Log for #openttdcoop.devzone on 15th April 2010:
Times are UTC Toggle Colours
00:16:36  <PeterT> Ammler: Issue created
00:16:44  <PeterT>
00:17:51  <PeterT> Ammler: damn, there is a mistake
00:18:04  <PeterT> can you delete join-part-no-op.diff please?
00:22:58  <Ammler> deleted
00:23:27  <Ammler> please do also note a usage example... :-)
00:23:38  <PeterT> what about it?
00:25:09  <Ammler> well, when do you use it?
00:25:43  <PeterT> I don't know
00:25:50  <PeterT> if you want a seperate testing channel
00:25:54  <PeterT> like #openttdcoop.bots
00:26:00  *** KenjiE20 has quit IRC
00:28:37  <Webster> Latest update from devactivity: Autopilot - Patch #896: Join and Part commands <" target="_blank">> || Autopilot - Patch #896 (New): Join and Part commands <>
00:31:53  *** SN1 has joined #openttdcoop.devzone
00:31:59  <PeterT> like this^
00:32:07  <PeterT> !part #openttdcoop.devzone
00:32:08  *** SN1 has left #openttdcoop.devzone
06:02:23  *** ODM has joined #openttdcoop.devzone
07:59:53  *** ODM has quit IRC
08:30:13  *** Doorslammer has joined #openttdcoop.devzone
08:53:58  <Webster> Latest update from devactivity: NFO Meta Language - Revision 67: Fix: Action6 was used incorrectly when setting the severity of a... <> || NFO Meta Language - Revision 66: Codechange: Use a param_list when parsing an error_block. <>
09:04:13  <planetmaker> Ammler, so... the old OpenGFX newgrfs. Do they go to legacy or obsolete?
09:04:20  <planetmaker> I'd put them now rather in obsolete?
09:04:46  <Ammler> hmm, shouldn't be needed at all
09:05:19  <Ammler> but why not discuss that in the ticket about?
09:05:48  <planetmaker> well... we don't want to supply them by default anymore with our grfpack, I think
09:05:49  <Ammler>
09:42:09  <Webster> Latest update from devactivity: #openttdcoop NewGRF package - Bug #542: Remove OpenGFX newgrfs <> || #openttdcoop NewGRF package - Bug #542: Remove OpenGFX newgrfs <>
10:27:42  <Hirundo> Yexo: Is it possible / recommended to insert the same action6 multiple times in the action list?
10:44:44  *** KenjiE20 has joined #openttdcoop.devzone
11:04:41  <planetmaker> Ammler, is the grfpack's repo not linked to redmine on purpose (mb)?
11:06:31  <Ammler> yeah :-(
11:06:45  <Ammler> it isn't public at all
11:07:07  <planetmaker> hm... could it be made visible only to loged in users?
11:08:02  *** ODM has joined #openttdcoop.devzone
11:09:47  <planetmaker> hm... only a general "public" option...
11:11:34  <planetmaker> hm... could we hack the system?
11:11:55  <planetmaker> Not allowing public acess to those files? Or does it work via a DB access, so that it'd work regardless?
11:16:57  <Ammler> well, we could close the project for members only
11:17:06  <Ammler> ah
11:17:12  <Ammler> you like the annoucements here?
11:19:20  <planetmaker> I liked a link to revisions actually
11:19:36  <planetmaker> but it's not really that important
11:19:44  <Ammler> trac is still there
11:20:01  <Ammler> or should
11:20:56  <Ammler>
11:34:26  <Webster> Latest update from devactivity: #openttdcoop NewGRF package - Bug #544 (Confirmed): Remove (e)GRVTS <>
11:37:41  <Yexo> Hirundo: it's possible, but it won't have the desired effect probably
11:39:06  <Hirundo> why not?
11:41:07  <Hirundo> desired effect = setting the severity in all ActionBs
11:41:48  <Yexo> ah, in that case it should work
11:46:49  <Yexo> Hirundo: currently your code assumes that every ID will be mapped to a constant number
11:47:05  <Yexo> in case of variables however it'll be mapped to something of the Variable class
11:48:14  <Hirundo> I had planned to pass an additional (lambda?) function that maps the value from the dictionary to an AST item, but I did not introduce that (yet)
11:48:25  <Yexo> ok
12:05:50  <Webster> Latest update from devactivity: NFO Meta Language - Revision 69: Feature: Support parameters for actionB error messages. <> || NFO Meta Language - Revision 68: Fix: Fix action6 for error messages, this time done properly (I ... <>
13:29:26  *** ODM has quit IRC
13:46:16  * Hirundo chases the nml equivalent of a memory leak: a free parameter leak
14:03:11  <Yexo> a missing "free_parameters_backup = free_parameters[:]" in one of the functions called from
14:10:59  <Hirundo> No, at least not in the functions I'm calling now
14:11:41  <Hirundo> It seems the parameters somehow leak in-between two blocks
14:12:16  <Hirundo> That is, at the end of block one the first free param is 0x40 (as it should), but at the beginning of the next block it is 0x41 or 0x42 :S
14:15:59  <Yexo> can you pastebin an exampel nml tha treprocudes it?
14:16:22  <Hirundo> I'll try to reproduce with 'trunk'
14:19:31  *** Yexo has quit IRC
14:19:47  *** Yexo has joined #openttdcoop.devzone
14:20:57  <Hirundo>
14:23:02  <Hirundo> Yexo: Question is; why does it use parameter 41 for the 2nd statement?
14:23:20  <Yexo> no idea, trying to find that out now
14:32:11  <Yexo> the "free_parameters = free_parameters_backup" at the end of parse_error_block doesn't work for some reason
14:34:32  <Hirundo> to me, it reports that it does work
14:36:56  <Hirundo> <- result of debug statements like this: print "error entry", to_hex(free_parameters[-1])
14:37:58  <Yexo> Hirundo: yes, now try printing the length also in
14:38:03  <Yexo> so just after each get_action_list
14:38:22  <Yexo> you'll notice that also error exit prints 40, in the main loop it'll be 41 again
14:39:13  <Hirundo> yes
14:41:09  <Webster> Latest update from devactivity: MiniGRFs - Revision 10: [basecosts] Release Presets 1.4 <>
14:42:04  <Yexo> replace "free_parameters = free_parameters_backup" with "free_parameters.extend([item for item in free_parameters_backup if not item in free_parameters])" and it does work
14:45:05  <Hirundo> Yes, but why ? what piece of python's black box magic am I missing?
14:45:27  <Yexo> it might have something to do with setting a global variable defined in another module
14:45:45  <Yexo> the new line doesn't set the global variable, it changes it
14:47:17  * Hirundo sees the light
14:47:42  <Yexo> ok if I commit that?
14:47:50  <Hirundo> yes
14:48:02  <Yexo> there might be a better solution, this is my first python project
14:48:39  <Hirundo> The assignment is valid only in the current module, I think
14:49:09  <Hirundo> So the second action D statement 'sees' the value left by the  actionD that was part of the error block, i.e. 41
14:50:47  <Yexo> yes, something like that
14:50:51  <Yexo> not 100% how it works
14:53:16  <Hirundo> <- regarding from-import
14:53:17  <Webster> Title: Importing Python Modules (at
14:56:39  <Webster> Latest update from devactivity: MiniGRFs - basecosts-presets.grf <>
14:57:33  <Ammler> planetmaker: I suggest making a grfpack 8.0 beta and make some tests with the archive
14:57:46  <Ammler> so we see if we removed too many
14:58:15  <Ammler> then we release 8.0 some weeks later
14:58:37  <Ammler> which should then hold for a year again :-)
14:58:53  <Ammler> or longer
14:59:33  <Ammler> meh, don't like beta, we simply release 8.0 and if it doesn't work, fix and release 8.1
15:00:58  <Yexo> Hirundo: ah, "And if name in X is changed to point to some other object, your module won’t notice" <- that was the problem
15:02:22  <planetmaker> Ammler, but we still should test it before we call it a release
15:02:30  <Hirundo> OK, improved ActionE coming in
15:04:39  <Ammler> planetmaker: we most probably can't
15:04:51  <Ammler> or will you try it on the 150 games?
15:05:01  <planetmaker> Ammler, how can't we? No. But on 'selected'
15:05:07  <Ammler> of course
15:05:27  <Ammler> we could for example on those which failed last time :-)
15:06:22  <Ammler>
15:07:23  <planetmaker> yep
15:07:28  <planetmaker> possibly some newer, too
15:09:00  <Ammler> we should ask Zephyris to upload the older versions to bananas
15:12:06  <Webster> Latest update from devactivity: NFO Meta Language - Revision 70: Fix: resetting the list with available parameters/labels didn't ... <>
15:18:41  <Yexo> Hirundo: are you working (or planning to work on) adding a lambda function to the id_list passed to reduce_expr? if not I'll give it a go
15:19:43  <Hirundo> You can give it a go, it should be a two line change I think
15:21:00  <Hirundo> hmm I have already done it, could not resist :)
15:21:45  <Yexo> ok :)
15:22:11  <Yexo> push it then, so I can work on varaction2 support
15:23:08  <Hirundo> Correction, it's 4 lines not 2
15:24:16  <Hirundo> Yexo: done
15:27:20  <Webster> Latest update from devactivity: Nutracks - Bug #865: Track costs are inconsistent and exploitable <> || NFO Meta Language - Revision 71: Feature: Make 'deactivate' accept more than one parameter, and a... <>
15:30:41  <Yexo> Hirundo: wouldn't a function per id_dict be better?
15:31:01  <Yexo> so make id_dicts an array of (id_dict, func) tuples
15:31:56  <Hirundo> Is it still possible to specify a default that way?
15:33:01  <Yexo> func = lambda x: ConstantNumerci(x) if not isinstance(id_dict, tuple) else id_dict[1]
15:34:54  <Hirundo> ok, go ahead
15:38:37  <Ammler> funny thing, andythenorth does also upload the release to tt-forums, so obviously his release tasks are bit more work... ;-)
15:38:59  <planetmaker> hehe :-)
15:41:05  <Ammler> my new backup script failed, somehow it synced the whole system :-/
15:42:23  <Webster> Latest update from devactivity: NFO Meta Language - Revision 72: Feature: lambda functions in reduce_expr. <>
15:57:28  <Webster> Latest update from devactivity: NFO Meta Language - Revision 73: Change: reduce_expr now accepts a list of (dict, func) couples s... <>
16:05:47  *** ODM has joined #openttdcoop.devzone
16:11:14  *** frosch123 has joined #openttdcoop.devzone
16:59:38  <Webster> Latest update from devactivity: NFO Meta Language - Revision 74: Feature: support references variables by name and make a start w... <>
16:59:43  <Yexo> Hirundo: I'd added constants for feature numbers now (feature numbers are now constant expressions), but I don't have a testcase for PropertyBlock yet
16:59:50  <Yexo> I hope I've fixed all cases in there
17:14:43  <Webster> Latest update from devactivity: NFO Meta Language - Revision 75: Feature: feature numbers are now parsed as constant expressions,... <>
17:41:32  *** frosch123 has quit IRC
17:46:37  *** Doorslammer has quit IRC
17:49:18  *** Doorslammer has joined #openttdcoop.devzone
18:07:37  *** Seberoth has joined #openttdcoop.devzone
18:08:15  *** Seberoth has quit IRC
18:10:44  *** Seberoth has joined #openttdcoop.devzone
19:20:19  <Webster> Latest update from devactivity: NFO Meta Language - Revision 76: Feature: add list of global varaction2 variables <>
19:22:12  *** Seberoth has quit IRC
19:25:55  *** Doorslammer has quit IRC
19:43:19  *** Seberoth has joined #openttdcoop.devzone
19:52:23  <Webster> Latest update from devactivity: NFO Meta Language - Revision 77: Fix: Varact2 variable 0x18 is callback info 2, not 1. <>
20:23:12  <Webster> Latest update from devactivity: NFO Meta Language - Revision 78: Feature: optimize expressions with a Variable and a constant mor... <>
20:45:23  *** frosch123 has joined #openttdcoop.devzone
21:09:23  <Webster> Latest update from devactivity: World Airliners Set - Feature #895: Standard helicopters support <>
21:26:25  <Ammler> planetmaker: you really should make that tar :-)
21:26:57  <Ammler> so people can use official zip for such things
21:27:52  <planetmaker> uh?
21:28:21  <Ammler> well, yexo or whoever suggested him to use bananas tar...
21:28:37  <Ammler> I would recommend the official zip
21:28:49  <planetmaker> yes. Ah
21:28:56  <Yexo> Ammler: I suggested the tar because that can be included in the .ipa bundle directly, as openttd can read inside a tar
21:28:57  <planetmaker> Well, he needs a format which can be read
21:29:00  <Yexo> with a zip that wouldn't work
21:29:09  <planetmaker> ^
21:29:19  <Ammler> Yexo: but maybe you can unzip a zip before?
21:29:36  <Yexo> yes, but then he'd have to extract the zip, and the risk he then removes files is higher
21:29:38  <Ammler> we do that on suse
21:29:47  <Yexo> if he includes the tar ocmpletely you can be sure all needed files are there
21:30:11  <Ammler> yes, that is why I think, we should have again a tar in the zip
21:30:12  <Yexo> same as he included a lot of AI tar, he didn't extract those, so the license files / readme files are included for them
21:30:24  <Yexo> why a tar in the zip?
21:30:28  <Yexo> why not a single tar only?
21:30:43  <Ammler> becasue windows can't unpack that
21:30:56  <Ammler> and so the readme is badly accessable
21:32:22  <Ammler> I do also wonder, you guys seems more focused to the license file than the readme
21:33:13  <Yexo> I'm not doing that intentionally, if I did so it was by accident
21:33:47  <Yexo> if you insisnt on having the tar + readme + licence in a zip then I'd inlcude the readme/license file twice, once in the tar and also in the zip
21:34:18  <planetmaker> Yexo, it's not done yet. But it has one advantage:
21:34:29  <planetmaker> people might download the zip file - and unzip it.
21:34:45  <planetmaker> then they have the text files to read - and if they re-distribute just the tar, they're still fine
21:34:55  *** frosch123 has quit IRC
21:35:06  <planetmaker> tar is a mostly unknown format by many non *nix people
21:35:10  <Yexo> planetmaker: yes, that's what I was saying ;)
21:35:17  <planetmaker> oh, ok :-P
21:35:19  <Rubidium> once upon a time... there was a zip with a readme, license and a tar (with a readme, license, grfs and obg)
21:35:29  <Ammler> mostly like we had before the Makefile upgrade
21:35:34  <planetmaker> and then I changed it, I guess :-)
21:35:37  <planetmaker> bad me :-)
21:35:40  <Ammler> :-)
21:35:46  <planetmaker> change request dully noted ;-)
21:36:19  <Ammler> well, new advantage is a common used Makefile framework
21:36:27  <Ammler> you can't have everything at once ;-)
21:36:33  <planetmaker> ah, well
21:36:50  <Rubidium> he should've just unzipped into a separate directory and there was not problem either
21:36:50  <planetmaker> though I'm quite sure I didn't touch the bundle* rules.
21:38:34  <Ammler> Rubidium: yes, same problem was with the suse guys...
21:38:51  <Ammler> but that is fixed now :-)
21:39:55  <Ammler> the only bad thing from my side was --enable-strip
22:12:22  *** ODM has quit IRC
22:42:36  <Webster> Latest update from devactivity: NFO Meta Language - Revision 80: Feature: support *,/,% operators for both varaction2 and actionD <> || NFO Meta Language - Revision 79: Feature: optimize some more combinations of a variable and a con... <>
22:43:48  *** Seberoth has quit IRC
22:47:10  *** Seberoth has joined #openttdcoop.devzone
23:38:54  *** Seberoth has quit IRC
23:48:14  *** KenjiE20 has quit IRC

Powered by YARRSTE version: svn-trunk