Monitor Calibration for Photography (at
07:45:15  <planetmaker> it has (a bit further down) proper adjustment graphics and instructions
07:45:34  <Alberth> it's an old beast, an 17" CRT that has already been considered as waste when I got it
07:46:36  <Alberth> so it's using its 2nd life already :)
07:47:08  <Alberth> besides I like a dark setting, all those paper-white windows of todays applications are giving me a headache
07:47:36  <Alberth> s/$/ otherwise/
07:47:36  <Brot6> Alberth meant: "besides I like a dark setting, all those paper-white windows of todays applications are giving me a headache otherwise"
07:47:42  <planetmaker> ah, yes. I used to use an old 17", too for my photos. Then I looked at the adjustments (colour, brightnes,...) at an LCD display. Oh, my god. They were grossly contrasted too much
07:49:08  <planetmaker> but still, it's worth to play with monitor settings with a page like I just linked. Maybe it can be improved
07:49:27  <planetmaker> When I setup this system I also spend quite some time finding good settings with that page or similar :-)
07:50:18  <Alberth> with photos it's important, I am mainly dealing with plain text
07:50:44  <Alberth> besides, if all goes accroding to plan, I should get a 'new' monitor next weekend
07:51:48  <planetmaker> :-)
07:58:16  <Brot6> FISH - Revision 821:e9f1e3ca20f9: Fix: prevent calculated model life overflowing 255 XandythenorthX @
08:10:50  <Brot6> FISH - Revision 822:05253272c68d: Change: set loading speeds for all currently converted ships XandythenorthX @
08:14:04  <dihedral> greetings
08:21:47  <Terkhen> hi dihedral
08:22:21  <dihedral> hello sir Terkhen
08:24:57  <planetmaker> juhu!
08:25:01  <planetmaker> Alberth: pull zbuild
08:25:31  <Alberth> s/pull/clone/ :)
08:25:31  <Brot6> Alberth: You did something wrong... Try s/you/me/ or tell me "help sed"
08:25:35  <planetmaker> yes
08:25:42  <Brot6> zBuild - Revision 0:fa402902f22e: Add: zbase, zbasebuild and opengfx as sub repos XplanetmakerX @
08:25:42  <Brot6> zBuild - Revision 1:a9b829e28bd8: Add: relative sub paths on server XplanetmakerX @
08:25:49  <planetmaker> you might need hg 2.0 or newer
08:26:15  <planetmaker> as it not only makes use of sub repos but also use of the subpath, a feature of subrepos introduced in 2.0
08:26:30  <planetmaker> which allows to specify the relative paths on a server for repos
08:28:12  <planetmaker> might also be that it only works via ssh properly right now.
08:29:04  <Alberth>
08:29:24  <planetmaker> hm
08:31:01  <planetmaker> ah
08:31:16  <planetmaker> :-) I need to define more relative paths... there are too many alternative names
08:31:33  <planetmaker> hg clone ssh:// zbuild works
08:31:42  <planetmaker> I'll make yours possible, too
08:34:56  <Brot6> zBuild - Revision 2:275dfb146b35: Add: Server paths when using XplanetmakerX @
08:36:31  <Alberth> nope
08:37:10  <planetmaker> nope?
08:37:12  <Alberth> perhaps it is still stuck on the previous revision?
08:37:24  <Brot6> zBuild - Revision 3:29b183956d60: Fix paths for user hg XplanetmakerX @
08:37:33  <planetmaker> I now pushed with the thing you tried first. Maybe try again what you tried first?
08:38:21  <Alberth> it works, except it's pulling an opengfx clone
08:38:39  <planetmaker> yes, it should clone all 4 repos actually
08:38:50  <planetmaker> the new one and the 3 subs
08:39:22  <planetmaker> that's the expected behaviour :-)
08:39:26  <Alberth> see you in an hour or so :)
08:39:30  <planetmaker> :-D
08:39:38  <planetmaker> I'm afraid that it can't be avoided
08:40:01  <planetmaker> hm... maybe... it can
08:40:08  <Alberth> probably only by stating a local file path
08:40:15  <planetmaker> but not sure ^^ is a good idea
08:40:37  <Alberth> unlikely imho
08:41:01  <planetmaker> yup. But you can "forget" the other clones now. Or maybe then link those dirs to the clones you get right now
08:41:42  <Alberth> where does it get dumped? inside zbuild?
08:41:57  <planetmaker> zbuild has 3 dirs: opengfx, zbase and zbasebuild
08:42:04  <planetmaker> which are the repos themselves
08:42:17  <Alberth> perhaps hg has some path/repo remapping thingie
08:42:47  <planetmaker> that's the [subpath] section in .hgsub in zbuild
08:43:06  <planetmaker> but I don't like to travers outside the repo itself for local folders
08:43:24  <planetmaker> I did the mapping for the server there
08:44:42  <planetmaker> for my part, I'll get an additional local clone of opengfx. But I'll only need zbase and zbasebuild inside zbuild
08:45:15  <planetmaker> but having this separate local clone of opengfx is a good thing as the revision of subrepos is stored inside zbuild
08:46:17  <Alberth> yeah, all path remapping point to subrepos; no way to say you can find "a clone of this and this repo at this location at the file system"
08:47:09  <Alberth> which somewhat defeats the idea of distributed VCS :p
08:47:35  <planetmaker> Probably a local mapping *could* be added. But ^^
08:48:16  <planetmaker> In this case it's a one-time thing to clone. Having opengfx twice (you don't need that either!) is not that hurtful, I think
08:48:44  <planetmaker> for my part I need two clones of opengfx, one default branch, one stable branch for working on it
08:48:59  <planetmaker> and then this in zbuild doesn't hurt me either :-)
08:49:22  <Brot6> Berries - Revision 35:fde06a5e2cde: Remove: old file, no longer needed XdihX @
08:49:40  <dihedral> wtf - i got highlighted....
08:52:59  <Alberth> planetmaker: It seems to me that I should be able to deicde where hg pulls its clones from, while cloning zbuild
08:53:29  <planetmaker> Alberth: I found it more important that neither of the 3 sub repos needed a single commit than requiring us two (or 3) to clone each repo anew
08:53:56  <planetmaker> Alberth: no, you should not. The repo state is part of the revision. As such the source is defined. Need be defined
08:54:33  <planetmaker> zbuild knows the revisions to checkout in each of the sub repos
08:54:46  <planetmaker> it can happen that it can't find them, if you pull from *somewhere*
08:55:20  <Alberth> planetmaker: what if I don't like devzone, and I have a better source (more reliable network eg). Imho a distributed VCS should be distributed
08:55:44  <planetmaker> Alberth: you can clone. And then go to the sub repos and pull -f from the sub repos of your choice
08:56:23  <Alberth> I can clone without pulling the subrepos?
08:56:33  <planetmaker> and... sub repos are a "feature of last resort". It *does* break some of the ways how things work. this is one
08:56:39  <planetmaker> no, you can't
08:56:58  <planetmaker> you clone zbuild means you get a clone of each subrepo
08:57:12  <planetmaker> otherwise it'd not be a clone
08:57:25  <Alberth> getting a clone of the subrepo is not the problem imho, not being able to chose the source is
08:58:44  <Alberth> but apparently it is broken
08:58:51  <planetmaker> what is broken?
08:59:27  <planetmaker> If I clone zbuild and specify to clone from local path the sub repos. But the server has further commits which I don't have locally. How can it then give me the correct revision for the sub repos for my clone of zbuild?
08:59:33  <planetmaker> It *must* pull from the path specified
08:59:36  <Alberth> suppose you use a subrepo X from another source, where I have no connection to, however, I know a different source for X
09:00:20  <Alberth> obviously, it should fail if X is not present, or it does not have the right revision
09:00:38  <Alberth> but it's a distributed VCS, including for the subrepos imho
09:01:04  <planetmaker> Alberth: that's exactly why you should get a fresh clone of *each* of the subrepos
09:01:15  <planetmaker> So that you got all repos as the person you pull from had them
09:01:29  <Alberth> why?
09:01:41  <planetmaker> consider the sub repos just paths. local to the main repo
09:01:54  <planetmaker> why do you want to pull the png folder of opengfx from somewhere else than the repo?
09:02:07  <Alberth> let's do one step back
09:02:12  <Alberth> single repo
09:02:24  <planetmaker> you get everything. All sub paths
09:02:40  <Alberth> no sub-paths, single repo, standard DVCS
09:02:55  <planetmaker> yes. You get the whole thing. All files. All dirs
09:02:58  <Alberth> I can pull, merge, update etc from anywhere and to anywhere
09:03:13  <planetmaker> Yes. And when you clone, you get all files as they are from there
09:03:21  <Alberth> no need to get some revision from a specific point
09:03:33  <planetmaker> exactly.
09:03:50  <planetmaker> But you get all files exactly as they are in the revision which you pull or clone
09:03:58  <planetmaker> And nothing else hapens with sub-repos
09:04:11  <planetmaker> you get dir; dir/repo1; dir/repo2
09:04:25  <planetmaker> and repo1 and repo2 are just normal paths and files
09:04:27  <Alberth> sure, but the place where I get a revision from does not matter
09:04:37  <Alberth> which is key to being DVCS
09:04:42  <planetmaker> just that it happens that there are also .hg folders in dir/repo1 and dir/repo2
09:04:53  <planetmaker> Alberth: it doesn't matter here either
09:05:04  <planetmaker> you ahve to consider zbuild one project
09:05:18  <planetmaker> which consists of all the 3 repos inside it. With all its files
09:05:29  <planetmaker> thus you get the whole repo always from one place
09:05:31  <Alberth> ah, ok
09:06:21  <planetmaker> the [subpath] thing in .hgsub is just for convenience to be able to distribute it easier
09:06:50  <planetmaker> usually there'd be all 3 repos in the main repo. Anew
09:06:57  <Alberth> yeah I see now
09:06:59  <planetmaker> the [subpath] just avoids having new copies on the server
09:07:01  <planetmaker> ok :-9
09:07:04  <planetmaker> :-)
09:08:05  <Alberth> not sure I like subrepos much :p
09:08:26  <planetmaker> what's wrong with them?
09:10:01  <Alberth> they don't let go of 'single repo with everything in it'
09:10:46  <planetmaker> Well... yes, they don't. But they can't exactly do that either
09:12:05  <planetmaker> when I created it, I moved zbase and zbasebuild dirs into zbuild. and copied opengfx there, too
09:12:16  <Hirundo> good morning
09:12:24  <planetmaker> then created the .hgsub and just pushed. It just pushed the 2 revisions of zbuild
09:12:26  <planetmaker> Hello Hirundo
09:12:51  <planetmaker> and you can still continue to work on each subrepo (zbase, zbasebuild, opengfx) as before
09:13:04  <planetmaker> But if you want the whole project - then of course you need all its parts
09:13:25  <planetmaker> I think it's like if you get a checkout of svn:// :-)
09:13:31  <planetmaker> without any path specified
09:13:47  <planetmaker> bla bla :-P
09:14:25  <Alberth> sure I need all parts, but I see no fundamental reason why I would have to shop for all parts in your store
09:14:45  <planetmaker> if you buy the product, you get it assembled
09:14:57  <planetmaker> you're free to replace any part
09:15:18  <planetmaker> why should it be different? Otherwise I might sell hot air :-P
09:16:19  <Alberth> well, we're not getting anywhere it seems, what do I do with zbuild?
09:17:06  <planetmaker> you get a checkout and continue with zbuild/zbasebuild, zbuild/zbase and zbuild/opengfx as you did with zbasebuild, zbase and opengfx before :-)
09:17:30  <planetmaker> I'll need to stick some Makefile into zbuild which calls zbuild/zbasebuild's makefile
09:17:45  <planetmaker> so that the CF then hopefully will start to do the right thing [TM]
09:18:44  <planetmaker> you can do the commit on the zbuild level, though
09:18:59  <planetmaker> that should update the whole thing properly
09:19:42  <Alberth> any news on the zephyris commit problems?
09:19:51  <planetmaker> yes. It should be solved
09:20:06  <planetmaker> failed transaction while pushing too big stuff without ssh
09:20:37  <Brot6> OpenGFX+ Trains renders - Revision 38:710261f8ceba: Add: Blender models and render run for early mon... XXotic750X @
09:20:51  <planetmaker>
09:20:52  <Webster> Title: Subrepository - Mercurial (at
09:23:02  <Brot6> zBuild - Revision 4:143a2f42cc64: Change: Bring subrepos up to date XplanetmakerX @
09:23:03  <Brot6> OpenGFX+ Trains renders - Revision 39:dd83a2951733: Update: Modernised monorail mail and armoured mo... XXotic750X @
09:23:12  <planetmaker> commits like this last one can be avoided, if you commit directly in zbuild
09:23:33  <planetmaker> the commit only did hg pull -u on each subrepo. So that they're all up to newest rev
09:24:21  <Alberth> I am reading the page you linked
09:25:08  <planetmaker> it cautions against the use of subrepos :-) But the only other way I see is to make it really one big repo. Not better really
09:30:40  <Alberth> a DIY solution comes very close to subrepos, it seems
09:31:23  <planetmaker> I don't think it's that bad :-)
09:31:40  <planetmaker> but what do you want really?
09:32:05  <planetmaker> you don't want one big repo. You don't want the repos referenced. What other solution is there?
09:32:17  <Hirundo> hmm.. where are NML nightlies of the last couple of days?
09:33:00  <planetmaker> The only other solution is to not do anything. Create a 4th repo as now. With no revision-logging of the "sub-repos" and hacking the CF to pull the "sub-repos"
09:33:11  <planetmaker> would be quite a lot worse, if you ask me
09:33:38  <planetmaker> as it would give no reproducable results
09:34:26  <Alberth> I have no problem with a big repo or with referenced repos. I just  expected a DVCS to be distributed
09:34:38  <planetmaker> it is?
09:34:57  <planetmaker> I don't quite see where the "distributed" breaks for you
09:35:11  <planetmaker> it's a repo with repos
09:35:18  <planetmaker> you get the main repo
09:35:25  <planetmaker> with all its contents
09:35:31  <Alberth> and I agree that you want to keep revisions of combinations of repos at the very least, for reproducable builds
09:36:03  <planetmaker> you can pull everything from everywhere. But if you want the main repo, you get it and its children always from the source from where you pull
09:36:10  <planetmaker> which is the same with any repo
09:36:15  <planetmaker> whether it has sub repos or no
09:36:17  <planetmaker> not
09:36:39  <Alberth> it has subrepos but it hasn't
09:36:54  <planetmaker> where doesn't have them?
09:36:57  <planetmaker> they're all there
09:37:03  <planetmaker> part of the big thing
09:37:04  <Alberth> subrepos are not distributed any more
09:37:14  <Alberth> which sucks imho
09:37:16  <planetmaker> how are they not distributed?
09:37:40  <Alberth> you get to decide where I copy them from, how is that distributed?
09:37:44  <planetmaker> you can - just as before - pull them from wherever you're happy
09:37:56  <planetmaker> but if you want the main repo, you get the whole thing
09:38:04  <planetmaker> that's got nothing to do with distributed or not
09:38:17  <Brot6> Berries - Revision 36:af2ac4ccd96a: Add: IrcMessageProvider and the like XdihX @
09:38:17  <Brot6> Berries - Revision 37:c579a5cd0063: Change: enable the irc module XdihX @
09:38:21  <planetmaker> you never pick for one repo where to pull from the individual files either, do you?
09:38:27  <Alberth> sure it is, imo
09:38:46  <Brot6> OpenGFX+ Trains - Revision 744:f37c515d6484: Update: 32bpp monorail mail and armoured sprites XXotic750X @
09:38:49  <Alberth> hg unit of working is 'repo'
09:38:51  <planetmaker> in that sense, every hg repo is a "central" one. You get all files from one person
09:39:44  <Alberth> I just expected that subrepos are also repos
09:39:51  <planetmaker> they *are* repos
09:39:57  <Alberth> but they are not from the main repo point of view
09:39:59  <dihedral> every hg repro is a distributed one, as you can get everything from everyone, until you decide to use hg in a cerntralized way
09:40:13  <dihedral> :-P
09:40:28  <planetmaker> from the main repo POV they're bound to it
09:40:36  <planetmaker> that's why they're sub-repos.
09:40:41  <Alberth> dihedral: but not when you decide to use a repo as subrepo
09:41:02  <planetmaker> Alberth: of course. You can clone every of the subrepos separately
09:41:03  <dihedral> then the main repos are central ;-)
09:41:20  <Alberth> (11:48:55) planetmaker: from the main repo POV they're bound to it <-- and this is where it goes wrong, a uniq UUID + revision is enough to get that data from anywhere
09:41:23  <dihedral> and you do not commit to those
09:41:42  <planetmaker> Alberth: exactly "is enough to get that data from anywhere" is not true
09:41:44  <Alberth> but I don't get that choice
09:42:17  <Alberth> planetmaker: why not? that's how repos recognize each other
09:43:05  <dihedral> why the heck can i not simply close an issue at the devzone??
09:43:18  <Alberth> you're not a developer/manager?
09:43:34  <planetmaker> Alberth: because *anywhere* might miss the exact revisions you need. Only by shipping it together you can get the repo with its full revision
09:43:41  <dihedral> Brot6, close #2295
09:43:44  <dihedral> :-(
09:43:47  <planetmaker> #2295
09:43:48  <Brot6> planetmaker: #2295 is "Feature Request #2295: Use PircBotX - Berries - #openttdcoop Development Zone"
09:44:29  <planetmaker> *history instead of last revision
09:44:35  <Alberth> planetmaker: that's my problem imho, you can create the same mess by modifying the original repo
09:44:47  <planetmaker> which same mess?
09:45:08  <Alberth> subrepos not having revisions that the main repo needs
09:46:05  <Alberth> planetmaker: the point is that revision availability is a distributed property as well
09:46:28  <Alberth> there is no such thing as 'I have revision X and nobody else has it'
09:46:39  <Alberth> s/$/ *ever/
09:46:39  <Brot6> Alberth meant: "there is no such thing as 'I have revision X and nobody else has it' *ever"
09:46:41  <planetmaker> dihedral: if you change it from "feature request" to something else you can edit the status again
09:46:57  <dihedral> and why can i not close a feature request?
09:47:13  <planetmaker> Alberth: without pushing the sub-repos, there is. And if you push the sub-repos on commit - then you can also clone them to ensure that you *will* have the proper revisions
09:47:18  <Brot6> OpenGFX+ Trains - Revision 745:73e4ff445e1e: Add: Sprites for early monorail mail and armoured XXotic750X @
09:47:42  <dihedral> thank you planetmaker
09:48:05  <planetmaker> getting the full repo from where you clone it from still doesn't stop you getting pulls for the sub-repos from wherever you want
09:48:15  <Brot6> OpenGFX+ Trains - Revision 746:57c6ab696229: Update: 32bpp model list XXotic750X @
09:48:20  <Brot6> Berries - Feature #2295 (Closed): Use PircBotX XdihX @
09:48:21  <planetmaker> it's imho actually a sign of being dvcs
09:48:55  <planetmaker> dihedral: but why you can't close a feature request... probably because it got confirmed and then no action taken :-)
09:49:17  <dihedral> i put it at 100% done and could still not close it :-)
09:49:17  * Alberth goes finding some lunch, much more productive
09:49:32  <Alberth> dihedral: just let the devs of the project handle it
09:49:35  <dihedral> Alberth, i feel with you :-) my tummy is empty too
09:49:37  <planetmaker> hehe :-)
09:49:47  <dihedral> Alberth, ....?
09:49:49  <planetmaker> sounds like a good idea
09:49:52  <dihedral> the devs of the project?
09:50:00  <dihedral> pffft
09:50:03  <dihedral> that is my project
09:50:07  <dihedral> :-P
09:50:18  <Alberth> oh, nice :)
09:50:42  <Alberth> reject it instead? :D
09:51:01  <dihedral> it was my issue :-P
09:55:48  <Alberth> I had that problem once with gforge (the sourceforge software), where I could not close my own issues as project owner; I had too many rights :)
09:56:57  <Alberth> the permission checking code was just broken for that case
10:03:57  <Alberth> someone missing an obiwan? I have one
10:05:19  <planetmaker> I'm sure I have some in stock. Shall we compare notes? :-)
10:13:06  <Brot6> OpenGFX+ Trains - Revision 747:b1e6c81d8ffb: Feature: Use 32bpp early wagon sprites for standard wag... XXotic750X @
10:14:58  * planetmaker now builds with new nml. Let's see how caching works :-)
10:16:02  <Alberth> it does
10:16:25  <planetmaker> :-) First time it has not time advantage yet :-) But... that's expected :-)
10:16:26  <Alberth> builds just take a 1-4 minutes
10:16:45  <Alberth> depending on how many new sprites you added
10:17:17  <planetmaker> I'll try to change things so that it builds as a separate base graphics set
10:17:30  <planetmaker> like I'll also rename the main pnml files to zbase_XXX
10:17:53  <planetmaker> and change the grfID to \FFzB
10:18:34  <Alberth> good idea, let's keep things separate for now
10:20:33  <planetmaker> do makefiles have a catch-all target? Is that "all"?
10:23:14  <planetmaker> he... zbase_base.grf takes longer than all other 5 grfs combined
10:24:24  <Hirundo> !logs
10:24:26  <Hirundo> @logs
10:24:27  <Webster> #openttdcoop IRC webstuff - IRC Log Viewer -
10:25:16  <Alberth> planetmaker: there are some hooks for VCS systems, so you can automagically pull stuff from them when needed, perhaps that's what you want
10:26:03  <Alberth> 'all' is often the first target, which is the default target of make
10:43:01  <planetmaker> nice. That's *so* much faster now :-)
10:46:25  * Alberth nods
10:47:28  <Alberth> even the non-cached version are faster
10:55:26  <Brot6> zBaseBuild - Revision 35:e7901db6cda6: Change: Use the set's own name zBase and adjust build config ... XplanetmakerX @
10:55:27  <Brot6> zBuild - Revision 5:c85f0c2f4925: Add: Prepare for building by the CF XplanetmakerX @
10:59:18  <Brot6> OpenGFX+ Trains - Revision 748:2be8d34a3c59: Feature: Use 32bpp early wagon sprites for standard and... XXotic750X @
11:02:12  <Brot6> FIRS Industry Replacement Set - Revision 2858:ec446cbc8e8f: Add: Afrikaans language XAlberthX @
11:02:42  <Brot6> FIRS Industry Replacement Set - Revision 2859:04ddf24bc6ea: Fix: String name in Swedish language fil... XAlberthX @
11:05:26  <Brot6> NewGRF Meta Language - Feature Request #2962: handling of substitute property XHirundoX @
11:08:49  <Brot6> zBuild - Bug #4094 (New): DevZone compile failed XcompilerX @
11:09:16  <planetmaker> of course it failed! I aborted it...
11:09:26  <Alberth> :)
11:09:48  <Brot6> zBuild - Bug #4094 (Rejected): DevZone compile failed XcompilerX @
11:09:48  <Brot6> zBuild - Bug #4094 (Rejected): DevZone compile failed XplanetmakerX @
11:12:33  <Brot6> zBuild - Revision 6:241b797c9c0a: Add: relative paths for CF XplanetmakerX @
11:14:15  <Brot6> zBuild - Bug #4095 (New): DevZone compile failed XcompilerX @
11:14:29  <planetmaker> really?
11:15:12  <Alberth> nah, it's just testing you
11:15:18  <Brot6> zBuild - Bug #4095 (Rejected): DevZone compile failed XcompilerX @
11:15:18  <Brot6> zBuild - Bug #4095 (Rejected): DevZone compile failed XplanetmakerX @
11:15:58  <Alberth> just keep the bug until it works, so it does not spam more issues?
11:16:09  <Alberth> or does that not work?
11:16:25  <planetmaker> maybe it doesn't then spam more. Worth a try :-)
11:17:07  <planetmaker> I don't like:
11:17:07  <planetmaker> 2012-07-25 11:14:03+00:00: clone zbuild
11:17:07  <planetmaker> 0 files updated, 0 files merged, 0 files removed, 0 files unresolved
11:17:35  <planetmaker> but it starts building. And issues the error before. I wonder why :-)
11:20:35  <Alberth> looks a bit empty :p
11:20:38  <planetmaker> <-- quite so, Hirundo
11:20:41  <Webster> Title: Transport Tycoon Forums View topic - Proposal: check for "last vehicle in chain" in var61 (at
11:21:40  <planetmaker> I'm glad that CETS uses 0 as normal. I don't like special cases without need :-)
11:21:53  <planetmaker> Thus CETS gives a compelling counter-argument
11:22:09  <planetmaker> especially as it's already "in the wild" (though not on bananas)
11:22:43  <planetmaker> thus also "don't break existing newgrfs without very good reason" also applies
11:23:23  <Hirundo> Even without CETS it's just stupid, nfo variables should be as generic as possible
11:23:24  <Hirundo> Creating a special case for MB makes life harder for everyone else
11:23:37  <planetmaker> absolutely
11:24:22  <Alberth> special cases for anyone are bad
11:25:33  <planetmaker> if a return value has a certain meaning, every special values complicate stuff needlessly
11:29:45  <Brot6> zBuild - Bug #4096 (New): DevZone compile failed XcompilerX @
11:34:28  <Brot6> zbuild: compile of r7 still failed (#4096) -
11:34:29  <Brot6> zbuild: compile of r7 still failed (#4095) -
11:35:07  <Brot6> zBuild - Revision 7:3b8e30aa8f4e: Fix (r6): typo XplanetmakerX @
11:37:05  <planetmaker> hm, doesn't yet fail where it failed last time :-)
11:38:37  <Alberth> you're making progress :)
11:39:16  <planetmaker> in typing filenames correctly :-P
11:48:38  <planetmaker> Ammler: the CF was in a state (before I started) that it thought that it's still running
11:48:51  <planetmaker> Ammler: that seems to be happening quite often as of late...
11:49:42  <planetmaker> Ammler: maybe force-kill and reset the CF  when the crontab asks it to start a new run?
11:50:03  <planetmaker> Ammler: not nice... but at least would somewhat cure the symptoms (though not the illness itself)
11:57:40  <dihedral> planetmaker, what to you use for the CF?
11:58:07  *** ODM has joined #openttdcoop.devzone
11:58:11  <planetmaker> it's a shell script, dihedral
11:58:19  <dihedral> \o/
11:58:33  <dihedral> ever thought of hudson / jenkins? :-P
11:58:49  <planetmaker> yes... but this is build on the way the build service of suse works
12:06:47  <Alberth> Hirundo:   nml r1941,  nml r1948
12:06:51  *** ODM has quit IRC
12:07:21  <Alberth> first build is always without cache files
12:07:32  <Alberth> it shows VERY nice speed improvements
12:12:33  <planetmaker> the CF still builds zBuild...
12:14:43  <Alberth> that does not look wrong :)
12:16:10  <planetmaker> indeed :-)
12:24:32  <Brot6> zbuild: compile of r7 still failed (#4096) -
12:24:39  <planetmaker> meh :-(
12:25:23  <Alberth> :(
12:26:29  <Alberth> Hmm, I have more obiwans, it seems
12:30:07  <Hirundo> Alberth: nice, thanks!
12:30:41  <Alberth> yw
12:32:17  <planetmaker> and... I don't exactly understand where it fails even...
12:32:34  <planetmaker> it build successfully. but seems to fail afterwards
12:32:39  <planetmaker> *built
12:37:37  <Brot6> OpenGFX+ Trains - Revision 749:be5f8e2b6da0: Bug #4075: Fixed pax, mail and valuables not being limi... XXotic750X @
12:37:37  <Brot6> OpenGFX+ Trains - Bug #4075 (Closed): Wagon speed limits XXotic750X @
12:40:19  <planetmaker> Ammler: can you help me?
12:40:50  <Brot6> Grapes - Revision 161:3b1ae26fa155: Cleanup: remove unused imports XdihX @
12:40:51  <Brot6> Grapes - Revision 162:25f40818a26a: Cleanup: braces XdihX @
12:40:51  <Brot6> Grapes - Revision 163:da164990c87b: Cleanup: unused private variable XdihX @
12:41:04  <planetmaker> Ammler: mind that I edited around in grf.spec to allow for the path to the Makefile in the subrepo. But... I don't see where it fails
12:44:13  <Brot6> FISH - Revision 823:ffcdb98f9196: Change: add some additional coastal tankers (also a joke) XandythenorthX @
12:46:20  <Brot6> NewGRF Meta Language - Bug #4097 (New): DevZone compile failed XcompilerX @
12:46:26  <planetmaker> ach?
12:47:06  <planetmaker> oh. regression failure :-)
12:47:10  <planetmaker> ^ Hirundo
12:48:24  <Brot6> fish: update from r781 to r823 done -
12:48:30  <planetmaker> uhm
12:48:37  <planetmaker> I did not ask for that...
12:49:43  <planetmaker> hm...
12:50:17  <Brot6> OpenGFX+ Trains - Revision 750:be46981ece18: Merge with default XXotic750X @
12:50:54  <planetmaker> regression does NOT seem to fail locally
12:51:42  <Alberth> python version?
12:52:50  <planetmaker> 2.7 vs. 2.6.1
12:53:14  <Brot6> NewGRF Meta Language - Bug #4097: DevZone compile failed XplanetmakerX @
12:55:11  <Brot6> OpenGFX+ Trains - Revision 751:8e6b2de3c315: Fix: Matched wagon speeds to default branch, 160 km/h XXotic750X @
12:57:33  <Alberth> nml regression works for me with python 2.7.3
13:02:56  <planetmaker> [  228s] [133/219] installing python-2.7.2-7.17.1 from logs
13:03:02  <planetmaker> so 2.7.2
13:35:22  <Brot6> firs: update from r2857 to r2859 done -
14:02:04  <Brot6> OpenGFX+ Trains - Bug #2284: Make use of special sprites for the purchase list XXotic750X @
14:41:14  <Brot6> OpenGFX+ Trains renders - Revision 40:ffb00a1f51a1: Update: Modernised monorail goods and rendered XXotic750X @
14:41:14  <Brot6> OpenGFX+ Trains renders - Revision 41:a745e8dfa26f: Add: Monorail early goods and rendered XXotic750X @
14:41:41  <Brot6> OpenGFX+ Trains renders - Revision 42:1f946b6564af: Add: Monorail working templates XXotic750X @
15:10:39  <Brot6> FIRS Industry Replacement Set - Revision 2860:fc1df1557dd6: Change: Update to Afrikaans translation ... XAlberthX @
15:20:40  <Brot6> firs: update from r2857 to r2860 done -
15:40:38  *** frosch123 has joined #openttdcoop.devzone
15:50:12  <planetmaker> quak
15:50:30  <frosch123> moin
15:50:40  <frosch123> quaks are unusual in this channel :)
15:50:56  *** Xotic750_ has joined #openttdcoop.devzone
15:51:25  <planetmaker> hehe, yes :-)
15:54:49  <Alberth> non-quak
15:56:46  <Hirundo> planetmaker: Is output of the failed regression test available somewhere?
15:58:27  <Alberth> Hirundo: follow the link in the issue:
15:58:45  <Alberth> it doesn't say much though :(
15:59:35  <Hirundo> I'd really like the .grfs that failed to match
15:59:43  <planetmaker> Hirundo: I don't have really more than what albert just linked. But I can give it a try to build locally on the server
16:01:03  <Hirundo> planetmaker: That'd be much appreciated
16:01:18  <Hirundo> Also, be sure to clear any cache cache files that might be around
16:02:27  <planetmaker> Hirundo: it always builds in a completely new chroot
16:06:49  <Brot6> OpenGFX+ Trains - Revision 752:7c3285b063ae: Add: 32bpp sprites for monorail goods early XXotic750X @
16:07:37  <Hirundo> Alberth: Aggregated profile data:
16:08:20  <planetmaker> hm, peculiar
16:08:38  <planetmaker> on the test VM, also with python 2.7 regression works...
16:08:43  <Alberth> aka LZ77 is the domnant time eater :)
16:09:10  <Hirundo> What used to take 56% now takes 3%
16:09:21  <Brot6> OpenGFX+ Trains - Revision 753:be68ca3c94ee: Update: Added monorail goods early models to the list XXotic750X @
16:09:21  <Hirundo> The other 44% (LZ77) now takes 39%
16:10:46  <Hirundo> planetmaker: Is the python 2.7.X sub version the same?
16:11:20  <planetmaker> need to check. this python doesn't report the sub version with --version. Why ever
16:11:58  <Alberth> -V
16:12:02  <Hirundo> There may be bugs in some python versions, else 2.7.1 and 2.7.2 would not exist
16:12:11  <planetmaker> Alberth: neither
16:12:38  <Alberth> planetmaker: you have strange pythons
16:12:46  <planetmaker> suse has. yes
16:12:47  <Hirundo> My python --version says 2.6.6
16:12:56  <planetmaker> mine says 2.6.1 here
16:14:44  *** ODM has joined #openttdcoop.devzone
16:17:04  <planetmaker> but I need to verify that it's the same as in the chroot
16:19:37  <planetmaker> hm... slightly different
16:25:00  <Brot6> OpenGFX+ Trains - Revision 754:37ea289de3e7: Update: Modernised 32bpp sprites for monorail goods XXotic750X @
16:31:19  <Hirundo> what is 'slightly different'?
16:33:48  <planetmaker> compare the version from the paste with [  107s] [62/219] installing libpython2_7-1_0-2.7.2-7.17.1
16:34:01  <planetmaker> and [  139s] [95/219] installing python-base-2.7.2-7.17.1
16:37:52  <Brot6> nml: compile of r1948 still failed (#4097) -
16:42:46  <Hirundo> So chroot seems to have version 2.7.2
16:43:33  <Hirundo> Could you inspect /usr/share/doc/packages/python/README, which is where the suse RPM claims to store the readme?
16:45:12  <planetmaker> Hirundo: not on the CF... the chroot is removed when the CF is done
16:45:33  <planetmaker> but I'll try to make the regress Makefile such that it prints the diff
16:46:03  <Hirundo> The CF has libpython2_7-1_0-2.7.2-7.17.1, right?
16:46:22  <Hirundo> Since that contains the substring '2.7.2', I assume that it's 2.7.2
16:46:30  <planetmaker> that's what I assume, too
16:48:59  <Hirundo> Your test environment (paste bin) says it's version 2.7-9.10.2, which led me here:
16:49:01  <Webster> Title: RPM OpenSuSE 11.X python 2.7 i586 rpm (at
16:52:14  <planetmaker> Hirundo: output like ? I'll commit to obtain that. Ok?
16:52:49  <planetmaker> actually... the diff itself would suffice, would it?
16:52:58  <planetmaker> hm
16:53:07  <Hirundo> The differences are in .grf files, not in the nf
16:53:07  <Hirundo> o
16:53:55  <Hirundo> If you have the failed output/006_vehicle.grf, I can binary diff that locally
16:54:15  <Hirundo> (006_vehicle.grf is the first file that contains real sprites, 001-005 don't)
16:54:41  <planetmaker> ah, interesting. Ok. Let's tackle the specs / publishing part then
16:58:32  <Alberth> xxd bla.grf > bla_hex.txt ?
16:58:51  <Alberth> or the nfo code instead?
17:01:31  <planetmaker> yes... the question is where in the build specs file that is done :-) I'm not too familiar with it. I'd just publish the regression folder
17:11:19  <Brot6> NewGRF Meta Language - Revision 1949:71d15c4b1fab: Change: Publish also the generated grf files from... XplanetmakerX @
17:19:39  <Brot6> nml: compile of r1949 still failed (#4097) -
17:20:00  <Alberth> A locally built Python 2.7.2 works for test 6
17:20:42  <planetmaker> he... damn, it should be 2 grfs... but they overwrite eachother :-)
17:21:06  <Alberth> you just need the generated one :)
17:21:25  <planetmaker> 2 are generated
17:21:32  <planetmaker> with different nml settings
17:21:38  <Alberth> ah, ok
17:21:43  <Hirundo> Only difference is cache/nocachee
17:22:13  <Hirundo> Which is a rather moot point for the CF
17:23:19  <Alberth> I wonder why it is not reproducable with a local 2.7.2
17:23:36  <Alberth> planetmaker: server is 32bit or 64bit?
17:24:30  <Brot6> OpenGFX+ Trains renders - Revision 43:06e349200ce5: Update: Modernised monorail refrigerated wagons XXotic750X @
17:24:30  <Brot6> OpenGFX+ Trains renders - Revision 44:d8dfb73ecc60: Add: Monorail early wagons XXotic750X @
17:25:16  <planetmaker> Linux dev 2.6.32-7-pve #1 SMP Mon Feb 13 07:33:21 CET 2012 x86_64 x86_64 x86_64 GNU/Linux
17:25:27  <Brot6> planetmaker: #1 is "Feature #1: Depot switches - Infrastructure Sharing - #openttdcoop Development Zone"
17:26:39  <Hirundo> The issue seems somewhere in the sprite cropping.. the first difference is in the ysize of some real sprite
17:27:20  <Alberth> hmm, I also have a 64 bit machine
17:28:47  <Brot6> NewGRF Meta Language - Revision 1950:0d39425fac77: Fix (r1949): Publishing one of the generated grfs... XplanetmakerX @
17:28:48  <Hirundo> grfcodec -d yields 'invalid sprite'
17:29:38  <planetmaker> hm
17:37:15  <Hirundo> grfcodec -d yields invalid sprite for regression/expected as well :S
17:38:47  <Brot6> nml: compile of r1950 still failed (#4097) -
17:43:08  <Brot6> OpenGFX+ Trains - Revision 755:6dc22d1dc2a4: Add: Monorail 32bpp early refrigerated wagon sprites an... XXotic750X @
17:44:15  <Brot6> OpenGFX+ Trains renders - Revision 45:9d45305c74aa: Update: Render of modernised monorail refrigerat... XXotic750X @
17:54:32  <Brot6> OpenGFX+ Trains - Revision 756:476586a2bb62: Update: 32bpp sprites for modernised monorail refrigera... XXotic750X @
17:55:50  <Alberth> good night all
17:56:03  <planetmaker> Hirundo: do you have the newest grfcodec?
17:56:08  *** Alberth has left #openttdcoop.devzone
17:56:09  <planetmaker> oh, good night, Alberth :-)
17:56:21  <Hirundo> planetmaker: probably not, I should get a nightly
17:56:27  <planetmaker> yes, you should
17:56:48  <planetmaker> even older nightlies might choke on valid real sprites in container v2
17:59:23  <Hirundo> openttd seems to handle both versions rather fine
17:59:32  <Hirundo> I'll continue looking at it later, away for now
18:01:40  <Brot6> nml: compile of r1950 still failed (#4097) -
18:01:58  <Brot6> OpenGFX+ Trains - Revision 757:9a932c52747b: Merge with default XXotic750X @
18:15:08  <planetmaker> Hirundo: the grf which differs now is published
18:40:11  *** Xotic750_ has quit IRC
19:35:48  <Brot6> zBase - Bug #4088: wrongly sloped rail and tunnel entrance XzephyrisX @
19:45:00  <Rubidium> aren't zbase pushes announced here?
19:47:11  <planetmaker> they are
19:49:24  <Rubidium> where's the last push then?
19:51:10  <Rubidium> damn... rebuilding with that many changed images takes a while
19:55:56  <Brot6> zBase - Bug #4098 (New): Arctic maglev too green XRubidiumX @
20:04:03  <Brot6> zBase - Revision 32:0819180fd1c4: Add: Tubular road, rail, monorail and meglev bridges. XzephyrisX @
20:08:03  *** frosch123 has quit IRC
20:51:19  <Brot6> zBase - Revision 33:83c4dc4db102: Add: Sawmill. XzephyrisX @
21:04:46  *** ODM has quit IRC
21:06:35  <Brot6> FISH - Revision 824:83fdf88b4b29: Fix: missing pngs XandythenorthX @
21:06:35  <Brot6> FISH - Revision 825:39bd4f9d9e9b: Change: add two inland tankers XandythenorthX @

Powered by YARRSTE version: svn-trunk