IRC logs for #openttd on OFTC at 2016-12-10
⏴ go to previous day
00:18:25 *** HerzogDeXtEr1 has joined #openttd
00:26:49 *** Biolunar has joined #openttd
00:34:44 *** Supercheese has joined #openttd
01:39:42 *** andythenorth has left #openttd
02:04:37 *** FLHerne_ has joined #openttd
02:45:12 *** FLHerne_ has joined #openttd
04:04:17 *** tokai|noir has joined #openttd
04:04:17 *** ChanServ sets mode: +v tokai|noir
04:05:48 *** Fatmice has joined #openttd
05:32:31 *** Supercheese has joined #openttd
07:46:23 *** orudge` has joined #openttd
07:46:24 *** ChanServ sets mode: +o orudge`
08:32:10 *** ToBeFree has joined #openttd
08:40:26 *** sla_ro|master has joined #openttd
08:43:15 *** CompuDesktop is now known as Compu
08:53:14 *** Progman has joined #openttd
09:03:15 *** BluesInTheNet has joined #openttd
09:49:25 <Rubidium> Eddi|zuHause: the same reason Americans park on the drive way and drive on the park way?
09:55:35 *** ChanServ sets mode: +v tokai
10:10:04 *** andythenorth_ has joined #openttd
10:34:04 *** Biolunar has joined #openttd
11:05:49 *** gelignite has joined #openttd
11:13:03 *** mescalito has joined #openttd
11:23:02 *** andythenorth has joined #openttd
11:36:38 *** MonkeyDrone has joined #openttd
11:37:46 *** Alberth has joined #openttd
11:37:46 *** ChanServ sets mode: +o Alberth
12:14:11 *** chomwitt has joined #openttd
12:18:10 *** Wormnest has joined #openttd
13:03:46 *** FLHerne has joined #openttd
13:17:50 <Rubidium> that's to balance with the noisiness of your children ;)
13:18:05 <andythenorth> there are 4 kids here right now
13:18:11 <andythenorth> they are playing mario
13:18:59 <andythenorth> they have friends :P
13:25:08 <andythenorth> can I crash ottd?
13:31:17 <andythenorth> so eh, how many strings can I put on the text stack?
13:31:38 <Eddi|zuHause> uhm... more than 4
13:31:54 <andythenorth> nml has them in 256-259
13:32:01 <andythenorth> which I count as 4
13:32:10 <Eddi|zuHause> you can put 2 strings per register
13:32:14 <Eddi|zuHause> since they are word sized
13:32:48 <andythenorth> and also that’s why my code doesn’t work right now
13:33:11 <andythenorth> so what do I do to stuff them?
13:33:43 <andythenorth> switch (FEAT_INDUSTRIES, SELF, ${industry.id}_extra_text_non_debug_2, [STORE_TEMP(string(STR_TEST_2) | STORE_TEMP(string(STR_TEST_2) << 16, 256)]) {
13:34:19 <Eddi|zuHause> something along the lines of STORE_TEMP(STR_2 | (STR_1 << 16))
13:34:36 <andythenorth> yes the one I posted had clown shoes on
13:34:47 <Eddi|zuHause> err switch 1 and 2
13:35:10 <Eddi|zuHause> i have some parameter magic buried in CETS somewhere
13:36:07 *** frosch123 has joined #openttd
13:36:22 <Eddi|zuHause> if you can decipher that...
13:55:59 *** tokai|noir has joined #openttd
13:55:59 *** ChanServ sets mode: +v tokai|noir
13:59:40 <andythenorth> I should pull NRT updates eh?
14:00:03 <frosch123> next is patching nml to add roadvehicle property 05
14:07:21 <Wolf01> About the editor, we will let designers lay every know roadtype then?
14:09:56 <frosch123> how does it work for industries?
14:10:05 <frosch123> can you plant oilrigs in 1920?
14:13:16 *** sla_ro|master has joined #openttd
14:13:19 <Wolf01> At least it is in the list
14:25:16 <andythenorth> I am putting strings on the stack
14:25:23 <andythenorth> and those strings need parameters from the stack
14:25:34 <andythenorth> this is getting hard to parse in my head :)
14:26:37 <DorpsGek> Commit by frosch :: r27687 /trunk/src (rail.h rail_cmd.cpp) (2016-12-10 14:26:29 +0100 )
14:26:38 <DorpsGek> -Fix: RailtypeInfo::alternate_labels leaked when reloading NewGRF. (adf88)
14:28:53 <DorpsGek> Commit by frosch :: r27688 trunk/src/rail_cmd.cpp (2016-12-10 14:28:47 +0100 )
14:28:54 <DorpsGek> -Fix-ish: Initialise new railtypes with the original railtype data, though strictly NewGRF could only have modified 'label' and 'alternate_labels' at the time AllocateRailType is called, which are reset anyway.
14:31:17 <Wolf01> Let's see how badly I can crash OTTD this time
14:32:00 <Wolf01> Mmmh, the engine pool doesn't like the editor
14:32:49 <Wolf01> Not so bad, I just make another function without the company shit
14:35:08 *** sim-al2 has joined #openttd
14:39:47 <Wolf01> Ok, I got a geyed out dropdown and it opens the industries window when released :P
14:40:56 <Wolf01> For the greyed part was easy, the problem now is why the items open strange windows
14:44:06 <Wolf01> Oh, shared clicked procs list, bad
15:11:31 <andythenorth> sometimes I wonder
15:11:43 <andythenorth> is there a better way than ‘start coding at the beginning’
15:12:32 <frosch123> people argue whether to do top-down or bottom-up or alternating between them :)
15:15:25 <andythenorth> I am always curious how to start with the design, without writing some code
15:17:25 <Alberth> think big blocks of functionality, and how to exchange data
15:17:54 <Alberth> or rather what kind of data each has, and what other blocks need access in what way to it
15:18:02 <andythenorth> that is predicated on knowing what the functionality is :)
15:18:26 <andythenorth> there’s always this assumption that the design can be known a priori
15:18:48 <Alberth> yep, an idea of suggested functionality for each block helps a lot :)
15:19:13 *** FLHerne has joined #openttd
15:19:21 <Alberth> I think nobody knows exactly how one reaches a design
15:19:32 <Alberth> so they abstract all the details away :)
15:20:08 <Alberth> I don't even know how I do it, let alone how you should do it in general
15:20:42 <Alberth> likely each person does things a little different
15:21:13 <Alberth> not a problem, as there are loads of good solutions, it doesn't matter much which good one you precisely pick
15:21:37 <Alberth> as long as you avoid bad ones, for your situation
15:23:36 <Alberth> in my experience, the problem with a design without coding it at least roughly, is that it's too easy to forget some crucial piece of data
15:24:04 <Alberth> which sometimes completely invalidates your design
15:25:13 <andythenorth> common findings I have from writing spike code...
15:25:24 <andythenorth> (1) not possible within spec (mostly a newgrf thing, not other places)
15:25:40 <andythenorth> (3) there’s a simpler way and most of the idea can be deleted
15:25:49 <andythenorth> (4) ‘this will take much longer than hoped’
15:26:07 <Alberth> I also always start at the wrong end of the problem :p
15:27:59 <Alberth> you can avoid writing code by doing much high level thinking what you need everywhere, but it seems more complicated than just writing rough pseudo code
15:28:12 <Alberth> or experimental code in python
15:28:54 <Alberth> start at the core, or the unknown parts, if that fails, no point in going further
15:34:04 <DorpsGek> Commit by frosch :: r27689 /trunk/src (3 files) (2016-12-10 15:33:58 +0100 )
15:34:05 <DorpsGek> -Fix: MSVC warnings.
15:35:24 *** FLHerne has joined #openttd
15:44:35 *** FLHerne has joined #openttd
15:48:24 <Wolf01> I really need a VCS which allows me to work backwards
15:48:52 <frosch123> sounds good, i would just stop at the star then :)
15:48:54 <Wolf01> I pushed the changes for the scenario editor
15:49:30 <Wolf01> It still open the standard road toolbar for every item on the dropdown
15:50:10 <Wolf01> But when you buid it it is the right type
15:51:00 <Wolf01> I need to change the widget definitions for the editor, because those weren't shared
15:52:35 <Wolf01> (which is strange because it is the same toolbar)
15:57:38 <Wolf01> Could be that RoadTypeInfo is not correctly set up in editor?
15:58:12 <Wolf01> The SetupRoadToolbar() is executed the same way of the normal game
16:03:59 <frosch123> Wolf01: are you on windows?
16:04:05 <frosch123> can you still build road and rail?
16:04:15 <frosch123> the win compile farm fails with the regression
16:04:19 <Wolf01> Seem so, I just built it
16:04:22 <frosch123> i already emrged the change to rnt
16:06:06 <andythenorth> how do I use ‘CARGO_SHORT'
16:06:17 <frosch123> you give two parameters
16:06:21 <frosch123> amount and cargo type
16:06:26 <frosch123> or cargo type and amount
16:06:37 <frosch123> i never know the order
16:12:26 <andythenorth> ach slow compile is slow
16:12:35 <andythenorth> fast FIRS is now slow FIRS, because I added more economies
16:13:42 <Wolf01> Eh, I look on what I need to work, seems easy, I start working on it, 90%... I need to change something else, save changes, revert changes, change the other thing, compile and test, do something stupid like try to add the saved changes, test, remove added changes... commit
16:14:15 <Wolf01> And usually I miss the last CTRL+Z
16:17:20 *** maciozo has joined #openttd
16:19:38 <Wolf01> The stupid is always in ambush inside my head
16:25:24 <michi_cc> Wolf01: git commit -i allows you to stage changes line by line, best way IMHO to split the complete changes into sensible commits (the various GUI tools all provide an equivalent function).
16:25:32 <andythenorth> is there a very abbreviated cargo format (’t’ instead of ‘tonnes’)?
16:26:05 <michi_cc> Wolf01: Or rather git add -i, we aren't at commiting yet :)
16:26:42 <frosch123> andythenorth: CARGO_TINY
16:26:50 <frosch123> or something like that
16:27:12 <Wolf01> Nah, the problem is that I need to reorder my actions edit->test->commit->more edit->more test->commit instead of edit->test->more edit->test->revert more edit->commit
16:27:16 <andythenorth> docs imply that doesn’t show units :)
16:32:49 <andythenorth> industry window text will just get longer :P
16:39:09 <andythenorth> ach cargo names are capitalised :P
16:39:26 <andythenorth> this time-saving automation idea is one of my worst ever :|
16:41:50 <andythenorth> there’s no formatter to lower-case all chars?
16:42:19 <Rubidium> andythenorth: but that fails for some language where words need to be capitalised
16:42:30 <Rubidium> you could, however, use cases
16:45:07 *** sla_ro|master has joined #openttd
16:46:06 <andythenorth> if this works, it possibly saves a lot of translator effort
16:46:12 <andythenorth> but it saves me nothing :D
16:46:28 <andythenorth> and the formatting is wasteful
16:48:31 <andythenorth> ach I could write my own formatters
16:49:31 <DorpsGek> Commit by frosch :: r27690 trunk/src/rail_cmd.cpp (2016-12-10 16:49:25 +0100 )
16:49:32 <DorpsGek> -Fix: Old compilers do not zero-initialise POD members when explicitly calling implicitly created default constructors.
16:57:24 *** FLHerne has joined #openttd
17:12:32 <goodger> automatic tabulating machines are noted for their speed and efficiency
17:30:15 *** welshdragon has joined #openttd
17:31:01 *** welshdragon has joined #openttd
17:31:26 *** welshdragon has joined #openttd
17:38:14 <andythenorth> ach I could process the lang files
17:38:22 <andythenorth> duplicating all the cargo names, and calling .lower() on them
17:41:07 *** welshdragon has joined #openttd
17:41:41 <andythenorth> nah, there must be a better way
17:47:42 <andythenorth> CARGO_LONG is better
17:47:49 <andythenorth> and one less word needed
17:55:04 *** welshdragon has joined #openttd
18:02:53 *** welshdragon has joined #openttd
18:05:37 *** welshdragon has joined #openttd
18:10:40 *** Progman has joined #openttd
18:30:22 <Wolf01> I need to fix all the cargo amount and costs of train fever, it's everything stupid as fuck... a boeing 757 for 55 passengers, the concorde 24, a ferry ship for 12
18:33:57 *** welshdragon has joined #openttd
18:36:46 *** ChanServ sets mode: +v Alberth
18:36:46 *** ChanServ sets mode: +v orudge`
18:36:46 *** ChanServ sets mode: +v DorpsGek
18:36:46 *** ChanServ sets mode: +v orudge
18:57:37 *** Wormnest has joined #openttd
19:22:08 *** FLHerne_ has joined #openttd
19:41:06 *** andythenorth has joined #openttd
19:45:46 <DorpsGek> Commit by translators :: r27691 trunk/src/lang/italian.txt (2016-12-10 19:45:38 +0100 )
19:45:48 <DorpsGek> italian: 9 changes by lorenzodv
19:46:18 <andythenorth> so I might need a way to deprecate a lot of translation strings in FIRS
19:46:26 <andythenorth> and I’d prefer not to remove them manually :P
19:47:05 <andythenorth> I could (a) write a script to rename them, or delete them from lang files
19:47:14 <andythenorth> (b) something else
19:48:33 <Alberth> find strings in the source?
19:51:02 <andythenorth> STR_EXTRA_ALUMINUM_PLANT will no longer need the ‘2t voor iedere 8.000l’ stuff
19:51:22 <andythenorth> and it needs removing, otherwise it will appear broken in game (duplicated info)
19:51:33 <andythenorth> just delete them from the lang files?
19:51:46 <Alberth> translations are done by Eints
19:52:01 <Alberth> it follows the base language
19:52:57 <Alberth> so no point in doing that manually, although the devzone machine may like you less afterwards :p
19:53:24 <Alberth> but you have a diff in the base language?
19:53:28 <andythenorth> I am confused about what the proposal is :D
19:53:36 <andythenorth> I don’t have the base lang diff yet, no
19:53:43 <Alberth> I am confused what your problem is :)
19:54:39 <andythenorth> so STR_EXTRA_MACHINE_SHOP
19:54:41 <Alberth> I'd suggest concentrate on base language only, for now
19:54:43 <glx> I guess eints auto removes strings
19:54:47 <andythenorth> is currently :{BLACK}Builds and services machinery.{}{}Cargo production:{} 8 crates per 8,000l petroleum fuels delivered{} 8 crates per 8t metal delivered
19:54:58 <andythenorth> it will become: :{BLACK}Builds and services machinery.
19:55:13 <andythenorth> the ‘Cargo production’ information is generated automatically
19:55:29 <glx> translation will be marked outdated
19:55:35 <andythenorth> this means that my choices are (1) delete some translation information (2) ship broken FIRS
19:56:01 <glx> but you can also remove the extra text yourself
19:56:25 *** alask0ud has joined #openttd
19:56:26 <andythenorth> there are 38 languages
19:57:00 <Alberth> ah, you want all languages changed, and not just base language
19:57:15 <Alberth> not sure if that's possible
19:57:25 <glx> then only touch base language and let translators do their work
19:57:34 <andythenorth> then FIRS is broken
19:57:36 <Alberth> I wouldn't know what text the first part is for all but a few languages
19:57:50 <frosch> i added the necessary stuff to give roadvehicles a roadtype to nml and ottd. anyone fancy to extend the testgrf with roadtypes of different compatibility, and also some roadvehicles with different type? :)
19:57:55 <glx> or remove strings from translations
19:57:58 <andythenorth> best thing I think is to delete all those strings, it’s a trivial script
19:58:04 <andythenorth> but that pisses off translators?
19:58:11 <andythenorth> it deletes their work with no way to get it back?
19:58:56 <Alberth> you have to consider what happens if the translation never gets updated
19:59:03 <frosch> andythenorth: if you fix the baselanguage, eints will fix the rest
19:59:21 <frosch> eints removes translations which do not exist in the baselanguage
19:59:30 <Alberth> frosch: he wants to rescue some parts of the old strings for all languages
19:59:36 <andythenorth> yeah, they’ll still exist, I’ve just edited the contents
19:59:39 <Alberth> which is complicated to say the least
19:59:42 <andythenorth> also I will happily do the NRT grf, I just need to finish a bit more of this first :)
19:59:58 <frosch> well, then they will show as outdated
20:00:04 <frosch> will they be much wrong?
20:00:38 <Alberth> :{BLACK}Builds and services machinery.{}{}Cargo production:{} 8 crates per 8,000l petroleum fuels delivered{} 8 crates per 8t metal delivered{}{}Cargo production:{} 8 crates per 8,000l petroleum fuels delivered{} 8 crates per 8t metal delivered
20:00:50 <Alberth> that's what you'd get
20:01:56 <Alberth> what if you make new strings, and replace the old texts with "obsolete?" and not use them in the grf, but keep them in the language file?
20:02:08 <frosch> Alberth: then i would try to cut the translations after the n-th {}
20:02:22 <andythenorth> Alberth: I wondered about new strings
20:02:23 <frosch> translators will still see the old translation in the eints history
20:02:29 <Alberth> well, for some languages that would work
20:02:33 <andythenorth> and the old ones are just unused
20:02:40 <andythenorth> and I blank the contents for base lang
20:02:48 <frosch> if you commit the translation changes before the baselang changes, then the translatiosn will still show up as "outdated" for validation
20:02:55 <frosch> i did just that yesterday in ottd
20:03:15 <Alberth> ^ that advantage of that andy, is I can copy/paste the relevant bit from the outdated version
20:03:32 <Alberth> which beats typing it again
20:04:24 <Alberth> hmm, even without changing the translation
20:05:07 <andythenorth> I think that’s the nicest solution
20:05:30 <Alberth> after a few weeks you could remove translation texts that still have the {}{} things
20:05:52 <Alberth> so they fall back to base language
20:08:48 <andythenorth> ach this won’t work anyway
20:09:17 <andythenorth> I’m up to register 261 already, and I have one more cargo to display
20:16:30 <andythenorth> 5 dwords is apparently not enough for this
20:17:56 <frosch> gamescripts complainted that 20 were not enough :p
20:18:42 *** Supercheese has joined #openttd
20:19:04 <andythenorth> I wonder if I can do it differently, with more duplication, and fewer reused substrings
20:19:43 <andythenorth> should only need 10 items on the stack
20:21:30 *** OsteHovel has joined #openttd
20:23:54 <andythenorth> I shouldn’t be storing the cargo labels like that?
20:25:20 <frosch> did you miscount the bits maybe?
20:25:31 <Alberth> IL) | GOOD << 16 <-- that looks wrong, add parentheses (GOOD << 16) ?
20:25:59 *** FLHerne has joined #openttd
20:26:00 <andythenorth> I took it from nml docs :)
20:26:19 <Alberth> ah, ok, maybe I am wrong :)
20:26:42 <andythenorth> but those have a typo in already
20:28:03 <Alberth> adding parentheses around shift operations wouldn't do harm
20:28:15 <andythenorth> no, but nor does it fix it :)
20:29:27 <andythenorth> it doesn’t only work with default cargos?
20:29:35 * andythenorth assumes not, but it appeared to work for coal
20:29:41 <andythenorth> and not cargos I added
20:29:58 <frosch> there is a GetCargoTranslation in the code
20:30:09 <frosch> it may not work, but at least it is intended to work :)
20:32:11 <andythenorth> coal works, FIRS cargos don’t
20:32:57 <andythenorth> STR_CARGO_UNIT_STONE :{WEIGHT} of stone
20:34:38 <andythenorth> I’ve pushed all this if anyone wants to try :P
20:34:52 <andythenorth> make clean, then make install TEST_INDUSTRY=lime_kiln
20:34:59 <andythenorth> is a quick compile
20:44:53 <frosch> what would be the expeted output?
20:49:29 <frosch> should the 53000 be a 8 ?
21:25:42 <frosch> it's the popular "pass parameters via global variables, and then break everything via recursions" issue
21:25:54 <andythenorth> the 53000 should be an 8 :)
21:28:22 <frosch> but i will look again tomorrow for a proper fix
21:30:05 <andythenorth> I have other things I could do right now
21:30:09 <andythenorth> like….NRT vehicles
21:30:41 <frosch> make sure to both pull nrt-ottd and nrt-nml :)
21:31:23 <frosch> roadvehicles now have a road_type / tram_type property to give them a road/tram label
21:31:49 <frosch> it is loaded by ottd, but has not yet any gameplay effect
21:34:22 *** welshdragon has joined #openttd
21:38:57 *** umgeher_ is now known as umgeher
21:55:28 *** goodger has joined #openttd
22:22:59 <andythenorth> I need 13 bytes on the text stack
22:23:05 <andythenorth> I’m allowed 12 :(
22:25:38 <andythenorth> actually I need 14
22:28:07 *** TheMask96- has joined #openttd
22:33:05 *** TheMask97- has joined #openttd
22:35:06 <andythenorth> but I use 14 words
22:35:16 <andythenorth> maybe I could stuff them more efficiently, dunno
22:41:19 <andythenorth> some of the values are in the range 1-8
22:41:25 <andythenorth> hardly needs a dword for that
22:52:35 *** sim-al2 has joined #openttd
22:54:04 *** andythenorth has left #openttd
23:01:04 *** Defaultti has joined #openttd
23:20:58 *** welshdragon has joined #openttd
23:24:08 *** FLHerne has joined #openttd
23:28:51 *** welshdragon has joined #openttd
23:36:34 *** Eddi|zuHause has joined #openttd
continue to next day ⏵
