Log for #openttdcoop.devzone on 27th September 2011:
Times are UTC Toggle Colours
00:51:34  *** supermop has joined #openttdcoop.devzone
04:11:25  *** supermop has quit IRC
07:49:07  *** hanf has joined #openttdcoop.devzone
08:42:03  *** hanf has quit IRC
10:22:35  *** hanf has joined #openttdcoop.devzone
11:41:12  *** hanf has quit IRC
12:01:19  * Hirundo plays with FIRS
12:02:08  <Hirundo> pity that the industry placement makes rivers are almost, but not quite, entirely unfit for transportation
12:02:36  <planetmaker> say again?
12:03:03  <planetmaker> why do you think that rivers are not fit for transport, given the industry placement?
12:04:35  <Hirundo> it's a 256^2 map with relatively few industries of the same type
12:04:38  <Hirundo> typically 1-2
12:05:23  <Hirundo> So the chances of having 2 industries along the same river that are 'compatible' are quite slim
12:05:49  <Hirundo> And in the meantime UKRS supplied me with 1000hp+ engines, so I'm good :-)
12:06:37  <planetmaker> :-D
13:54:18  <planetmaker> Yexo, did you have a chance to look at ?
13:54:42  <Yexo> yes, I just forgot to comment (so here goes):
13:55:05  <Yexo> I don't like the autodetection of .dep output when given to the "-o" option
13:55:26  <Yexo> IIRC gcc uses -M (or -MM?) as name, perhaps copy that instead of introducing "--dep"?
13:56:24  <Yexo> instead of "self.skip_sprite_checks" just make skip_processing do "return False" and override the function in OutputDEP (as opposed to setting that variable)
13:58:36  <planetmaker> nml does auto-detection of all dep output. And so does gcc, if not -MT is given
13:59:21  <planetmaker> and --dep follows the nml style of output: --grf --nfo --nml etc
14:00:00  <Yexo> planetmaker:  "-M Instead of outputting the result of preprocessing, output a rule suitable for make describing the dependencies of the main source file. The preprocessor outputs one make rule containing the object file name for that source file, a colon, and the names of all the included files, including those coming from -include or -imacros command line options. "
14:00:27  <planetmaker> yes
14:00:37  <planetmaker> but that says nothing about the filename
14:00:39  <Yexo> I also don't really like the "out.grf" as default value, but I can't think of a better alternative
14:00:44  <Yexo> other than disallowing it
14:00:57  <Yexo> planetmaker:  it wasn't about the filename, I just meant use "-M" instead of "--dep" as option name
14:00:58  <planetmaker> could use stdout
14:01:19  <Yexo> stdout would be nice
14:01:22  <Yexo> since input is on stdin
14:01:23  <planetmaker> I don't mind really the name of the option... but -M is not really intuitive
14:02:07  <planetmaker> ok, but so that I understand you: you (only) mind the default filename when nml gets its input from stdin?
14:02:34  <Yexo> yes
14:03:10  <Yexo> there is really no good default value for it, so perhaps disallow it completely?
14:03:13  <planetmaker> ok :-). Making a default method which returns False and True for the depcheck is easy, too
14:03:17  <Yexo> or add another option to specify the name?
14:03:31  <planetmaker> Yes, I guess I'll (first) just disallow it...
14:03:53  <planetmaker> The option to actually specify a name was something I kept in mind all the time - which would correspond to -MT with gcc
14:04:11  <planetmaker> But which I didn't want to include in the main step (it's IMHO something separate)
14:04:12  <Yexo> ah, yes :)
14:05:36  <planetmaker> but ok, I'll for now disallow -M when called from stdin
14:05:51  <Yexo> I do agree that --dep is more descriptive than -M, but I'm still in favor of copying gccs naming since it makes it easier to copy the other options too
14:06:11  <Yexo> -MG for example might be useful too when you create the png files from gimp files
14:06:16  <planetmaker> yes, agreed. I swayed this way and that - and chose --dep ;-)
14:06:30  <planetmaker> thus I don't really mind either way
14:07:14  <planetmaker> hm, what was -MG?
14:07:22  <Yexo> In conjunction with an option such as -M requesting dependency generation, -MG assumes missing header files are generated files and adds them to the dependency list without raising an error. The dependency filename is taken directly from the #include directive without prepending any path. -MG also suppresses preprocessed output, as a missing header file renders this useless.
14:07:51  <Yexo> although it seems like that we don't check the existance of png files anyway during dep check
14:07:55  <Yexo> so we won't need it
14:09:00  <planetmaker> hm... yes. When files are script-generated
14:09:12  <planetmaker> not png but pnml files
14:09:24  <planetmaker> like for cets
14:09:40  <Yexo> but nml isn't involved yet at that step
14:09:54  <Yexo> that's only for the gcc preprocessor, nml only ever sees the final nml file
14:10:05  <planetmaker> hm, yes
14:10:34  *** hanf has joined #openttdcoop.devzone
14:10:49  <planetmaker> anyway, thanks for the feedback. I'll implement those changes then.
14:11:13  <planetmaker> Seems I got the gimp thing now working properly, too. And rewrote 25% of the makefile framework :-P
14:11:37  <planetmaker> with the added benefit that 'make clean' doesn't build deps anymore :-)
14:11:52  <Yexo> nice :)
14:12:13  <Yexo> with nml creating deps we have to do the gcc preprocessing before we can create all deps
14:12:36  <planetmaker> Yes, I know
14:13:13  <planetmaker> <-- if you want to try with ogfx+rv
14:13:18  <planetmaker> ^ also Terkhen
14:13:26  <planetmaker> it requires the NML patch, too
14:13:39  <Terkhen> I'll check it in a few hours :)
14:13:47  <planetmaker> np :-)
14:19:38  <planetmaker> Yexo, alternatively we need to teach nmlc the include command ;-)
14:20:28  <Yexo> yes, that would make #2929 easier too
14:20:28  <Brot6> Yexo: yes: #2929 is "NewGRF Meta Language - Feature Request #2929: More detailed error location - #openttdcoop Development Zone"
14:21:05  <planetmaker> hm, but nmlc already reports the file the error is in
14:21:16  <planetmaker> what would I want more than filename:linenumber?
14:21:42  <Yexo> the "in file included from ... included from ..." that gcc does
14:21:56  <planetmaker> fair enough :-)
14:21:59  <planetmaker> I didn't think of that
14:22:42  <planetmaker> that usual is not a big limitation, though
14:22:47  <planetmaker> +ly
16:23:39  * Terkhen is installing virtualbox again
16:37:36  <Terkhen> hmm... I lost my stored VM state
16:38:30  <Terkhen> planetmaker: can you remind me the urls of the patches?
16:39:17  *** hanf^ has joined #openttdcoop.devzone
16:41:49  <Terkhen> heh, it seems that I didn't turn off the machine (instead of saving the state) since august 16th
16:42:27  *** hanf is now known as Guest11763
16:42:27  *** hanf^ is now known as hanf
16:42:37  <Terkhen> 177 changesets for FIRS :P
16:44:52  *** Guest11763 has quit IRC
17:21:07  <Brot6> vactrainset: compile of r1 still failed (#3044) -
17:48:04  <planetmaker> <-- Terkhen
17:48:09  <planetmaker> the NML patch didn't change
17:48:34  <planetmaker>
17:48:45  <Terkhen> yes, but I couldn't remember its name
17:48:47  <Terkhen> thanks :)
17:52:22  <planetmaker> hm, why did you ever remove virtualbox? :-)
17:55:53  <Terkhen> I moved the VM to a different PC
17:56:03  <Terkhen> I moved the "hard drive" but I forgot to take the saved state
17:56:11  <planetmaker> ah
17:56:12  <planetmaker> :-)
17:56:25  <Terkhen> and it's not important enough to wait for it, I only lost some checkouts
17:57:25  <hanf> is something like this correct for a spritegroup def?
18:00:59  <hanf> I ask because it doesn't show the intermediate steps
18:01:46  <Terkhen> why is loaded using 4 different sprites?
18:02:20  <hanf> hurf I thought they were the other way around
18:02:37  <hanf> well that would explain it. :V
18:03:38  <Terkhen> I'm not sure either, I never coded gradual loading spritegroups :)
18:04:03  <Terkhen> planetmaker: it compiled fine :)
18:04:17  <planetmaker> great news :-)
18:04:42  <Terkhen> hanf: <--- according to this, they can have more than one sprite
18:04:43  <planetmaker> thanks for testing
18:04:51  <Terkhen> planetmaker: thank you for coding it :P
18:04:54  <hanf> Terkhen: they can, it was just a real dumb typo
18:05:01  <Terkhen> ok :)
18:46:47  *** JVassie has joined #openttdcoop.devzone
19:02:07  *** andythenorth has joined #openttdcoop.devzone
19:02:18  <Brot6> clientpatches: compile of r22964 still failed (#2964) -
19:07:15  *** frosch123 has joined #openttdcoop.devzone
19:07:30  <Brot6> openttd-vehiclevars: update from r22963 to r22964 done -
19:09:08  <Brot6> serverpatches: compile of r22964 still failed (#2966) -
19:11:02  <Brot6> 32bpp-ez-patches: compile of r22964 still failed (#2446) -
19:50:21  *** andythenorth has quit IRC
21:19:00  *** frosch123 has quit IRC
22:10:29  *** hanf has quit IRC
22:59:02  *** JVassie has quit IRC

Powered by YARRSTE version: svn-trunk