Log for #openttdcoop.devzone on 20th March 2010:
Times are UTC Toggle Colours
00:01:17  <andythenorth> that's expected
00:01:34  <andythenorth> next: a bigger test!
00:01:38  <andythenorth> then: enhancements!
00:03:06  <DJ_Nekkid> the interesting thing is, the number of them are quite similar in 1920
00:03:43  *** frosch123 has quit IRC
00:04:23  <DJ_Nekkid> average is 128...
00:05:10  <DJ_Nekkid> 128/20 is roughly 6 :)
00:15:06  *** andythenorth has left #openttdcoop.devzone
00:52:28  *** KenjiE20 has quit IRC
01:00:32  <Ammler> PeterT: that is newgrf pack, no baseset replacement
01:00:41  <Ammler> but can finally be as big...
01:00:46  <PeterT> Sure sounded like one
01:01:00  <Webster> Latest update from devactivity: OpenGFX - Bug #848 (Confirmed): Makefile doesn't replace {{TAG}} anymore <> || OpenGFX - Feature #208: Reorder ogfx_extra to make it easier to maintain <>
01:08:47  *** tneo has quit IRC
01:09:27  *** tneo has joined #openttdcoop.devzone
01:26:13  <Rubidium> you're not thinking of releasing ogfx this weekend, right?
01:26:44  <Ammler> I don't
01:26:57  <Rubidium> okay, that's good
01:26:59  <Ammler> I just prepared it it bit
01:27:27  <Ammler> why would it be bad?
01:27:59  <Rubidium> for releasing osfx around the same time
01:28:08  <Rubidium> and that still needs some makefile stuff
01:28:28  <Ammler> oh well, 0.2.2 would be a kind of RC
01:28:59  <Rubidium> make[1]: *** No rule to make target `-.pcx', needed by `sprites/nfo/extra/extra-openttd-gui.pnfo'.  Stop.
01:29:01  <Ammler> but if pm isn't around, it doesn't matter
01:29:02  <Rubidium> :(
01:29:47  <Ammler> yeah, I guess, I commit my workaround...
01:29:57  <Ammler> make bundle_* works
01:30:06  <Ammler> it is a bit strange
01:30:41  <Rubidium> just don't try your luck with fancy characters
01:30:51  <Rubidium> I fear + doesn't quite work well on Windows
01:31:08  <Rubidium> I seem to remember copy a+b+c d
01:33:31  <Ammler> pushed...
01:33:41  <Rubidium> ah, yes... VFAT/FAT32 has "+" as reserved character [Wikipedia]
01:33:47  <Ammler> wasn't aware + is bad :-)
01:34:15  <Ammler> and it the silly thing was that make bundle_tar worked
01:34:16  <Rubidium> If you want a filename that works on your Macintosh or Windows computer and also when uploaded, restrict yourself to these characters: ${}^[]`=,;`abcdefghijklmnopqrstuvwxyz._-0123456789.html [Wikipedia]
01:34:26  <Ammler> so I thought, it is a bug of the Makefile...
01:35:47  <Rubidium> personally I'd say, limit yourself to abcdefghijklmnopqrstuvwxyz._-0123456789
01:35:47  <Ammler> he, I would never use the first 10
01:35:53  <Ammler> yes :-)
01:37:33  <Ammler> the list misses space, not that I would use it, just wondering...
01:37:45  <Ammler> as you see a lot of filenames with it
01:38:20  <Ammler> Rubidium: openttd should also restrict to that for the saves and screens :-)
01:39:23  <Ammler> people don't rename the files before uploading...
01:42:16  <Rubidium> oh, make clean removes too much in a source bundle
01:42:29  <Rubidium> source bundle as in .tar.gz
01:42:39  <Rubidium> is there already a bug report about that?
01:42:47  <Ammler> not sure
01:43:09  <Ammler> there is a bug report about no .hg files needed there
01:44:15  <Ammler> also a silly one:
01:45:14  <Rubidium> so.. then I'll make a ticket
01:46:31  <Ammler> yeah, please do.
01:46:49  <Webster> Latest update from devactivity: OpenGFX - Revision 375: Fix (r366): the buildsystem seems not to like +- in a file name? <>
01:52:40  <Rubidium> oh, have "you" though about testing all stuff with a tagged version, i.e. with a "real" release scenario?
01:53:11  <Ammler> yes, that is why I like to make the release :-)
01:53:35  <Rubidium> but you can test 99% of it without doing the actual release
01:53:46  <Rubidium> just locally tag it, test stuff and then rollback the tagging
01:54:20  <Rubidium> something to play with tomorrow though
01:55:30  <Ammler> oh, yes, I did that,
01:55:39  <Ammler> so far worked...
01:55:56  <Ammler> well, except readme update
01:56:37  <Ammler> that is already posted too
01:57:22  <Ammler> but, tests with obs I liked to make with real release...
02:00:46  <Ammler> last time I tried with nightly, building failed on suse factory
02:01:42  <Ammler> [NFORENUM] ogfxe_extra.nfo
02:01:43  <Ammler> Error: invalid compression, data diff at 275 of 576 bytes, trying without it for this sprite
02:01:53  <Webster> Latest update from devactivity: OpenGFX - Feature #850 (New): Makefile target to test md5 checksums <> || OpenGFX - Revision 376: Fix (last): remove the bad file <> || OpenGFX - Bug #849 (New): make clean removes too much with source tarball <>
02:02:05  <Ammler>
02:17:06  <Webster> Latest update from devactivity: 2cc train set - Bug #851 (New): Nightly builds are failing <>
02:32:22  <Webster> Latest update from devactivity: 2cc train set - Bug #851: Nightly builds are failing <>
03:44:37  *** Frankr has quit IRC
06:47:11  *** ODM has joined #openttdcoop.devzone
07:20:12  *** ODM has quit IRC
07:23:12  *** ODM has joined #openttdcoop.devzone
07:27:40  *** DJ_Nekkid has quit IRC
07:42:54  *** DJ_Nekkid has joined #openttdcoop.devzone
07:49:12  *** andythenorth has joined #openttdcoop.devzone
08:04:12  <Webster> Latest update from devactivity: OpenGFX - Feature #850 (Rejected): Makefile target to test md5 checksums <" target="_blank">> || OpenGFX - Bug #848 (Feedback): Makefile doesn't replace {{TAG}} anymore <" target="_blank">> || OpenGFX - Revision 377: Fix (r316) [#848]: Replacement of some {{TAG}} didn't work properly. Used... <> || OpenGFX - Feature #850 (Rejected): Makefile target to test md5 checksums <> || OpenGFX - Bug #848 (Feedback): Makefile doesn't replace {{TAG}} anymore <>
08:36:17  <Webster> Latest update from devactivity: OpenGFX - Feature #771 (Feedback): bundle md5 for source releases and check for them. <> || OpenGFX - Bug #849 (Feedback): make clean removes too much with source tarball <" target="_blank">> || OpenGFX - Revision 378: Feature [#771]: Add check for md5sums <> || OpenGFX - Bug #849 (Feedback): make clean removes too much with source tarball <>
08:38:36  <planetmaker> moin
08:48:57  <andythenorth> he's back :)
08:48:59  <andythenorth> morning
08:53:54  <planetmaker> yes. Semi. I'll be 'more' back on Monday, but yes :-)
08:54:21  <planetmaker> how are things going, andythenorth ? :-)
08:54:35  <andythenorth> ummm
08:54:38  <andythenorth> frustratingly
08:54:44  <andythenorth> :|
08:54:52  <planetmaker> I see you started a big test concerning lifetime of industries?
08:54:59  <planetmaker> oh... frustrationg? :-(
08:55:53  <andythenorth> industry closing code has a fundamental flaw
08:55:59  <planetmaker> I truely wished I could give you more of a helping hand than I'm doing right now... but I'd need a clone of myself :S
08:56:14  <andythenorth> ottd has a fundamental flaw in industry placement
08:56:25  <andythenorth> make is slow :P
08:56:30  <andythenorth> it's a hard knock life!
08:56:47  <planetmaker> hm... in which way is the placement / closing code flawed?
08:56:57  <planetmaker> haha :-)
08:57:24  <andythenorth> planetmaker: there is a check for secondary industries: if they accept the same cargo, they cannot be built near each other
08:57:38  <andythenorth> it made sense before newgrf industries, but now looks stupid
08:57:43  <planetmaker> uh? In principle? that's bad, yes
08:57:54  <planetmaker> :-(
08:58:05  <andythenorth> not only that, but the error message to players is really abrupt and unhelpful
08:58:08  <andythenorth> I have two choices
08:58:26  <andythenorth> 1. no secondary industries that accept same cargo
08:58:32  <planetmaker> hmpf
08:59:01  <andythenorth> 2. write a complicated action 2 to check for other industries and replace the unhelpful error with a better one
08:59:08  <andythenorth> or 3. get a lot of player complaints
08:59:24  <andythenorth> 2 is a lot of work for no benefit other than to prevent 3.
09:00:20  <planetmaker> hm... what error message do we get when placing an industry?
09:01:35  <andythenorth> Can't construct this industry type here...too close too another industry
09:01:38  <andythenorth> doesn't say which
09:01:55  <andythenorth> I've seen it for months, I thought it was a mistake with FIRS :|
09:02:11  <andythenorth> yesterday I checked all FIRS mistake :|
09:02:13  <planetmaker> hm, ok, that's what I'd have thought, too
09:02:37  <planetmaker> and it always happens, if they share an accepted cargo? I'd consider it a bug...
09:03:27  <andythenorth> me too
09:03:56  <andythenorth> apparently it's by design
09:10:45  <andythenorth> planetmaker: the prevention of industry construction in this way is to avoid bug reports from players due to the way stations distribute cargo
09:11:03  <andythenorth> i.e. two industry at one station, but only one will accept cargo
09:11:45  <planetmaker> hm, ok, also understandable. Thanks for that piece of information.
09:12:23  <andythenorth> it's the most fricking stupid thing I've ever seen in the game though, it makes me sad :)
09:12:40  <andythenorth> the limit is 14 tiles.  A 6 tile station will bridge both the industries
09:13:10  <andythenorth> and the limit is ignored in the scenario editor, so in scenarios industry can be built next to each other
09:13:15  <planetmaker> :-) Especially 64-tile stations :-D
09:13:34  <andythenorth> so it (a) confuses players - it's totally at odds with the 'allow industry close to each other' option
09:13:35  <planetmaker> Well, SE of course should be different. That's for a game designer
09:13:47  <planetmaker> But I agree with your last statement
09:13:47  <andythenorth> (b) it doesn't fix the problem it claims to
09:14:02  *** Seberoth has joined #openttdcoop.devzone
09:14:21  <planetmaker> hm... I guess it needs some thorough solution which may well be much more than a quick fix.
09:14:50  <planetmaker> like a thorough concept with compatibility accounted for and then implemented
09:15:49  <andythenorth> yep
09:16:27  <andythenorth> I am not the best person for that at the moment, I am somewhat glum :|
09:16:50  <andythenorth> the game's industry location code seems to have about 6 problems with it - frosch listed them somewhere
09:17:04  <andythenorth> so it could become too big a problem for anyone to bother to fix
09:18:04  <andythenorth> meanwhile, can we speed up make :D
09:18:38  <andythenorth>
09:20:24  <planetmaker> - we ban the use of hg push <-- that doesn't make sense. What do you mean instead of 'push'?
09:20:47  <andythenorth> so in workflow, authors never type 'hg push'
09:21:03  <planetmaker> make must not mess around with pushing changes IMHO
09:21:11  <planetmaker> also it cannot
09:21:16  <andythenorth> instead we use a shell script that runs dep check, makes the grf, and does the push ?
09:21:20  <andythenorth> not make
09:21:21  <planetmaker> that's conceptually wrong
09:21:56  <andythenorth> hmm :)
09:22:13  <planetmaker> out-sourcing the dep check to a shell script being called by make  - that might be an idea.
09:22:29  <planetmaker> but not replacing make by a shell script
09:22:32  <andythenorth> I think Rubidium showed that the dep check seems slow with hg
09:22:37  <andythenorth> compared to svn?
09:22:48  <Rubidium> no, just the version check
09:22:49  <andythenorth> for OTTD anyways
09:22:51  <planetmaker> I briefly read that.
09:22:53  <andythenorth> ah
09:23:01  <planetmaker> oh, hey Rubidium :-)
09:23:11  <Rubidium> findversion for svn is more than an order of magnitude faster than for hg
09:23:42  <andythenorth> planetmaker: forgive my ignorance, what does the dep check do?
09:24:06  <planetmaker> it checks which files are needed in order to compile the newgrf
09:24:24  <andythenorth> what will happen if a file is missing?
09:24:25  <planetmaker> so that make will complain, if one #include is missing
09:24:36  <andythenorth> and what will happen if we don't dep check?
09:24:38  <planetmaker> and fail
09:24:53  <planetmaker> it won't fail then and just produce a newgrf with missing features / industries / code
09:24:57  <andythenorth> ah
09:25:04  <andythenorth> yup
09:25:05  <planetmaker> that's why I consider it very important
09:25:08  <andythenorth> I see the point of that
09:25:26  <Rubidium> removing the depcheck 'only' causes make to not recompile when one of the included files is changed
09:26:05  <planetmaker> yes. So there's also the chance to not test what you thing you're testing
09:26:23  <planetmaker> except if you re-build everything every time even if it's not needed
09:26:31  <planetmaker> we could do that...
09:26:46  <andythenorth> if it was faster, I'd buy that for a dollar
09:27:29  <planetmaker> ok, dully noted. Don't expect me to revise it in the next two ... four weeks though.
09:27:40  * andythenorth considers learning about make
09:28:15  <planetmaker> Getting out a new OpenGFX has way higher priority than a few seconds shaved off of make
09:29:06  <planetmaker> and the momentum for that project has to be utilized to some advancement there even beyond the release :-)
09:29:08  <andythenorth> planetmaker: I have 93 minutes 'free' over the next 4 weeks if make gets 11s faster.  Should I invest them in make?? :D
09:29:27  <planetmaker> well... not sure
09:30:08  <planetmaker> if you like, you can. But don't expect to understand the concept behind my makefiles in 93 minutes, if you don't know about make yet :-)
09:30:20  <andythenorth> is knocking out the dep check a one line thing?  I would like to do it locally.  I can put a ticket in for FIRS 0.1 to restore it before release
09:30:52  <planetmaker> andythenorth: just edit your local makefiles and don't commit it
09:31:12  <planetmaker> but don't wonder if you don't test stuff which you thought you tested
09:31:21  <andythenorth> my risk :)
09:31:31  <andythenorth> what do I edit?
09:32:40  <planetmaker> even removing the dep check won't restore the old speed. For me it reduced run time from 22s to 17s, though
09:32:48  <planetmaker> while the old run time was 11s
09:32:54  <andythenorth> I'll take 5s
09:32:58  <andythenorth> I build a lot
09:34:14  <planetmaker> replace the last line of scripts/Makefile.common by $(_V) touch $@
09:35:09  <planetmaker> and remove Makefile.dep manually once
09:37:22  <andythenorth> 15s :)
09:37:28  <andythenorth> hmmm
09:37:34  <andythenorth> how to avoid committing that?
09:37:57  <planetmaker> I nearly always explicitly state which files I want to commit
09:38:22  <planetmaker> hg ci -m "blubber blah" sprites/nfo/file1.pnfo sprites/pcx/graphics1.pcx ...
09:38:38  <planetmaker> there's probably also a way to tell that you want to exclude a file. But dunno
09:39:03  <andythenorth> I'll look into it
09:39:06  <Webster> Latest update from devactivity: OpenGFX - Feature #208 (Closed): Reorder ogfx_extra to make it easier to maintain <>
09:42:28  <planetmaker> However you're certainly free to modify FIRS' Makefiles or revert that. But I see myself not really able to maintain a dozen always ever so slightly different Makefiles. That's a pain. And was my motivation to change them to one single system
09:42:56  <andythenorth> I don't want to introduce new problems ;)
09:43:48  <planetmaker> one of the advantages also using it in more than one project: problems surface faster :-)
09:44:26  <planetmaker> like the speed. I agree that a factor of two slower is not really nice - though it didn't strike me as crucial either I have to admit ;-)
09:44:53  <andythenorth> i build often :)
09:45:12  <planetmaker> well. As do I.
09:45:56  <planetmaker> But having made sure that I don't test the wrong thing (which I already did a few times) seems to me worth the trade-off
09:46:07  <andythenorth> working with registers and text strings, I get a lot of silly issues like 'byte should be word' etc :)
09:46:15  <planetmaker> And that easily eats those 93 minutes you quote
09:46:25  <planetmaker> it can eat days actually
09:46:44  <andythenorth> can we have cheap, fast *and* good please?
09:46:45  <andythenorth> :P
09:47:29  <planetmaker> you have it ;-)
09:48:06  * andythenorth wonders if there is a bash way to make hg commit fail unless a file name is specified
09:48:12  <planetmaker> it probably would also speed up things, if I didn't calculate the VPATH
09:48:32  <planetmaker> andythenorth: there's probably a hg way to make it fail ;-)
09:49:54  <planetmaker> via commit hooks or so
09:52:42  * andythenorth gets excited about commit hooks.
09:52:47  <andythenorth> but doesn't find the answer yet
09:54:11  <andythenorth> hmm
09:54:11  <andythenorth>
09:54:14  <Webster> Latest update from devactivity: Example NewGRF Project - Feature #450 (Closed): remove target 'bananas' <>
09:55:32  <planetmaker> anyway... my fridge is still empty and I guess I should fill it :-) See you later!
09:58:48  <andythenorth> enjoy :)
10:10:36  <Webster> Latest update from devactivity: FIRS Industry Replacement Set - Feature #852 (New): Makefile.common may be missing dep check if A... <>
10:24:15  <Ammler> andythenorth: add it to your .hgignore
10:24:54  <andythenorth> Ammler: will that prevent hg bringing in changes from the remote repo though?
10:26:03  <Ammler> well, to your local hgignore
10:26:16  <Ammler> not the ignore file of the repo
10:27:05  <andythenorth> Ammler: where is my local ignore (or I can google if you're busy)
10:27:30  <Rubidium> planetmaker: regarding #849, that seems fine
10:27:31  <andythenorth> found it
10:27:33  <Ammler> man hgignore
10:28:02  <Ammler> but I aslo once started a wiki page
10:28:17  <andythenorth> Ammler: I'll look on the wiki
10:28:29  <andythenorth> I want to be sure I edit only my local .hgignore!
10:28:37  <Ammler>
10:31:12  <Rubidium> haven't tested it yet though
10:32:31  <Ammler> if we don't release this weekend, I will make a fake release for me...
10:35:32  <Rubidium> you could make a release candidate :)
10:35:39  <Rubidium> planetmaker:
10:37:42  <Ammler> Rubidium: yes, I thought 0.2.2 to be one
10:37:57  <Ammler> and then release 0.3 after it...
10:38:22  <andythenorth> hmm
10:38:41  <andythenorth> my local ignore isn't ignoring the file I want to ignore....
10:38:51  <Rubidium> but you're very far from the milestone you set for 0.3.0
10:40:20  <Ammler> yes, quite...
10:40:51  <andythenorth>
10:40:53  <andythenorth> meh
10:41:25  <Ammler> looks fine?
10:41:44  <andythenorth> pdq2s-macbook-3:.hg andy$ hg st | grep scripts
10:41:44  <andythenorth> M scripts/Makefile.common
10:42:28  <Ammler> how did you add that ignore file to your mercurial?
10:42:46  <andythenorth> [ui]
10:42:46  <andythenorth> ignore = /Users/andy/Documents/OTTD graphics/FIRS/firs_build/.hgignore_local
10:43:22  <Ammler> wow, space in the path :-o
10:43:41  <andythenorth> hmm
10:43:42  <Ammler> you might need to quote that...
10:44:12  <Ammler> and give it a name like ignore.firs
10:44:38  <andythenorth> hg doesn't like quotes for paths :o
10:44:43  <Ammler> but be aware, you do also ignore it for your other repos
10:45:08  <Ammler> or did you edit hgrc in firs?
10:45:16  <andythenorth> I edited hgrc in firs
10:46:06  <Ammler> I wonder, how you survived that long with a space in the path
10:46:38  <Ammler> thought, that is a windows only behavior
10:46:57  <andythenorth> Ammler: I think that the space is fine
10:47:10  <andythenorth> hg shows an error if it can't reach the file
10:47:16  <andythenorth> with the space there is no error
10:47:33  <andythenorth> and none of the usual escapes work anyway :|
10:47:54  <Ammler> hmm, prefix with "*"
10:48:11  <Ammler> *Makefile.common
10:48:18  <andythenorth> tried that, doesn't work :)
10:48:25  <andythenorth> tried changing the syntax
10:48:41  <Ammler> what if you add it testwise to the repo hgignore
10:48:45  <andythenorth> it's starting to consume the time that will be saved by having a faster makefile
10:48:53  <andythenorth> :D
10:49:05  <andythenorth> so I think I'll stop now and rely on the devzone ticket
10:50:30  <Ammler> Rubidium: we can also release 0.2.3 if needed...
10:51:57  <Ammler> without sprites, I see no ticket from 0.3 we can resolve now...
10:52:05  <Ammler> well the wiki, but...
11:00:35  <andythenorth> Ammler: what milestones are you working on?  I am *so* sick of FIRS right now, I want to work on something else
11:13:19  <Ammler> andythenorth:
11:13:40  <Ammler> bit you are always welcome to contribute something to opengfx
11:14:49  * andythenorth looks for tickets
11:15:19  <Ammler> you like to draw or to code?
11:15:46  <Ammler> molave just posted a big sprite sheet on tt-forums
11:16:32  <andythenorth> Ammler: I don't normally use opengfx...I am looking at
11:16:47  <Ammler>
11:16:49  <andythenorth> that i seems a strange choice for the windowshade icon ?
11:17:08  <Ammler> ?
11:17:54  <Ammler> well, eddi complained about it looking like a error sign, I did agree and made the ticket
11:18:06  <andythenorth> I agree
11:18:24  <andythenorth> where else is it used besides windowshade button?
11:18:30  <andythenorth> ah, get info
11:18:37  <Ammler> your painting style is more like opengfx than ttd
11:19:19  <andythenorth> nope, that's zephryis you're thinking of :)
11:19:27  <Ammler> sorry, you
11:19:29  <andythenorth> although I imagine it's highly subjective :)
11:19:58  <Ammler> specially antialising...
11:20:07  <andythenorth> a lot of my graphics use techniques directly copied from the original style ;)
11:20:16  <andythenorth> anyway, it's a weird use of that 'i' icon
11:20:39  <Ammler> yeah, would be nice, if you can do something about..
11:20:53  <andythenorth> and it makes window menu bars big
11:21:04  <Ammler> I really recommend you not using opengfx...
11:21:17  <andythenorth> seems silly to introduce windowshade to save screen space...then increase the size of menu bars :P
11:21:21  <Ammler> if you use it too long, ttd base set become really ugly
11:21:56  <andythenorth> :P
11:22:17  <Ammler> maybe a simple "pimp up" of ttd base set might be an idea for a new project?
11:22:38  <andythenorth> what would you change?
11:22:54  <Ammler> maglev tracks
11:23:02  <Ammler> snow tiles
11:23:16  <Ammler> water
11:25:56  <andythenorth> nice water in opengfx
11:26:01  <andythenorth> the waves go the wrong way though!
11:26:09  <andythenorth> wind is from top right of game!
11:26:50  * andythenorth appreciates some large buttons and smaller cursor of opengfx.  Seems more accurate
11:27:09  <andythenorth> just easier to user.
11:27:54  <Ammler> yes, the guy needs some time to get familiar but is going well then
11:27:58  <Ammler> gui*
11:28:59  <andythenorth> Ammler: this windowshade icon - why not just use the one from the standard base set?  It's a recent addition far as I know, so it ought to be GPL.
11:29:48  <Rubidium> isn't that shade button in opengfx already for a long time?
11:29:56  <andythenorth> hmm
11:30:01  <andythenorth> do I need to update opengfx
11:30:01  <Ammler> isn't that the same?
11:30:03  <andythenorth> that might help
11:30:06  <Ammler> :-D
11:30:19  <Ammler> now, I get you :-)
11:30:28  <andythenorth> sorry
11:31:52  <andythenorth> that's better :)
11:31:54  <Rubidium> andythenorth: you were using pre 0.2.1?
11:32:02  <andythenorth> possibly
11:32:10  <andythenorth> I just now got the latest from Bananas
11:32:50  <andythenorth> seriously improved :)
11:33:23  <Ammler> there isn't much viewable changes since then...
11:33:29  <Ammler> well, the houses :-)
11:33:59  <andythenorth> haven't found any black squares yet
11:34:12  <Ammler> you don't since 0.2
11:34:47  <Ammler> that would be a awful bug
11:35:14  <andythenorth> Ammler: I think the 'i'  icon is fine in red
11:35:51  <Ammler> then comment that in the ticket, so we have other opinions about
11:35:57  <Ammler> :-) else we change it...
11:42:39  <andythenorth> Ammler: there is a basic error with the brown fences on slopes
11:42:56  <andythenorth> how do I change source?  Do I check out the whole thing?
11:43:16  <Ammler> well, might be easiest
11:43:18  <Webster> Latest update from devactivity: OpenGFX - Feature #741: toolbar info sprite blue instead red? <>
11:43:27  <Ammler> but feel free to commit a patch
11:43:37  <Ammler> whatever you prefer
11:44:17  <Ammler> if you like, we can add you as dev :-)
11:44:55  <Ammler> I like the blue i :-)
11:45:03  <andythenorth> I'll checkout
11:47:29  <andythenorth> hmm
11:48:03  <andythenorth> big project
11:48:29  <andythenorth> Ammler what is the workflow for modifying graphics?
11:48:33  <andythenorth> source -> export pcx?
11:48:38  <andythenorth> or modify pcx directly?
11:49:01  <Ammler> well, depends if the source file is "more" than the pcx
11:49:07  <Ammler> else we have only the pcx
11:49:35  <Ammler> just a "converted" png isn't needed as source
11:50:00  <andythenorth> I guess fences are just in a landscape pcx somewhere then
11:50:17  <Ammler> but specially photoshop or gimp files are very welcome...
11:50:29  <Ammler> or 32bpp pngs
11:50:59  <Ammler> andythenorth: I fear so, we still couldn't convince zeph to commit his working archive
11:51:21  <andythenorth> ho hum
11:52:31  <Ammler> it's a open ticket: ;-)
11:56:42  <planetmaker> [11:35]	<Rubidium>	planetmaker: <-- what purpose does that serve, Rubidium?
11:58:18  <Ammler> planetmaker: why do you make that at all?
11:58:46  <Ammler> why not simply use the native error or md5sum -c?
11:58:48  <planetmaker> "that"= ?
11:58:59  <Rubidium> planetmaker: that make fails check if there's a problem
11:59:18  <Ammler> of*
11:59:22  <Rubidium> i.e. so anything calling it gets easily notified of "something went wrong" without the need for grepping the output
11:59:56  <planetmaker> ah, so issuing an error / changing the return value. Good idea, thanks
12:00:48  <Ammler> if md5sum -c fails, building of the package fails
12:01:03  <Ammler> does that still happen with that make target?
12:01:04  <andythenorth> how do I test opengfx?
12:01:08  <andythenorth> I've used make install
12:01:19  <planetmaker> andythenorth: select it in the game options
12:01:25  <planetmaker> of OpenTTD
12:01:28  <andythenorth> so I can't apply it to an existing game?
12:01:29  <andythenorth> ok
12:01:32  <Ammler> yes
12:01:51  <planetmaker> it's not a newgrf. It's a base set, so different ways :-)
12:02:02  <Ammler> I have a testsave were I build things I modify and compare the changes
12:02:39  <Ammler> andythenorth: you can apply it to an existing game...
12:02:52  <Ammler> but you need to do it before you load the game
12:02:55  <planetmaker> andythenorth: savegames have nothing to do with an existing savegame
12:03:03  <planetmaker> they work with each base set
12:03:14  * andythenorth has some troubles with make install
12:03:18  * andythenorth makes some tea
12:03:18  <planetmaker> they MUST work with each base set
12:03:19  <Ammler> but it might be recoomed to test without any newgrf
12:03:37  <planetmaker> (that's actually the difference to a newgrf)
12:04:13  <Ammler> planetmaker: will there be a release of opengfx this weekend?
12:04:18  *** KenjiE20 has joined #openttdcoop.devzone
12:04:25  * andythenorth makes the fences worse not better :|
12:04:40  <Ammler> andythenorth: did you find the fence?
12:04:49  <Ammler> which sprite# is it?
12:05:02  <planetmaker> Ammler: I didn't really plan a release this weekend.
12:05:21  <planetmaker> There's IMHO little point to do one this weekend and another the next
12:05:27  <Ammler> planetmaker: then I make a rc :-)
12:05:27  <andythenorth> Ammler: 893 in landscape031.pcx
12:05:33  <planetmaker> Ammler: what for?
12:05:41  <Ammler> for me, a private rc
12:05:55  <Ammler> to test the new build system...
12:06:00  <Ammler> on obs
12:06:10  *** frosch123 has joined #openttdcoop.devzone
12:06:15  <planetmaker> I mean... if we manage to keep the momentum we can have a new release in another 4 weeks or so.
12:06:44  <planetmaker> but if it's only testing the build system tag it locally and see what you get
12:06:54  <planetmaker> that's what I do, too
12:06:58  <planetmaker> and then rollback
12:07:01  <planetmaker> when done
12:07:24  <Ammler> well, I like also to test my new upcomming chrooted newgrf compiler...
12:08:01  <Ammler> but yes, I won't push the tag
12:09:17  <planetmaker> new compile farm or what is 'newgrf compiler'?
12:09:36  <planetmaker> or is like everybody replacing the makefiles? ;-)
12:09:38  <andythenorth> Ammler: also sprite 894
12:09:45  <Ammler> I have also still a open issue that opengfx doesn't compile on suse factory anymore
12:10:00  <andythenorth> Ammler: the fences don't line up perfectly in original base set either
12:10:41  <Ammler> planetmaker: it will be a real chroot
12:11:23  <Ammler> you could also theoretically choose the distro you like to build for...
12:11:28  <planetmaker> I mean... if you're happy, there could be a 0.2.2 and then another 0.2.3. But in terms of testing I don't entirely see the need.
12:11:43  <Ammler> no, it is fine
12:11:58  <Ammler> that is why I asked you
12:12:30  <planetmaker> Making a release eats sufficient time which I'd like to devote otherwise more productively :-)
12:13:22  <Ammler> I prepared it a bit already ;-)
12:13:25  <Webster> Latest update from devactivity: OpenGFX - Revision 379: Change [Makefile]: Fail for 'check', if md5sums don't match (Rubidium) <>
12:13:30  <Ammler> I hope it wasn't useless :-P
12:14:02  <planetmaker> I saw that changelog commit. I didn't check yet, though :-)
12:14:10  <Ammler> also I need to check, if I can use that make target or still have to do md5sum -c self :-)
12:14:31  <planetmaker> well, that's why it's on feedback
12:15:55  <Ammler> doesn't mac have md5sum -c or why do you it that way?
12:16:09  <andythenorth> these fences need to be smaller, or their bounding box larger.  it's a maths fail :)
12:16:22  <andythenorth> (by the artist drawing them, not in the nfo)
12:16:52  <planetmaker> Ammler: it does. Somewhat. But I didn't quite figure out how to use it.
12:17:00  <Ammler> andythenorth: no bounding box changes possible...
12:17:23  <andythenorth> ok.  well the fence needs to fall by 8px in the y direction to match the slope
12:17:27  <andythenorth> it only falls by 6px
12:17:30  <Ammler> planetmaker: just run it, it does automatically set error flag
12:17:55  <planetmaker> md5 -c ogfx1_base.grf
12:17:57  <planetmaker> md5: illegal option -- c
12:17:58  <planetmaker> usage: md5 [-pqrtx] [-s string] [files ...]
12:18:15  <planetmaker> and the man page is not really enlightening me.
12:18:31  <planetmaker> though it interestingly quotes an option -c
12:19:26  <Ammler> <-- at the bottom
12:19:57  <Ammler> <-- if it matches
12:20:22  * andythenorth bodges the fences
12:20:25  *** DJ_Nekkid has quit IRC
12:20:29  <andythenorth> think anyone will notice
12:20:30  <andythenorth> ?
12:20:31  <Ammler> andythenorth: :-)
12:20:53  <Ammler> did you already blew ottd today?
12:21:13  <planetmaker> Ammler: that line doesn't work for me. No -c as it seems
12:21:46  <Rubidium> what the F went wrong with Factory?
12:21:59  <Ammler> Rubidium: something with nforenum
12:22:04  <andythenorth> Ammler: I didn't blow up ottd today
12:22:19  <andythenorth> FIRS is on ice until I decide where to go with it next
12:22:30  <Ammler> [NFORENUM] ogfxe_extra.nfo
12:22:31  <Ammler> Error: invalid compression, data diff at 275 of 576 bytes, trying without it for this sprite
12:22:33  <Rubidium> Ammler: what about the south?
12:22:42  <Ammler> south?
12:22:49  <Rubidium> argh...
12:22:58  <Rubidium> stupid tab completetion
12:23:05  <Rubidium> a -> Ammler, A -> andy :(
12:23:21  <Rubidium> andythenorth: go to the south with it
12:24:21  * andythenorth cannot go south from here
12:24:25  <Ammler> planetmaker: did you see the wrapping of readme?
12:25:23  <planetmaker> Ammler: I saw those compression warnings. I don't understand them nor do I get them here when running make
12:25:40  <Ammler> planetmaker: on suse factory only
12:25:49  <Rubidium> Ammler: I think it has more to do with grfcodec
12:25:50  <Ammler> works fine with suse 11.2
12:25:52  <planetmaker> what nforenum do you use
12:25:57  <Rubidium> broken upx?
12:25:59  <Ammler> Rubidium: yep
12:26:08  <Ammler> I disabled that
12:26:35  <Ammler>
12:27:04  <planetmaker> Ammler: what wrapping of the readme do you mean?
12:27:18  <Ammler> newline after ~80charrs
12:28:54  <Rubidium> maybe a broken compiler then?
12:29:00  <Webster> Latest update from devactivity: OpenGFX - Bug #781: farm hedges not properly aligned <> || OpenGFX - Bug #849 (Closed): make clean removes too much with source tarball <>
12:29:07  *** DJ_Nekkid has joined #openttdcoop.devzone
12:30:03  <Rubidium> because erroring within the graphics data can't be caused by nforenum
12:30:21  <Rubidium> and it's actually the verification of the compression
12:31:00  <Ammler> hmm, I should test 0.2.1 on factory
12:31:23  <Ammler> at least it worked on it as I commited to obs
12:34:50  <planetmaker> doh. never mind me with "what wrapping" ;-) Yes, I saw that. It's fine with me
12:37:33  <Ammler> yes, definitly a factory issue
12:37:37  <Ammler> 0.2.1 fails now too
12:38:26  <Ammler> ah, just because the missing md5sum check, it didn't complain
12:38:39  <Ammler> and packed the failed grfs ;-)
12:39:01  <Ammler>
12:39:41  <Ammler> he, the md5sum check isn't good for openttd itself ;-)
12:39:52  <Ammler> only*
12:41:05  <Ammler> well, soon, you should get the same issues with debian/ubuntu too, I assume...
12:41:25  <Rubidium> oh... might have a reason for the compiler failing
12:41:58  <Ammler> could update grfcodec...
12:54:18  <planetmaker> Ammler: we could replace md5sum by shasum - that seems to be the same for linux and macos ;-)
12:54:34  <planetmaker> and shasum -c blubber.sha works
12:54:48  <Rubidium> only for those checks ofcourse :)
12:55:29  <planetmaker> :-) true
12:55:52  <planetmaker> but I guess that's stupid
12:56:52  <Ammler> planetmaker: maybe use md5sum -c for linux and "your command" for mac?
12:57:08  <planetmaker> that's also stupid. Bad maintainability
12:57:11  <Ammler> so you can at least fail building...
12:57:39  <planetmaker> what doesn't work with the current approach?
12:57:55  <Ammler> didn't test it yet, will tell you...
12:58:03  <Ammler> does it tell, which grf failed?
12:58:12  <planetmaker> thanks to Rubi's diff it will fail the check, if sums differ
12:58:47  <planetmaker> ingo@devera:~/ottd/grfdev/opengfx> make check
12:58:49  <planetmaker> Differences in md5sums:
12:58:50  <planetmaker> 1c1
12:58:52  <planetmaker> < 93cf3409245f35dd29c60f79315c9fa6  ogfx1_base.grf
12:58:53  <planetmaker> ---
12:58:55  <planetmaker> > a3cf3409245f35dd29c60f79315c9fa6  ogfx1_base.grf
12:58:56  <planetmaker> make: *** [check] Error 1
12:58:58  <planetmaker> ^ on my suse
12:59:22  <Ammler> well, nvm, if "someone" doesn't like make check, nobody stops him to use use md5sum -c
12:59:50  <planetmaker> so you know what fails
13:00:14  <Ammler> ok :-)
13:00:30  *** Seberoth has quit IRC
13:00:49  <Ammler> use diff -u
13:01:51  <Ammler> hmm, maybe not :-)
13:02:31  <Rubidium> Ammler: take a look the grfcodec feature thread for a patch that might fix the suse error
13:02:52  <Ammler> oh, your post with the 3 patches?
13:03:38  <Rubidium> no, the one after that
13:08:00  *** Seberoth has joined #openttdcoop.devzone
13:13:17  <Ammler> openSUSE Factory is like beta of next release (currently openSUSE 11.3)
13:13:27  <Ammler> or like trunk in openttd :-)
13:15:14  <Ammler> and exactly this weekend, they build a milestone snapshot, so obs is very occupied, needs around 4 hours to get something build there
13:17:18  <Webster> Latest update from devactivity: FIRS Industry Replacement Set - Feature #852 (Rejected): Makefile.common may be missing dep check... <" target="_blank">> || FIRS Industry Replacement Set - Feature #852 (Rejected): Makefile.common may be missing dep check... <>
13:20:51  *** ODM has quit IRC
13:44:04  <Ammler> I hate tt-forums file download
13:44:32  <Ammler> our ugly php download script on bundles is 100 times better
13:51:35  <Rubidium> lies
13:52:18  <Rubidium> Saving to: `download.php?'
13:52:27  <Rubidium> that's not that helpful
13:53:03  <Rubidium> even though one would think that one would get a proper filename because the URL is nice
13:54:54  <Ammler> well i have an alias on wget :-)
13:55:15  <Ammler> alias wget="wget --content-disposition"
13:55:48  <Ammler> Rubidium: you can't do it better, afaik
13:56:18  <Rubidium> it works fine for (our) flyspray
14:00:19  <Ammler> well, like the DevZone, I guess
14:00:42  <Ammler> Rubidium: your patch doesn't work :-(
14:01:12  <Rubidium> doesn't work or doesn't fix the factory problem?
14:01:23  <Ammler> fix the factory issue
14:01:46  <Ammler> applied to 2306
14:06:47  <Ammler> afaik, dalestan is a suse user too
14:12:13  <Ammler> hmm, but maybe we should replace the download.php with a logfile analyzer
14:38:31  *** Seberoth2 has joined #openttdcoop.devzone
14:46:03  *** Seberoth has quit IRC
15:23:01  <Webster> Latest update from devactivity: FIRS Industry Replacement Set - Revision 675: Change: set correct industry types prop 0B for vari... <> || FIRS Industry Replacement Set - Revision 674: Change: commented experimental protection period of... <> || FIRS Industry Replacement Set - Feature #836: Is code needed to prevent annoying random industry ... <>
16:26:46  <Ammler> planetmaker: btw. I asked for real names in tt-forums :-)
16:26:51  <Webster> Latest update from devactivity: FIRS Industry Replacement Set - Bug #535 (Closed): Primary/secondary industry types <> || OpenGFX - Revision 380: Fix [#781]: Improve visual continuation of adjacent fences around farm la... <>
16:27:05  <andythenorth> Ammler: I don't want to do my real name :)
16:27:26  <planetmaker> Ammler: I know :-)
16:27:45  <planetmaker> andythenorth: then you changed your mind. You once agreed to be credited with your real name ;-)
16:27:53  <andythenorth> oh well
16:28:06  <andythenorth> what's done is done
16:28:09  <planetmaker> at least it's in the credits section for a few releases
16:28:29  * planetmaker is without real name there. Currently
16:28:31  <planetmaker> ;-)
16:29:02  <Ammler> we should split the list and add the anonymous contributors at bottom in a comma list
16:29:36  <Ammler> and change the others with nick in brackets
16:30:10  * andythenorth removes FIRS issues from milestone by putting them in backlog :P
16:31:04  <Ammler> andythenorth: I am also not sure, if it is legal to use nickname for gpl work...
16:31:20  <Ammler> (beside it is ugly)
16:31:34  * andythenorth likes a bit of privacy
16:32:13  * andythenorth adds to tickets by moving them in from future milestone :o
16:33:39  <Ammler> andythenorth: whois :-)
16:33:51  <andythenorth> planetmaker: you're clearly pretty busy for some time?  I'll bounce some FIRS tickets out of the 0.1 release...
16:34:42  <planetmaker> andythenorth: unfortunately true
16:34:53  <andythenorth> life is life
16:41:55  <Webster> Latest update from devactivity: FIRS Industry Replacement Set - Bug #316 (Closed): Conflicting industry types <> || FIRS Industry Replacement Set - Bug #316: Conflicting industry types <>
16:44:27  *** ODM has joined #openttdcoop.devzone
16:47:16  *** Seberoth2 is now known as Seberoth
16:57:56  <Webster> Latest update from devactivity: FIRS Industry Replacement Set - Bug #715 (Closed): Furniture Factory - large building sprite trun... <> || FIRS Industry Replacement Set - Revision 677: Fix: Furniture Factory had a truncated sprite <> || OpenGFX - Bug #848 (Closed): Makefile doesn't replace {{TAG}} anymore <> || FIRS Industry Replacement Set - Revision 676: Change: fixed misplaced sprites in furniture factor... <>
17:22:27  <Ammler> Rubidium: it is something with the compression
17:22:31  <Ammler> using -u works
17:22:46  <Ammler>     -u        Save uncompressed data (probably not a good idea)
17:23:58  <Ammler> oh, quite a big grf now :-)
17:24:22  <Ammler> -rw-r--r-- 1 abuild abuild 5.7M Mar 20 17:22 ogfx1_base.grf
17:33:11  <Ammler> planetmaker: do you need more info for the issue?
17:39:45  <Ammler> [Generating]
17:39:46  <Ammler>         zip warning: name not matched:
17:41:02  <Ammler> we might consider using grfcodec with -u in general?
17:41:44  <Ammler> opengfx is 500kb less
17:46:08  <Rubidium> IIRC the compression speeds up the drawing by skipping over the uninteresting bits
17:48:58  <Ammler> so there are other disadvantages then just disk wasting...
17:55:49  <planetmaker> the newgrf wiki tells that compression makes drawing slower
17:56:22  <planetmaker> Ammler: issue: I need an idea why it happens. It actually also happens for me here locally
17:56:36  <planetmaker> But I simply don't see *why*. I didn't touch that part of the makefiles
17:56:58  <planetmaker> And I have no clue how to fix that :-(
17:57:08  <planetmaker> except going for another zip programme - which works
17:58:13  <frosch123> btw. ottd ignores the "tile compression" flag for drawing. basically "8bpp-simple" is draw all sprites without compression, while "8bpp-optimised" is draw all sprites with compression
17:58:58  <planetmaker> frosch123: and what's the difference between compression and no compression?
17:59:07  *** ODM has quit IRC
18:00:14  <planetmaker> I don't quite get the difference between the two things you just stated ;-)
18:00:17  <frosch123> shall i point you to the ogfx dev thread?
18:00:45  <planetmaker> :-D
18:01:01  <frosch123> hmm, otoh, maybe i just got your question wrong
18:02:02  <planetmaker> If I understood compression flag wrong it's means how things are stored or shall be stored in memory and what to expect when reading from the file
18:02:09  <frosch123> yeah, i was talking about the 01 / 09 spritetype compression, you were talking about grfcodec -u
18:02:17  <planetmaker> Where valid values are 01, 03 and 09 and something different usually not used
18:02:49  <planetmaker> uhm... not quite :-)
18:03:04  <planetmaker> I was thinking in terms of the pcx compression flags
18:03:26  <planetmaker> but I understood your statement in the direction of how ottd handles the sprites as a function of that flag
18:03:36  <planetmaker> And I didn't get f(pcx compression flags)
18:03:56  <planetmaker> what advantage does each offer?
18:04:01  <frosch123> <- that is about 01, 03, 09 and friends
18:04:02  <Webster> Title: Transport Tycoon Forums • View topic - [8bpp] Graphics Replacement Project - OpenGFX (at
18:04:36  <frosch123> the grfcodec -u flag has no influence on drawing speed, it only increases file size
18:04:38  <planetmaker> right, thanks :-)
18:04:54  <planetmaker> There indeed are pieces in your posting I didn't recall - and which answer my question :-)
18:04:58  <Ammler> I can build btw. opengfx in suse Factory with 11.2 rpm
18:05:02  <frosch123> well, otoh, if your spritecache is too small...
18:06:00  <Ammler> [18:56] <planetmaker> Ammler: issue: I need an idea why it happens. It actually also happens for me here locally <-- it works on the server witz p7zip
18:06:21  <planetmaker> frosch123: pixelline = horizontal line of pixels?
18:06:45  <planetmaker> Ammler: yes, that's another programme and you also supply your own calling flags.
18:06:47  <frosch123> yup, "scanline" usually :p
18:07:17  <planetmaker> :-) Yeah, but it was an unusual word - so I thought I rather make sure than introduce yet another mis-understanding :-)
18:09:25  <Ammler> frosch123: so you disagree to Rubidium, who wouldn't recommend using -u
18:09:42  <frosch123> i would not recomend -u either
18:09:55  <Ammler> but only because of file space waste :-)
18:10:10  <frosch123> but -u has only an effect when loading sprites, which happens not that often if your spritecache is big enough
18:10:51  <Ammler> but shouldn't -u speed that part up, as openttd then doesn't need to "uncompress" first?
18:11:14  <PeterT> teehee, he called you "Rubi"
18:11:15  <Webster> Title: Transport Tycoon Forums • View topic - GRFCodec feature discussion thread (at
18:11:29  <frosch123> Ammler: that depends on your disk, your diskcache, your cpu cache, ...
18:11:40  <frosch123> there is more data invoved, but the operations are easier
18:12:14  <frosch123> with the assumption that the cpu is faster than the memory, compression is better
18:21:14  <planetmaker> hm... but CPU is usually the bottle neck - not memory.
18:27:21  <planetmaker> he... I only see it now. A bit older posting, that with the explanation on the compressions ;-)
18:27:54  <PeterT> Is there a huge difference between decoding a GRF or downloading the source from HG?
18:28:00  <Ammler> yes
18:28:14  <planetmaker> comments :-)
18:34:08  <planetmaker> <-- hm... I should consider the fix wrt remake for my makefiles, too, eh?
18:40:09  <frosch123> hehe, after i saw that diff i very quickly checked ttdviewer, but i was lucky :p
18:44:03  <planetmaker> :-)
18:52:31  <Ammler> planetmaker: that might be the issue we had with bundle_*
18:52:49  <Ammler> as I use -j3
19:20:31  <Webster> Latest update from devactivity: OpenGFX - Feature #741: toolbar info sprite blue instead red? <>
19:30:21  <Ammler> he, I guess
19:30:28  <Ammler> I found the Factory issue
19:30:32  <Ammler> no boost
19:33:21  <Ammler> hmm, no that was the grf chroot, grfcodec chroot has boost installed :-(
19:35:47  <Ammler> maybe gcc45?
20:23:20  *** frosch123 has quit IRC
20:23:56  *** frosch123 has joined #openttdcoop.devzone
21:37:10  *** Seberoth has quit IRC
22:44:15  *** Seberoth has joined #openttdcoop.devzone
23:28:26  *** frosch123 has quit IRC
23:29:15  *** Seberoth has quit IRC
23:50:53  *** ODM has joined #openttdcoop.devzone

Powered by YARRSTE version: svn-trunk