IRC logs for #openttd on OFTC at 2012-02-04
⏴ go to previous day
00:20:42 *** roboboy has joined #openttd
01:25:34 *** roboboy has joined #openttd
03:17:53 <Rhamphoryncus> michi_cc: My primary thing would be schedule offsets, but I was considering falling back to unitnumber if the offset was somehow the same. However, that should ONLY be for importing old save games; active games should always maintain it perfectly sorted.
03:30:37 *** roboboy has joined #openttd
05:22:49 *** Xrufuian has joined #openttd
05:56:22 *** Eddi|zuHause has joined #openttd
07:16:58 *** andythenorth has joined #openttd
07:18:05 *** cmircea has joined #openttd
07:32:12 *** andythenorth has joined #openttd
07:38:55 *** sla_ro|master has joined #openttd
07:59:15 *** |Jeroen| has joined #openttd
08:04:58 *** _11Runner_ has left #openttd
08:13:37 * andythenorth adventures again in the makefil
08:14:47 <andythenorth> with limited success :(
08:25:30 *** tokai|noir has joined #openttd
08:25:31 *** ChanServ sets mode: +v tokai|noir
08:31:39 *** Phoenix_the_II has joined #openttd
08:36:22 <planetmaker> andythenorth: makefile principle is easy:
08:36:42 <planetmaker> target_file: files_it_needs_present_to_be_build
08:36:50 <planetmaker> rules_to_create_file
08:37:55 <andythenorth> does the dep check rely on hg?
08:38:03 <andythenorth> i.e. a file hg has forgotten won't be seen by dep check
08:38:26 <andythenorth> or the path rather than file
08:39:06 <Ammler> no, it works also without, but a file not in the repo will be missing on remote :-)
08:39:41 <andythenorth> I have been moving files in the BANDIT repo so it makes more sense
08:39:59 <andythenorth> but make is now failing
08:40:05 <andythenorth> I have the rule bandit.nml: sprites/nml/bandit.pnml
08:40:17 <andythenorth> and sprites/nml/bandit.pnml is present
08:40:22 <andythenorth> but bandit.nml is not built
08:41:03 <andythenorth> although....maybe now is the time to eliminate the .pnml step anyway
08:43:43 <Eddi|zuHause> the easiest way to test whether such a rule is actually evaulated is "touch sprites/nml/bandit.pnml"
08:45:19 <andythenorth> seems I've fixed it
08:45:25 <andythenorth> the structure is a bit more logical I hope
08:45:32 <andythenorth> [for repo filesystem]
08:46:47 <Eddi|zuHause> i'm sure there are plenty of inconsistencies left :)
08:47:01 * andythenorth has....questions :P
08:48:11 <Ammler> also moving around is not supported by Redmine, so do it as less as possible :-)
08:48:36 <andythenorth> but I will try not to
08:48:43 <andythenorth> it's often tmwftlb
08:49:02 <Ammler> you just will lose history view there
08:49:51 <andythenorth> I have a few inconsistencies to figure out
08:49:59 <andythenorth> filename extensions is one
08:51:24 <andythenorth> what's a good extension for a file that contains nml with templating markup?
08:51:59 <Eddi|zuHause> sounds about right
08:52:40 <Ammler> also you might consider replacing the targets with wildcards: %.nml: %.pnml
08:52:43 <andythenorth> the convention for the template module I use is .pt (page template), but it's not enforced
08:53:06 <andythenorth> Ammler: so they are decoupled from the extension?
08:55:40 <Ammler> the future goal is move the content of Makefile.in to the makefile framework, right?
08:55:58 <andythenorth> I hadn't thought that far ahead
08:56:12 <andythenorth> it's kind of not my decision either ;)
08:56:45 <Ammler> well, if you prepare a nice patch, planetmaker would sure welcome it, I guess
08:57:46 <Ammler> at least the parts you share with cets
08:57:55 <planetmaker> I could add a variant w/o gcc indeed
08:58:42 <andythenorth> what would need to be generic?
08:58:50 <andythenorth> - build script name
08:59:07 <andythenorth> - config data file name
08:59:37 <Eddi|zuHause> maybe take the prepocessing part out of Makefile.nml, and make it a Makefile.cpp
08:59:39 <planetmaker> but renaming the main file to tnml IMHO is stupid
08:59:45 <Eddi|zuHause> then you could add a Makefile.python
08:59:56 <Eddi|zuHause> and it would be arbitrarily combinable by the config
08:59:57 <planetmaker> it's not a template
09:00:02 <andythenorth> planetmaker: I'm open to suggestions
09:00:06 <planetmaker> it's still a file which needs pre-processing. Thus pnml
09:00:22 <planetmaker> template is something which is being re-used
09:00:39 <andythenorth> different semantics I guess
09:00:46 <andythenorth> for the templating module, it's a template
09:00:59 <planetmaker> basically I think your whole last commit there was... wrong by my definition of what's a template
09:01:04 <andythenorth> it's called by PageTemplateLoader class
09:01:18 <andythenorth> I'm happy to agree a convention
09:01:30 <andythenorth> I don't really have strong opinions on this stuff, just whatever makes sense to other people
09:01:47 <planetmaker> nml to-be- pre-processed
09:01:59 <andythenorth> using [pre-processor of choice]
09:02:05 <planetmaker> thus everything is a pnml.
09:02:12 <andythenorth> I'll just flip them
09:02:24 <andythenorth> anybody else rabidly disagree with planetmaker ?
09:02:25 <planetmaker> Except those parts which are reused several times in the coding of individual vehicles
09:02:26 <Eddi|zuHause> i do think it's sensible to keep extensions for cpp-processed and python-processed nml separate
09:02:28 <andythenorth> before I play commit tennis?
09:03:10 <andythenorth> I considered cnml
09:03:14 <andythenorth> for chameleon-processed nml
09:03:20 <andythenorth> but that's specific to my templater
09:03:28 <andythenorth> "other python templaters are available"
09:03:30 <Ammler> shouldn't it be possible to use both in one project?
09:04:24 <Eddi|zuHause> this is one of those questions where you shouldn't just make your mind up in 2 minutes and commit it
09:04:48 <andythenorth> second time around :P
09:05:12 <Ammler> how do you create target line, if you the source depends on different preprocessor?
09:05:56 <andythenorth> Ammler: not sure
09:06:08 *** Phoenix_the_II has quit IRC
09:06:15 <andythenorth> the pipeline I had was python build script -> cpp -> nml
09:06:24 <andythenorth> I've cut the cpp step afaict
09:06:48 <Ammler> so you just run a file through both
09:06:58 <andythenorth> same as CETS is doing
09:07:05 <planetmaker> well. At least tnml is for source code templates which are re-used
09:07:12 <andythenorth> I've eliminated all the cpp now though
09:07:46 <planetmaker> a CTT or the main file never are re-used in a project, they're one-time code snippets
09:08:43 <Ammler> so it isn't possible to setup makefile rules depending on source available, e.g. I need %.nml and have either %.cnml or %pnml which need either python or cpp?
09:09:06 <andythenorth> planetmaker: seems you'd like the semantics to refer to how file is used, rather than what module uses the file?
09:09:14 <andythenorth> which is fine, just want to be sure
09:09:17 <planetmaker> that's not possible. You'd need to add a configure step. Or individual rules, Ammler
09:09:27 <planetmaker> andythenorth: of course
09:10:02 <planetmaker> .cpp is also not about what's in it. Not about the module using it
09:10:06 <planetmaker> it's about the language it contains
09:10:14 <andythenorth> what about the case where I have a template, but only re-use it once? :o
09:10:27 <planetmaker> then it's silly to use a template tbh
09:10:38 <planetmaker> it's over-engineered then
09:10:59 <peter1138> if you re-use it once, then you've used it twice
09:11:29 <andythenorth> I wonder if it's confusing though
09:11:31 <planetmaker> well, yes :-) For two usages it makes sense :-)
09:12:46 <andythenorth> authors using python will be calling Template() class or PageTemplateLoader() or whatever
09:12:46 <andythenorth> even to setup things which are singletons
09:12:46 <andythenorth> I'll leave it as tnml with the agreement to change it when we have agreement
09:16:30 <Rubidium> andythenorth: make it ntl
09:16:50 <Rubidium> nml template language, or rather newgrf meta language template language ;)
09:17:26 <Ammler> andythenorth: pnml is not because it needs cpp, it is because it needs preprocessing, same as your tnml needs
09:18:34 <andythenorth> I'm only not changing it right now because of what Eddi reminded me - don't decided in 2 mins then commit in haste
09:18:37 <Ammler> I don't think, it is a good idea to invent new names just for good readability
09:18:49 <andythenorth> I'd be fine with .pnml as well
09:18:56 <planetmaker> tnml is really unfortunate. Functunally it's the same as the pnml in the other projects. And additionally it has already a meaning in the other projects different from this tnml
09:19:18 <planetmaker> and... you already went for the hasty renaming
09:19:39 <planetmaker> so the damage is done and now festers
09:19:57 <Ammler> we wait for hg 2.1, where you can push such things as draft :-p
09:20:58 <andythenorth> as the templater by convention uses .pt :P
09:21:20 <Ammler> but I fear redmine will not be able to handle hg phases
09:21:26 <Eddi|zuHause> permutation(l,m,n,p,t)
09:21:41 <planetmaker> you use python for pre-processing. then eddi's suggestion to use pynml is a quite good one
09:22:04 <andythenorth> I kind of like it too
09:22:20 <andythenorth> but the point was made that .pnml is not specific to cpp?
09:23:10 <Eddi|zuHause> that's kind of an unfounded argument
09:23:29 <Eddi|zuHause> it was only used for cpp in the past
09:23:39 <Eddi|zuHause> so it was never defined
09:24:06 <Eddi|zuHause> so now it's exactly the right time to define it
09:24:16 <Ammler> Eddi|zuHause: the method you use to preprocess should not define the source
09:24:49 <Eddi|zuHause> yes, that's the entire point of extensions
09:24:56 <Eddi|zuHause> specify the program it should be used with
09:25:14 <Ammler> but makefile can't handle it anyway
09:25:40 <andythenorth> I don't have text files called .letter .angry_note_to_wife .tax_evasion_plan
09:26:43 <Eddi|zuHause> but you have xml files called .odf
09:26:53 <Eddi|zuHause> and RIFF files called .avi
09:27:17 <andythenorth> and mac folders called .app
09:27:22 <Eddi|zuHause> yes, a zipped xml
09:27:34 <andythenorth> is this bikeshedding? :D
09:27:51 <Rubidium> vcxproj is a xml as well ;)
09:27:58 <andythenorth> what should the precedence order be for .tnml and .pnml?
09:28:01 <Rubidium> on the other hand, it's a text file too
09:28:29 <andythenorth> all .tnml files are processed. but not all .pnml files are templates used more than once
09:28:36 <Eddi|zuHause> extensions are used to let meta-tools like filemanagers or makefiles make assumptions about the content without looking at the content
09:29:16 <planetmaker> andythenorth: pnml does NOT indicate template
09:29:38 <Eddi|zuHause> i don't use tnml in CETS
09:29:47 <planetmaker> it indicates need for pre-processing
09:29:51 <Eddi|zuHause> it's all just pnml to me
09:31:28 <planetmaker> and my idea behind tnml was that those are files which are included multiple times
09:32:25 <planetmaker> and yes, again as eddi says: extensions are meant to give one-glance idea of what's inside
09:32:25 *** TGYoshi has joined #openttd
09:32:50 <planetmaker> in terms of file type / language contained
09:34:13 <andythenorth> so repeat_inclusion > processing?
09:34:24 <andythenorth> i.e. a pnml file is a tnml file if used > once?
09:35:37 <andythenorth> it made total sense with cpp
09:35:42 <andythenorth> there were no problems with the convention
09:35:44 <Rubidium> hahaha... in some town somewhere in Belgium the preschools apparantly sent a letter to everyone "of the right birth year" to start preschool, so everyone whom's birth year ended in 09 got the letter
09:36:03 <planetmaker> haha @ Rubidium :-)
09:36:09 <planetmaker> I can imagine the result :-P
09:36:20 <Rubidium> including a 102 year old guy ;)
09:36:38 <Rubidium> I fear he's not going to finish his second round of education
09:37:09 <planetmaker> also in the sports classes he might not be up to some challanges
09:37:50 <planetmaker> he might also get marks deducted for "bad" handwriting ;-)
09:38:12 <andythenorth> planetmaker: it seems that to avoid bikeshedding this, I should just follow the convention established
09:39:16 <planetmaker> I don't really mind which you follow. But if you use another, it's bad to use the same extensions with different meaning
09:39:30 <Eddi|zuHause> Rubidium: these y2k bugs never die out
09:39:46 <andythenorth> planetmaker: agreed on extension collisions
09:40:13 <andythenorth> so either .tnml and .pnml, or .pynml
09:40:14 <Eddi|zuHause> a few years ago i think it was in finland, where a 106 year old got the notification for school-enrollment
09:40:21 <andythenorth> I'll decide later :P
09:40:27 <andythenorth> classic decision by committee :P
09:42:08 * andythenorth just thought of an evil practical joke
09:42:23 <andythenorth> swap someone's smiley set to invert :( and :)
09:42:29 <andythenorth> assuming they use graphic smileys
09:42:53 <planetmaker> so you changed the files in a split second to clash with "standard" and decide later whether to use something sane? :-P
09:43:04 <andythenorth> standard andythenorth haste :P
09:43:15 <andythenorth> no point repeating the mistake twice
09:43:40 <planetmaker> yeah. Let it rest, forget and build on... a bad decision :-P
09:43:49 *** Alberth has joined #openttd
09:43:49 *** ChanServ sets mode: +o Alberth
09:44:11 <andythenorth> planetmaker last time I asked the question about extension you weren't here
09:44:27 <andythenorth> I was told 'stop asking dumb questions, just decide something and write code' effectively
09:45:15 <andythenorth> advice varies by person giving it :D
09:45:39 <planetmaker> the question in the first place is: why does it need a change at all?
09:45:44 <planetmaker> And that wasn't IMHO asked at all
09:46:04 <planetmaker> but your game, your call
09:46:32 <andythenorth> why does the extension need a change?
09:46:52 <andythenorth> only that I had files in my repo that started life as cpp and migrated to python
09:47:01 <andythenorth> and the extension use had no consistency
09:47:10 <andythenorth> it's just cleanup
09:48:56 <planetmaker> "cleanup"? Interesting definition
09:49:48 <andythenorth> we seem to be at cross purposes
09:49:59 <andythenorth> I need to use correct extensions according to a convention
09:50:07 <andythenorth> I needed to change them
09:50:14 <andythenorth> my first attempt was...hasty
09:50:41 <Alberth> first one always is ;)
09:51:11 <andythenorth> the makefile project convention of 'template' suffers an unfortunate naming collision with python templates
09:51:12 <Alberth> it's more 'an exploration of possibilities', I think, in general :)
09:51:16 <andythenorth> which confused me a bit
09:51:35 <andythenorth> we now have two entities in play, with the same name
09:51:50 <andythenorth> python template modeules/classes, which load template files
09:52:02 <andythenorth> and templates, which are template files used more than once
09:52:27 <andythenorth> I'm sorry if the decision was wrong or annoying
09:52:35 <andythenorth> I'm just confused by what is the right thing to do
09:52:52 <planetmaker> just use unique extensions, if you must use something different
09:53:29 <planetmaker> anyway, RL is calling. See you later
09:58:28 *** chester has joined #openttd
10:00:37 *** Chris_Booth has joined #openttd
10:01:43 *** Progman has joined #openttd
10:17:10 *** Brianetta has joined #openttd
10:21:44 *** OpenTTD_Johannes has joined #openttd
10:22:41 <OpenTTD_Johannes> I have a map save for the title-image contest. Do I need to register in the forum to submit?
10:23:48 <Alberth> yes, otherwise you cannot post
10:23:56 <Zuu_> That would make it easier for pm to organize it.
10:24:16 <OpenTTD_Johannes> Don't send me spam after registration :P
10:24:36 <Zuu_> If you email he would need to monitor his email too for saves.
10:24:43 *** mahmoud has joined #openttd
10:24:51 *** tensai_cirno has joined #openttd
10:24:59 <Alberth> OpenTTD_Johannes: did not happen with me so far
10:25:18 <Alberth> be sure to disable the email address display in your preferences
10:25:35 <OpenTTD_Johannes> Only kidding. Thanks for the advice, I'll upload it soon! :)
10:27:30 <Zuu_> You will only soon get addicted to the forums. ;)
10:27:37 *** frosch123 has joined #openttd
10:30:33 <frosch123> yay, it's not even noon, and i am already here :p
10:32:32 <Ammler> Alberth: do you have any clue, how fedora handles "recommends" with package manager (rpm)?
10:33:34 *** JVassie has joined #openttd
10:34:23 <Eddi|zuHause> # it's so cold it's so cold it's so cold it's so cold
10:36:16 <Alberth> Ammler: it has 'recommends'? never noticed it
10:37:13 <OpenTTD_Johannes> The perfect weather for playing openttd all day :)
10:39:30 <V453000> Elukka: AMAZING website :D :D :D
10:40:22 <frosch123> Elukka: you are living in a weird climate, if it snows at -18°C
10:40:50 <V453000> it says snowing even here, and no snowing :) but still F cold
10:45:36 <Ammler> Alberth: it seems not having it, that is why I ask, how do you Fedora guys handle that?
10:45:40 <OpenTTD_Johannes> Ummm is openttd interested in C/C++-Coders?
10:46:14 <Alberth> OpenTTD_Johannes: the top-left corner of the save game stays where it is, so perhaps if you move that (in your screenshot at least), it may be better for small screens
10:46:40 <Alberth> OpenTTD_Johannes: as well as squirrelers :)
10:46:51 <OpenTTD_Johannes> okay Alberth, I'll fix it, thanks!
10:47:39 <Alberth> OpenTTD_Johannes: in the game options you can switch to smaller window sizes so you can see how it looks
10:47:56 <OpenTTD_Johannes> Oh? Squirrel is an animal, right? Sorry, I am bad in English... :(
10:48:48 <Eddi|zuHause> yes, but it's also a scripting language
10:50:21 <OpenTTD_Johannes> Ahhh, okay, I usually only code in C++. But I have time left, and if there was anything that needed to be done, I could try it.
10:50:31 *** tokai|mdlx has joined #openttd
10:52:23 *** FLHerne has joined #openttd
10:52:47 <planetmaker> just back from shopping I updated the titlegame page
10:54:45 *** DayDreamer has joined #openttd
10:55:53 <planetmaker> ^ OpenTTD_Johannes, thus you can just browse the screens
10:57:10 <Alberth> too much temperate climate :(
10:59:02 <planetmaker> Alberth: we had arctic and desert already :-)
10:59:13 <planetmaker> kinda time for a temperate ;-)
10:59:27 <planetmaker> But then, I know what you mean
10:59:38 <planetmaker> and somewhat feel with you
11:00:40 *** OpenTTD_Johannes has quit IRC
11:02:16 <Alberth> I'd like that, but it is too controversial I fear to ever get enough votes for
11:05:12 <frosch123> it's funny how the screenshots randomly switch between english and german :p
11:05:34 <planetmaker> :-) Different times they were generated
11:05:41 <planetmaker> And ... I switched languages, I guess
11:05:42 * Alberth watches a screenshot intently to see it change
11:12:33 *** lmergen has joined #openttd
11:21:23 *** andythenorth has joined #openttd
11:24:52 <andythenorth> Eddi|zuHause: .gnml - generated nml?
11:26:35 <andythenorth> it's used in CETS ;)
11:27:00 <Alberth> xnml could be fun too ;)
11:27:22 <andythenorth> sounds like it would be xml :P
11:27:30 <andythenorth> mine contains some xml :P
11:27:37 <andythenorth> but it's not a valid xml doc
11:27:48 <andythenorth> fun with extensions :o
11:28:05 <andythenorth> I want to use .pynml for templates
11:28:39 <andythenorth> but I'm kind of reluctant to make the change without some agreement
11:29:08 <Alberth> it contains python code?
11:29:30 <andythenorth> and I can't find the extension in use by other projects
11:29:55 <Alberth> the number of 'nml' projects is not so large :)
11:30:45 <Alberth> I'd just use .py, but that is much less exciting
11:32:11 <planetmaker> andythenorth: what is a template in your context?
11:33:18 <andythenorth> it's a file used by chameleon templating module
11:33:41 <andythenorth> although I was considering just now writing a python module that aliased all references to 'template' to something else :)
11:33:50 <andythenorth> but then the docs will be harder to use ;)
11:36:38 <andythenorth> I also need to think about the dir I have called 'templates'
11:37:01 <andythenorth> as that's a convention for the python module, but collides with our convention for newgrfs
11:37:19 <andythenorth> 'templates' in HEQS currently means something quite different to templates in 'BANDIT'
11:37:32 <andythenorth> those quote marks were wrong :P
11:43:02 *** KindOne has joined #openttd
11:43:57 <andythenorth> maybe I don't need them in a dir
11:44:02 <andythenorth> a bigger project might
11:44:59 <andythenorth> Alberth: would you have time to review some more code?
11:45:06 <andythenorth> it might help figure out the structure too
11:49:33 <Alberth> sure, just need to get a coffee
11:51:06 <Eddi|zuHause> andythenorth: yes, .gnml is the generated files. that is to easily do "rm -rf *.gnml" on clean
11:51:52 <Eddi|zuHause> and "**.gnml" in hg
11:51:57 <Alberth> -f ? you protect them?
11:52:23 <Eddi|zuHause> Alberth: that was just an example
11:53:42 <Eddi|zuHause> actually... the makefile doesn't currently seem to do that
11:56:41 <Alberth> I normally explicitly calculate which files to delete: for i in $(OBJFILES) $(CXXFILES:.cpp=.d) table/strings.h $(TARGET); do if test -f $$i; then rm $$i; fi; done
11:58:24 <CIA-1> OpenTTD: rubidium * r23878 /trunk/ (6 files in 4 dirs): -Prepare: 1.2.0-beta4
12:00:13 *** yorick_ has joined #openttd
12:02:05 <CIA-1> OpenTTD: rubidium * r23879 /tags/1.2.0-beta4/: -Release: 1.2.0-beta4
12:04:09 <Eddi|zuHause> Alberth: right, but not all .gnml files have a makefile rule
12:05:01 <Eddi|zuHause> and that fails to remove files whose original got removed
12:05:54 <Ammler> does distcc or ccache have advantage for the final binary?
12:06:18 <Alberth> Eddi|zuHause: nah, CXXFILES=config_reader.cpp fileio.cpp language.cpp main.cpp map.cpp math_func.cpp ... :)
12:07:32 <Alberth> Ammler: other than building it possibly faster, not that I am aware of
12:07:37 <Eddi|zuHause> Alberth: but if one of those files is removed from the repository (moved, split, obsoleted, etc.), then it won't be in CXXFILES anymore
12:08:20 <Eddi|zuHause> Alberth: the repository makes sure the .cpp file is removed, but the compiler files remain
12:14:59 *** sla_ro|master has joined #openttd
12:17:45 *** |Jeroen| has joined #openttd
12:17:50 *** Brianetta has joined #openttd
12:18:32 <andythenorth> Alberth: might be worth getting a checkout
12:18:54 <Alberth> ah, you are not frozen :)
12:20:28 <andythenorth> I am building lego forest police with a toddler
12:20:35 <andythenorth> but I can multi-task ;)
12:21:13 <andythenorth> most of my code is now written with some > 0 amount of child-based brain contention
12:24:26 <andythenorth> worth reading the templates
12:24:33 <andythenorth> the syntax is easy enough
12:24:39 <andythenorth> I learnt it 7 years ago without docs :)
12:25:09 <andythenorth> I'm not convinced I need the 'vehicle.properties' dict
12:25:21 <andythenorth> I could just create instance variables on vehicle
12:25:52 <andythenorth> but I don't know an easy way to do that without manually writing out every var
12:26:14 <andythenorth> and afaik, programmatic creation of vars is frowned upon anyway, due to possible collisions
12:27:42 <andythenorth> could just do a self[i] = j for [i,j in properties.items()]
12:32:25 <andythenorth> Alberth: if you look at the templates, I should mention that where you see <tal:foo ...> the foo part is arbitrary
12:33:00 <Alberth> you could make properties queryable from the instance
12:34:06 <Alberth> I sort of skipped lines with "<tal: ......>" :)
12:35:13 <Alberth> (I am considering to have a go at a WT3 REST server)
12:36:49 <Alberth> BANDIT_vehicles_config.py is generated?
12:37:14 <Alberth> it could be moved to a ConfigParser file :)
12:39:59 <andythenorth> BANDIT_vehicles_config.py is generated from my web server
12:40:17 <andythenorth> although I want to try building a browser app to generate it locally
12:40:30 <andythenorth> using pyramid framework, or just simple python wsgi
12:40:40 <andythenorth> tal is template attribute language
12:40:49 <andythenorth> chameleon is a tal implementation
12:41:18 <andythenorth> it's one of about 4 or 5 widely used python templaters for web stuff
12:41:19 <Ammler> wsgi is just a way deploy...
12:41:39 <andythenorth> well it's a little more
12:42:27 *** valhallasw has joined #openttd
12:43:49 <andythenorth> Alberth: if I've understood your goal with WT3 correctly, you might also use either pyramid (or a similar lightweight framework), or just write a from-scratch python app, using wsgi for the http server interface etc
12:44:26 <Ammler> just meant, using wsgi should not be up to the dev :-)
12:45:10 <andythenorth> you might use paste as server to dev locally, but not maybe in production :)
12:45:39 <Alberth> yeah, smaller frameworks are better, but you need authentication and utf-8 support
12:46:30 <Ammler> we don't use wsgi to deploy hg or paste
12:48:50 <andythenorth> what's in front of hg? anything?
12:50:14 <Mazur> I don't know the table by heart, but I believe it's Au.
12:51:29 <andythenorth> Alberth: ConfigParser?
12:51:50 <andythenorth> so that's a module...
12:54:26 <Alberth> andythenorth: the INI file parser module of Python
12:55:15 <andythenorth> is it overkill? :D
12:55:21 <andythenorth> my config file is a dict...
12:56:59 <Alberth> it is less cluttered with quotes, commas, and curly brackets, and more human readable, but that's all
12:58:07 <andythenorth> those are all the things I want to get rid of, (except for human readable)
12:58:25 <andythenorth> currently I use the web cms to abstract from those things
12:58:33 <andythenorth> let's see about ConfigParser then
12:58:42 <Eddi|zuHause> ini format sounds like a sensible solution
12:59:03 <andythenorth> "You can use this to write Python programs which can be customized by end users easily."
12:59:40 <Alberth> I used it as file format for state machines :p
13:00:02 <andythenorth> can it handle equivalent of list structures?
13:00:18 <andythenorth> foo = (value1,value2,value3) etc
13:01:25 <Alberth> it's just strings, so after reading, you can do some string processing, like [val.strip() for val in split(line, ',')]
13:01:41 <Alberth> hmm, line.split(',') of course
13:01:42 <andythenorth> so this is basically sane name value parits
13:01:55 <andythenorth> it has utilities for modifying the config file as well
13:02:15 <andythenorth> that's the main reason I'm using a web cms: I can batch add / remove / change attributes
13:02:17 <Alberth> you can create and write a config file too :)
13:02:38 <andythenorth> is writing a dict then importing it just a bit dumb then?
13:02:46 <andythenorth> it's not very human readable for starters
13:02:55 <andythenorth> I tried pretty print, but it was tmwftlb
13:03:26 <Alberth> import is inherently dangerous, as you execute the imported file
13:03:41 <andythenorth> easy attack vector
13:03:57 <Alberth> but as machine-readable format, it would be fine imho
13:03:58 <andythenorth> especially as my instructions are 'paste here [stuff from the interwebs]'
13:04:50 <Alberth> json would be another option, which has support for different data structs as well
13:05:02 <andythenorth> I considered JSON too
13:05:10 <andythenorth> the INI style file is very simple
13:06:22 <Alberth> the only disadvantage imho is that it does not give you a true dict, you have to fiddle with method calls a bit to get all data you want
13:06:42 <andythenorth> the dict arrives at Truck class anyway and is parsed
13:06:47 <andythenorth> it's only an intermediate format
13:06:54 <andythenorth> I'm not convinced it's necessary
13:07:05 <andythenorth> especially as I am not happy with vehicle.properties in the templates
13:08:32 <andythenorth> I'll rewrite the CMS to output INI style format
13:08:48 <andythenorth> in case you point out anything else that should be done first
13:12:47 <andythenorth> Alberth: you might raise your eyebrows at how I am getting some global constants into the templates
13:16:51 *** Rhamphoryncus has joined #openttd
13:17:09 <andythenorth> __setattr__ and __getattr__ are new to me
13:17:21 <andythenorth> although they match a slightly more abstracted version used in my cms
13:21:03 <Rubidium> @topic set 1 1.1.5, 1.2.0-beta4
13:21:03 *** DorpsGek changes topic to "1.1.5, 1.2.0-beta4 | Website: *.openttd.org (translator: translator, server list: servers, wiki: wiki, patches & bug-reports: bugs, revision log: vcs, release info: finger) | Don't ask to ask, just ask | 'Latest' is not a valid version, ever | English only"
13:21:14 <andythenorth> so this is a mapping into the properties dict?
13:21:27 <andythenorth> rather than creating instances of variables directly on the object?
13:22:10 <Alberth> the indenting of build_bandit.py is not PEP8 compliant :)
13:22:11 <Alberth> but otherwise mostly fine. Not sure what to change further tbh
13:22:31 <Alberth> yes, it redirects attribute access to the properties dict
13:22:54 <Alberth> at the cost of not being able to make your own variables, as you can see in the init
13:25:27 <Alberth> I would have moved all the assignments to below the classes, and put all the stuff in a run() function, but that's not a big issue
13:25:51 <andythenorth> might be worth considering
13:25:59 <andythenorth> I have no problem binning code
13:27:57 <Alberth> together with if __name__ == '__main__': run() it allows you to import the file without running the code
13:28:37 <CIA-1> OpenTTD: michi_cc * r23880 /trunk/src/core/alloc_type.hpp: -Add: A simple smart pointer.
13:28:41 <CIA-1> OpenTTD: michi_cc * r23881 /trunk/src/spriteloader/grf.cpp: -Codechange: Move GRF sprite decoding into a separate function.
13:28:45 <CIA-1> OpenTTD: michi_cc * r23882 /trunk/src/ (newgrf.cpp newgrf_sound.cpp newgrf_sound.h sound.cpp): -Codechange: Delay parsing of NewGRF sound effects until first usage.
13:28:50 <CIA-1> OpenTTD: michi_cc * r23883 /trunk/src/newgrf.cpp: -Fix: [NewGRF] It wasn't possible to import sounds from a NewGRF later in the load order.
13:28:53 <CIA-1> OpenTTD: michi_cc * r23884 /trunk/src/spritecache.cpp: -Codechange: Store recolour sprites in memory directly during GRF loading.
13:28:57 <CIA-1> OpenTTD: michi_cc * r23885 /trunk/src/ (landscape.cpp spritecache.cpp): -Codechange: Use the GRF sprite loader for then mapgen sprites as well.
13:29:01 <CIA-1> OpenTTD: michi_cc * r23886 /trunk/src/ (base_media_base.h base_media_func.h gfxinit.cpp): -Codechange: Allow limiting the MD5 file hash to the first x bytes of the file.
13:29:06 <CIA-1> OpenTTD: michi_cc * r23887 /trunk/src/ (12 files in 2 dirs): -Feature: [NewGRF] Support for container version 2.
13:29:11 <CIA-1> OpenTTD: michi_cc * r23888 /trunk/src/table/ (misc_settings.ini settings.ini): -Change: Move the min/max zoom settings to the "misc" section so they are valid before sprites are loaded.
13:29:18 <CIA-1> OpenTTD: michi_cc * r23889 /trunk/src/ (9 files in 3 dirs): -Codechange: Centralise sprite resizing in one place. (peter1138)
13:29:20 <CIA-1> OpenTTD: michi_cc * r23890 /trunk/src/spriteloader/grf.cpp: -Add: [NewGRF] Support for RealSprites with multiple zoom levels.
13:30:49 *** FLHerne has joined #openttd
13:34:33 <andythenorth> Alberth: I have two other things I'm not convinced by:
13:34:43 <andythenorth> - structure of /src, specifically use of /templates dir
13:34:49 <andythenorth> - global_constants.pt
13:39:18 <andythenorth> I'm not sure nesting into /templates is needed
13:40:06 <Alberth> global_constants.py somewhat screams INI file, but on the other hand, this may be easier to manage, and they change very little probably
13:40:19 <Alberth> better a directory too much than too few
13:41:21 <Alberth> it reduces clutter in the src directory, and I found it useful for understanding the repo structure
13:41:45 <andythenorth> global_constants.pt is basically a cheat to make CONSTANTS appear in the template namespace
13:41:54 <andythenorth> it's one extra place to find when looking where things are defined
13:42:07 <andythenorth> and it's one extra piece of code to write when adding a global var
13:42:24 <Alberth> you seem to have more trickery around the template stuff
13:42:39 <andythenorth> maybe I could make globals() an object
13:42:49 <andythenorth> and then use __setattr__ and __getattr__ for names
13:43:18 <andythenorth> the template namespace only contains what is passed to it, or what it imports
13:43:21 <Alberth> in my view the template stuff is broken if it needs global vars
13:43:32 <andythenorth> it doesn't need them particularly
13:43:46 <andythenorth> I am trying to preserve the ability to use things like SOLO_TRUCK_TYPE_NUM
13:43:55 <andythenorth> without accessing another obj or structure
13:44:04 <andythenorth> but that might not be a valid goal
13:44:12 <andythenorth> magic constants :o
13:44:29 <Eddi|zuHause> what's a "container version 2"?
13:44:57 <Eddi|zuHause> in the commitgasm
13:45:09 <andythenorth> probably replaces version 1 :P
13:45:21 <andythenorth> is it vehicles in vehicles? :D
13:50:13 <michi_cc> Eddi|zuHause: The magic stuff that enables multiple zoom levels and colour depths per sprite and allows stripping of unneeded zoom levels/colourdepths without affecting network compatibility.
13:50:18 <Alberth> in Python I only have global constants, but that may be due to the sort of things I normally program
13:51:08 <andythenorth> I considered putting them on the vehicle objects
13:51:23 <andythenorth> I don't really like code where you have to read 10 loc to find the actual point a var is defined though
13:51:49 <andythenorth> e.g. foo = truck.foo = global foo = from globals import stuff{foo:10}
13:52:10 <andythenorth> annoys other people :)
13:52:15 <Eddi|zuHause> don't read my code then
13:53:08 <andythenorth> I should have probably explained that differently :)
13:53:12 <andythenorth> one line assignments are fine
13:53:33 <andythenorth> but playing 'hunt the actual value' in 3 or 4 different files is dull
13:53:41 <frosch123> michi_cc: nice summary ^^
13:56:47 *** Phoenix_the_II has joined #openttd
14:09:05 *** Aygar[Work|afk] has joined #openttd
14:18:57 <andythenorth> Alberth: can config parser sort out types?
14:19:09 <andythenorth> or will I be calling int() and such a lot?
14:21:35 <Alberth> no, it only does strings
14:22:54 <Alberth> but otherwise, that's why loading text is so much safer than importing :p
14:24:02 *** lmergen has joined #openttd
14:24:26 <andythenorth> I ruled out JSON, it has all the same scaffolding issues - braces, etc
14:26:56 * andythenorth ponders, split lists on | or ,
14:27:04 <andythenorth> , might appear validly in strings
14:28:00 <Alberth> the latter messes up ini file formatting though
14:28:57 *** theholyduck has joined #openttd
14:29:34 <Eddi|zuHause> <andythenorth> , might appear validly in strings <-- that's why i used tsv and not csv
14:30:04 <SpComb> psv fixes all the delimiter issues!
14:30:51 <andythenorth> I used to use || for passing lists in url query strings
14:30:52 <valhallasw> csv supports escaping
14:31:06 <valhallasw> a,b,"boo,hoo",d,etc
14:31:24 <valhallasw> at least, escaping is generally accepted by parsers
14:31:32 <SpComb> valhallasw: and " in values? :P
14:32:03 <valhallasw> at least, that's what excel does
14:32:09 <SpComb> CSV is great, because it supports escaping in a variety of different fashions, all depending on what generated it and what's parsing it!
14:32:20 * Alberth ponders using a scanner for this kind of things
14:33:14 <valhallasw> >>> x.writerow(["a", ",", "'", '"'])
14:33:29 <valhallasw> SpComb: everybody uses the excel dialect
14:33:49 <SpComb> python's csv is cofigureable for different dialects
14:34:06 <SpComb> and excel uses something bizarre like UTF-8 CSV
14:34:25 <SpComb> which severely messes things up
14:34:56 <SpComb> (and python's csv module does neither unicode nor NULs)
14:35:06 *** Cybertinus has joined #openttd
14:35:23 <SpComb> pain and suffering, etc
14:36:21 <Alberth> but you may want something more light-weight :)
14:36:28 <valhallasw> however, creating /another/ new format is even worse ;-)
14:36:54 <valhallasw> andythenorth: you could also consider YAML
14:37:10 <valhallasw> or just go enterprise with XML directly
14:39:35 <Alberth> xml doesn't do text-int conversion automagically either
14:40:38 <Alberth> bleh, the scanner debug code got removed from nml :(
14:44:26 *** TWerkhoven[l] has joined #openttd
14:51:52 <Eddi|zuHause> so, is this how passing "static" parameters to strings is supposed to work? i didn't find any documentation about it: name: string(STR_NAME_ENG_pruss_E_Abt4,string(STR_NAME_ENG_PASSENGER),string(STR_NAME_ENG_ACCELERATED),string(STR_NAME_ENG_FOUR_AXLE));
14:52:28 <Eddi|zuHause> the language file contains this:
14:52:29 <Eddi|zuHause> STR_NAME_ENG_pruss_E_Abt4 :{STRING} - {STRING} (Abteilwagen, {STRING})
14:56:00 <Eddi|zuHause> and what happens if there's too few (or too many) parameters?
14:56:48 <Hirundo> demons may fly out of your nose
14:57:28 <Hirundo> I wouldn't know really, string handling is 99% Yexo's work
14:58:24 *** Biolunar has joined #openttd
14:58:33 <Hirundo> Alberth: what 'scanner debug code' ?
15:00:18 <Hirundo> frosch123: Is there any relationship between real sprite 'info version 7 / 32' and grfv7/grfv8?
15:01:02 <Alberth> Hirundo: a function that runs the scanner only, and dumps the found tokens. Would have been nice to show andy
15:01:20 <frosch123> Hirundo: there is no relation between grf version, nfo version and grf container version
15:01:25 <frosch123> all combinations are possible
15:01:40 <frosch123> however, only grf container 2 plus nfo32 allow usage of multiple zoom levels or 32bpp
15:02:05 <frosch123> conversion to grf container 1 or nfo < 32 will strip 32bpp and zoom levels, but still work
15:02:34 <Hirundo> Obviously, I should have checked the forum before asking questions :-)
15:02:44 <michi_cc> NFO version is only really relevant for grfcodec though.
15:02:53 <frosch123> you asked your question 2 minutes before the forum post :p
15:03:33 <Hirundo> NML can output both grf (needs container v2) and nfo (needs nfo v32)
15:06:59 <Eddi|zuHause> nmlc: "src/../src/pruss/E_Abt4.gnml", line 126: First parameter of string() must be an identifier. <-- that's probably the wrong error message :p
15:07:47 <Eddi|zuHause> (or maybe i'm doing something else wrong)
15:08:42 <Eddi|zuHause> ah, think i found it
15:13:04 <Eddi|zuHause> openttd: /home/johannes/spiele/OpenTTD/trunk/src/strings_func.h:82: int64 StringParameters::GetInt64(WChar): Assertion `this->offset < this->num_param' failed.
15:13:09 <Eddi|zuHause> that's probably not right...
15:15:12 <Eddi|zuHause> Removed NewGRF: GRF ID 4A430201, filename: newgrf/test/ng_frails.grf (matches GRFID only). Gamelog inconsistency: GrfID was never added! <-- err...??
15:15:40 <Eddi|zuHause> (that is probably not related, though :p)
15:15:59 <Hirundo> loaded as static grf?
15:16:53 *** theholyduck has joined #openttd
15:18:45 <Hirundo> frosch123: Currently the GRFTempEngineData::refitmask_grf is set only when setting the xor mask, which I plan to deprecate / remove in NML
15:19:48 <Hirundo> Might it make sense, to also set refitmask_grf when setting the allowed / disallowed cargo lists?
15:21:44 *** frosch123 is now known as frosch
15:22:46 <Eddi|zuHause> anyway... conclusion: NML happily applies too many or too few parameters, but openttd crashes when displaying the string in the vehicle menu
15:22:53 <Eddi|zuHause> it probably shouldn't do that
15:24:11 <andythenorth> openttd is very fussy about strings changing :P
15:24:18 <andythenorth> most times I crash it is because I changed a string :)
15:24:35 <andythenorth> valhallasw: XML has the same 'way too much scaffolding problem' as others
15:24:56 <Eddi|zuHause> for varying values of "same" :)
15:26:23 <andythenorth> could be prettier
15:26:43 <andythenorth> but I like to try INI format
15:26:52 <andythenorth> as intermediary at least
15:27:00 <andythenorth> easier for others to work with
15:27:10 <Hirundo> frosch: Why is the refit mask not translated when parsing the property?
15:27:12 <andythenorth> and I'm not doing write = anonymous on my cms :P
15:27:28 *** Chris_Booth has joined #openttd
15:30:05 <andythenorth> Alberth: anything wrong with this to set attrs on an object?
15:30:05 <andythenorth> for i,j in properties.items():
15:30:06 <andythenorth> self.__setattr__(i, j)
15:30:25 <andythenorth> I could probably play the one line game on it :P
15:30:33 <Eddi|zuHause> what's wrong with self[i]=j?
15:30:45 <andythenorth> TypeError: 'Truck' object does not support item assignment
15:30:51 <andythenorth> tried that first
15:31:04 <Eddi|zuHause> self should be derived from dict, then
15:32:35 <andythenorth> "python is so bracing"
15:32:45 <Rhamphoryncus> Deriving from dict is usually pretty silly
15:33:09 <Rhamphoryncus> andythenorth: use setattr()
15:33:30 <andythenorth> rather than __setattr__()
15:34:09 <Rhamphoryncus> __setattr__ is the backend API to be used by setattr() and normal assignment
15:34:20 <andythenorth> setattr() is not available to my class
15:34:28 <Rhamphoryncus> it's a global function
15:34:28 <andythenorth> AttributeError: 'Truck' object has no attribute 'setattr'
15:34:37 <Rhamphoryncus> A builtin more specifically
15:34:58 <Rhamphoryncus> Think of it as a keyword that happens to be build as a function
15:37:24 <andythenorth> not available to me for some reason ;)
15:38:01 <valhallasw> andythenorth: setattr(self, 'attr', 'value')
15:38:04 <Rhamphoryncus> are all the builtins missing? len, zip, int, float, str, list, dict..
15:38:29 <valhallasw> andythenorth: you attributeerror suggests you're doing self.setattr
15:38:36 <andythenorth> that is the error
15:38:45 <valhallasw> setattr(self, 'attr', 'value') < do that instead
15:38:53 <Rhamphoryncus> It's a builtin. It's not a method.
15:44:29 <Alberth> (16:35:04) andythenorth: Alberth: anything wrong with this to set attrs on an object? <-- except for breaking the data hiding concept and the 'objects take their of their own data' concept, absolutely nothing
15:45:36 <Alberth> imho if you have dynamic entries, use a dict rather than an object.
15:46:37 <Eddi|zuHause> there's probably also "named dicts" or similar
15:47:12 <Eddi|zuHause> that you can access by both blah["i"] and blah.i
15:47:16 <andythenorth> Alberth: so the solution you posted for mapping into a dict is better?
15:48:20 <andythenorth> what about just explicitly writing out all the attributes I want on the object?
15:48:29 <andythenorth> I don't trust dynamic attributes either
15:48:39 <Alberth> but then I am generally very worried about changing state in ways I cannot find again
15:49:32 <andythenorth> I don't mind explicitly writing them out
15:49:37 <Rhamphoryncus> Eddi|zuHause: no named dicts. Easy to do, but it's frowned upon. Pick one approach and stick to it.
15:50:14 <andythenorth> writing out explicitly makes it easy to understand the interface
15:50:22 <andythenorth> but it's quite a lot of loc
15:50:25 <Alberth> andythenorth: that's the alternative, but you loose the option to iterate over them like in a dict
15:50:37 <andythenorth> I have no requirement for that at the moment
15:51:10 <andythenorth> this way there are fewer entities
15:51:18 <Alberth> you need to list the attributes in 'properties' too, which costs also loc
15:51:34 * andythenorth thinks manually write them out
15:52:26 <Alberth> in need of some warm food
15:52:49 <SamCat> I live in California... explain to me... what is this 'cold' you speak of?
15:53:17 <Alberth> no 'weather in europe' news items there?
15:53:39 <planetmaker> you do the conversion to °F, I can't be bothered ;-)
15:54:24 <Rubidium> planetmaker: it's simply... just < 0°F and < 0°C
15:55:11 <frosch> planetmaker: 0°F is the coldest thing which that fahrenheit guy knew
15:55:16 <Rubidium> or at least it was around 08:00 for me
15:55:18 <frosch> (and thought is possible)
15:55:21 <frosch> currently it is colder :p
15:55:26 <planetmaker> iirc it*s T(°F) = (T(°C) - 32) * 4/9 or so
15:55:34 <SamCat> I guess I'm kinda just tired
15:55:42 <SamCat> I've been playing with trains instead of sleeping like I ought to have
15:55:57 <Rubidium> planetmaker: so, it can be both less than 0 °F and less than 0 °C, right?
15:56:00 <valhallasw> planetmaker: other way around
15:56:02 <frosch> in other words, the lower fixpoint of the fahrenheit unit is one of the most stupid things in human science
15:56:47 <planetmaker> yes, I know. I just failed to derive the formula from that knowledge :-P
15:56:54 <valhallasw> frosch: it was 'the lowest temperature'. Which was wrong, but hey.
15:57:03 <valhallasw> it's as arbitrary as 0 and 100°C are
15:57:31 <frosch> celsius is very useful for cooking and weather
15:57:32 <SamCat> that would be on the freezing and boiling point of water!
15:57:40 <planetmaker> that is clearly defined as freezing and boiling temperature of destilled water at normal pressure, valhallasw
15:57:49 <frosch> 100°F is useful when caring for cows
15:57:51 <planetmaker> °F is like "cold mixture" and "body temperature"
15:58:25 <planetmaker> And for added benefit 1K = 1°C
15:58:32 <planetmaker> as differences go. Not absolute
15:58:38 <SamCat> yeah, but having celsius based on water is awesome because everything else in the metric system is based on water
15:58:38 <frosch> so, while 100°F might be more useful than 100°C, 0°F is completely stupid
15:58:39 <Elukka> <@planetmaker> iirc it*s T(°F) = (T(°C) - 32) * 4/9 or so
15:58:41 <valhallasw> planetmaker: and 1°R = 1°F
15:58:43 <SamCat> okay, a lot of stuff is based on water
15:58:51 <frosch> planetmaker: thats what R is for :p
15:58:51 <Elukka> no, these days it's "20 c in f" written in google :p
15:58:53 <SamCat> a cubic centimeter of water weighs a gram, for example
15:58:58 <Rubidium> frosch: at standard atmospheric pressure ;)
15:58:58 <frosch> but neither R not C have a ° :p
15:59:09 <Elukka> google is terribly useful for unit conversions
15:59:15 <planetmaker> valhallasw: your point being?
15:59:20 <valhallasw> frosch: nor K you mean ;-)
15:59:46 <Elukka> it can do esoteric stuff too, i once asked google what "200 times the speed of light in AU/s" was :P
16:00:13 <planetmaker> °C - and K of course foremost - goes quite well with the thermodynamic definition of temperature
16:00:16 <valhallasw> planetmaker: my point is: 1°C=1K is not a specific advantage - we could use Rankines as well
16:00:23 <frosch> hmm, i wondered what's so special about 33
16:00:30 <valhallasw> how does °C go well with thermodynamics?
16:00:33 <valhallasw> K goes well and R goes well
16:00:45 <SamCat> yes, but the intervals of K were derived from C so there
16:00:53 <frosch> why does °N define 33°N = boiling of water? :o
16:01:25 <valhallasw> why do we use kelvins or rankines at all? why not just joules!
16:01:30 <planetmaker> South of it it's hot, frosch
16:01:48 <SamCat> because joules are a measure of energy whereas kelvins are a measure of average energy
16:01:50 <planetmaker> valhallasw: then you didn't understand the concept of external and internal thermodynamic properties
16:01:54 <SamCat> (average thermal energy)
16:02:17 <valhallasw> SamCat: how is 'average thermal energy' not an energy?
16:02:21 <Eddi|zuHause> frosch: what's so special about 100?
16:02:23 <Alberth> Rubidium: it's B! >>> chr(ord('K') & ord('R')) --> 'B'
16:02:34 <Elukka> last time there was talk of thermodynamics on this channel someone was arguing the mass of a certain amount of liquid changes with temperature, and volume doesn't...
16:02:40 <planetmaker> with the same argument, valhallasw, you could argue away the concept of (mass) density when we have mass itself
16:02:46 <Eddi|zuHause> well, technically, Celsius wanted 0 to be the boiling point, and 100 the freezing point
16:02:55 <Eddi|zuHause> but it got switched around quite fast
16:03:00 <SamCat> valhallasw: because something with lots of mass has more energy in it than something with less mass?
16:03:04 * andythenorth should maybe learn how to use pdb :P
16:03:17 <valhallasw> SamCat: so? in terms of units it's still energy
16:03:18 <frosch> Eddi|zuHause: i did not ask why 60°Rø = boiling water
16:03:21 <Alberth> andythenorth: 'print' works too :)
16:03:25 <Rubidium> Alberth: I didn't say is, I said make (although I should have said made)
16:03:26 <frosch> i know about traditional usages of 100 and 60
16:03:30 <planetmaker> valhallasw: temperature is NOT energy
16:03:33 <valhallasw> it's a different physical quantity
16:03:34 <planetmaker> that's fundamentally wrong
16:03:39 <SamCat> valhallasw: if I apply a blow torch to a pin it's going to get really hot really fast but if I apply it to a ton of steel it's barely going to do anything
16:04:05 <SamCat> valhallasw: you're applying the same amount of energy (joules) but because you're applying it to more mass you get less rise in temperature (kelvin)
16:04:11 <valhallasw> planetmaker: yes, and newtonmeters and newtonmeters can be completely different things. So what?
16:04:38 * andythenorth looks for something like getattrs
16:04:39 <valhallasw> (namely: energy and torque)
16:05:03 <valhallasw> SamCat: yes, I understand. The fact that they are different physical quantities does not mean we cannot use the same units.
16:05:32 <SamCat> SamCat: saying kelvins are the same as joules is like saying a volt is the same as an amp
16:05:52 <SamCat> valhallasw: yes, it does, beacuse you can't describe something's temperature by saying how many joules are in it
16:05:56 *** FLHerne has joined #openttd
16:06:13 <SamCat> valhallasw: not only would you need to know the amount of mass but you would need the substance's thermal coefficient
16:06:23 <valhallasw> SamCat: wrong. I can't describe somethings temperature by saying how many *energy* is in it.
16:06:34 <valhallasw> the *units* are seperate from the *physical quantity*
16:07:05 <SamCat> valhallasw: but joules are a measure of energy...
16:07:19 <Eddi|zuHause> valhallasw: next you're trying to tell us that mass is energy as well :p
16:07:25 <valhallasw> joules are *generally used* as a measure of energy
16:07:34 <Rubidium> Eddi|zuHause: that'd be useful
16:07:39 <valhallasw> Eddi|zuHause: that's what theoreticians do generally, yes.
16:07:40 <planetmaker> valhallasw: you cannot describe the temperature by the energy content. You need many other parameters to succeed
16:07:53 <valhallasw> planetmaker: I'm not doing that. I'm describing temperature in Joules.
16:07:53 <SamCat> vahlallasw: saynig something has 100 joules is like saying you have 100 liters of water... saying something is 100K is like saying the water is at 100 mm of hg
16:07:58 <valhallasw> That's something completely different.
16:08:02 <Rubidium> today it's the weather will 42kg in the afternoon and a mere 30kg at night
16:08:23 <SamCat> you know... when I made the joke about someone explaining cold to me I didn't exactly expect a debate about thermodynamics ;)
16:08:26 <Rubidium> and my grammar is totally sucking at the moment
16:08:48 *** lmergen has joined #openttd
16:08:52 <Alberth> SamCat: we are always highly off-topic here :p
16:09:08 <Eddi|zuHause> well, there _are_ those crazy physicists who define c=1, and thus E=m
16:09:24 <valhallasw> SamCat: saynig something has [Energy] is like saying you have [Volume] of water... saying something is [Temperature] is like saying the water is at [Pressure].
16:09:26 <Alberth> Eddi|zuHause: they live in a different universe :)
16:09:32 <SamCat> valhallasw: you can't describe temperature with units of energy for the same reason you can't describe speed with just units of distance or just units of time or why you can't describe pressure with just units of mass
16:10:04 <Eddi|zuHause> Alberth: ah well, we have plenty of those :p
16:10:07 <SamCat> vahlallasw: yeah... that's what I said... now I'm completely confused ;_;
16:10:30 <SamCat> valhallasw: nevermind, I misread what you said
16:10:40 <valhallasw> SamCat: it's the difference between *units* [J] and *physical quantity* [Energy] that's important
16:10:56 <valhallasw> you can use the same *unit* (J) for a *different* physical quantity (Temperature)
16:11:05 <SamCat> valhallasw: right... joules are a measure of thermal energy, why?
16:11:14 *** theholyduck has joined #openttd
16:11:16 <valhallasw> just like you use Nm's for both energy and torque
16:11:21 <Eddi|zuHause> valhallasw: so you're also saying power = loudness?
16:11:40 *** Cybertinus has joined #openttd
16:11:45 <valhallasw> Eddi|zuHause: no, I'm not. Those are different physical quantities
16:11:47 <Eddi|zuHause> for sure, if i get speakers with more watts, they're louder
16:11:55 <valhallasw> but if you want to use the same units, go ahead
16:12:17 <SamCat> Eddi: no, you've got less efficient speakers. there's a difference!
16:12:22 <Eddi|zuHause> but it's the same thing, if i put more energy in, it gets warmers
16:12:34 <valhallasw> SamCat: I can also use Joules to measure Torque
16:12:54 <SamCat> valhallasw: you can use joules to measure torque... but torque is not temperature
16:12:54 <valhallasw> people will look at me as if I'm crazy, but that's a different story
16:13:40 <SamCat> valhallasw: actually you can't... because torque is usually a description of power, not energy and joules are energy not power. You need Watts or Newtons for that
16:14:31 <Eddi|zuHause> SamCat: actually, if i apply the right torque to your arm, it gets warm :p
16:14:44 <valhallasw> SamCat: again, the difference between units and physical quantities
16:14:52 <valhallasw> Joule = kg m^2/s^2
16:15:00 <Eddi|zuHause> we call that "Brennessel"
16:15:25 <valhallasw> Newton metre = kg m^2/s^2
16:16:29 <valhallasw> but although you use the same units for energy and torque, they are clearly different physical quantities
16:16:52 <andythenorth> it's long but explicit. No 'wtf is that value coming from?'
16:17:11 <andythenorth> tempted to align all the = though :P
16:17:54 <SamCat> Valhallasw: the joule is an SI unit that is DEFINED AS a measure of energy. You can therefore not use it to describe the concentration of said energy in a system (temperature for example) or to describe power without considerations of other variables
16:18:19 <SamCat> oh hey look! something OpenTTD related!
16:20:08 <SamCat> by the way... the reason I came here was to ask: how can you tell when an industry has reached that point in its production where it's incapable of increasing production? I've been playing a larger map than my usual 64*64 where there are multiples of industries but they're not going away and they don't increase production either even when it says they have
16:20:15 <valhallasw> SamCat: as far as I know, it's only defined as J=kg m^2/s^2 - the rest is convention
16:21:00 <valhallasw> and inventing a new temperature scale is breaking with convention anyway ;-)
16:21:31 <SamCat> because I see a lot of industries that are producing 40 units of something per month which is kinda where I assumed the point of no return was
16:21:33 <planetmaker> Point is though that the Kelvin is one of the 7 base units
16:22:14 <planetmaker> SamCat: there's no general way to tell
16:22:39 <SamCat> planetmaker: damn... so how come the damn things won't sod off already?!
16:23:07 <planetmaker> depends on the industry. Any newgrf there?
16:23:26 <SamCat> yes, but no industries
16:23:31 <planetmaker> the last instance of an industry actually might never close
16:23:46 <SamCat> I know that last bit but this is in a map with at least two of everything
16:24:05 <planetmaker> and "no industries" is... not clear. There's quite a few newgrfs which are not industry newgrfs on first sight :-) But which actually are
16:24:27 <CIA-1> OpenTTD: rubidium * r23891 /trunk/src/video/sdl_v.cpp: -Fix-ish: compile failures with SDL 1.3
16:24:41 <planetmaker> And even then, the industry might need several years before it decides to close down
16:25:09 <SamCat> the newgrfs I was using: ukrs2, ukrs2+,finescale tracks, uk town set, bob's random british vehicles, BATS, use-able default aircraft, city stations, rural stations, suburban stations, industrial stations renewal, generic cars, and FISH
16:27:34 <planetmaker> I'm not sure about uk towns
16:28:07 <planetmaker> I simply don't know it enought
16:28:27 <SamCat> my understanding is it just adds a bunch of new house graphics but I should double check
16:29:10 <andythenorth> would it be useful if a newgrf reported what features it implements?
16:29:17 <planetmaker> let's assume it doesn't do anything. Then it's just the default industries where there's basically a random chance to increase or decrease production - or not change
16:29:22 <andythenorth> (action 0 features)
16:29:26 <planetmaker> Thus... it's quite random how long it takes, SamCat
16:29:44 <planetmaker> andythenorth: no. It does that already
16:30:14 <andythenorth> did I miss that :o
16:30:35 <planetmaker> andythenorth: feature byte
16:30:41 <planetmaker> you did not miss that ;-)
16:30:48 <SamCat> planetmaker: thanks... it's just frustrating when you see it reporting increases and decreases by percentage of an industry which is stuck at minimum and unsalvigable
16:31:13 <SamCat> by the way... would it make sense in the code to have it check to see if the production value is at 1 and, if so, to add one instead of trying to multiply it?
16:31:25 <SamCat> or subtract one instead of trying to multiply it
16:32:08 <planetmaker> SamCat: it's up to the newgrf
16:32:22 <andythenorth> I was thinking of a window in newgrf config
16:32:22 <planetmaker> and up to the economy you enabled
16:32:35 <andythenorth> This grf modifies: (list)
16:32:55 <SamCat> yeah, that would be nice Andythenorth
16:33:01 <planetmaker> andythenorth: iirc that's part of grf-topia
16:37:53 * andythenorth is unconvinced by
16:37:53 <andythenorth> from global_constants import * #import all stuff from constants for easy reference in python scripts
16:38:22 <andythenorth> means it's guesswork where the actual values / objects in global_constants are coming from
16:38:28 <Eddi|zuHause> import * is an antipattern
16:38:37 <Eddi|zuHause> or at least code smell
16:38:52 <andythenorth> better to be explicit
16:39:05 <Eddi|zuHause> just import global_constants
16:39:13 <Eddi|zuHause> and use global_constants.blah everywhere
16:39:42 <Rhamphoryncus> SamCat: Do I understand this right? An industry can drop to an internal output value of 1, which gets multiplied up when displayed to the user (which is why I've never seen 1), but all the modifiers aren't sufficient to change it any further?
16:39:59 <andythenorth> I have a non-useful obsession with trying to keep lines short
16:40:18 <andythenorth> e.g. self.non_refittable_classes = standard_class_refits['default']['disallow']
16:40:21 <andythenorth> self.non_refittable_classes = global_constants.standard_class_refits['default']['disallow']
16:40:34 <SamCat> Rhamphoryncus: that's my understanding of how it works, yes, but I'm not one of the devs so I can't be sure
16:40:35 <andythenorth> but it makes for bad reading
16:41:57 <Eddi|zuHause> Rhamphoryncus: an industry produces every 255 ticks, which is 8 or 9 times per month, the GUI will only ever display the sum of all these. so the lowest an industry can produce per month is 8
16:42:01 <kermie> I need help installing openttd 1.0.5
16:42:13 <Eddi|zuHause> kermie: don't install 1.0.5
16:42:14 <kermie> + I am beginner with ubuntu
16:42:16 <Rhamphoryncus> Eddi|zuHause: *nods*
16:42:27 <SamCat> kermie: because it's old
16:42:29 <Eddi|zuHause> kermie: because that is outdated by over a year
16:42:35 <Eddi|zuHause> kermie: we have 1.1.5 now
16:42:38 <Rhamphoryncus> SamCat: I'd definitely call it a bug for an intended increase/decrease to have no effect
16:42:45 <Eddi|zuHause> kermie: and 1.2.0-beta4
16:42:48 <SamCat> I think the 1.0.5 is the one in the ubuntu repos
16:43:04 <kermie> I currently have 1.0.0 installed
16:43:10 <Eddi|zuHause> kermie: so, what did you try, and what fails?
16:43:40 <kermie> Let me delete it and try to download 1.1.5 and install it again and i'll tell you what happens
16:43:44 <SamCat> download the .deb and install it ;)
16:43:53 <SamCat> that's what I did on my computer
16:44:23 <planetmaker> you don't even need to install it actually
16:44:28 <kermie> An older version is available in a software channel Generally you are recommended to install the version from the software channel, since it is usually better supported.
16:44:31 <planetmaker> much easier to update without insall
16:44:51 <kermie> That's why i used ubuntu repos
16:45:22 <SamCat> ubuntu's repos are outdated last I chicked
16:45:36 <SamCat> if you install the .deb ubuntu will recognize that it's a newer version of the same program, too
16:46:19 <kermie> version 1.0.0.0-2 is installed
16:46:43 <SamCat> uninstal the old one first?
16:46:44 <kermie> I'll try to uninstall it first
16:46:47 *** JVassie has joined #openttd
16:48:20 * andythenorth concludes that trying to make GLOBAL_CONSTANTS_LIKE_THIS is tmwftlb
16:48:35 <andythenorth> in python templates
16:49:47 <SamCat> kermie: yes, it removes config files as well as the app itself
16:51:10 <Eddi|zuHause> andythenorth: if you have a module as prefix, there's no use syntactically separating the constants by capitalizing them
16:51:13 <kermie> i did not select complete removal
16:51:27 <kermie> i have removed (normal) all 3 entries of openttd installed
16:51:32 <kermie> I have installed it succesfully
16:51:52 <andythenorth> Eddi|zuHause: I was trying to preserve the CPP method, I quite like it. This is specific to use in the template files, not the python script ;)
16:52:00 <andythenorth> but it's just wrong
16:52:25 <SamCat> kermie: are you sure you're using the right .deb?
16:52:31 <SamCat> kermie: what version of ubuntu are you using?
16:52:48 <andythenorth> ${FOO} is not much better than ${global_constants.foo}
16:53:39 <kermie> SamCat: You are using Ubuntu 10.04 LTS - the Lucid Lynx - released in April 2010 and supported until April 2013.
16:54:34 <kermie> SamCat: openttd-1.1.5-linux-ubuntu-lucid-i386.deb
16:54:44 <SamCat> Kermie: okay... I'm using 11.10 here...
16:54:59 <kermie> SamCat: I've also updated ubuntu
16:55:09 <kermie> Maybe it's because i have no GFX
16:55:10 <SamCat> have you tried running it from a terminal and seeing if it pukes any error messages at you?
16:55:24 <kermie> SamCat: i don't know how to do that
16:55:26 <SamCat> ooooh yeah... you need the gfx
16:55:38 <kermie> I have the original GFX can i use those?
16:55:42 *** kkb110_ has joined #openttd
16:55:44 <kermie> The new ones are just... ugly
16:55:49 <SamCat> open a terminal and type openttd
16:55:55 <SamCat> yeah, you can use the original
16:56:06 <SamCat> I like the new ones...
16:56:37 <SamCat> cool, install graphics set
16:56:47 <kermie> What should i get from my windows ?
16:57:03 <Eddi|zuHause> all .grf files and sample.cat
16:57:23 *** theholyduck has joined #openttd
16:57:23 <Eddi|zuHause> and the .gm files if you want music
16:57:33 <kermie> all grf sample.cat and .gm
16:57:54 <Eddi|zuHause> the grfs and sample.cat in ~/.openttd/data
16:58:07 <planetmaker> in ~/.openttd/baseset
16:58:07 <Eddi|zuHause> the gms in ~/.openttd/gm
16:58:17 <Eddi|zuHause> planetmaker: also in 1.1.5?
16:58:36 <kermie> how do i reach ~/.openttd
16:58:52 <Eddi|zuHause> kermie: select "show hidden files" in your filemanager
16:59:30 <Eddi|zuHause> ~ is your home directory
16:59:45 <Eddi|zuHause> /home/<username>/
16:59:47 <kermie> bin boot cdrom dev etc home
17:01:19 <kermie> They can not escape the craving of ottd
17:01:49 <kermie> content_download/data you mean
17:02:51 <Eddi|zuHause> now put the .grfs in
17:03:20 <Eddi|zuHause> well, they are on your original cd :)
17:03:31 <Eddi|zuHause> (if you have the windows version)
17:03:41 <Eddi|zuHause> (they're not on the dos version)
17:03:52 <kermie> Why can't i alt tab from the game
17:04:34 <Eddi|zuHause> that has always worked...
17:04:55 *** Phoenix_the_II has quit IRC
17:07:37 <kermie> Thanks for the help but i'm dissapointed about 2 things
17:08:14 <kermie> And scrolling fast with right click throughout the land skips and isn't smooth
17:08:43 <kermie> Thanks again for the help. I'm going back to windows :D
17:22:45 * andythenorth just decided that debugging python-driven newgrfs was way too hard
17:23:08 <andythenorth> ...found I have wrong version of grf in game
17:24:26 <SamCat> *sighs* I wish industries would close once they become unrecoverable
17:26:21 <Alberth> nothing the magic bulldozer cannot fix :)
17:26:49 <SamCat> yeah, but then I have to use cheats
17:28:28 <Rubidium> are you using an industry NewGRF?
17:30:09 <andythenorth> so .pynml ok then?
17:30:20 <andythenorth> for nml files templated with python
17:38:35 *** [1]Mark has joined #openttd
17:38:36 *** Mark is now known as Guest1544
17:38:36 *** Guest1387 is now known as Mark
17:47:25 <andythenorth> || sucks as a separator for lists in config files
17:50:24 *** theholyduck has joined #openttd
18:01:35 *** lmergen has joined #openttd
18:13:06 <planetmaker> andythenorth: ";"
18:13:27 <andythenorth> , might turn up in strings sadly
18:13:32 <andythenorth> | less so I gope
18:14:17 <planetmaker> short for "guess and hope" ;-)
18:16:24 * andythenorth wonders if python has equivalent of setdefault() for casting something to an int
18:16:31 <andythenorth> try: except: seems overkill
18:17:09 <andythenorth> calling int('') is not valid for obvious reasons
18:19:46 <Alberth> andythenorth: explicit is better than implicit :)
18:20:05 <andythenorth> I'm not sure this is wise
18:20:23 <andythenorth> but it would catch a user who set num trailers = 2, but forgot to define their capacities :P
18:20:49 <andythenorth> the issue is that I end up calling int('') on vehicles where trailer_capacities = ''
18:21:02 <andythenorth> which can be valid, or a user mistake
18:21:30 <andythenorth> I think I'll use your method
18:23:37 <Alberth> I'd throw an error back to the user rather than guess what he meant
18:24:54 <andythenorth> I think it's best to write a function that cleanly parses config file strings into lists
18:26:42 <Alberth> you might as well use a proper parser such as ply :p
18:29:34 <andythenorth> let the errors pass?
18:29:43 <andythenorth> it's not a nice pattern, but a common one :P
18:29:43 *** Mark is now known as Guest1547
18:29:43 *** [1]Mark is now known as mark
18:30:35 <Eddi|zuHause> hm... so the string(STR, PARAM1, ...) thing doesn't seem to work in translations
18:30:49 <Alberth> "except:" is wrong by definition, you catch ALL errors, including your typos in that try line
18:31:14 <andythenorth> so I could check for txt=''
18:31:21 <andythenorth> but that would only catch one kind of error
18:31:37 <Eddi|zuHause> i find it sometimes hard to find an exception specific enough to catch only what i want to catch
18:31:42 <Alberth> I hate the "absorb all errors" pattern
18:31:43 <andythenorth> I could check the string is in the literals
18:31:56 *** FLHerne has joined #openttd
18:32:00 *** LordAro has joined #openttd
18:33:29 <Eddi|zuHause> clarification for the above: string(STR, PARAM1, ...) doesn't work if STR doesn't exist in the translation file
18:33:55 <Eddi|zuHause> which is probably a correctable error...
18:34:09 <Alberth> andythenorth: that's what '\\d+' does
18:35:06 <Eddi|zuHause> Alberth: tried r'\d+' yet? (raw string literals disable the whole escaping mess)
18:35:34 <andythenorth> regular expressions scare me
18:35:36 <Alberth> Eddi|zuHause: until you try r'\' :p but yeah, I know
18:35:45 <andythenorth> incidentally, why can't I import digits?
18:36:04 <Alberth> from string import digits
18:36:15 <andythenorth> I'm doing it wrong
18:37:56 <Eddi|zuHause> to me it looks like a simple "is already at the end of the string" check is missing
18:38:43 * andythenorth didn't think that parsing a list would be this hard :o
18:45:32 <CIA-1> OpenTTD: translators * r23892 /trunk/src/lang/ (french.txt latvian.txt spanish.txt unfinished/tamil.txt):
18:45:32 <CIA-1> OpenTTD: -Update from WebTranslator v3.0:
18:45:32 <CIA-1> OpenTTD: french - 65 changes by OliTTD
18:45:32 <CIA-1> OpenTTD: latvian - 23 changes by Parastais
18:45:32 <CIA-1> OpenTTD: spanish - 1 changes by Terkhen
18:45:33 <CIA-1> OpenTTD: tamil - 15 changes by aswn
18:45:44 <andythenorth> maybe config parser should be a separate module
18:46:59 *** _11Runner_ has joined #openttd
19:03:04 <Eddi|zuHause> what happened to the idea to make nmlc ignore the value of the magic pink palette entries?
19:04:18 *** FLHerne has joined #openttd
19:07:36 <michi_cc> Today's nightly will be delayed by a few minutes till we sort out a code problem :)
19:09:04 *** HerzogDeXtEr has joined #openttd
19:19:46 <andythenorth> I could add comments
19:21:34 <CIA-1> OpenTTD: michi_cc * r23893 /trunk/src/spriteloader/grf.cpp: -Change: [NewGRF] Extended chunked sprite encoding to work for bigger sprites.
19:21:54 <michi_cc> Thank you for your patience, we now resume or regular nightly service :)
19:25:19 <Alberth> thank you for your nice announcements :)
19:26:46 <Alberth> andythenorth: looks like it needs a default entry, which you use as base
19:27:26 <planetmaker> Eddi|zuHause: iirc it was considered a nice idea. But no-one wrote code so far
19:27:31 <Alberth> or just default properties
19:30:14 <andythenorth> you refer to things that are duplicated so could be defaults?
19:30:18 <andythenorth> or something else
19:31:54 <Alberth> yes, either seperately for each property, or per vehicle (eg 'default = <section-name>')
19:32:02 <__ln__> anyone been to Falkland Islands?
19:32:23 <Alberth> so you just need to list the differences
19:33:01 <Alberth> and aligning the '=' would make it much more readable
19:33:07 <andythenorth> mostly the options should be different, as each vehicle should be unique
19:33:15 <andythenorth> but there are some that are choices from a limited list
19:34:12 <Eddi|zuHause> falkland islands has a mine field. and the penguins are light enough to not trigger the mines, so they actually now have a protected area :)
19:35:10 * andythenorth ponders a way to do 30 * ' '
19:35:17 <andythenorth> to get 30 spaces
19:35:25 <Alberth> andythenorth: ok, not much use thus
19:35:53 <Alberth> andythenorth: ' ' * 30
19:36:03 <Eddi|zuHause> Alberth: doesn't python do that natively?
19:36:09 <Eddi|zuHause> andythenorth: i meant
19:36:33 <andythenorth> I was going to do ' '.join([0:30]) or such
19:36:42 <andythenorth> guess that's silly
19:36:43 <Alberth> oh, even 30 * ' ' works :o
19:37:05 <Alberth> Eddi|zuHause: I thought you had to put the string first
19:38:33 *** Biolunar has joined #openttd
19:42:36 <andythenorth> there is but one small problem
19:42:58 <andythenorth> this is the script on the cms that outputs the config file
19:43:08 <andythenorth> making it output in your format is not hard
19:43:21 <andythenorth> it just means changing the last few lines
19:44:08 <Alberth> ok :) I just grouped similar setting next to each other in blocks
19:44:13 <andythenorth> could write a python script in the repo to do the reformat
19:44:18 <andythenorth> but that's a bit magical
19:44:39 <andythenorth> now I'll have to change the cms script :)
19:44:49 <andythenorth> "look what you made me do"
19:45:03 <Alberth> make a ini-file template :p
19:45:23 <andythenorth> maybe you could?
19:45:29 <andythenorth> you can learn tal that way :P
19:45:34 * andythenorth will do it though
20:02:45 *** mahmoud has joined #openttd
20:10:14 *** Keyboard_Warrior has joined #openttd
20:12:07 <CIA-1> OpenTTD: michi_cc * r23894 /trunk/src/spriteloader/grf.cpp: -Fix (r23893): Use the right variable, not the one that's always going to be 0.
20:12:22 <michi_cc> All the effort, and still a slightly broken nightly :(
20:15:23 *** eQualizer has joined #openttd
20:28:43 *** Phoenix_the_II has joined #openttd
20:33:05 <Hirundo> hmm... how does the 32bpp 'mask' thingie work?
20:33:30 *** tparker has joined #openttd
20:34:28 *** Keyboard_Warrior is now known as theholyduck
20:35:48 <Hirundo> you specify a palette colour, and the pixel is recoloured as if it were that colour but with intensity / alpha adjustments?
20:37:20 <michi_cc> If the pixel has a mask with color index 0, the blitter simply takes the rgb colour. If it is non-zero, the rgb part is converted to a brightness value which is then used to modulate the palette colour from the mask pixel.
20:39:17 <michi_cc> Very simplified pseudo-code: colour = (mask == 0) ? rgb : pal2rgb[mask] * max(r,g,b) / 64
20:51:04 <andythenorth> took longer than it should have due to errors by me :P
20:51:33 *** Markavian` has joined #openttd
21:00:27 <Alberth> looks nice, perhaps add a space behind the '=' ?
21:03:48 *** Progman has joined #openttd
21:04:16 *** theholyduck has joined #openttd
21:13:51 *** peteris has joined #openttd
21:19:04 *** Belugas has joined #openttd
21:19:04 *** ChanServ sets mode: +o Belugas
21:21:28 *** cmircea has joined #openttd
21:23:33 *** Brianetta has joined #openttd
21:37:41 <andythenorth> Alberth: thanks for the config file tip, it's a good format
21:38:07 <andythenorth> also - Eddi|zuHause - thanks, what you did with CETS encouraged me that python is a good route
21:38:23 <andythenorth> + I couldn't have hooked up the makefile alone without the copy-paste from CETS
21:39:23 *** tparker has joined #openttd
21:40:43 <andythenorth> if a file doesn't get processed by python, but is included by python, should the extension be .nml? or .pynml?
21:40:52 <andythenorth> before I rush into a commit ;P
21:41:22 <planetmaker> if it's pure nml... either .nml or .pnml, I'd say
21:41:56 <andythenorth> disable default vehicles
21:42:03 <andythenorth> I'm not even sure they should be separate files
21:42:26 <andythenorth> one of them is 2 loc, 50% of that is comment :P
21:44:21 <xahodo> I've got an assertion failed when I choose advanced settings in the main menu.
21:45:14 <Rubidium> sounds like you patched OpenTTD
21:46:53 <Rubidium> michi_cc: ^ (good luck)
21:47:34 <xahodo> In r23874 advanced settings works as normal.
21:50:54 * andythenorth wonders if .pylng is good for .lng files templated with python?
21:51:17 *** theholyduck has joined #openttd
21:54:12 *** Phoenix_the_II has quit IRC
21:58:14 <andythenorth> do I need to put a GPL notification in every file?
21:58:16 <Alberth> perhaps use a .pytemplate or so, instead of the zillion different extensions?
21:58:46 <andythenorth> you mention that after i commit :P :o :D
21:59:46 <Alberth> haven't you yet learned that 8 minutes is nothing in IRC time? ;)
22:00:19 <andythenorth> I think I'll leave them be for now
22:00:29 <andythenorth> I think the lesson for me here is that I should have used .pt
22:00:36 <andythenorth> which is what all the docs refer to
22:00:50 <andythenorth> and the same as I type for these files every day of the week at work :P
22:01:06 <Alberth> it looks like a translation :)
22:02:04 <Alberth> yeah, not sure why I associate it with translations, perhaps the gnutext thngie uses it
22:03:00 <andythenorth> it's quite widely used, 7 or 8 different meanings it seems
22:14:18 <CIA-1> OpenTTD: michi_cc * r23895 /trunk/src/table/ (misc_settings.ini settings.ini): -Revert (r23888): Broken and not even needed anymore in the container format 2 implementation that was committed.
22:14:22 <CIA-1> OpenTTD: michi_cc * r23896 /trunk/src/spritecache.cpp: -Fix (r23884): Don't free unallocated memory.
22:18:59 <CIA-1> OpenTTD: michi_cc * r23897 /trunk/src/ (6 files in 2 dirs): -Feature: [NewGRF] 32bpp sprites in GRFs.
22:19:04 <CIA-1> OpenTTD: michi_cc * r23898 /trunk/ (8 files in 4 dirs): -Remove: PNG sprite loader.
22:34:47 <Hirundo> r23898 does not appear in the hg repo
22:36:18 *** kkb110_ has joined #openttd
22:39:00 <michi_cc> No idea, it does appear on the git repo though.
22:39:23 <michi_cc> So it can't be the complete convert process that is stuck.
22:59:00 <Elukka> well, what's notable with them?
22:59:11 *** Markavian has joined #openttd
23:00:14 <Elukka> the ryanair one looks like it's been stuck for a while but i'm not sure that's particularly weird...
23:00:59 <Elukka> i imagine there might be issues with snow
23:01:56 <__ln__> and the ezy one doesn't look like it's been stuck for a while?
23:09:51 *** HerzogDeXtEr has joined #openttd
23:22:53 *** Markavian` has joined #openttd
23:28:58 *** Keyboard_Warrior has joined #openttd
23:55:38 *** tensai_cirno has joined #openttd
continue to next day ⏵
