Log for #openttdcoop.devzone on 19th January 2011:
Times are UTC Toggle Colours
00:15:24  <Brot6> Polish PKP Set - Bug #2183 (New): .grf doesn't belong to the repo (Ammler) @
01:07:49  *** KenjiE20 has quit IRC
03:13:13  *** thgergo has quit IRC
04:07:37  *** Lakie has quit IRC
07:48:42  <planetmaker> moin
07:49:19  <planetmaker> Ammler: be sure, I also commit via svn to the openttd repository ;-)
07:49:41  <planetmaker> and also rest assured: everyone does.
09:14:28  *** ODM has joined #openttdcoop.devzone
09:19:06  *** andythenorth has joined #openttdcoop.devzone
09:21:50  <Brot6> Polish PKP Set - Bug #2183: .grf doesn't belong to the repo (BarthVader) @
09:26:38  *** andythenorth has quit IRC
09:42:33  *** andythenorth has joined #openttdcoop.devzone
10:31:55  *** thgergo has joined #openttdcoop.devzone
10:50:40  <Ammler> planetmaker: michi__cc doesn't
10:52:05  <Ammler> it doesn't make sense to develop with hg and then commit with svn
10:52:53  <planetmaker> Ammler, of course he does
10:53:37  <planetmaker> I also develop with hg and commit with svn. It's not like it matters
10:55:39  <Ammler> :-) ok, if you think so
10:55:58  <planetmaker> did he tell you that he uses hg (or rather git as it is his preferred vcs)?
10:56:23  <Ammler> I was just wondering, maybe someone had already experience with hgsubversio
10:57:02  <planetmaker> Ammler, I have so many repos of openttd... it doesn't matter to have one which I only commit to (and possibly add a last polish to a patch)
10:57:30  <planetmaker> actually it's IMHO a great safety measure. I cannot accidentially commit crap
11:00:14  <Ammler> for tests, you can still use other branch or mq, that does not neccessary need to be a other vcs :-P
11:00:39  <planetmaker> so even if I w/could commit by another method I'd have a separate repo.
11:01:40  <planetmaker> well, that's what I do. But then in order to commit it's the same effort to commit via hg than it is by svn. So...
11:01:52  <Ammler> ok, as said, was just wondering
11:03:17  *** andythenorth has quit IRC
11:12:14  *** andythenorth has joined #openttdcoop.devzone
11:48:09  <DJNekkid> Rubidium: thanx :D
11:53:51  *** andythenorth has quit IRC
12:54:53  *** andythenorth has joined #openttdcoop.devzone
13:10:44  <michi_cc> planetmaker: I commit using git svn dcommit, but only from one repository and not directly from where I develop/hack around.
13:11:14  <michi_cc> Ammler: Can a hg tag contain spaces? The stuff apparently fails when using mercurial queues.
13:13:17  <Ammler> michi_cc: I guess not
13:13:36  <Ammler> but you can tag one changeset with multiple tags
13:14:19  <Ammler> hmm, it can have spaces
13:14:33  <Ammler> hg tag "hallo welt" is possible
13:15:38  <Ammler> why does it fail with mq?
13:16:56  <michi_cc> Output of hg id -t is "030_Codechange_Uniform_road_vehicle_state_functions.diff qtip tip" according to Terkhen, which is != "tip" and thus becomes the version.
13:17:23  <michi_cc> As it seems "qtip" will always be in the output, would filter that out
13:18:42  <Ammler> michi_cc: just tip
13:18:58  <Ammler> then we don't need to query for tip again
13:19:26  *** andythenorth has quit IRC
13:19:52  <Ammler>
13:21:32  <michi_cc> Well, it's goint to fail if somebody has a tag containing tip. But if tags can contain spaces, comparing just the string after the last space can fail as well.
13:22:32  <Ammler> well, it mainly need to support the openttd tags, which are numbers and beta/rc
13:23:05  <Ammler> or use tip$
13:24:06  <Ammler> does findversion work with your svn mirror btw.?
13:24:14  <Ammler> I mean, do you get the svn revision?
13:24:30  <Ammler> as you don't have (svn rX) in your msg
13:24:54  *** DayDreamer has joined #openttdcoop.devzone
13:25:41  <michi_cc> It does as the script also contains a line looking for git-svn-id stuff if the test for (svn rX) fails.
13:26:02  <michi_cc> Paste is broken..
13:26:30  <Ammler> he?
13:26:44  <michi_cc> Anyway, would be my proposal then
13:27:01  <michi_cc> -> Internal Server Error
13:27:32  <Ammler> 502?
13:27:53  <michi_cc> 500
13:29:57  <Ammler> hmm
13:30:42  <Ammler> the pastebin script is from here:
13:30:43  <Webster> Title: New Paste | LodgeIt! (at
13:33:58  <Ammler> then it might make sense to add the svnrev detection to the hg too, so it wouldn't need a patched hgsubversion
13:35:14  <Ammler> could you wait with committing that "tip fix" :-)
13:35:26  <planetmaker> <michi_cc> planetmaker: I commit using git svn dcommit, but only from one repository and not directly from where I develop/hack around. <-- well, still sounds like committing to the svn repo ;-)
13:36:32  <Ammler> planetmaker: yes, of course, what did you mean?
13:36:46  <Ammler> I never meant you push to a hg repo :-P
13:37:18  <planetmaker> that's what I understood. Sorry, if that was wrong.
13:37:37  <Ammler> hg clone|pull|push svn:// works just fine
13:38:03  <Ammler> well, I never tested push :-P
13:38:30  <planetmaker> well. I'm not sure that even with the help of hgsubversion I could have released the 1.1.0-beta2 ;-)
13:38:52  <DJNekkid> are any of you guys into the music genre 'dubstep' ?
13:50:25  <Ammler> michi_cc: is it right, if the svn rev detection would be fixed to show the revision which is tagged instead the changeset itself, git would then be wrong?
13:54:08  <michi_cc> Most SVN tags in ottds repo are not empty, so the tagged revision is not equal to the tag changeset. 1.1.0-beta1 might tag the revision 21614, but the contents of 21615 (the actual tag) is not the same. So calling it 21614 is wrong.
13:58:01  <Ammler> no 1.1.0-beta1 is 21615 also with hg
14:05:17  <Ammler> the problem is that hgsubversion does only branch, if the tag has branch, else it simply tags
14:05:28  <Ammler> s/branch/changes/
14:06:00  <Ammler> the fix might be, that hgsubversion does always branch before tag
14:06:21  <Ammler> (I need to formulate that somehow logical for the hgsubversion guys)
14:08:15  <michi_cc> I'd call it don't skip empty svn revs but instead create an empty hg commit for exact mirroring of svn repositories.
14:20:25  <Ammler> since hgsubversion does support tags with changes, this should really doable
14:21:03  <Ammler> michi_cc: your convert is from a unmodified git-svn?
14:21:13  <Ammler> so I can use that as a reference
14:21:35  <michi_cc> yeah
14:35:25  <Ammler> michi_cc: and that url you posted to your mirror will stay, then I use it that in the bugreport. (
14:37:23  <michi_cc> Ah, sorry, beta3 is now (which will stay). I had a small error in faking that I actually cloned from and not locally from filesystem which required changing some commit messages (which in turn changes the hash)
14:38:39  <michi_cc> There were some commits that had a wrong git-svn-id: line, which I had to fix so it is actually possible to use a clone with git-svn.
14:42:48  <Ammler> I used so this should not matter :-)
14:43:13  <michi_cc> Yes, that one will of course not change :)
14:43:56  <michi_cc> (and the other hashes neither now, I just had a wrong config var after copying the repo from my home machine to the server)
14:48:20  *** KenjiE20 has joined #openttdcoop.devzone
14:49:47  <Ammler>
14:49:48  <Webster> Title: durin42 / hgsubversion / issues / #250 - empty tag does not show the same revision number as svn or git-svn Bitbucket (at
14:49:49  <Brot6> Webster: Title: #250 is "World Airliners Set - Feature #250: Boeing 737-500 Nfo - #openttdcoop Development Zone"
14:50:01  <Ammler> hmm
14:54:37  <michi_cc> Ammler: Is there anything more you want to change for hg in or can I commit the mq tag fix?
14:54:56  <Rubidium> michi_cc: how does that fix look like?
14:55:02  <Rubidium> s/how/what/
14:55:33  <DJNekkid> Rubidium: unless you saw it further up; thanx :D
14:55:46  <Ammler> michi_cc: I would like to add the svn revision grapping from metadata svnrev
14:55:55  <Rubidium> DJNekkid: I saw it
14:56:06  <michi_cc> Rubidium:
14:56:20  <michi_cc> Also adds some quotes in case we ever get git/hg tags with spaces
14:57:13  <DJNekkid> you know what for? :P
14:57:51  <Rubidium> DJNekkid: the new version of nforenum?
14:58:03  <Ammler> my guess is something with railtypes
15:04:43  *** DayDreamer has quit IRC
15:45:22  <DJNekkid> sorry, dinner and all ...
15:45:30  <DJNekkid> FS 4393 :)
15:45:56  <DJNekkid> :D :D :D
15:55:59  <Ammler> missing http://
16:00:42  <DJNekkid> my browser dont require that :)
16:05:28  <DJNekkid> Rubidium: according to is property 17 (intro date) a word. And according to the text is it in long format.
16:05:52  <DJNekkid> all other places are the log format dates in days since year0
16:05:59  <DJNekkid> and in dword format
16:06:34  <Rubidium> DJNekkid: it is the dword of days since year 0
16:06:38  <DJNekkid> i assume the text in the list on the top is wrong, and the W should be a D ?
16:07:03  <Rubidium> yeah, it's wrong
16:07:14  <DJNekkid> oki, goodie :D
16:07:23  <Rubidium> please fix it, and possibly make it more clear if needed
16:07:47  <DJNekkid> is it the same account as in the openttd-pages?
16:07:52  <Rubidium> no
16:07:55  <DJNekkid> im not sure if i have an account there...
16:07:58  <DJNekkid> could make one tho :)
16:15:09  <Ammler> there is a bug in hg log -k "(svn r", it finds also
16:16:27  <planetmaker> DJNekkid, and when you're at it: update the railtypes page.
16:16:43  <planetmaker> I added the previously used ones by NuTracks, but I lost track of your changes
16:17:10  <DJNekkid> not wierd, i've done two, but i think its set in stone now :)
16:17:48  <planetmaker> I don't know your changes nor did I really look at it. I just saw you changed *something*
16:18:12  <planetmaker> So if you want to allow for some kind of compatibility: please update what you use. And remove those which you don't use anymore
16:20:07  <DJNekkid> im about to :) And, i think my changes are "better"
16:20:12  <DJNekkid> compared to the old ones
16:20:23  <DJNekkid> instead of x160 its now xMED
16:20:45  <DJNekkid> x120 is xLOW, x230 is xHIG
16:24:25  <Ammler> hmm
16:26:13  <planetmaker> sounds more flexible
16:27:42  <DJNekkid> yea, thats what i thought as well
16:28:08  <DJNekkid> plus, i plan to have speed parameters for the different 'kinds'
16:28:18  <DJNekkid> one for very slow, one for slow, etc
16:28:46  <Ammler> michi_cc:
16:29:03  <Ammler> Rubidium: planetmaker, hg detection mod ^
16:31:57  <Ammler> <-- tested here, clone with "clean" hgsubversion
16:32:44  <michi_cc> Ammler: For what do you need the grep "^(svn r[0-9]*)" for? What breaks without?
16:32:59  <Ammler> [17:15] <Ammler> there is a bug in hg log -k "(svn r", it finds also
16:33:08  <Ammler> hg log -k does not allow regex
16:33:42  <Ammler> so if you have a clone without the (svn rx) in the message, rev_nr would not be empty
16:33:54  <michi_cc> Okay, so that has to go to into the vbs as well.
16:34:48  <Ammler> I guess, the missing grep is to blame me as I asked to add to -f for mq support
16:35:14  <Ammler> there was a grep before that patch
16:35:35  <Ammler>
16:36:40  *** Lakie has joined #openttdcoop.devzone
16:36:55  *** andythenorth has joined #openttdcoop.devzone
16:37:18  <michi_cc> Ammler: The vbs passes -l 1 to log whereas the sh uses head -n 1. How does -l 1 interact with -k, i.e. if the top commit is not from svn, does it get the first that matches svn or none?
16:38:08  <DJNekkid> Ammler: im gonna reinstall my server ... any tips on distros?
16:38:17  <Ammler> head -n1 could be replaced by "{desc|firstline}\n"
16:38:17  <michi_cc> git currently takes the first one that has svn info, which IMHO makes sense because when you do custom development in a hg clone you probably still want the svn revision where you started for NewGRF compatibility.
16:38:38  <Ammler> DJNekkid: :-P
16:38:51  <Ammler> use what you like, it doesn't really matter
16:39:36  <Ammler> use debian/centos if you like to keep it stable, use fedora/suse/ubuntu, if you prefer bleeding edge
16:40:27  <Rubidium> and use gentoo when you want to be bleeding
16:40:30  <Ammler> from the openttd view, every distro is supported
16:40:55  <Brot6> 2cc train set - Feature #2184 (New): HŽ Class 51 (Voyager1) @
16:41:33  <DJNekkid> i've tried some googleing, and it seems like ubuntu is popular...
16:41:52  <Ammler> yes, a reason to not use it :-P
16:42:18  <DJNekkid> and suse also seems popular...
16:42:23  <Ammler> :-D
16:42:32  <DJNekkid> it looked like thoose two were the "main" ones
16:43:08  <Ammler> I would use ubuntu for GNOME and suse for KDE
16:43:29  <DJNekkid> with debian, centos and fedora for the more 'geeky' people
16:43:37  <DJNekkid> can be more custum and things...
16:44:15  <Ammler> nah, fedora is quite nice too
16:44:18  <DJNekkid> i just need to finish backing up my backups :)
16:44:42  <Ammler> do you plan to use desktop?
16:44:45  <DJNekkid> yes
16:44:52  <DJNekkid> remotely as i am now
16:44:59  <Rubidium> DJNekkid: Debian (stable) only majorly updates every 2 to 3 years, the rest of the time it's security fixes
16:45:20  <Rubidium> so e.g. Debian stable has OpenTTD 0.6.2
16:45:51  <Ammler> debian is for setting up something and then keep that running
16:45:53  <DJNekkid> iirc were 6.2 about as playable as ttdpatch?
16:45:56  <Ammler> like centos
16:46:13  <Rubidium> so if you want anything that's recent choosing one of the slow releasing/stable distros is not what you want
16:46:30  <Rubidium> DJNekkid: 0.6.2 is probably more playable than the last ttdpatch stable
16:46:46  <Rubidium> 0.3.5... that's "interesting" to play
16:46:53  *** andythenorth has quit IRC
16:47:15  <Ammler> Rubidium: there it starts the advanture with building already
16:47:35  <DJNekkid> i guess i'll go for suse then :)
16:48:21  <Ammler> yeah, then you can at least blame someone, if it isn't good :-P
16:48:24  <planetmaker> <Rubidium> DJNekkid: 0.6.2 is probably more playable than the last ttdpatch stable <-- hehe :-)
16:48:48  <planetmaker> distributing blame is always good. ;-)
16:49:11  <DJNekkid> i did 'install' patch about a year and a half ago
16:49:22  <Ammler> since I run debian as hypervisor, I can't say, suse only anymore :-)
16:49:32  <DJNekkid> i were pulling my hairs and swearing and cursing...
16:49:46  <Ammler> well, not debian, but almost
16:49:55  <DJNekkid> no autorail, no shortcuts, atleast not the ones i were used to
16:50:17  <DJNekkid> iirc limited foundations
16:50:21  <planetmaker> but you got custom bridge heads! and cargodist.
16:50:35  <DJNekkid> cargodist in patch?
16:50:43  <Rubidium> 0.6.2 has shortcuts and autorail
16:50:44  <planetmaker> and some other signals.
16:50:47  <Ammler> custom bridge heads for roads is the only thing I would miss
16:51:09  <Ammler> maybe enhanced tunnels entries a bit
16:51:16  <DJNekkid> Rubidium: i might be thinking of 0.4.x
16:51:19  <Rubidium> actually 0.3.5 has autorail and shortcuts as well ;)
16:51:26  <planetmaker> DJNekkid, I can't tell for sure, I somewhen stopped investing time in getting it configured to run.
16:51:35  <planetmaker> but it's what I read
16:52:01  <DJNekkid> there is cargodist for openttd also, just not official yet, and it is easy to get running :P
16:52:24  <Ammler> since I know openttd, I run ttdpatch only for comparision and once to play canadian scenario
16:52:38  <DJNekkid> brb
16:52:43  *** DJNekkid has quit IRC
16:53:09  *** DJNekkid has joined #openttdcoop.devzone
16:53:25  <planetmaker> why canadian scenario? won't that work with ottd?
16:53:36  <Ammler> hmm, now it might
16:54:27  <Ammler> but at that time it wasn't, the canadian set demo save
16:54:45  <michi_cc> Ammler: Does look alright?
16:54:51  <DJNekkid> suse = opensuse, yes?
16:55:04  <DJNekkid> or vica verca
16:57:12  <Ammler> suse is company behind opensuse, like redhat->fedora
16:58:13  <Ammler> michi_cc: looks fine
16:58:30  <Ammler> are you aware, that git does add -trunk in the rev?
16:59:41  <michi_cc> It does, but if you re-clone my repository you'll find that I've renamed the trunk branch back to master (and that is filter out already)
16:59:55  <Ammler> ok :-)
17:00:27  <michi_cc> There's just too much stuff that assuming master is the main branch so calling trunk trunk doesn't really work out.
17:01:04  <DJNekkid> oh crap, it requires a DVD... i only  have CD's ... no web installs... *search*
17:01:57  <Ammler> oh well, why no netinstall?
17:03:04  <DJNekkid> i could netinstall ... i just cant find the link :P
17:04:06  <Ammler>
17:04:07  <Webster> Title:" target="_blank"> Download openSUSE 11.3 (at" target="_blank">
17:05:17  <DJNekkid> thats where im at
17:05:59  <DJNekkid> but there are only the DVD option
17:06:10  <Ammler> nah, go down, Network
17:06:58  <DJNekkid> oh, it were such :P
17:07:28  * DJNekkid turns his head away from ammler, waiting for a serious smack
17:07:39  <Ammler> thanks :-P
17:23:32  *** andythenorth has joined #openttdcoop.devzone
17:31:46  <Ammler> Rubidium: just in case, will future svn releases still have changes in the tag?
17:39:01  *** frosch123 has joined #openttdcoop.devzone
17:45:06  <Rubidium> some
17:46:01  <Rubidium> (or probably most)
17:46:34  <Rubidium> but likely not all
17:57:46  *** DanMacK has joined #openttdcoop.devzone
18:44:37  *** Lakie has quit IRC
18:44:49  <Brot6> MailAI - Revision 8:e252aa1eebda: Getting ready for beta version launch on bananas. (Hephi) @
18:47:10  *** andythenorth has quit IRC
19:01:11  <Brot6> 32bpp-ez-patches: compile of r21861 still failed (#2069) -
19:03:56  <Brot6> clientpatches: update from r21488 to r21488 done -
19:05:02  <Brot6> serverpatches: compile of r21861 still failed (#2080) -
19:18:58  *** andythenorth has joined #openttdcoop.devzone
19:51:53  <DJNekkid> hmm ...
19:52:10  <DJNekkid> that screen after i press the "install-button" should not take too long?
19:52:19  <DJNekkid> its stuck on '1 pixel'
19:52:25  <DJNekkid> been like so for 5ish minutes
19:53:11  <DJNekkid> (suse)
20:03:45  <DJNekkid> hmm
20:03:56  <DJNekkid> after reboot it and trying again a few times it seemed to work
20:04:35  <Ammler> hehe
20:05:38  <DJNekkid> go figure
20:10:22  <DJNekkid> and nice, it seems like it can detect the old raid :D
20:11:18  <DJNekkid> or ... did it
20:14:05  <DJNekkid> yea, it did :D
20:16:33  <Hirundo> andythenorth: Could you tell me which cargos are in the CC_EXPRESS cargo class without looking it up?
20:16:44  <Hirundo> *which FIRS cargos
20:16:55  <andythenorth> I can look it up for you :P
20:17:38  <andythenorth> goods
20:17:42  <andythenorth> food
20:17:46  <andythenorth> milk
20:17:49  <andythenorth> fruit + vegetables
20:17:53  <andythenorth> fish
20:17:58  <andythenorth> mnsp
20:17:59  <andythenorth> ensp
20:18:03  <andythenorth> fmsp
20:18:12  <andythenorth> is all
20:18:13  <Hirundo> are you looking it up now or not?
20:18:23  <andythenorth> yes that was a lookup
20:18:28  <andythenorth> from current tip
20:19:16  <Hirundo> Point was not that I needed to know, but that cargodist offers separate destination types for EXPRESS while no-one knows what cargos that actually are :)
20:20:21  <andythenorth> hmm
20:20:30  <andythenorth> I haven't tried a cargodist game, so I can't 100% picture the issue yet
20:20:44  <andythenorth> is it something where help/changes are needed to resolve it?
20:22:09  <Hirundo> no, at least not by you
20:22:20  <Hirundo> unless you want to fork cargodist, that is :)
20:22:34  <andythenorth> I'm ok for that right now
20:26:59  <Brot6> Polish PKP Set - pkp_set.grf (BarthVader) @
20:44:54  <Brot6> Polish PKP Set - Revision 6:79945c8341f8: Fix: Removed GRF from repository (BarthVader) @
21:13:29  *** frosch123 has quit IRC
21:28:53  *** andythenorth has quit IRC
21:34:27  <Brot6> Polish PKP Set - Bug #2183 (Closed): .grf doesn't belong to the repo (Ammler) @
21:34:27  <Brot6> Polish PKP Set - Bug #2183 (Closed): .grf doesn't belong to the repo (BarthVader) @
21:41:11  *** andythenorth has joined #openttdcoop.devzone
21:59:08  *** andythenorth has left #openttdcoop.devzone
22:37:13  *** DanMacK has quit IRC
23:27:50  *** ODM has quit IRC

Powered by YARRSTE version: svn-trunk