Log for #openttdcoop.devzone on 30th November 2011:
Times are UTC Toggle Colours
00:17:11  *** Zuu has quit IRC
06:51:39  *** JVassie has joined #openttdcoop.devzone
06:56:56  *** JVassie_ has joined #openttdcoop.devzone
07:01:08  *** JVassie has quit IRC
07:01:37  *** JVassie has joined #openttdcoop.devzone
07:04:59  *** JVassie_ has quit IRC
07:09:17  *** Zuu has joined #openttdcoop.devzone
07:25:36  *** JVassie has quit IRC
08:10:12  *** ODM has joined #openttdcoop.devzone
08:15:56  *** Zuu has quit IRC
12:54:28  *** ChillCore has joined #openttdcoop.devzone
12:54:47  <ChillCore> Good afternoon all.
12:55:46  <planetmaker> hi ChillCore
12:55:56  <ChillCore> Anyone knows why the compile farm was not triggered this night?   I can see no recent activity here
12:56:06  <ChillCore> hello planetmaker
12:57:04  <ChillCore> Do I need to push before a certain hour in order for the CF to do its thing in the wee hours that follow?
12:57:58  <planetmaker> no
12:58:36  <planetmaker> My only guess could be that you pushed the exact same tag as was present already 3x
12:58:51  <planetmaker> thus no need for update was recognized *whereever*
12:58:57  <planetmaker> and you didn't want to have nightly compiles ;-)
12:59:17  <planetmaker> but I'm not sure
12:59:53  <ChillCore> No nightlies because the test builds do so against the current revision. Which will fail.
13:00:58  <ChillCore> Ammler said that the tag is only for detecting against which revision the repo should be build and that the CF is triggered by new tags regardless of which one.
13:02:53  <Ammler> planetmaker: using same tag should be no issue
13:03:04  <Ammler> it uses the hash to detect
13:03:30  <ChillCore> I checked my repo and all seems fine, i renamed a patch, the right way hg qrename from trunk, and the series file is correct too
13:03:35  <ChillCore> hi Ammler
13:03:48  <Ammler> ChillCore: let me check...
13:04:17  <ChillCore> Thank you as I am puzzled. :)
13:05:12  <Ammler> hmm, maybe there is indeed an issue with tag and hash
13:07:03  <ChillCore> hmm, it was not an issue before ... last thing that happened in regards of my repo is planetmaker fixing the continues building without me pushing.
13:07:41  <ChillCore> before me pushing yesterday that is.
13:07:59  <Ammler> hmm, it has same hash, silly
13:09:21  <Ammler> I use the hash of the main repo, why?
13:10:31  <ChillCore> ... my fault?  ^^^ question to yourself again? :P
13:11:01  <Ammler> :-P
13:11:23  <Ammler> well, maybe you have a clue, why I used the hash of the mainrepo, instead the ash of the mq
13:11:43  <Ammler> anyway, I changed that and triggered a check
13:11:49  <Ammler> we will see, how that will work
13:12:26  <Ammler> ChillCore: the devzone itself should create the release right after you push the tag
13:12:42  <Ammler> if not, you can complain, you do not need to wait the night
13:13:26  <Ammler> the openttd cf just does the build, if devzone changes the repo
13:13:49  <ChillCore> OK ... So I commit something and tag? right? that should not take too long ... plenty of coding style issues to fix. :P
13:14:02  <Ammler> <-- this repo
13:14:22  <Ammler> well, it is building right now
13:14:33  <Ammler> and as you see on the .mq repo, it is changed 2 mins ago
13:14:46  <Ammler> so this night, you will have a new bundle for sure
13:16:35  <ChillCore> Ah Ok ... then I continue as usual and instead of announcing the new binaries I'll wait until they are there (or not) to do so. ;)
13:17:06  <Brot6> Chill's PatchPack - Bug #3299 (New): DevZone compile failed (compiler) @
13:18:15  <ChillCore> hmmm. checking.
13:19:32  <ChillCore> I'm pretty sure that is not on my side.
13:19:57  <ChillCore> + something
13:24:25  <Ammler> hmm, git binary diffs are not supported
13:24:27  <Ammler> hmm
13:24:28  <Ammler> hmm
13:24:30  <Ammler> hmm
13:24:51  <Ammler> planetmaker: what does it need to support git diffs?
13:24:58  <Ammler> does it need git?
13:25:44  <planetmaker> no
13:25:48  <Ammler> or whoever knows it
13:25:49  <planetmaker> just a change in .hgrc
13:25:58  <Ammler> oh,
13:26:07  <Ammler> you mean, that is not just for creating?
13:26:12  <planetmaker> or support where in what way?
13:26:27  <planetmaker> creation just needs a change of .hgrc
13:26:37  <Ammler> yes, and appling?
13:26:43  <planetmaker> patching might need svn 1.7 patch
13:26:46  <planetmaker> not sure
13:27:01  <Ammler> svn?
13:27:14  <Ammler> you confuse something?
13:27:19  <planetmaker> well, tell me, what do you want to do? Apply a git patch?
13:27:24  <planetmaker> what doesn't work where?
13:27:29  <Ammler> hg push
13:27:36  <planetmaker> eh?
13:27:36  <Ammler> check the chillpp log
13:27:53  <Ammler> qpush*
13:28:20  <Ammler> hmm
13:29:20  <planetmaker> ah, I see.
13:29:34  <planetmaker> Yes, that definitely is a too old version of patch
13:29:44  <planetmaker> "My" patch works just fine with binary git patches
13:29:51  <Ammler> we use patch?
13:29:53  <Ammler> where?
13:30:20  <Ammler> the patches shold be applied via qpush, afaik
13:31:11  <Ammler> ah, for the svn stuff
13:32:00  <Ammler> there is a patch tool which works with binary diffs, are you sure?
13:32:22  <Ammler> patch 2.6.1 is instaled
13:33:44  <Ammler> I fear, we need to skip the binary stuff for the svn patch
13:34:01  <planetmaker> eh? That's not true
13:34:21  <planetmaker> working with binary diffs *does* work locally
13:34:33  <planetmaker> Thus *something* is just not working as it could on the server
13:35:26  <planetmaker> my current openttd patch queue happily diffs openttd.grf and all is fine with qpop and qpush
13:36:06  <Ammler> yes, that is not patch
13:36:19  <Ammler> hg has no issue with binary diffs
13:36:23  <planetmaker> well. But obviously it fails. Whatever the internal thing is
13:36:24  <Ammler> we speak about patch
13:36:39  <Ammler> patch does not work with binary diffs, afaik
13:36:58  <planetmaker> that's why I said: svn 1.7 might work. If I remember reading correctly
13:37:07  <Ammler> patch is not part of svn
13:37:12  <Ammler> patch is patch
13:37:16  <planetmaker> it brings an updated patch.
13:37:27  <planetmaker> iirc you err
13:37:46  <Ammler> iirc you confuse hg patch tool with native patch tool :-P
13:39:14  <Ammler> we use tumbleweed to build on devzone, so there is hardly newer software available :-P
13:40:13  <planetmaker> I'm quite sure I don't. But back to page 1:
13:40:21  <Ammler> :-)
13:40:24  <planetmaker> a) what does the CF use to patch the thing there?
13:40:34  <Ammler> patch 2.6.2
13:40:39  <Ammler> 1*
13:40:44  <planetmaker> why? why not use hg?
13:41:15  <Ammler> hmm
13:41:21  <planetmaker> like hg qpush -a && hg qfinish -a
13:41:32  <Ammler> that works on a svn repo?
13:41:38  <planetmaker> no, doesn't
13:41:50  <Ammler> you know, the whole point here is to create a svn patch
13:42:02  <planetmaker> what for?
13:42:14  <planetmaker> why?
13:42:15  <Ammler> for the poor svn guys
13:42:30  <Ammler> but as you ask, my suggestion to skip it might be good then
13:42:39  <planetmaker> yes ;-)
13:42:46  <planetmaker> provide git style patch.
13:42:50  <planetmaker> and be done ;-)
13:42:51  <Ammler> that is there
13:42:53  <Ammler> already
13:42:59  <planetmaker> exactly
13:43:01  <Ammler> svn patch is additional service
13:43:36  <planetmaker> they should use proper tools ;-)
13:43:57  <Ammler> well, youself speak sometimes otherwise
13:44:11  <Ammler> e.q. we provide .gz archives for opengfx
13:45:10  <Ammler> sometimes, we invest more effort to make it nice for others, sometimes you tell others they should use proper tools, what now? :-P
13:45:42  <Ammler> anyway, binary diffs doesn't work, and you just confirmed it now
13:46:02  <Ammler> so we remove the whole svn part
13:48:46  <planetmaker> "Subversion < 1.6 doesn't have patch support. It looks like Subversion 1.7 will allow applying patches and the git/hg extensions to unified diff are on our TODO list."
13:48:56  <planetmaker> so who has an issue has a too old svn. Issue solved
13:49:22  <Ammler> "on our TODO" sounds not like it in
13:49:47  <planetmaker> Dec 22 '09
13:50:11  <Ammler> yes, svn 1.7 is not out yet, is it?
13:50:22  <Ammler> at least not with suse
13:50:52  <planetmaker> <-- please read
13:50:53  <Webster> Title: Apache Subversion 1.7 Release Notes (at
13:50:58  <planetmaker> And if suse doesn't have it, it's slow
13:51:00  <planetmaker> very slow
13:51:03  <Ammler> 1.6.17
13:51:08  <planetmaker> even macports has it already
13:51:27  <planetmaker> so one can have obviously much newer software ;-)
13:51:41  <Ammler> I am sure, there is exotic factory repo with it
13:52:06  <Ammler> but there might be a reason suse does not use it
13:52:16  <Ammler> Rubidium: which svn does debian use?
13:53:39  <ChillCore> I found a package in synaptic ... could be a solution, maybe not. description here:
13:54:03  <Ammler> lol
13:54:14  <Ammler> ChillCore: all fine :-P
13:54:37  <planetmaker> oh, Ammler
13:54:38  <Webster> Title:" target="_blank"> Install package home:stsp / subversion17 (at" target="_blank">
13:54:42  <Yexo> ChillCore: but using that you still can't create svn diffs
13:54:56  <Ammler> planetmaker: as said, some exotic factory repo might have it
13:55:13  <Ammler> I don't think, you need to search in home repos
13:55:20  <Ammler> check devel:tools:scm first
13:55:31  <ChillCore> Yexo: I have never used it ... I just knew it existed and read the discription. ;)
13:56:51  <planetmaker> well, SVN 1.7 is already 1.5 months old, Ammler
13:56:56  <Brot6> chillpp: update from r22555 to r22555 done (15 warnings) -
13:57:00  <Ammler> ah, just
13:57:09  <Ammler> I thought it is from 2009?
13:57:09  <planetmaker> and we already have 1.7.1 ;-)
13:57:18  <planetmaker> the message I quoted
13:57:23  <planetmaker> well, nvm.
13:57:33  <Ammler> well, then suse factory is also on 1.7
13:57:47  <Ammler>
13:57:48  <Webster> Title: Sources of subversion (Project devel:tools:scm:svn) - openSUSE Build Service (at
13:58:17  <Ammler> but a tool like svn needs a bit more time than 1 month to get in standard repo
13:58:38  <planetmaker> yes, but I thought we use "bleeding edge"?
13:59:38  <planetmaker> thus using an old svn kind caught me by surprise :-)
13:59:42  <Ammler> well, "bleeding working edge" :-P
14:01:24  <planetmaker> 1.5 months is half the release cycle of hg...
14:01:28  <Ammler> planetmaker: anyway, it would not solve our issue
14:01:46  <Ammler> so just removing svn might be easiest
14:02:02  <Ammler> or does tortoisesvn also have svn 1.7 already?
14:02:02  <planetmaker> it would. As it understands the unified diff. Thus no "special" svn diff is needed any longer
14:02:08  <planetmaker> probably
14:02:16  <planetmaker> as tortoise is just an add-on
14:02:24  <Ammler> well, if it will support it, it is enabled quite fast
14:02:31  <Ammler> also adding the factory repo would be easy
14:02:46  <Ammler> as we use already factory for openttd tools
14:03:06  <Ammler> ChillCore: do you want svn patch?
14:04:54  <Ammler> ChillCore: also should you not patch the findversion to have a nicer version string?
14:05:01  <Ammler> (and version.vbs)
14:06:03  <ChillCore> bof ... I am using hg myself and those that really still want svn patches can change the patchfile manually or patch with p1 from the console instead of using tortoisesvn.
14:06:32  <planetmaker> btw... patching file Makefile.msvc
14:06:32  <planetmaker> Hunk #1 FAILED at 1.
14:06:32  <planetmaker> 1 out of 2 hunks FAILED -- saving rejects to file Makefile.msvc.rej
14:06:33  <Brot6> planetmaker: #1 is "Infrastructure Sharing - Feature #1: Depot switches - #openttdcoop Development Zone"
14:07:05  <ChillCore> I could change the version string if you really want that ... You just need to tell me how to do it properly instead of hacking rev.cpp ;)
14:07:12  <Rubidium> Ammler: still a 1.6 version, though I'm not sad about it. 1.7 caused some mayhem at work :(
14:07:36  <planetmaker> hm, it does / did? :S
14:08:19  <ChillCore> hgrevision - chippvxx in the titlebar would be nice ... :)
14:08:50  <ChillCore> ^^^ Ammler
14:08:54  <planetmaker> might be too long
14:10:16  <ChillCore> I have no clue on how long the string may be ... but I have seen similar before (maybe not in lenght)
14:10:38  <Ammler> I would rather use the svn rev as the hg hash
14:11:02  <Ammler> chillpp22456-n
14:11:18  <ChillCore> could do, if I change the revision for each build then.
14:11:28  <Ammler> that is -n for :-P
14:12:21  <ChillCore> I need really the chipp verion in there too if I change to svn revisions. Else muchos trouble ahead
14:12:26  <Ammler> everyone can work with svn revs, but hg hash is only useable for same hg repo
14:12:59  <Ammler> ChillCore: yes, again -n   :-P
14:13:37  <ChillCore> I missed that part reading it the the first time :P
14:14:13  <planetmaker> Ammler, wrong: the hash are unique. The hg rev is non-unique
14:14:31  <Ammler> planetmaker: svn rev
14:14:51  <planetmaker> that's why the hg hash is just as good ;-)
14:14:54  <Ammler> which is also in the git repo and hg repo
14:15:03  <planetmaker> (or even 'better' than the svn rev as suggested)
14:15:06  <Ammler> how useful is the hg hash, if you work with git?
14:15:23  <planetmaker> how useful is the svn rev?
14:15:48  <Ammler> he?
14:15:58  <Ammler> you can update every repo to that svn rev
14:16:17  <planetmaker> neither hg nor git exactly know the svn rev
14:16:17  <Ammler> as the svn rev is part of those "converts"
14:16:29  <Ammler> of course they do
14:16:54  <planetmaker> in a hg checkout you just have the svn revs in the commit messages
14:18:28  * ChillCore goes wheee: 15 warnings and he is only getting 3 himself. :( Anyway he is moving to a 64bit compiler soon hoping that solves this issue.
14:18:50  <planetmaker> and all you can easily do is like hg log -k 'rXXXXX'
14:18:56  <planetmaker> with XXXXX being the svn rev
14:20:42  <Ammler> planetmaker: but you can
14:20:55  <Ammler> tell me how you update the svn repo to an hg ash
14:21:48  <planetmaker> same way. hg log -rHASH
14:21:59  <planetmaker> and then svn up the rev I got from the commit message
14:24:19  <Ammler> hg log on a svn rev?
14:24:20  <Ammler> how?
14:24:28  <Ammler> svn repo*
14:24:45  <Ammler> :-(
14:24:48  <Ammler> oh
14:25:17  <Ammler> wrong smilie, well, you know you failed :-P
15:01:11  <Brot6> TransRapid Track Set - Bug #3300 (New): doesn't work on ttdp (foobar) @
15:06:00  <ChillCore> I'll be away from the computer for a little while but I'll leave the window open so I can read back if needed.
16:59:42  <Brot6> TransRapid Track Set - Bug #3300: doesn't work on ttdp (foobar) @
17:07:10  * ChillCore is back but forgot to say so.
17:08:11  <Ammler> imagine, everyone would tell when he leaves and returns :-P
17:08:50  <ChillCore> true ... I'll keep quit now. ;)
18:41:57  <Rubidium> ChillCore: regarding the OpenTTD compile farm; it will only perform rebuilds when the .mq has been changed which is done by #openttdcoop's server. Not sure when that happens normally, but it might be useful to schedule that some 15 minutes before the OpenTTD CF runs
18:44:00  <Rubidium> in any case, it looks like it changed it around 13:45 today, so the next OpenTTD CF run might build the new version
19:03:12  <Brot6> clientpatches: compile of r23380 still failed (#2964) -
19:05:01  <Brot6> serverpatches: compile of r23380 still failed (#2966) -
19:07:04  <Brot6> 32bpp-ez-patches: compile of r23380 still failed (#2446) -
19:10:54  *** JVassie has joined #openttdcoop.devzone
19:13:08  *** frosch123 has joined #openttdcoop.devzone
19:16:09  <frosch123> hmm, i would have liked to congrat pm on his work...
19:16:27  <frosch123> but i need to revert the first page i looked at :s
19:19:27  <ChillCore> Rubidium: Thank you for the answer. The change this afternoon was done manually by Amm*ler so it will take another push (after tomorrow morning) to see if things are really fixed now. No biggie really; I just find it a bit annoying/embarrassing to keep bothering you all like this.
19:19:29  *** andythenorth has joined #openttdcoop.devzone
19:19:47  <andythenorth> planetmaker: Ammler how do I get bundles server to build BANDIT?
19:19:49  *** hanf has joined #openttdcoop.devzone
19:19:58  <andythenorth> apologies if I've missed a really obvious doc or setting
19:20:01  <ChillCore> Rubidium: I'll keep those 15 minutes in mind. Thank you
19:21:22  <Ammler> andythenorth: check .devzone files
19:21:46  <Ammler> if you have nothing special, you simply need to define type and enable
19:22:05  <Ammler> echo "nml" > .devzone/build/type
19:22:19  <Ammler> touch .devzone/build/nightlies/enable
19:22:25  * andythenorth forget ls -la
19:22:31  <andythenorth> specifically the 'a'
19:22:36  <Ammler> touch .devzone/build/releases/enable
19:23:54  <Ammler> Rubidium: yes, the devzone failed to apply the patches because there was no change of svn rev, changed that to hash now...
19:24:26  <Ammler> hmm, not svn rev, tag name
19:24:48  <andythenorth> Ammler: echo "nml" > .devzone/build/type fails
19:24:56  <andythenorth> .devzone/build/type: No such file or directory
19:27:58  *** JVassie_ has joined #openttdcoop.devzone
19:29:30  <Brot6> DACH Trains - Revision 26:895d1121c97a: fixed white pixels in KISS and FLIRT graphics (officercrockey) @
19:30:57  *** JVassie_ has quit IRC
19:31:02  <andythenorth> Ammler: can I just copy .devzone from another repo?
19:31:05  <andythenorth> I have them there
19:33:18  <Brot6> DACH Trains - Bug #3174 (Closed): white pixels (officercrockey) @
19:35:19  *** JVassie has quit IRC
19:36:36  <Ammler> andythenorth: obviously you need to create those directories
19:36:52  <Ammler> andythenorth: copy from another repo just if you do not have other files
19:36:58  <Ammler> e.g. do not copy from firs
19:37:22  <Ammler> I would just run the commands I pasted
19:37:30  <Ammler> (with creating the dirs)
19:45:27  *** JVassie has joined #openttdcoop.devzone
19:47:30  <andythenorth> I have two backup.push files - one in repo root and one in .devzone
19:47:33  <andythenorth> does that matter?
20:11:47  <Brot6> Japanese Trains - Revision 29:ec78d99db4d6: Add 9700 (steam) and update 9600 (steam) (dandan) @
20:11:47  <Brot6> BANDIT - Revision 23:fc30d52753bf: Add: .devzone files (andythenorth) @
20:12:30  *** JVassie_ has joined #openttdcoop.devzone
20:19:21  *** JVassie has quit IRC
20:21:15  *** Zuu has joined #openttdcoop.devzone
20:30:23  *** JVassie has joined #openttdcoop.devzone
20:34:52  *** ChillCore has quit IRC
20:35:53  *** JVassie_ has quit IRC
20:44:01  *** JVassie_ has joined #openttdcoop.devzone
20:50:23  *** JVassie has quit IRC
21:13:06  <Brot6> Japanese Trains - Revision 30:2781d262700d: Add 1800 and 2100 Tank engines (dandan) @
21:13:06  <Brot6> Japanese Trains - Revision 31:8f10f6434e79: Forgot to add the png files (dandan) @
21:14:53  <Brot6> Japanese Trains - Revision 32:57639fd596bc: Forgot to add some more graphics before (thanks oberh... (dandan) @
21:18:47  <Brot6> Japanese Trains - Bug #3244 (Closed): Compilation error - missing sprite file(s?) (dandan) @
21:35:06  <Brot6> FISH - Revision 707:65cf0ef9b156: Change: test version of gen 1 large coaster (andythenorth) @
21:37:29  <Brot6> SuperLib - Revision 26:1ccf6eb68f5d: Fix (r25): use foreach instead of just 'for' (Zuu) @
21:42:38  <Yexo> Zuu: why a ruby script and not simply a shell script?
21:43:00  <Rubidium> to please the Japanese?
21:44:19  <Zuu> Yexo: You mean for the make tar script?
21:44:32  <Yexo> yes
21:45:18  <Zuu> Because I used to use ruby for coding before and it can just be double clicked in windows. A .sh can't do that without extra work setting up the necessary bindings.
21:46:29  <Yexo> but "cp" doesn't work in windows by default, does it?
21:46:47  <Zuu> No, but I have C:\cygwin\bin in my PATH.
21:46:52  <Yexo> aha
21:47:21  <Zuu> That is one of the essential things I "fix" on a new computer :-)
21:49:26  <Zuu> The new script to translate SuperLib for AI to GS is for example in python. I think I'll publish/commit it when I've creared it up a bit.
21:53:37  *** frosch123 has quit IRC
21:54:41  <Yexo> in _SuperLib_Airport::GetAirportTile I think the Valuate / KeepBottom(1) can be left out
21:55:03  *** andythenorth has quit IRC
21:55:21  <Yexo> I'm browsing through the source right now to see what can be added to the OpenTTD API
21:57:05  <Zuu> IIRC it has something to do with that it want to find the north tile.
21:57:38  <Yexo> in that case you need to sort by item, not valuate it
22:06:10  <Zuu> Thanks for the tip, I'll take a look at the actual code when I'm done fiddling with my nogo tranlator :-)
22:23:15  <Brot6> SuperLib - Revision 27:a6421387bd16: Add the nogo translator (Zuu) @
22:23:15  <Brot6> SuperLib - Revision 28:5652aecc04c4: Only copy wanted files instead of copying everything and the... (Zuu) @
22:24:55  <Brot6> SuperLib - Revision 29:bdd8e79c8ab3: Fix: forgot some changes as I tested the script elsewhere (Zuu) @
22:26:38  <Zuu> Regarding Airport::GetAirportTile, it is just an habit that I use an ItemValuator instead of sorting by item. But you are right that sorting by item would be faster.
22:37:42  <Brot6> SuperLib - Revision 30:6d68c368aad6: Simplify and speed up Airport::GetAirportTile a bit (suggest... (Zuu) @
22:59:41  *** ODM has quit IRC
23:21:54  *** JVassie_ has quit IRC
23:22:26  *** hanf has quit IRC
23:32:51  *** Zuu has quit IRC

Powered by YARRSTE version: svn-trunk