Log for #openttdcoop.devzone on 22nd June 2010:
Times are UTC Toggle Colours
00:02:23  <planetmaker> very nice. Version awareness is now working
00:07:51  <Brot6> Swedish Rails - Revision 64:2b7e6e9d5a8c: Change: Use depot introduction year as good as an OpenT... (planetmaker) @
06:12:56  *** ODM has joined #openttdcoop.devzone
06:36:11  <Brot6> NewGRF Meta Language - Bug #1033 (New): internal error with wrong paletted file (planetmaker) @
08:00:20  *** ODM has quit IRC
08:45:45  *** Seberoth has joined #openttdcoop.devzone
09:28:52  <Ammler> OpenMSX 1.0 :-)
10:10:22  *** KenjiE20 has joined #openttdcoop.devzone
10:34:25  <Rubidium> planetmaker: regarding OMSX 1.0, when you've got a list of music you want to include you should explicitly ask Jim whether he agrees with your licensing
10:34:34  <Rubidium> for those songs ofcourse
10:36:34  <planetmaker> Rubidium, I asked him in my e-mail to him
10:37:31  <planetmaker> explaining that the set is licensed under both, GPL v2 and CCSP 1.0 and whether he would allow me to include some of his work
10:39:59  <Ammler> well, I hope you fix that issue with 1.0 :-)
10:40:05  <Rubidium> planetmaker: at least add something like to the source repository; it makes it very clear for everyone that contact has been made to include third party material that can't be found elsewhere under that license
10:40:21  <Ammler> removing the ugly CC license
10:43:37  <Ammler> I think you have enough songs now to make the set GPL...
10:44:38  <planetmaker> <-- you mean like that, Rubidium ? :-)
10:45:13  <Ammler> I wonder, if people are aware, if they allow you to use those that they also give it "open" for everyone
10:46:56  <planetmaker> Ammler, well, there's actually no need to remove the other license IMHO
10:47:11  <Ammler> what is the reason to keep it?
10:47:18  <planetmaker> flexibility for others
10:47:25  <planetmaker> and there's not much source
10:47:31  <Ammler> if I fork it, can I remove one?
10:47:35  <planetmaker> yes
10:47:47  <Ammler> and why do you support that?
10:47:57  <Ammler> why not simply make it GPL?
10:48:19  <Ammler> I thought, you did it for cases if later someone would disagree with GPL
10:48:30  <planetmaker> yes, I did :-)
10:48:50  <Ammler> so since you can now make a complete set with GPL, shouldn't it be fine
10:49:14  <Rubidium> planetmaker: yes-ish. He vaguely said that it is okay to use, although I am not sure (based on that email) that he agrees with it being licensed under those licenses
10:49:15  <planetmaker> It should
10:49:54  <planetmaker> ^ @ Ammler
10:50:39  <planetmaker> Rubidium, I read his e-mail like that all he asks is credits - whatever the license.
10:50:50  <Ammler> planetmaker: My opinion was, that we try to host GPL projects only...
10:50:59  <planetmaker> But sure... it's not a big deal to ask him about specific songs once the decision is made
10:51:19  <planetmaker> Ammler, it is GPL
10:51:36  <planetmaker> it's even more free.
10:51:41  <Ammler> hmm, someone can still fork it and publish his set with CC
10:51:58  <planetmaker> Ammler, there's no source code
10:52:06  <planetmaker> except my makefile
10:53:03  <Ammler> from GPL, I know what I have, CC is just a big complicate hassle
10:53:25  <Ammler> sfx should be a good example
10:53:29  <planetmaker>  Non-Commercial Sampling Plus: People can take and transform pieces of your work for non-commercial purposes only. Non-commercial copying and distribution (like file-sharing) of the entire work are also allowed.
10:53:39  <planetmaker>  Sampling Plus: People can take and transform pieces of your work for any purpose other than advertising, which is prohibited. Non-commercial copying and distribution (like file-sharing) of the entire work are also allowed. Hence, "plus".
10:53:51  <planetmaker> sorry, the latter only ^
10:54:37  <Ammler> non-commercial :-(
10:54:53  <planetmaker> gpl allows you to do it despite.
10:54:59  <Ammler> isn't it the same as the sfx one?
10:55:05  <planetmaker> so if you want to do it commercially, use GPL
10:55:19  <planetmaker> yes, it's the same as OpenSFX
10:55:38  <Ammler> well, it is your set...
10:55:54  <Rubidium> if you want it to be used commercially, use CC0 :)
10:55:58  <Ammler> I just don't get it, why you would like to support CC also after a full release :-(
10:56:07  <planetmaker> I don't see the point of now removing CCSP, Ammler
10:56:19  <planetmaker> Having two _alternative_ licenses is no limitation
10:56:34  <Ammler> it does "limit" constribution
10:56:53  <planetmaker> Especially as the the music (which is important here) can with both licenses be re-used for our purposes
10:57:02  <Rubidium> it *might* fuel speculations/thoughts about one of them being removed in the future which could be troublesome
10:57:04  <planetmaker> Thus the CCSP is more limiting than the GPL
10:57:09  <Ammler> if someone later likes to contribute CC only, you could start a new CC set
10:57:36  <planetmaker> Rubidium, how do you mean?
10:57:46  <Ammler> for me, it is just a big confusing hassle now
10:59:24  <Ammler> imo, it would be easier, if you make to packages, one with GPL, the other with CC
10:59:31  <Ammler> two*
10:59:55  <Rubidium> planetmaker: imagine you know nothing about *your* intentions and you see OpenSFX being licensed under CCSP because there isn't enough freeer content. Then you see OpenMSX that isn't finished and licensed under both GPL and CCSP which might make you think they can remove the GPL in case they can't find enough free music.
11:00:00  <Ammler> distributing one package with both licenses is somehow silly
11:00:56  <Rubidium> that fear might encourage not adding the package to the distribution because you don't want to (re)move it when the GPL license is removed in an update and because you don't want to distribute an old version
11:04:35  <Ammler> Rubidium: would it make sense to write all sample authors and ask, if they would agree to publish the sample with with GPL?
11:04:52  <planetmaker> Ammler, no, they agreed already
11:04:52  <Ammler> did you try such already?
11:04:58  <planetmaker> oh. OpenSFX
11:05:01  <Ammler> planetmaker: samples
11:05:04  <Ammler> yes :-)
11:05:20  <Rubidium> Ammler: good luck finding current contact information of some of them :(
11:05:28  <planetmaker> ok, point taken about dual licenses. GPL it is then for 1.0
11:05:59  <Ammler> planetmaker: maybe you make a final release with both
11:06:11  <Ammler> and then after that a new release with GPL only
11:06:25  <Rubidium> Ammler: that makes it even less transparent
11:06:27  <planetmaker> that then doesn't make sense :-)
11:06:58  <Rubidium> removing/changing a license would, imo, warrant a major version number increase
11:07:20  <Ammler> yes, so release 0.9 with both, and release 1.0 with GPL only :-)
11:07:49  <planetmaker> there won't be a 0.9 ;-)
11:07:53  <planetmaker> next one is 1.0 :-P
11:08:18  <planetmaker> people can still then resort to the current one, if they need both licenses.
11:08:27  <Ammler> or a nightly
11:08:32  <planetmaker> yeah
11:25:03  <Ammler> I have a lot missing instruments from those redfarn midis
11:31:42  <planetmaker> hm,
12:28:16  <Brot6> Swedish Rails - Revision 65:06be5ea59356: Change: Allow for debuging purposes also the de-activat... (planetmaker) @
12:57:19  <Brot6> Swedish Rails - Revision 66:f0688621a403: Feature: In the absence of railtypes replace the defaul... (planetmaker) @
13:25:21  <Brot6> NewGRF Meta Language - Bug #1033 (Closed): internal error with wrong paletted file (planetmaker) @
13:25:21  <Brot6> NewGRF Meta Language - Bug #1033 (Closed): internal error with wrong paletted file (yexo) @
13:25:21  <Brot6> NewGRF Meta Language - Revision 449:d3bcb91a9595: Fix (closes #1033): don't raise strings but use... (yexo) @
14:47:59  <Hirundo> hmmm... I may do a quick hg merge on IS2 to see what happens, but I think I'll drop the patch queue thingy
14:51:24  <planetmaker> right. No worries
14:52:54  <Brot6> Swedish Rails - Revision 67:6e7989063e01: Codechange: Move all sprite templates to their own file (planetmaker) @
15:42:12  <Hirundo> PeterT: Can you compile openttd?
15:42:19  <PeterT> yes
15:42:51  <PeterT> Hirundo: I can.
15:43:24  <Hirundo> I will upload a synced version of IS2 to the devzone when there are no compile errors
15:43:30  <Hirundo> So you can have fun with it :P
15:44:44  <PeterT> wait, so you will upload a synced version now, so I can test its compile?
15:45:13  <Hirundo> I'll test for compilation
15:48:14  <Ammler> Hirundo: could we move the patch queue to a .hg/patches path on the repo
15:48:22  <Ammler> so someone else could qclone it
15:48:49  <Hirundo> My latest updates ignore the patch queue, since it's quite a hassle
15:49:11  <Hirundo> Merging trunk with a patch queue takes about 5x longer than with a normal repo
15:49:12  <Ammler> so you went back again to the branch/fork?
15:49:23  <Ammler> oh :-o
15:49:28  <Hirundo> I just update the 'old' IS repo
15:49:33  <Ammler> ok
15:49:52  <Ammler> doesn't sound that well :-(
15:50:35  <Hirundo> Problem is, that three way merge tools don't work for patch queues
15:50:58  <Hirundo> And updates to earlier patches re-break later patches
15:51:13  <planetmaker> kinda
15:52:09  <Ammler> so we keep our old script, which does make patches from the repo, instead making repo from patches :-)
15:52:59  <Hirundo> It'd be nice if that worked
15:53:09  <Ammler> let me check, if it is still active
15:53:18  <Ammler> and ping me, when you finished the pull
15:53:25  <Hirundo> pull ?
15:53:27  <Ammler> so I could trigger it manually
15:53:30  <Ammler> push :-)
15:54:05  <Ammler> 19 19 * * * cd /home/ottdc/hg-repos/ && ./ 1>/dev/null 2>&1
15:54:05  <Hirundo> I don't know what script you are referring to, but I doubt that it works
15:54:14  <Ammler> should still work :-)
15:54:31  <Hirundo> I'm not updating patch repo in any way
15:54:32  <Ammler> the script which made the patches in bundles
15:54:53  <Ammler> <--those
15:55:22  <Hirundo> ah ok
15:55:31  <Ammler> nothing to do with mq
15:55:41  <Hirundo> Sorry, got confused a little
15:55:46  <Hirundo> Pushing now
15:58:49  *** frosch123 has joined #openttdcoop.devzone
15:58:51  <Brot6> Infrastructure Sharing - Revision 15817:967587da7155: (svn r20009) -Fix [FS#3893]: sign sorting w... (Rubidium) @
15:58:51  <Brot6> Infrastructure Sharing - Revision 15818:bdca06c70a48: [IS] Update to trunk r19456. (Hirundo) @
15:58:51  <Brot6> Infrastructure Sharing - Revision 15819:0a863fd21485: [IS] Update to trunk r20009 (Hirundo) @
15:58:51  <Brot6> NewGRF Meta Language - Revision 450:36b15172ccf6: Codechange: use new operator table in (yexo) @
15:58:55  <Brot6> NewGRF Meta Language - Revision 451:ff0c84e7f22f: Codechange: use new operator table in expressio... (yexo) @
15:58:59  <Brot6> NewGRF Meta Language - Revision 452:3ca949db7663: Codechange: use new operator table in action2va... (yexo) @
15:59:03  <Brot6> NewGRF Meta Language - Revision 453:e4539cf9c507: Codechange: use new operator table in (yexo) @
15:59:07  <Brot6> NewGRF Meta Language - Revision 454:d1f44c3468c8: Codechange: use new operator table in the rest ... (yexo) @
16:01:00  <Brot6> is2: new trunk patch:
16:01:17  <Ammler> hmm
16:02:41  <Ammler> Hirundo: I have 2 heads
16:03:04  <Rubidium> then kill one :)
16:03:04  <Ammler>
16:03:14  <Ammler> hmm, seems like something else is broken
16:03:18  <Hirundo> Interesting, since I have three
16:03:20  <Ammler> you seemd to merge
16:04:20  <Hirundo> Currently, I have one tip and the heads of the closed 2.0 and 2.1 branches
16:06:07  <Brot6> NewGRF Meta Language - Revision 455:17fe3c9d94af: Fix (r452): missed one instance (yexo) @
16:06:32  <Ammler> hg log -k "(svn r20009)" --template="{rev}\n"
16:06:34  <Ammler> 15817
16:07:15  <Ammler> I guess, that breaks the patch creating :-)
16:08:29  <Hirundo> hg log -k svn -l 1 works here
16:10:04  <Ammler> that gives the same :-)
16:10:36  <Ammler> how would you create a patch?
16:12:20  <Rubidium> hg diff -r `top`:`svn head` ?
16:14:04  <Ammler> Rubidium: r15819 is svn head
16:14:27  <Ammler> hmm, I try again, maybe there is something broken from before
16:15:00  <Brot6> is2: new trunk patch:
16:15:21  <Hirundo> hg diff -r XXXX:tip
16:15:39  <Ammler> ok, there was something broken, now it seemed have worked
16:15:45  <Hirundo> But the current repo is broken, hg messed up the file adding again so infrastructure.cpp/infrastructure_func.h are missing
16:16:33  <Ammler> yes, but that is a bug of you :-P
16:16:51  <Ammler> the script should work now
16:17:10  <Hirundo> done
16:17:14  <Brot6> Infrastructure Sharing - Revision 15820:93c43658147b: Fix: infrastructure.cpp and infrastructure_... (Hirundo) @
16:17:14  <Brot6> NewGRF Meta Language - Revision 456:6e193265f8e3: Codechange: introduce that has a table... (yexo) @
16:17:31  <Yexo> I hope the nigthlies have not been started yet, without ^^ nml will fail again
16:17:40  <Brot6> is2: new trunk patch:
16:17:40  <Yexo> should've been the first commit, forgot a hg add
16:18:22  <Ammler> Yexo: one minute :-P
16:18:57  <Brot6> nml: update from r428 to r456 done -
16:18:58  <planetmaker> Yexo: feature request: not operator
16:19:08  <planetmaker> I missed it a few times :-)
16:19:12  <Yexo> planetmaker: (1 - expr)
16:19:14  <Yexo> but yes :)
16:19:26  <planetmaker> sure I know the ways around :-)
16:20:03  <Rubidium> (0 == expr) is safer
16:20:44  <Yexo> yes, but with the current code also a lot more inefficient
16:21:41  <Brot6> swedishrails: update from r57 to r67 done (4 errors) -
16:21:43  <Brot6> Following repos didn't need a nightlies update: 2cctrainset (r557), 32bpp-extra (r36), airportsplus (r50), bros (r12), comic-houses (r70), firs (r1010), fish (r375), heqs (r320), newgrf-makefile (ERROR r100), newgrf_makefile (ERROR r100), nmts (r16), nutracks (r69), ogfxplus (ERROR r30), opengfx (r460), openmsx (r57), opensfx (r94), snowlinemod (r12), worldairlinersset (r643)
16:21:49  <Ammler> that took long :-)
16:22:16  <Yexo> nml is very slow with encoding real sprites
16:22:22  <Yexo> so swedishrails took long
16:22:30  <Ammler> 2010-06-22 18:19:02+02:00: start building...
16:22:32  <planetmaker> yes... 1:45m here
16:22:42  <Ammler> 2010-06-22 18:21:41+02:00: end building...
16:23:07  <planetmaker> and most probably 2:00m of that is real sprites encoding
16:23:50  <Ammler> ogfxplus (ERROR r30) <-- btw. :-)
16:24:16  <Yexo> Ammler: it builds fine with the latest nml
16:24:17  <planetmaker> hm
16:24:28  <Yexo> yesterday it failed because of old nml version, when you rebuild again it'll succeed
16:24:37  <Ammler> hmm, so I need to add possibiltiy to retrigger same version?
16:24:43  <planetmaker> yup
16:24:43  <Yexo> yes
16:24:44  <planetmaker> currently
16:24:46  <Ammler> :-P
16:25:39  <Ammler> planetmaker: you could simply touch a file :-P
16:26:27  <planetmaker> what would it help?
16:26:31  <planetmaker> touch != diff
16:26:45  <Ammler> it just needs one commit
16:27:38  <Ammler> planetmaker: wrap the readme :-P
16:29:16  <Ammler> I need the possiblity to retrigger building if a buildrequired changes
16:29:18  <Yexo> I haven't updated the readme when I added the small airport sprites, that could be a commit
16:29:53  <Ammler> but how shall I detect that?
16:30:14  <Ammler> I would need to check all repos, if they use the package
16:30:33  <Yexo> a way to manually trigger a revision to be rebuld would be easier
16:30:44  <Yexo> maybe only make that possible if it failed
16:30:47  <Ammler> that is easy
16:30:52  <Ammler> just add a commit
16:31:43  <planetmaker> :-)
16:32:05  <Ammler> but maybe we should trigger nightly rebuild on every push, if the last nightly build failed
16:32:12  <Ammler> so you don't have to wait for the next day
16:34:31  <Ammler> well, adding a commit could also mean for example run touch .devzone/build/nightlies/enable
16:35:05  <Yexo> you can't commit a "touch"
16:35:16  <Ammler> are you sure?
16:35:20  <Yexo> file modification date is not stored in the repo
16:35:28  <Ammler> hmm
16:35:51  <Yexo> I really think it's a non-issue
16:36:38  <Ammler> make mrproper should remove
16:36:39  <Ammler> ?
16:36:41  <Ammler> ? parser.out
16:36:42  <Ammler> ?
16:36:44  <Ammler> ? sprites/ogfxplus.nml
16:37:06  <planetmaker> hm, the debug mode w/o railtypes was a good idea
16:37:14  <Yexo> is not generated at all
16:37:16  <Brot6> Swedish Rails - Revision 68:4f100d943f0b: Feature: Add custom fences, controlled by parameter 2 (planetmaker) @
16:37:17  <planetmaker> It looks totally screwed
16:37:25  <Yexo> maybe only by a very old version that still used grfcodec
16:37:47  <planetmaker> ogfxplus used it very early, yes
16:37:58  <planetmaker> when there was no grf output in nml
16:38:28  <Ammler> yeah, touch doesn't work
16:38:36  <Ammler> so you need to make a real change
16:39:12  <Ammler> "fixing" readme is one :-)
16:39:26  <planetmaker> yeah
16:39:39  <planetmaker> current ogfxplus has at least one real bug, too
16:39:45  <planetmaker> with wagon cargos
16:39:58  <Ammler> I think, we don't need a special retrigger function
16:40:30  <Ammler> the only thing I could think about is that it does try to build nightly on every push until it has a working one
16:40:54  <planetmaker> :-)
16:41:45  <Rubidium> it needs a dependency chain and retriggers when a dependency has changed! :)
16:42:03  <Ammler> [18:29] <Ammler> I need the possiblity to retrigger building if a buildrequired changes
16:42:27  <Rubidium> then do it! :)
16:42:30  <Ammler> :-P
16:43:02  <Rubidium> can't quite help you with it as OpenTTD doesn't quite need it
16:43:25  <Yexo> Ammler: as a workaround you could also try to build a failed nightly every night even if there were no commits, that way if a dependency was update it'll be used
16:43:45  <Ammler> oh, it shouldn't be that hard, I just need "hg cat .devzone/build/require | grep nml"
16:44:35  <Ammler> and then adding it to the QUEUE, QUEUE is already supported
16:45:00  <Ammler> since someone could trigger release at same time
16:45:34  <Ammler> I think about  :-P
16:48:33  <Ammler> I could also install the full OBS
16:48:53  <Ammler> then we have native repo support :-P
16:49:03  *** ODM has joined #openttdcoop.devzone
16:51:06  <Ammler> the question is, shall we also rebuild successful packages?
16:51:11  <Yexo> planetmaker: using swedishrails with "no fences" results in 1 white pixel for every fence
16:51:26  <planetmaker> which version?
16:51:35  <Yexo> you use "1, 26" as x,y position but that is a white pixel in fences.png
16:51:46  <Yexo> r68
16:51:54  <planetmaker> ok, I'll look, thx
16:53:29  <planetmaker> indeed... somehow got *completely* wrong coordinates
16:53:57  <Yexo> and maybe add another option "4 = Use no fences for swedishrails, default fences for other railtypes"
16:54:19  <planetmaker> that's present. Option1
16:54:27  <planetmaker> oh.
16:54:30  <planetmaker> no, it's not ;-)
16:55:02  <Ammler> hmm, if you start adding other things then rails, you might also add a parameter to disable swedish rails
16:55:09  <planetmaker> good idea, though, Yexo
16:55:31  <planetmaker> hu, Ammler ?
16:55:52  <Ammler> swedish tunnels with normal rails
16:56:03  <Ammler> or swedish depots with normal reails etc...
16:56:28  <planetmaker> why disable rails in a rail set?
16:56:45  <Ammler> how can I use the depots without the rails?
16:56:51  <planetmaker> you can't
16:56:58  <Ammler> that's why :-)
16:57:11  <planetmaker> well, this is a rail set. Not a depot replacement
16:57:37  <Ammler> ok, you learned from other set authors :-P
16:57:58  <planetmaker> and the multitude of depots is also only available, if you code it as rail types
16:58:18  <planetmaker> Ammler: it would just be a lot of work to include that here
16:58:55  <Ammler> yeah, it is ok, I see why others also don't do it
16:58:56  <planetmaker> if someone wants just a depot replacement, it's easy to take the templates, the png and call - depending upon a parameter - the proper actionA
16:59:30  <Ammler> it isn't because they want you to use it like they made, it is because it makes it complicated to code
16:59:55  <planetmaker> it'd add at least two additional layers of iffs
17:00:10  <Ammler> I thought, such things are easy with nml
17:00:18  <planetmaker> and checking againf for parameters in a different manner
17:00:34  <planetmaker> well, they are easier to write. Much easier
17:01:06  <planetmaker> But still... I don't see much gain it it.
17:01:10  <Ammler> hehe
17:01:25  <planetmaker> it'd add ifs in many places which make maintenance much more complicated
17:02:42  <Ammler> I see
17:09:01  <planetmaker> I perfectly understand that one would want especially the old depots here in combination with other rails :-)
17:21:50  <Brot6> Swedish Rails - Revision 69:3edf12a6eb03: Fix: No fences also means no white dots (planetmaker) @
17:24:00  * planetmaker ponders another release of swedish rails
17:24:10  <planetmaker> hm... no
17:24:13  <planetmaker> depots need updates
17:26:49  *** Alberth has joined #openttdcoop.devzone
17:38:51  <planetmaker> hmpf... crashed openttd by updating the swedishrails grf :-(
17:41:28  <Brot6> NewGRF Meta Language - Revision 457:835590c792a0: Add: check if expressions are supported by acti... (yexo) @
17:42:54  <Alberth> hai hai
17:43:03  <Yexo> hello Alberth
17:43:05  <planetmaker> moin Albert
17:43:09  <planetmaker> +h
17:46:01  <planetmaker> catch you all later
18:05:41  <Alberth> NML: "src/header.pnml", line 16: Operator '==' not supported in parameter assignment  <-- is this a new restriction?
18:05:56  <Alberth> param[10] = ((ttd_platform == PLATFORM_TTDPATCH) | ((ttd_platform == PLATFORM_OPENTTD) & (openttd_version < version_openttd(1, 0, 0)))) * 3 | (param[0]); // Use TTD ground sprites
18:06:21  <Yexo> no, it's an error in the new data tables
18:07:39  <Brot6> NewGRF Meta Language - Revision 458:ccfe01365c5c: Fix: comparison operators are of course support... (yexo) @
18:08:22  *** Alberth1 has joined #openttdcoop.devzone
18:08:22  *** Alberth is now known as Guest812
18:08:22  *** Alberth1 is now known as Alberth
18:08:57  <Alberth> I knew I should not have pulled the new version :p
18:10:01  <Yexo> well, it isn't in the regression test
18:12:56  <Brot6> NewGRF Meta Language - Revision 459:515a2cde45c6: Fix: token -> self.token (yexo) @
18:13:27  <Alberth> of course not
18:14:09  *** Guest812 has quit IRC
18:31:21  <Brot6> NewGRF Meta Language - Revision 460:d30e2e79979a: Codechange: refactor the varaction2 code to use... (yexo) @
18:50:13  <Brot6> NewGRF Meta Language - Revision 461:2f2425ae18bd: Add: operators <= and >= (yexo) @
19:12:22  <Brot6> NewGRF Meta Language - Revision 462:29e2f49a9132: Fix: action7 would in some cases be used instea... (yexo) @
19:16:39  <Brot6> NewGRF Meta Language - Revision 463:41393d6a1a76: Revert (r462): don't commit/push too soon, the ... (yexo) @
19:38:53  *** Seberoth has quit IRC
19:40:50  *** Seberoth has joined #openttdcoop.devzone
19:50:56  <Alberth> hmm, 'speedup' from 2:11 to 12:06... not good :p
19:51:24  <Alberth> and it was such a nice change :(
19:53:43  <Yexo> ouch
19:54:07  <Yexo> swedishrails doesn't work currently, that's know
19:57:11  <Alberth> yeah, I read that it breaks openttd, but I am only interested in encoding speed at the moment, so that is not a problem, I think
19:57:42  <Yexo> I ment encoding swedishrails with nml
19:57:52  <Yexo> < and > operators are broken
19:57:56  <Yexo> or rather no longer implemented
20:03:53  <Alberth> that happened after r459?
20:04:40  <Yexo> yes, r461 (but before that they were off-by-one)
20:06:23  <Alberth> I will not update for a while then
20:12:24  <Brot6> NewGRF Meta Language - Revision 464:6e0ceb594cbf: Fix: there actually still was code implementing... (yexo) @
20:30:27  <planetmaker> hi again
20:30:55  <Alberth> hi
20:31:32  <Alberth> does swedishrails have a standard for line endings of text files?
20:31:40  <Alberth> (in the repo)
20:31:47  <Alberth> if yes, what is it?
20:32:25  <Yexo> too many fixes of fixes tonight :(
20:32:32  <Yexo> anyway, everything should work again
20:32:40  <Alberth> too much in a hurry :)
20:33:06  <Brot6> NewGRF Meta Language - Revision 465:2247ca97e122: Add: really implement <= and >= for actionD now (yexo) @
20:34:15  <planetmaker> [22:31]	<Alberth>	does swedishrails have a standard for line endings of text files? <-- everything should usually be unix
20:34:38  <planetmaker> is there different? Then I'd wonder where it comes from as I don't have windoze
20:35:50  <Alberth> ok, then docs/license.ptxt and lang/default.lng are messed up
20:36:08  <planetmaker> he
20:36:22  <Yexo> lang/default.lng is probably indirect from me
20:36:30  <planetmaker> probably
20:36:39  <planetmaker> and license from someone / somewhere else
20:36:51  <planetmaker> maybe directly downloaded from the website
20:37:04  <Alberth> quite likely
20:37:10  <Yexo> if ogfx+ has the same problems then there is a commit to it's rebiuld tomorrow :p
20:37:17  <planetmaker> :-)
20:40:51  <Brot6> NewGRF Meta Language - Revision 466:0e4dd1582aad: Fix: basestring != Identifier. (Hirundo) @
20:43:21  <Brot6> NewGRF Meta Language - Revision 467:ec4d4c2127b6: Fix: (r466) Although not strictly necessary (it... (Hirundo) @
20:50:50  <Brot6> NewGRF Meta Language - Revision 468:9e63ddd129bb: Fix: print the offending operator if an unsuppo... (yexo) @
21:00:24  *** frosch123 has quit IRC
21:12:01  <Alberth> good night
21:12:32  <PeterT> good night Alberth
21:13:09  <Yexo> gn Alberth
21:14:16  <Alberth> Yexo: speeding up lz77 is not going well, every change slows it down :(
21:14:26  <planetmaker> good night Alberth
21:14:44  *** Alberth has left #openttdcoop.devzone
21:14:54  <Brot6> NewGRF Meta Language - Revision 469:3e958d5dc62d: Feature: support for all comparison operators i... (yexo) @
21:17:50  <Yexo> planetmaker: with the latest nml this works now:
21:18:30  <planetmaker> ah, nice :-)
21:18:41  <Yexo> should be >= actually, not >
21:19:03  <Hirundo> nice work Yexo
21:19:17  <Yexo> thanks
21:20:16  <planetmaker> indeed
21:22:46  <Yexo> missing operators: xor, min, max in actionD code
21:22:51  <Yexo> and shift for everything
21:23:31  <planetmaker> what about not?
21:25:20  <Yexo> also missing, but there is no trace of that in the code at all :p
21:25:25  <Hirundo> not(expr) = min_unsigned(expr, 1) ^ 1
21:25:52  <Yexo> min/max are not supported in actionD
21:26:25  <Yexo> and logical and/or is also missing
21:26:40  <Hirundo> does a < b ? a : b work in actionD?
21:26:52  <Yexo> yes :)
21:27:09  <Yexo> of course there are workaround like that, it's just not yet implemented :)
21:27:52  <Yexo> not(expr) = expr ^ 1 if we're sure that expr is already a boolean expression (either 0 or 1)
21:27:56  <Hirundo> Implementation may be equal to that workaround, although evaluating a or b twice isn't optimal
21:28:12  <Hirundo> min_unsigned(expr, 1)
21:28:16  <Hirundo> does just taht
21:29:18  <Yexo> when the original is not(a == b) you can leave out that step as == already returns 0/1
21:31:08  <Rubidium> what operations do you have?
21:31:37  <Rubidium> because "expr == 0" kinda equals "!expr"
21:31:43  <Yexo>
21:32:08  <Rubidium> and not(a == b) should probably be rewritten to a != b :)
21:32:53  <Yexo> yes, but ==/!= is more expensive then xor 1, at least in varaction2
21:39:53  <Yexo> expr/expr is the smallest way to do expr->boolean in with actionD
21:40:15  <Yexo> the alternative is res = 0; expr != 0? res=1
21:40:23  <Yexo> that uses 2 actionD's and an action7
21:51:37  *** ODM has quit IRC
22:55:39  <Brot6> Swedish Rails - Revision 70:2a71fcb4d0fd: Change: Increase height of modern depots slightly (planetmaker) @
22:55:39  <Brot6> Swedish Rails - Revision 71:ec6bc1b97824: Simplify version query for depots a bit (Yexo) (planetmaker) @
23:05:28  <Yexo> planetmaker: logical not operator implemented (syntax: !expr)
23:05:40  <Brot6> NewGRF Meta Language - Revision 470:c42aa0588242: Add: is_boolean() function to expressions (yexo) @
23:05:41  <Brot6> NewGRF Meta Language - Revision 471:773e1468915b: Fix: define precedence rules for <= and >= oper... (yexo) @
23:05:41  <Brot6> NewGRF Meta Language - Revision 472:9720450b559f: Add: code to force an expression to boolean (yexo) @
23:05:42  <Brot6> NewGRF Meta Language - Revision 473:814025534f08: Feature: support logical and/or operators (&& a... (yexo) @
23:05:45  <planetmaker> great :-)
23:05:46  <Brot6> NewGRF Meta Language - Revision 474:570f8ea07e1d: Feature: logical not operator (yexo) @
23:05:49  <Brot6> NewGRF Meta Language - Revision 475:c168c35e743e: Change: optimize the logical not operator somew... (yexo) @
23:06:04  <planetmaker> thanks a lot
23:10:14  <Yexo> and also logical operators (&& and ||), they'll force each of their parameters to boolean before doing a logical and
23:10:33  <Yexo> 2 && 1 will result in 1, while 2 & 1 will result in 0
23:11:07  <Yexo> 2 || 1 will be 1, while 2 | 1 will be 3
23:24:19  <planetmaker> oh nice :-)
23:26:16  <Brot6> NewGRF Meta Language - Revision 476:81236136e62f: Fix: debug print for BinOp (the operator part) ... (yexo) @
23:28:24  <Brot6> NewGRF Meta Language - Bug #985 (Closed): conditional expressions (yexo) @
23:56:36  <Brot6> Swedish Rails - Revision 72:7aca12e6a155: Feature: Add company fences with some company colour an... (planetmaker) @
23:57:10  *** KenjiE20 has quit IRC

Powered by YARRSTE version: svn-trunk