IRC logs for #openttd on OFTC at 2012-11-17
⏴ go to previous day
00:07:43 *** George|2 has joined #openttd
00:07:43 *** George is now known as Guest5870
00:07:44 *** George|2 is now known as George
01:31:20 *** BadBrett has joined #openttd
01:33:10 *** HerzogDeXtEr1 has joined #openttd
02:05:58 *** valhallasw has joined #openttd
02:14:40 *** Djohaal has joined #openttd
02:55:47 *** valhallasw has joined #openttd
02:56:58 *** Zeknurn has joined #openttd
03:23:17 *** Biolunar_ has joined #openttd
06:39:10 *** andythenorth has joined #openttd
06:40:35 *** andythenorth has joined #openttd
07:03:38 <andythenorth> don't remove cargos from a newgrf, then reload it on a running game
07:03:43 <andythenorth> with the cargo payment window open :P
07:07:19 <Supercheese> Wait, how do you make OTTD reload grfs from disk while running? Doesn't it lock the .grf files so you can't overwrite them?
07:12:31 <andythenorth> my shell ignores that
07:13:14 <andythenorth> mv foo.grf openttd/data/foo.grf
07:13:30 <Supercheese> izzat command line?
07:13:33 <andythenorth> doing this is a source of openttd crashes though
07:13:56 <andythenorth> equivalent to command line I guess
07:19:58 <Ammler> don't you get a crash.log?
07:22:34 *** oskari89 has joined #openttd
07:31:12 * andythenorth is closing FIRS bugs
07:40:46 <andythenorth> can't figure out how I broke the *old* old spritelayouts though
07:41:07 <andythenorth> there are pnml versions that were already deprecated, before I converted to python
07:41:13 <andythenorth> which are now broken :P
07:49:18 *** sla_ro|master has joined #openttd
07:50:05 <andythenorth> might just rm it
07:50:18 <andythenorth> it's a FIRS parameter
07:55:20 <andythenorth> ok, so nobody else knows either, and nobody's complaining about me removing it
07:56:54 <Terkhen> andythenorth: it's for scenarios
07:57:27 <Terkhen> if someone creates a scenario with FIRS, carefully placing industries and so on, he does not want new industries to appear
07:57:29 <andythenorth> strikes me as (1) buggy
07:57:34 <andythenorth> (2) out of scope for a newgrf
07:57:42 <andythenorth> this should be done by the GS
07:58:10 <andythenorth> or an advanced setting in ttd
07:58:17 <V453000> I cant help myself but I still see GS as Gay Ass
07:58:30 <V453000> (I read it that way, the feature is nice)
07:58:32 <Terkhen> I don't think that GS can do that
07:58:41 <andythenorth> that's because GS is unfinished
07:58:58 <andythenorth> I think more and more that the stock answer 'do it in newgrf' has significant problems
07:59:03 <andythenorth> in certain specific cases
07:59:11 <Terkhen> andythenorth: NewGRFs may make assumptions about how industries are founded
07:59:19 <Terkhen> which limits what GS can or should do
07:59:24 <Terkhen> it's our usual scope problem, yes
07:59:25 <andythenorth> that's one of the problems
07:59:45 <andythenorth> newgrfs being highly unpredictable makes it very hard to define a useful GS API
07:59:56 <andythenorth> yet gameplay should be in the domain of the GS
08:00:05 <Terkhen> NewGRFs do a lot of stuff that, if we were to remake the specs today, we would not want them to do
08:00:28 <Terkhen> it's a classic of our discussions over here
08:00:50 <andythenorth> so remove the parameter?
08:00:54 <andythenorth> or can someone fix it?
08:01:14 *** Eddi|zuHause has joined #openttd
08:01:35 <andythenorth> [because I don't know how it should behave]
08:01:50 <Terkhen> andythenorth: I would remove it... given that FIRS development process obsoletes scenarios very quickly it is not a great idea to carefully make a scenario with a soon to become old FIRS version
08:02:08 <andythenorth> planetmaker? +/- 1
08:04:16 <planetmaker> I'm undecided. +-0 :-P
08:04:47 <andythenorth> that doesn't help the bug count :)
08:05:36 <andythenorth> shall I just rename it 'prevent industries opening' ?
08:05:49 <andythenorth> and mark the behaviour as correct o_O
08:05:52 <planetmaker> that's the purpose, I think?
08:07:08 <andythenorth> should it prevent industries during map gen though?
08:07:44 <Supercheese> For a scenario with industries already placed, there should be no further industry placement, no?
08:08:14 <Supercheese> well, that's not 'map gen'
08:08:53 <andythenorth> so it should allow map gen to build industries?
08:09:31 *** Devroush has joined #openttd
08:09:34 <Supercheese> If the parameter is only intended for use with scenarios, then that behavior is acceptable
08:09:40 <Ammler> yes, you have another switch to control that
08:11:26 <planetmaker> yes. mapgen not influenced by the switch
08:11:48 * Supercheese had read that as "not allow".
08:12:36 <andythenorth> it checks game_mode == GAMEMODE_GAME
08:13:28 <andythenorth> or maybe another nml switch is needed
08:14:55 <planetmaker> mostly that switch is needed to ignore placement restrictions for player funding
08:15:11 <andythenorth> I'll add another
08:15:14 *** valhallasw has joined #openttd
08:15:26 <andythenorth> although I am +0.7 on removing this parameter altogether :P
08:28:18 <andythenorth> haven't run the game for very long to see if it builds :P
08:28:44 <andythenorth> I can't get oil rigs to build in FIRS
08:28:51 <andythenorth> with recent revs
08:33:56 *** mahmoud has joined #openttd
08:55:31 *** mattfury has joined #openttd
08:56:05 <mattfury> got any gfx to work on?
08:56:13 <mattfury> photoshopping models would be fun
08:56:25 <mattfury> can use simcity gfx model creator :p
08:56:51 <mattfury> simcity 2000 that is, what do you guys use?
08:58:02 <Supercheese> I like taking renders in Sketchup and further editing/polishing them
08:58:08 <Supercheese> but to each their own
08:58:56 *** Alberth has joined #openttd
08:58:56 *** ChanServ sets mode: +o Alberth
09:02:10 <V453000> pixel-art is teh way mattfury :P
09:03:40 <V453000> otoh Pikka did some serious magix with renders and pixels in aviators :)
09:05:27 <Supercheese> Zbase is all rendered, IINM
09:06:28 <mattfury> do you torrent at all?
09:06:44 <mattfury> i dropped my 2tb hdd and its not workingg :O
09:07:07 <mattfury> ]i had the usb pluggged in the whole time
09:07:19 <V453000> zbase indeed is rendered, but the visual appearance¨ also shows that
09:08:46 <Supercheese> well, I'm off to bed
09:11:50 *** Hyronymus has joined #openttd
09:19:50 <mattfury> V453000, what editor??
09:20:01 <V453000> I personally use photoshop
09:20:16 <mattfury> have you put in tunnel maglevs yet?
09:22:00 *** RavingManiac has joined #openttd
09:22:03 <V453000> I dont understand what do you mean by that, mind rephrasing? :)
09:23:13 <planetmaker> V453000, there's a railtype like vacuum tube (rails). Maybe that?
09:23:39 <V453000> is that what tunnel maglev means?
09:24:11 <planetmaker> How can I know? :-)
09:29:00 <planetmaker> it's shocking how different the game looks when switching between opengfx and zbase :-)
09:29:30 <V453000> im sorry I will not comment zbase any further :)
09:29:45 <andythenorth> zbase is pretty awesome
09:30:11 <Rubidium> only andythenorth's graphics look somewhat out-of-place in there
09:31:25 <planetmaker> :-) they work acceptably until zi2. At zi4 it shows quite different
09:32:06 <planetmaker> but anyway, I heart rumors about 32bpp firs :D
09:36:00 <mattfury> that hurts my eyes thinking about it
09:36:30 <Eddi|zuHause> you confused that with ppb :p
09:43:39 <mattfury> i know an easier way of making money than programming lols
09:43:52 <mattfury> survey sites that pay you
09:47:32 <mattfury> what does this mean?
09:47:53 <mattfury> i dont know where it came from o.o'
09:47:59 <andythenorth> how much can I earn?
09:48:06 <andythenorth> for how many hours work?
09:48:40 <planetmaker> andythenorth, of course 10% of the amount found on an abandoned nigerian bank account of the deceased cousing or so. Several million GPB for sure
09:49:12 <planetmaker> you "just" need to pay a small transaction fee in advance. Deal?
09:50:08 <andythenorth> planetmaker: it might be a deal, but can you send me a guide explaining how it works?
09:52:08 <planetmaker> andythenorth, sure. I send you some forged PDF which lists a fake account somewhere. And another account where you need to transfer, say 3000€ as transaction fee so that I can start processing the transfer of the BIG money to you
09:52:21 <planetmaker> of course you have to transfer the fee in advance and w/o any security given
09:52:48 <planetmaker> but you can totally trust me ;-)
09:53:08 <andythenorth> how about I pay you in some diamonds that I have from an ancient relative? But you'll need to come to UK to get them, and then take them to somewhere like Amsterdam to get them traded
09:55:22 <planetmaker> sounds also like a deal :-P
09:56:58 <mattfury> wow /b/ just made my day
09:58:55 <mattfury> in street language = fock the police, or file transfer protocol lols
10:26:00 <andythenorth> FIRS could compile faster :(
10:30:17 <Terkhen> rewrite key parts of nml in C
10:36:12 *** Progman has joined #openttd
10:39:08 <Eddi|zuHause> i'd like to have CETS compile without needing 16GB of ram...
10:39:39 <planetmaker> CETS compiles for me with much less. Though I didn't meter the ram usage
10:40:20 <Eddi|zuHause> yes, it's something around 1GB
10:40:29 <Eddi|zuHause> but it's waaaay too much
10:41:06 <andythenorth> is resolving all the identifiers still one of the issues?
10:41:18 * andythenorth wonders if they could be scoped, allowing partial compiles
10:41:28 <andythenorth> [scoped something other than global]
10:42:59 <Eddi|zuHause> the last time i measured CETS, the bottleneck was in ply building the parse tree
10:43:41 <Eddi|zuHause> (everything else i could resolve circumventing NML)
11:02:17 *** Progman has joined #openttd
11:03:38 <Alberth> of course you can have partial compiles, just do some extra administration.
11:03:38 <Alberth> that is, see a set of sprites as a blob of bytes, with some 'holes' in it where you need to fill in the various IDs. One source of IDs are of its own definitions, another source are the IDs defined by other blobs.
11:05:56 <Alberth> give each ID a number such that they do not conflict, fill in the holes, write it all out to a single file, and you're done
11:09:50 <andythenorth> and to figure out the IDs?
11:14:25 <Alberth> IDs are independent, so it does not look a very complicated problem at first sight, to me
11:14:47 <Alberth> nml also does that afaik
11:16:06 <Alberth> ie just assign a 'random' free ID, and make it available for use again after its last use
11:17:14 <andythenorth> so track used IDs
11:19:32 <andythenorth> I'm out of my depth with this :P
11:20:09 <andythenorth> it's tempting to write a python script that generates 1 grf per industry, then opens the nfo, strips the header stuff, merges it to one file and runs grfcodec
11:20:22 <andythenorth> use multiprocessing library on that
11:20:39 <andythenorth> and try and figure out deps to avoid unchanged industries
11:21:35 <andythenorth> I could dump comments to split on :P
11:22:13 <andythenorth> performance is getting worse as I add more spritelayouts, spritesets etc
11:22:13 *** Biolunar has joined #openttd
11:22:24 <Alberth> the problem is a) lack of a clear understanding which type of IDs need to be passed from one blob to the next (for me, at least), and b) total lack of tooling for building upon, so you're back to square 1, mostly
11:22:25 <andythenorth> for snow, animation, graphics that change by date
11:23:08 <andythenorth> I wonder if I could just track IDs manually
11:23:22 <andythenorth> nfo FIRS did it, even with extensive use of includes
11:23:55 <Alberth> between industries you don't have many IDs that you want to keep, I guess
11:27:47 <andythenorth> afaict, there is nothing in industry code that needs to be a single instance, shared by all industries
11:28:13 <Alberth> cargo labels? or are they not an ID?
11:28:26 <andythenorth> not sure on labels
11:28:41 <andythenorth> afaict, they depend on the cargotable
11:28:49 <Alberth> but even if they are, they are a fixed set for all industries
11:28:58 <andythenorth> they can be templated in wherever they are needed :P
11:29:07 <andythenorth> iirc, it's valid to declare a cargo table multiple times :P
11:30:24 <andythenorth> this is assuming that the identifier resolution is the slow part :P
11:31:00 <Alberth> no, it's about not having to build everything again every time
11:31:42 <Alberth> it assumes that merging the blobs together is fast
11:32:41 <andythenorth> FIRS does use nml identifiers of other industries for things like location checks, but the nfo IDs are also known and could be used
11:33:39 <Alberth> you can perhaps even assign a unique number to each industry, which makes it all quite trivial to do
11:34:51 <andythenorth> they have that already ;)
11:38:46 <andythenorth> nml already has an nfo flag
11:39:03 <andythenorth> combine that with 'this may not be valid for a complete grf' flag :P
11:39:16 <andythenorth> then pass fragments of code to nml
11:41:45 <Alberth> wouldn't nml complain about IDs that you use but not define?
11:41:56 <planetmaker> andythenorth, frankly the optimizations you talk of should not go to the NewGRF but to NML itself
11:42:43 <Alberth> I think you need to make an nml-linker for this
11:43:54 <Alberth> planetmaker: nfo linking would be interesting for all people coding in nfo
11:44:13 <Alberth> but the puzzle is then to modify grfcodec :(
11:45:14 <Alberth> (or build a new nfo encoder, but I don't know how complicated that is)
12:12:40 <Eddi|zuHause> <Alberth> of course you can have partial compiles, just do some extra administration. <-- i once thought about how to do that... all i needed was a way to insert an action C into the nfo output "here end the headers", then i can make a complete GRF for each vehicle, and combine them by removing everything up to this special action C
12:13:37 <Eddi|zuHause> but nobody wanted to implement action C support for nml ;)
12:14:42 <Alberth> I doubt somewhat that the NFO is that simple in structure when generated from nml
12:15:12 <Alberth> and like andy, you have a simple case with fixed IDs, it seems
12:15:26 <Eddi|zuHause> "header" would be stuff like action 14, parameter definitions, cargo translation table, ...
12:15:46 <Eddi|zuHause> yes, the vehicle IDs are known, and each vehicle is fairly isolated
12:16:25 <Eddi|zuHause> this would be a system specially crafted for the use case, not something generic
12:16:40 <Alberth> adding an action C is not so difficult if you don't care about the place
12:17:31 <Alberth> adding hacks for anything 'production' is wrong, imho
12:17:38 <Eddi|zuHause> well, nml should not do too much reorganizing
12:17:43 <planetmaker> well. Still. Isolated. That then rather should be done by NML directly. Instead of then requiring a detour via nfo export
12:18:33 <Alberth> tbh, I never understood why nml did not use grfcodec
12:18:55 <Alberth> ie you have a perfectly working assember, why not use it?
12:20:04 <planetmaker> you didn't have support for various graphics types other than pcx and png. you don't have automatic conversion from palettes. you will need a 2nd programme
12:20:35 <planetmaker> nml can use any graphics type supported by PIL. grfcodec only png and pcx
12:20:48 <Eddi|zuHause> there's really no reason why grfcodec couldn't do palette conversion
12:21:23 <Alberth> yeah, extending grfcodec seems much more useful to me than doing that in nml
12:21:56 <andythenorth> how do I get FIRS makefile to use grfcodec?
12:22:36 <Eddi|zuHause> why would you do that?
12:23:04 <Eddi|zuHause> i kinda hacked that into CETS makefile, it wasn't officially supported
12:24:30 <andythenorth> want to compare times
12:24:52 <andythenorth> nml got faster, iirc the grfcodec advantage diminished
12:28:48 *** Devroush has joined #openttd
12:29:04 <andythenorth> think it was the spritecache
12:31:06 <andythenorth> FIRS is 40s without grfcodec
12:31:13 <andythenorth> 10s of that is now python pre-processing :(
12:32:56 <Eddi|zuHause> preprocessing of CETS: real 0m4.328s user 0m1.627s sys 0m0.445s
12:34:21 <V453000> nuts takes like 1 minute too after I updated nml yesterday :) was like 10-15 minutes before
12:34:56 <Ammler> the new gui to chose value is menu instead a list with arrows and textfiled anymore, is it possible to change that?
12:36:27 <Eddi|zuHause> oh, when you click on a setting you mean
12:36:43 <Eddi|zuHause> i don't think you can change that
12:37:21 <andythenorth> 15 mins to compile NUTS? :o
12:38:08 <Ammler> Eddi|zuHause: newgrf parameter settings
12:38:23 <V453000> yeah, with the old nml it took about 15 mins
12:38:27 <andythenorth> not rendering any industries in FIRS cuts python step to 0.5s
12:39:27 <V453000> illustrative image :p
12:39:39 <Eddi|zuHause> Ammler: the action 14 defines what is used
12:39:51 <Eddi|zuHause> (system kinda missed some keystrokes)
12:39:57 <Ammler> V453000: I mostly miss the "doubleclick" textfield
12:40:07 <Ammler> Eddi|zuHause: that was my question
12:40:09 <V453000> oh wow that isnt there
12:40:18 <V453000> yeah that is quite bad
12:41:19 <NGC3982> Running a ninety minute workout while a bit hangover. Yes or no?
12:42:29 <Eddi|zuHause> NGC3982: never been in that situation :p
12:43:15 <Ammler> V453000: it is IMO worth a bugreport
12:43:15 <Eddi|zuHause> Ammler: from what to where? which context? which version?
12:43:25 <NGC3982> I feel that I need it, but only if it's doing any good :-\.
12:43:44 <Eddi|zuHause> Ammler: words. need. to. describe.
12:43:45 <Ammler> Eddi|zuHause: check the screen from V453000
12:44:20 <Eddi|zuHause> i've seen the image. i still have no clue what you're trying to say
12:44:22 <Ammler> the menu looks nice if you don't have a big list of values
12:44:51 <Ammler> but it should not replace the doubleclick textfield
12:45:40 <Ammler> I wonder, why that conflict anyway
12:46:54 <V453000> ... you cant doubleclick a value to edit it
12:47:31 <Eddi|zuHause> oh, that... yeah, that should still be possible... if it's not, make a bug report
12:48:41 <Eddi|zuHause> hm... python now at 48% mem (~2GB)
12:48:48 <Zuu> I mostly missed a text filter for the BaseCostMod parameter list. To actually find the parameter to change.
12:49:01 <V453000> I thought exactly the same Zuu
12:50:10 <planetmaker> I don't quite see how you need to enter the text manually there, Ammler
12:50:58 <Ammler> planetmaker: double click on the value in 1.2
12:51:22 <Ammler> or don't you see why I miss it for basecosts.grf?
12:51:48 <planetmaker> not exactly. As a value of X does not exactly explain itself to a user
12:51:57 <planetmaker> while the named strings are self-explanatory
12:52:09 <Ammler> well, the question is does it conflict
12:52:29 <Ammler> regular users know the values
12:52:30 <planetmaker> in 1.2 you don't have a quick way to get a setting not adjacent to the current. Now you just pick from a list where you have all
12:53:25 <planetmaker> There's certainly not much more than two dozen people who know them all by heart :-)
12:54:22 <Ammler> as said, the question is, does the doubleclick confict with something else that it got removed?
12:54:44 <Ammler> it's not like you have to use it if not there
12:55:13 <Ammler> hmm, how is done with numbers?
12:55:43 <Alberth> that should work afaik
12:56:15 <Alberth> with arbitrary text strings, it makes no sense to type them manually, imho
12:56:22 <Warod> TrueBrain: Btw, have you considered some ready to use CDN for ottd distribution?
12:56:26 <andythenorth> I wonder if opening the same template over and over again is bad for IO on FIRS
12:56:27 *** Eddi|zuHause has joined #openttd
12:56:58 <Ammler> yes, numbers still allow the doubleclick
12:57:14 <Alberth> andythenorth: unlikely, the file system cache will keep the file in memory
12:57:32 <Ammler> to this quite much looks like a bug as you added the menu and forgot the old feature
12:57:37 <TrueBrain> Warod: I have, but I couldn't find any that complies to our wishes, which are a bit exentric I have to admit :)
12:58:25 <V453000> andythenorth: what is the maximum production / town population for a recycling depot?
12:58:31 <Warod> TrueBrain: Heh. Do you know anything about CloudFlare?
12:58:44 <andythenorth> V453000: not sure it has any cap
12:58:55 <andythenorth> there is a spec limit, probably 65k tons
12:59:52 <Warod> TrueBrain: They seem to have quite sensible CDN network.. with kinda nice ideas of how they've build their infra. I kind of like that and would like to test what they actually offer for free with "We never charge for bandwidth"...
13:00:25 <Zuu> Ah, so for NewGRFs you need to double click to get the manual input while AI + GS only need single click. That is why I was confused about the talk about double clicks...
13:01:13 <Ammler> hmm, if you readd the doubleclick textfield, you could also add one on the main newgrf gui with the parameters
13:02:19 <Ammler> basically would allow grf parameter editing like it was on the very first newgrf gui
13:02:59 <Alberth> that only had numbers :p
13:03:42 <Alberth> and for numbers it works, so we do have the old functionality covered :)
13:03:45 <Ammler> how do you edit parameters on a grf without a14?
13:04:00 <Zuu> Btw, double click on an AI/GS in their window opens the select AI/GS window. (if anyone would aim to make them more consistent)
13:04:19 <TrueBrain> Warod: it lacks a lot of details on how it works ...
13:04:23 <TrueBrain> guess they will tell after signup :P
13:04:30 <TrueBrain> but you are going to test it? Lemme know how it works out :D
13:05:14 <Alberth> Ammler: don't know, I mostly don't edit grf parameters
13:05:24 <Zuu> Ammler: You get a question first for how many parameters to set, then you set the parameters blindly as eg. param 1, 2, ...
13:05:48 <Ammler> Alberth: I guess then it isn't a bug, rather a misisng feature
13:06:11 <Zuu> You could try "City Stations" NewGRF which afaik should be on banans.
13:06:15 <Warod> TrueBrain: Well.. it seems it's targeted to be primarly a "ordinary website" accelerator. Not for actually serving mainly downloads or "image rich" content. :P
13:06:26 <Warod> "Using an account primarily as an online storage space, including the storage or caching of a disproportionate percentage of pictures, movies, audio files, or other non-HTML content, is prohibited. "
13:06:45 *** Eddi|zuHause has joined #openttd
13:06:53 <planetmaker> hehe, quite. We need mirror only exclusively for binary
13:06:58 <TrueBrain> I should finish my mirror software, and find more mirrors :)
13:07:10 <TrueBrain> should be easy with the new stuff ...
13:07:20 <Alberth> Ammler: you could also wonder why you don't have numbers, as your parameters look very number-ish to me
13:07:21 <TrueBrain> I pay 2B ISK for it
13:07:29 <Warod> TrueBrain: I'm going to set up a more generic way of mirroring stuff.. after that openttd can have .fi mirror
13:07:44 <TrueBrain> awesome; yes please ;)
13:07:51 <planetmaker> Alberth, yes. But then you need to explain what the numbers mean. And that the default of 8 means "unchanged" is not exactly clear
13:08:12 <Alberth> planetmaker: yeah, we need fractions :)
13:08:20 <planetmaker> thus providing strings which explain / replace the numbers seems the way to go
13:08:41 <Alberth> what about 100 == unchanged?
13:08:53 <Warod> TrueBrain: There are also some other opensource projects that look for mirroring capacity.. so I think it should be done properly and in a generic enough way to provide mirroring not only to ottd, but other projects as well.
13:09:12 <TrueBrain> I fully understand :D
13:09:21 <planetmaker> Alberth, that's the NewGRF specs. Not the individual NewGRF's decision
13:10:07 <Warod> TrueBrain: The point being, those projects are "too small" to get university supported mirroring etc.
13:10:15 <planetmaker> of course one could apply an offset and scaling conversion...
13:10:26 <Yexo> planetmaker: in theory the newgrf could accept 100 as unchanged and change it to 8 before passing it on to OpenTTD
13:10:26 <Alberth> you cannot translate a parameter value from 100 to 8 in the newgrf? :o
13:10:29 <planetmaker> but even that needs explanation
13:10:43 <planetmaker> Alberth, of course :-)
13:10:48 <Ammler> Alberth: yeah, using numbers again would be a workaround
13:10:53 *** Eddi|zuHause has joined #openttd
13:11:01 <Yexo> on the other hand: how would *2 work? 200? what if the user enters 150?
13:11:01 <Ammler> but I don't see why it can't work like on 1.2
13:11:06 <TrueBrain> Warod: open source is open source ;)
13:11:14 <planetmaker> but it only changes the problem to explain those numbers instead of the current ones
13:11:18 <Ammler> there you had text and numbers either
13:11:26 <Alberth> Yexo: 100 == 100% which everybody understands
13:11:38 <planetmaker> Alberth, how is then 1/512 explained in %?
13:11:48 <planetmaker> you can only change factors of 2
13:11:54 <Yexo> Alberth: yes, but if you allow free text there will be users entering 150% and wondering why it becomes either 100% or 200% since anything in between isn't supported
13:11:55 <planetmaker> powers of two ataully
13:12:16 <Yexo> the current fixed options prevent that
13:12:41 <Alberth> Yexo: you'd need checking and error handling of course, just like the other numbers
13:13:00 <Yexo> both of these are hard for a NewGRF
13:13:08 <planetmaker> Alberth, current scheme requires no error handling and checking. A set of (sequential) numbers is allowed, that's it
13:13:34 <planetmaker> and these numbers currently are at the same time the values acceptable by the specs. And each has a (textual) explanation
13:13:40 <Alberth> simple and effective in most cases
13:14:39 <planetmaker> and from a UI perspective I don't see entering the values in a text field as faster than selecting the entry from the drop down. Both is at least two clicks
13:14:46 <Alberth> k, so 100 is not going to fly :)
13:15:10 <planetmaker> well. 10000 might do the skaling trick with low values. But still :-)
13:15:15 <Alberth> most users can quicker select a value from the drop-down, I'd say
13:16:40 <planetmaker> which is what ottd 1.3 does :-)
13:17:04 <Zuu> Isn't the query string entry just a way for knowledgable people to enter raw values.
13:17:55 <Eddi|zuHause> oooh... it's done...
13:18:01 <Zuu> Typically a parameter with labels would use integer values from Min to Max and the raw values wouldn't make much senes them self. But someone could use the GUI first and then take a look at the raw value to see how it mapps.
13:18:31 <Eddi|zuHause> it only took half an hour of swap hell
13:18:37 <planetmaker> but then... where is that mapping needed (except maybe in server config)
13:19:16 <Zuu> planetmaker: someone might want to set all aparams to some value and prefer to enter it via keyboard than looking through that long list of labels.
13:20:20 <planetmaker> Zuu, but he still has to click. Either way
13:20:24 <Zuu> The keyboard way needs less precision work with the mouse which some may have problem with.
13:20:53 <planetmaker> text / number entry needs at least a double click. drop-down selection two single clicks
13:21:03 <Zuu> I would guess 99.9% would use the GUI way to pick one of the labels.
13:21:16 <planetmaker> that's what I believe, too.
13:21:37 <planetmaker> I'm not opposed to number entry. But for that reason I don't consider it a high priority on my anyway too long list of things
13:22:10 <Zuu> Oh, the drop-down doesn't actually require you to hold the mouse button down as one would expect in OpenTTD.
13:22:47 <Eddi|zuHause> Zuu: none of the dropdowns do. only the main toolbar has that behaviour
13:22:59 <planetmaker> he. strange, actually, Eddi|zuHause :-)
13:23:28 <Zuu> I was extrapolating from the main toolbar behaviour and though all drop-downs in OpenTTD had that behaviour.
13:23:36 <Eddi|zuHause> that's because they are buttons with extra behaviour, not actual dropdowns
13:23:53 <tycoondemon> old scool ttd dos rocks!!
13:24:18 <Eddi|zuHause> "old school" would be TT, not TTD
13:24:28 <planetmaker> you might want to look for #tycoon channel, tycoondemon ;-)
13:24:37 <tycoondemon> indeed, I thought of that whilest typing
13:24:44 <tycoondemon> but I only played ttd so
13:25:01 <Eddi|zuHause> i only really played TT
13:25:20 <Zuu> I have been thinking about adding a button of half the main toolbar button height with a arrow downwards at the places where you can hold and open a menu. A click on that button could open a sticy menu that doesn't need you to hold down the mouse button.
13:25:44 <tycoondemon> I do still have a pc with ttd installed and windows 98
13:26:06 <planetmaker> Zuu, that might make sense. And also at the same time indicate that there's more to the button
13:26:21 <planetmaker> especially needed e.g. for the rail and road buttons
13:26:23 <Eddi|zuHause> i probably have a laptop with TT installed, haven't tried it in years
13:26:32 <Eddi|zuHause> and a 40MB hard disk
13:26:35 <Zuu> Yeah, I think it is useful to show that there are more on most of the buttons.
13:27:02 <planetmaker> Zuu, otoh it could simply be changed by modifying the baseset graphics :-)
13:27:37 <Eddi|zuHause> hm... random thought: an action14-y way to randomize the parameters each time the grf is added to the list/preset is selected/game is started
13:27:37 <Zuu> The visual part possible, but not the actual click behaviour.
13:27:55 <tycoondemon> those where the times when defragmentation had real effects
13:28:09 <tycoondemon> sorry for off topic ill be quiet
13:28:36 <planetmaker> Eddi|zuHause, yes... but it need an entry per setting whether it can be randomized or not
13:28:59 <planetmaker> randomizing parameters dealing with support of other newgrfs would seem funky ;-)
13:29:00 <Eddi|zuHause> yes. per-parameter randomization set by a flag in the a14
13:29:18 <Eddi|zuHause> defining how and in what range the parameter is randomzied
13:30:02 <planetmaker> Eddi|zuHause, but... the newgrf can make it such. Assign one value of the parameter to mean "random" and the others their respecitve fixed setting
13:30:26 <planetmaker> I'm not exactly sure how then to distinguish a wanted random from a "I want this parameter setting now"
13:31:21 *** LordAro has joined #openttd
13:31:24 <Eddi|zuHause> planetmaker: but there is no "random action D" or something
13:31:58 * LordAro waves hai to Alberth and the rest
13:32:01 <planetmaker> Eddi|zuHause, that sounds like something which could be amended then :-)
13:34:19 <planetmaker> so you want variable 0x26 which returns the seed? :-)
13:34:44 <planetmaker> or 0xA6 for actionD
13:35:03 <Eddi|zuHause> or something-ish
13:35:30 <planetmaker> that makes sense from my POV
13:35:48 <Eddi|zuHause> the question is whether that is already available during the activation stages
13:46:38 <planetmaker> he... I should first compile :-)
13:49:08 <Eddi|zuHause> err... Message: Assertion failed at line 122 of /mnt/disk2/spiele/OpenTTD/trunk/src/thread/thread_pthread.cpp: err == 0
13:49:18 <Eddi|zuHause> when trying to add CETS...
13:49:34 <Eddi|zuHause> (without above patch)
13:51:24 <Eddi|zuHause> hm no, didn't pull oberhümers commits yet
13:53:38 <planetmaker> ok... worked. any more info, Eddi|zuHause ?
13:55:17 <Eddi|zuHause> started game, added cets afterwards
13:55:46 <planetmaker> bad, bad Eddi|zuHause ;-)
13:57:58 <Eddi|zuHause> something with switch blitter
13:59:50 * andythenorth wonders what to do next
14:00:35 <Alberth> translate FIRS to Dutch?
14:00:44 <planetmaker> still, I'm quite unable to reproduce that, Eddi|zuHause
14:01:35 <planetmaker> do you (only) get that when switching newgrfs during game?
14:01:40 <planetmaker> or also when starting a new one?
14:02:03 <Eddi|zuHause> starting new game works, and after that, i can remove and add cets
14:02:07 <andythenorth> fix the string :P
14:02:20 <Eddi|zuHause> so it has to do with enabling 32bpp blitter during a game
14:02:32 <Yexo> planetmaker: given the "SwitchBlitter" it might be necessary to play with opengfx and add a 32bpp newgrf later
14:02:50 <Eddi|zuHause> i have original-windows
14:02:52 <planetmaker> just tested that, Yexo
14:03:14 <Eddi|zuHause> but that shouldn't matter, only the switch from 8bpp to 32bpp
14:03:48 <planetmaker> Eddi|zuHause, which openttd?
14:11:28 <andythenorth> so....the way FIRS supply boost works is:
14:11:42 <andythenorth> - over a 3 month moving average, amount of supplies delivered is tallied
14:11:57 <Rubidium> a nuisance when station ratings are not near 100 ;)
14:12:09 <andythenorth> - 30t delivered = 2x production boost, 120t = 4x boost
14:12:41 <andythenorth> how to explain this in industry window text (if at all)
14:12:56 <Rubidium> as the chain looses too much to get the 30t back to the industries
14:13:16 <Rubidium> to get the 2x boost and with the 4x boost it's even harder
14:13:26 <andythenorth> this 'improved' supplies mechanic still has flaws
14:14:00 <Alberth> that's why it is not called 'ultimate' :p
14:14:09 <andythenorth> I never play without 'improved' station ratings on
14:14:16 <andythenorth> so I hadn't noticed the issue :P
14:15:31 <andythenorth> the supplies mechanic is also very volatile
14:15:53 <andythenorth> production can rise steeply, enough to cause network jams...then fall horribly due to network jam
14:16:06 <andythenorth> then network james again due to low production :P
14:16:17 <planetmaker> it's basically the same mechanic now as in ECS. Just two steps instead of 5
14:16:26 <planetmaker> well. not same. similar
14:16:45 <andythenorth> I've never played ECS :)
14:16:53 <planetmaker> in ecs you need to supply in proper amounts the cargos, iirc
14:17:04 <planetmaker> did I scare george?
14:17:28 <Rubidium> actually, in the second to last we also had oil to transport and could easily get that done in time with three oil wells and those oil well went into gun-ho pretty soon and stayed (or is it staid?) there
14:17:39 <andythenorth> in the second to last, we also had planes :P
14:17:51 <andythenorth> in the last game, I didn't see anyone delivering supplies
14:17:57 <planetmaker> we had 5 oil wells finally, Rubidium
14:18:06 <planetmaker> we funded two half-way
14:18:09 <Rubidium> but they didn't transport the eng. supplies to 'my' oil wells
14:18:14 <andythenorth> the last game had no planes, no quick RVs, and only slow ships
14:18:32 <andythenorth> mind - I am happy to tweak the supplies mechanic
14:18:38 <andythenorth> I am not convinced it's done yet
14:18:47 <Rubidium> what I did was make the oil trains have 7 oil wagons and 2 eng. supplies
14:19:17 <Rubidium> and transport the petrol to the thing that makes eng. supplies and farm supplies
14:19:19 <planetmaker> I saw that. seemed to work nicely?
14:19:24 <peter1138> last time i tried to play with ships the running cost was impossibly high
14:19:39 <Rubidium> and then transport the eng. supplies back to the station where the oil is dropped
14:20:26 <Rubidium> in the second to last game one wagon was not enough to get rid of all eng. supplies, in the last one most trains went back without eng. supplies at all
14:22:04 <Rubidium> so I'd surmise that with the FIRS station rating the supplies mechanics work quite well, without it they fail horribly in early-ish games since you can't get the station ratings high enough
14:22:54 <andythenorth> remove the parameter for station rating? :P
14:22:57 <andythenorth> default to 'improved' :P
14:23:04 <Rubidium> though... maybe, now I think about it, I shouldn't have gone for max TE but for max speed trains (regardless of acceleration), just to get more cargo transported
14:23:28 <andythenorth> I went for 3x fastest engine
14:23:29 <Rubidium> later I also spend a few hundred thousand just to build those lovely company statues in the important towns
14:23:57 <Alberth> andythenorth: rename to 'easy', and the 'disable station rating' to 'challenging' :)
14:23:59 <andythenorth> now I want to play again :P
14:24:07 <andythenorth> Alberth: not a bad idea :)
14:24:19 <andythenorth> Alberth: those strings look ok
14:24:19 <Rubidium> andythenorth: not a bad idea ;)
14:26:50 <andythenorth> 100% = easy, 'FIRS' = 'normal', default = 'challenging'
14:26:58 <andythenorth> this should be an ottd thing though :P
14:27:04 <andythenorth> newgrfs are doing too much :)
14:28:06 <Alberth> uh oh, andy has found out about the secret strategy to move the entire game to NewGRF :p
14:30:28 <Zuu> I think the 'basic' economy in FIRS is fairly complicated still.
14:31:26 <Eddi|zuHause> i'm still uncertain about the new narrower vehicle templates :/
14:31:41 <planetmaker> narrower in CETS?
14:34:04 <Eddi|zuHause> the argument was that a wagon is 4m high and 3m wide, 3m=6px and 4m=8px according to the rough CETS scale
14:34:10 <Eddi|zuHause> but it doesn't "look right"
14:34:23 <Eddi|zuHause> or, we need to do everything for 2x zoom
14:35:59 <andythenorth> Zuu: it can be more basic
14:36:13 <andythenorth> but there are limits with the current industries
14:36:35 <andythenorth> it might need new industries introducing
14:36:54 <planetmaker> Eddi|zuHause, I'd not mind using 2x zoom as "default" ;-)
14:37:02 <planetmaker> CETS could support that anyway
14:37:39 <Eddi|zuHause> oberhümer didn't want to draw that much when i asked that the last time...
14:39:29 <andythenorth> Zuu: too many industries / cargos? Or too many steps in chains?
14:40:13 <Eddi|zuHause> planetmaker: in that case, we should have a way to automatically downscale and 8bpp-ify sprites
14:40:31 <Eddi|zuHause> so we only have 32bpp-double.png files
14:41:07 <planetmaker> oh. there I see no automatic way, Eddi|zuHause. At least none which doesn't suck
14:41:51 <Zuu> andythenorth: too many industries, chains, cargos etc. to allow grasping everything without reading up all chains.
14:42:32 <Zuu> Eg. lot of information to keep in mind and then map down to a OpenTTD game in order to find out what to build.
14:42:42 <Eddi|zuHause> Zuu: i suggested the basic economy be the default industry plus something introducing the supplies mechanism
14:43:12 <Rubidium> planetmaker: you don't really need a perfect -> 8bpp conversion
14:43:15 <planetmaker> Eddi|zuHause, that doesn't exactly work. E.g. for a lack of powerplant :-)
14:43:39 <planetmaker> Rubidium, sure, one doesn't. That's why I said "one that doesn't suck" :-P
14:44:23 <Eddi|zuHause> planetmaker: downscaling 2x->1x is the more important problem, i suppose
14:44:46 <Rubidium> just render the sprites ;)
14:45:18 <planetmaker> In 32bpp I'd do that ^
14:46:12 <Eddi|zuHause> i was gonna add some basic pixa rendering... adding details should be a pixel art thingie...
14:48:16 <planetmaker> andythenorth, they still look big. e.g. tropic I'd cut at least fruit plantation and metal workshop
14:48:39 *** sla_ro|vista has joined #openttd
14:49:01 <planetmaker> or alternatively forest, sawmill, furniture factory
14:49:10 <planetmaker> we need goods :-)
14:51:48 <andythenorth> I like the idea of one per climate
14:52:24 <Zuu> Would it make sense to allow a GS to provide a list of setting+value pairs which will be enforced in new games that use this GS? I'm unsure if all or only eg. world gen settings should be allowed to be specified. I mainly have in mind eg. a GS that enforce zero towns + zero industries and take care of that itself. The world gen dialog would then need to indicate if such an enforcement is in place.
14:53:35 <Zuu> For settings that are changable in the game, a GS can sort of enforce them already by calling the API function to change them. But for world, gen this need to be handled before the game is even created.
14:54:47 <Zuu> An alternative to having this information provided by info.nut is to have a new method of the main class that is called when a user clicks "Generate" which will allow the GS to make setting changes.
14:55:02 <planetmaker> Yes-ish, Zuu. But... IMHO that only makes sense with re-writing the GUI so that newgrf, ai and adv. settings become a sub-window of the new game dialogue
14:55:02 <Zuu> The usability of the later is probably worse.
14:56:24 <Zuu> Agreed, that a such rewrite would be useful along with this, and perhaps it is a good idea to start there rather than adding more things for that rewrite to handle.
14:59:21 <Zuu> Though, the approach that the GS applies its settings (or give a suggestion that the player is confronted with) as you click on Generate wouldn't really conflict with the current layout of things.
15:10:57 <planetmaker> agreed, it's not a conflict. And both can be done independent
15:11:29 <planetmaker> I just mentioned that a GUI restructuring will make things easier to understand for the user
15:11:57 <planetmaker> like you quit playing your map. Configure a new one without being aware that you configured a GS. And then you wonder about either strange results or inaccessible options
15:12:31 <planetmaker> as they're configured from completely different windows which neither need be visited to generate a new map.
15:13:45 <planetmaker> but by all means. These consideration should not stop such feature
15:14:13 <planetmaker> Though I've not really yet developped an opinion in how far GS should steer map generation
15:14:36 <planetmaker> it might collide with the (new) scenario format. Or it might not
15:14:47 <planetmaker> maybe Terkhen has an opinion on that :-)
15:29:02 <Zuu> Yeah, it is a question what powers it should get.
15:29:43 <Zuu> How mayn did know that the save button in the main toolbar have a different sprite when hold down?
15:29:55 <Zuu> I didn't notice until seeing it in the code. :-)
15:43:19 <Terkhen> Zuu, planetmaker: I don't see this suggestion colliding with the new scenario format; since the new scenario format allows to enable or disable data for certain layers, it could for example disable the "towns data file" on load if the specified GS wants to take care of that setting
15:43:39 <Terkhen> the problem would be that said preference would need to be known before tha map is generated, and therefore, before the GS starts
15:45:19 <Zuu> It probably need to go into info.nut a bit similar to the configuration options. Which implies that I don't think that it would be possible for a GS to add an option that controls it preference of OpenTTD settings.
15:46:37 <Zuu> If you want to make it really complicated, you add this minimum-setting requirement to NewGRFs too and add some code to check for conflicts etc.. :-)
15:46:50 <Terkhen> sounds like a recipe for conflict :P
15:47:43 <Terkhen> as long as GS don't decide on terrain settings, I don't see any conflict with the new scenario format
15:47:49 <Zuu> Another option is that a such gs could use your scenario format to specify just what it want to impose and leave everything else for the player to decide?
15:48:00 <Terkhen> if GS try to decide over terrain settings, you would have the same problems with normal heightmaps, though
15:48:37 <Zuu> Eg, I could ship a hidden GS and a visible scenario and to play this GS, a user will have to play the scenario which loads the GS.
15:48:41 <Terkhen> I have not written anything with regard to GS in the format yet, but I would like to set up the option of requiring a specific GS somehow
15:48:52 <andythenorth> in principle I think GS should be able to control pretty much everything. Practically though, it might not work :(
15:49:47 <andythenorth> Terkhen: GS should require the scenario :)
15:49:51 <andythenorth> flip the hierarcy
15:50:19 <Terkhen> andythenorth: how would you identify that the scenario is the one that the GS wants?
15:51:36 <andythenorth> package them in one tar?
15:51:43 <andythenorth> new bundle format?
15:52:03 <andythenorth> 'scripted scenarion'
15:52:10 <andythenorth> yeah, maybe not :)
15:52:39 <Terkhen> I guess that you could bundle a GS inside the scenario tar
15:52:45 <Terkhen> but I don't think that's a good idea
15:52:51 <andythenorth> in railroad tycoon - each challenge is map, script, vehicles, everything
15:53:08 <andythenorth> it strikes me that some of the things scenario authors want to do - specific control over lots of stuff - is GS
15:53:34 <Terkhen> although, maybe scenario specific GS belong inside the scenario tar
15:53:36 <andythenorth> although I don't really care for the goal of creating detailed scenarios which evolve over time in a precise way
15:54:34 <Terkhen> I'll add a note to the scenario notes so I don't forget about this :P
15:54:44 <Terkhen> but the format implementation is still very green
15:54:54 <Terkhen> I always find me doing other stuff instead :)
15:59:33 *** supermop has joined #openttd
16:01:20 <Eddi|zuHause> gimp is rubbish... i can't do ctrl+s anymore to save the png :(
16:01:54 <supermop> pay adobe that cash money dude
16:01:57 <Eddi|zuHause> need to do ctrl+e now
16:02:05 <andythenorth> it's GPL right? :P
16:03:50 <Zuu> Have they released a single-window Gimp that work with wacom yet?
16:04:58 *** Djohaal has joined #openttd
16:05:02 <Zuu> Hmm.. Doesn't look like that. Better stay it 2.6 until they get that resolved. :-)
16:17:55 *** Djohaal_ has joined #openttd
16:18:34 *** mahmoud has joined #openttd
16:19:29 <andythenorth> did we get rid of the silly 'waiting to be processed' text at industries yet?
16:25:31 *** fonsinchen has joined #openttd
16:41:54 <Eddi|zuHause> it's GPL, right? :p
17:02:50 *** Djohaal has joined #openttd
17:05:02 *** DanMacK has joined #openttd
17:11:11 <supermop> so what were you saying about my sheds -
17:11:25 <supermop> what would you like me to fix?
17:13:30 <DanMacK> Are you planning any upgrades?
17:13:49 <supermop> i mean , yes in the sense that the set is broken and incomplet
17:14:05 <DanMacK> Water andcoaling towers come to mind
17:14:07 <supermop> but I don't know if it is a fruitful endeavor
17:14:32 <DanMacK> You've got the only roundhouse grf out there
17:15:03 <supermop> the current version doesn't even have the roundhouse in it
17:15:05 <DanMacK> Yes its eye candy but damned good eye candy
17:16:02 <supermop> well the nfo got to the point where i could make sense of it as it was so big
17:16:30 <supermop> so i started over from scratch writing just the bare needed for a few types of sheds
17:17:02 <supermop> get a bit of anxiety about trying to re write for all of those 'yard' tiles
17:17:21 <supermop> especially the roundhouse as it has weird offsets
17:17:41 <supermop> i never understood what i was doing - my mind doesn't work that way
17:18:10 <DanMacK> That my friend is why I draw and leave coding to others
17:18:14 <supermop> so it was always just cargo cult stuff - copying and pasting stuff with no understanding of if or why i needed it
17:18:34 <supermop> i tried asking for people to code, but no one offered
17:18:50 <supermop> also NML still doesn't do stations
17:19:12 <supermop> and few people are volunteering to code huge sets in nfo for strangers
17:20:08 <andythenorth> supermop: I don't code my own nfo stations either :P
17:26:00 *** FLHerne has joined #openttd
17:39:57 <supermop> anyway so don't expect a lot of progress anytime soon
17:49:50 <andythenorth> can I nest ternary operator expressions in nml?
17:50:00 <Eddi|zuHause> so... why the hell must the pen have odd size?
17:50:06 <andythenorth> I need to return 1 of 3 possible spritesets in a spritelayout
17:50:15 <Eddi|zuHause> stupid program...
17:50:53 <Alberth> andythenorth: try it?
17:51:15 <andythenorth> even trying to write it boggles my brain :P
17:51:19 <andythenorth> maybe I just use a switch :P
17:51:27 <Alberth> andythenorth: ok if I commit the dutch language changes?
17:51:33 <andythenorth> Alberth: sure :)
17:51:56 <Eddi|zuHause> andythenorth: sure you want the ternary operator for that?
17:52:07 <andythenorth> Eddi|zuHause: really no
17:52:34 <Eddi|zuHause> andythenorth: as in, do you want to return the computed action2 result, or do you want to use the cases?
17:52:50 <andythenorth> I want to use this in a spritelayout
17:53:01 <Eddi|zuHause> that was not the question
17:53:22 <andythenorth> I don't really have a model of what logic is doing in spritelayouts
17:53:27 <andythenorth> I can't map it to nfo
17:54:31 <andythenorth> I'll just use a switch :)
17:55:01 <Eddi|zuHause> formulate it in words: what should the result be in the 3 cases, and what is the input data?
17:55:19 <andythenorth> input data is animation frame number
17:55:42 <andythenorth> result is (spriteset_1, spriteset_2, spriteset_3)[pick one]
17:56:05 <Eddi|zuHause> then you should use the switch-cases
17:56:38 <Eddi|zuHause> "switch(do_something_with_frame_number) { 1: spriteset_1, 2: spriteset_2, spriteset_3 }"
17:59:26 <Yexo> <andythenorth> I can't map it to nfo <- at the time you were writing nfo this was not yet possible :p
17:59:36 <andythenorth> that explains it :P
18:00:32 <Yexo> the computations are all done in varaction2 switches and stored in temporary vars. The spritelayouts has a reference to the temporary vars
18:00:50 <Yexo> like "take the x-offset from var 80 instead of hardcoded here"
18:01:05 <andythenorth> that makes sense
18:01:32 <andythenorth> can I chain to a custom switch I've written from a spritelayout?
18:02:13 <Yexo> a spritelayout is always the end of the cain
18:03:07 <andythenorth> I'll just use a switch earlier, it's fine
18:03:17 <andythenorth> doing it in the spritelayout is tidier somehow :P
18:03:25 <andythenorth> but actually makes my brain hurt a bit more
18:14:23 *** oskari89 has joined #openttd
18:44:51 <supermop> andy, do you have passenger stuff in chips?
18:45:21 <DorpsGek> Commit by translators :: r24753 /trunk/src/lang (catalan.txt spanish.txt) (2012-11-17 18:45:13 UTC)
18:45:22 <DorpsGek> -Update from WebTranslator v3.0:
18:45:22 <supermop> the concret platforms have white lines like a british passenger station, but cranes appear on them
18:45:23 <DorpsGek> catalan - 5 changes by Bassals, arnau
18:45:24 <DorpsGek> spanish - 1 changes by Terkhen
18:59:50 *** sla_ro|master has joined #openttd
19:14:02 *** andythenorth has joined #openttd
19:34:10 *** Djohaal has joined #openttd
19:36:45 *** Djohaal has joined #openttd
19:44:09 *** DanMacK has joined #openttd
19:50:11 <andythenorth> early FIRS industries - production limits?
19:50:26 <andythenorth> as a way to enforce players moving to newer industry types?
19:51:54 <FLHerne> andythenorth: Yes, definitely
19:52:40 * FLHerne has been delivering a four-track mainline's worth of grain to one post mill recently :D
19:53:00 <FLHerne> andythenorth: Weren't you trying to figure that out a few months back?
19:53:13 <andythenorth> what should the limit be?
19:53:31 <FLHerne> I even looked up production for post mills, should be about 100-200 t/month IIRC
19:54:40 <Alberth> can you accept more, but not produce more?
19:55:31 <FLHerne> I'd rather it didn't. That would be irritating :P
19:55:46 <andythenorth> so reject cargo instead
19:56:00 <andythenorth> which industries should do this?
19:56:57 <FLHerne> Windmill, Smithy Forge, any of the 'town' industries
19:57:42 <FLHerne> Could you limit EngSup acceptance based on current production?
19:58:51 <andythenorth> so split windmill from grain mill
19:58:56 <andythenorth> add a new 'early brewery'
19:59:02 <andythenorth> and limit all town industries?
19:59:20 <FLHerne> Do :P . One small scrapyard shouldn't be willing to buy the entire global production of pit props :P
20:00:07 <FLHerne> andythenorth: The brewery you have looks quite old. Perhaps a more 'industrial' one?
20:00:12 <supermop> why not have new industries be more efficient
20:00:39 <FLHerne> Also, early version of the Sugar Refinery would be nice, unless I forgot something
20:01:24 <andythenorth> what about early versions of machine shop?
20:01:28 <supermop> have modern ones be about 8t -> 6t, old ones 8t -> 3t
20:02:07 <FLHerne> And a 'Market', which would accept Fish, Food, Goods, Alcohol(?), Fruit, Milk, blah in small quantities
20:02:15 <andythenorth> also early oil refinery?
20:02:16 <FLHerne> Oh, except acceptance limits. Drat.
20:02:47 <andythenorth> when would the market close?
20:02:49 <FLHerne> andythenorth: Both would be nice. What would an early refinery look like, though? Also agree with supermop
20:03:03 <FLHerne> andythenorth: Never. There still is one here :P
20:03:22 <andythenorth> there is room for another 15 industries
20:03:24 <FLHerne> Perhaps randomly select cargoes to accept out of plausible ones?
20:03:29 <andythenorth> so a lot more could gain 'early' versions
20:03:31 <FLHerne> That might be confusing though
20:04:45 <andythenorth> Zuu: how do AIs handle industries that stop accepting cargo when limit is reached?
20:04:49 <FLHerne> I'd still like to see a cannery (takes metal and [fish and/or livestock])
20:05:14 <FLHerne> Fishing Harbours as the only destination for fish is a bit boring
20:05:37 <FLHerne> Perhaps not livestock actuall, that would be weird :P
20:06:27 <andythenorth> ok, so conditional orders can reroute trains depending on station acceptance, right?
20:06:29 <Zuu> AFAIK an AI can check for "temporary" blocking of cargo acceptance. However, it doesn't have a clue how long the temporary block will last or for what reason it have happened.
20:06:56 <andythenorth> Zuu: but it could count cargo delivered, then work out the limit?
20:07:41 <andythenorth> also GS could avoid setting cargo goals for industries that have limits?
20:07:44 <Zuu> counting cargo deliveried is not very easy for an AI.
20:07:56 <Zuu> it doesn't have access to the CargoMonitor that GS can use.
20:08:19 <andythenorth> what should the limits be? 100t? 1,000t?
20:08:51 <Zuu> An AI would have to estimate amount of delivered cargo by transport distance and vehicle fleet size and possible queue length at loading station.
20:09:05 <andythenorth> Zuu: doesn't sound very fun :P
20:09:29 <Zuu> No, it is probably quite difficult to make an accurate measurement.
20:09:30 <Eddi|zuHause> the AI could check when each of the vehicles unloads?
20:09:47 <Zuu> If the AI doesn't do anything else, it can do that.
20:10:05 <andythenorth> give the AI capability to detect which industry newgrf is in use?
20:10:11 <andythenorth> and to read the permanent storage?
20:10:24 <Zuu> But there is no callback from OpenTTD on tho unload event, so you need to spend lots of processing time to scann vehicles very often then.
20:12:29 <Zuu> Actually, AI have AIIndustry::GetStockpiledCargo(industry_id, cargo_id)
20:12:42 <Zuu> But it have no idea what the limit is.
20:13:17 <Eddi|zuHause> we need CB18 for industries then :)
20:13:27 <andythenorth> might not use Stockpile anyway
20:13:28 <Zuu> So it can check GetStockpiledCargo every now and then to detect industries which use stockpiling to detect industries that may have the risk of getting full.
20:13:36 <andythenorth> should all FIRS industries have limits?
20:14:25 <Eddi|zuHause> only "small" ones (early economy)
20:14:32 <Zuu> You could formalize stockpiling in the specs to include also the upper limit as a formal field which NewGRF sets and AIs can read.
20:14:55 <andythenorth> Eddi|zuHause: what's the rationale for that?
20:15:57 <Zuu> Or we use the scp-ish approach to allow NewGRFs to add custom fields to eg. industries. Eg. define "stockpile_max" and pass the value 1500.
20:16:02 <Eddi|zuHause> andythenorth: early/small industries should have the same efficiency in inpuz/output ratio, but they should not scale well to high inputs
20:16:26 <Eddi|zuHause> whether you do that by input limits or output limits is your choice
20:16:36 <Eddi|zuHause> output limits would be easier for the player
20:17:05 <Eddi|zuHause> i.e cargo would still be accepted, but there will be no processed cargo for it
20:17:57 <Eddi|zuHause> Zuu: CB18 is supposed to be the AI<->NewGRF interface
20:18:06 <andythenorth> how do I get more newgrf debug output from ottd?
20:18:11 <Eddi|zuHause> but it's only implemented for stations
20:18:19 <andythenorth> I have an industry refusing to appear in minimap
20:18:28 <Zuu> I have no clue when you talk in things like CB<num> instead of human readable variable names :-)
20:18:32 <Eddi|zuHause> there's also a spec for trains, but it's useless, so it was never implemented
20:20:21 * andythenorth omitted the action 0 block :P
20:25:43 <andythenorth> I have a 180 frame animation, and a 3-sprite spriteset
20:25:49 <andythenorth> iron_works_spriteset_iron_pigs_1_0((1/60)* animation_frame)
20:27:21 <andythenorth> that needs a modulus?
20:27:54 <FLHerne> andythenorth: Can industries change size while existing?
20:31:05 <Eddi|zuHause> andythenorth: just move the /60 to the end
20:31:30 <Eddi|zuHause> andythenorth: there are no fractional numbers, so 1/60 == 0
20:31:42 <Eddi|zuHause> so you're multiplying the frame number by 0
20:31:52 <Eddi|zuHause> which obviously always results in 0
20:32:07 <andythenorth> integer maths :P
20:32:35 <Zuu> Maybe visually by changing sprites?
20:33:55 <Eddi|zuHause> andythenorth: common newbie error :p
20:34:14 <andythenorth> done it many times before :P
20:34:28 <andythenorth> if I put a float in, nml actually warns me
20:34:48 <andythenorth> but not if I do the n/x expression
20:35:10 *** fonsinchen has joined #openttd
20:35:12 <Zuu> Eg. an early age industry could have flat ground with some scrap laying around, and then as production goes up, it get restructed into a larger building that now goes all way out to the edge.
20:36:09 <Eddi|zuHause> Zuu: similar to the headquarter, yes
20:37:10 <Eddi|zuHause> it should be fairly simple, just have a moving average of production rates or something, and exchange the graphics based on that
20:38:32 <Zuu> Though that would probably work best on an industry set with not so many industries. As industries changing graphics will help getting you lost.
20:38:36 <Eddi|zuHause> but the industry can't simply add tiles around it, they have to be reserved from the start
20:39:44 <Eddi|zuHause> do industry tiles have storage? animation? can the construction stages be influenced?
20:40:32 <Zuu> For user friendlyness it is good if all stages of an industry follow the same pattern. Or that at least it is easy to see which type of industry each type is without clicking on it.
20:41:51 <andythenorth> industry tiles don't have storage
20:41:55 <andythenorth> they do have animation
20:42:00 <andythenorth> but the animation is typically in use
20:42:10 <andythenorth> changing the graphics is trivial
20:42:50 <Eddi|zuHause> i mean that the change of graphics will not be instant, but during the tileloop in multiple steps
20:43:47 <Alberth> ECS does that quite sanely, it adds buildings in the same style, so you can still recognize the industry
20:50:44 <andythenorth> Eddi|zuHause: so "This industry can only produce 1,000t per month. Any additional cargo delivered will be accepted, but won't result in any output"
20:51:07 <Eddi|zuHause> andythenorth: yes. something like that
20:51:29 <FLHerne> andythenorth: But accepting unusable cargo would lead to inefficiency :-(
20:52:04 <andythenorth> is it more efficient to leave it on trains?
20:52:23 <FLHerne> andythenorth: Breaks any sort of cargodest :P
20:52:50 <FLHerne> If you leave it on trains, it'll go somewhere else eventually
20:53:03 <andythenorth> the packets get re-routed?
20:53:06 <FLHerne> s/breaks/leads to waste when combined with/ :P
20:53:37 <andythenorth> YACD doesn't re-route cargo packets
20:53:51 <FLHerne> I believe CDist at least reroutes packages that have sat for ages
20:54:02 <FLHerne> I'm not totally sure though
20:54:11 <andythenorth> so the decision that needs to be made then: is YACD or CargoDist going into trunk?
20:54:27 <FLHerne> At any rate, intermittent acceptance of cargo would dramatically reduce the amount sent
20:55:57 <michi_cc> YACD only changes cargo destination if the dest vanishes, it doesn't look at temporary acceptance changes.
20:56:21 <andythenorth> FLHerne: so which one goes into trunk? YACD? Or CargoDist?
20:56:25 <andythenorth> FIRS can't support both it seems
20:56:45 <andythenorth> I guess these limits will also need a parameter to disable them?
20:56:54 <michi_cc> So if you get both you're screwed? :p
20:57:33 <FLHerne> CargoDist would appear to be a lot closer (or at least more finished) - it's almost at current trunk rather than being 5k revisions behind :P
20:57:51 <FLHerne> Parameter actuall sounds sane in that case
20:58:15 <FLHerne> Both for 'limits' and 'behaviour for limits'
20:59:33 <andythenorth> is the limit per cargo?
20:59:35 <michi_cc> Almost no code paths YACD interacts with have received significant changes (except the settings.ini stuff and savegame rev) so that's not a useful measurement.
20:59:37 <andythenorth> or total of all cargos?
20:59:59 <andythenorth> is that another parameter?
21:00:23 <FLHerne> michi_cc: 'Someone cares about updating it' is relevant though :P
21:00:41 * FLHerne would love to see either in trunk
21:01:02 <FLHerne> CDist preferably, but tbh either is better than the stupid pax we have now
21:01:53 <michi_cc> YACD gets an update if I get a good idea to solve an implementation problem.
21:03:58 <FLHerne> michi_cc: Is that the thing with working out if cargo can actually get somewhere?
21:04:17 <andythenorth> FIRS gets some limits on some industries if I can figure out amounts, how to not break AIs, how to not break GS, how to not break *Dist, and how not to piss off players who hate the idea
21:04:44 <michi_cc> It's the 'some stuff is too slow' thing.
21:05:25 <FLHerne> michi_cc: Computers? Make it optional :P
21:05:54 <Zuu> FLHerne: Optional computers? So there will be a board game variant too?
21:06:22 <andythenorth> optional is lame :P
21:06:29 <Eddi|zuHause> andythenorth: certainly allowing unlimited acceptance and producing no output will be the least problematic for AIs
21:06:43 <FLHerne> I still have both YACD and CDist on my PIV, both seem to work fine
21:06:58 <andythenorth> YACD would use about 45% CPU
21:07:03 <andythenorth> and kill my battery
21:07:38 <Zuu> Eddi|zuHause / andythenorth: accepting all cargo but producing no output will probably not be a too bad situation for AIs.
21:07:46 <FLHerne> andythenorth: Optional might be necessary, because either behaviour is going to annoy a reasonable number of people :P
21:08:08 <Zuu> At least I don't see that it would break CluelessPlus. :-p
21:08:13 <andythenorth> "optional because the implementation is fundamentally inefficient" is invalid
21:08:23 <andythenorth> YACD had perfomance issues on my laptop
21:08:33 <FLHerne> andythenorth: Oh, I was thinking of FIRS again :P
21:09:03 <andythenorth> FIRS is having options removed, not added :P
21:09:37 <FLHerne> andythenorth: Says the person who just added Economies? :P
21:10:07 <andythenorth> so why won't oil rigs build?
21:10:12 <andythenorth> what did I break?
21:13:56 <Eddi|zuHause> try to build one manually?
21:14:44 <Eddi|zuHause> "fails" is not a problem description
21:15:01 <andythenorth> definitely something I've done
21:15:11 <andythenorth> returns 'site unsuitable'
21:15:19 <andythenorth> didn't do that 400 revs ago :P
21:16:38 * andythenorth waits for compile :P
21:16:58 <fonsinchen> Cargodist will reroute cargo if the destination stops accepting it.
21:17:21 <fonsinchen> And it will only start routing cargo to a destination if the destination has been accepting it for some time.
21:17:41 <fonsinchen> Last time I tried with FIRS it was a pain.
21:18:24 <Eddi|zuHause> i can see pain with ECS, but i've not experienced any problems with FIRS
21:18:28 <fonsinchen> Because some industries need some kind of "bootstrap" deliveries to start producing anything
21:18:39 <fonsinchen> Like the stuff with fishing boats and passengers
21:18:54 <andythenorth> sounds like a bug :o
21:19:07 <Eddi|zuHause> you're probably mixing up ECS and FIRS
21:19:31 <FLHerne> fonsinchen: I've being playing with those two for years now
21:19:35 <andythenorth> secondary industry needs cargo to produce
21:19:47 <FLHerne> Seems to work very well :-)
21:19:50 <andythenorth> is that the issue?
21:20:02 <FLHerne> fonsinchen: CDist is yours, right?
21:20:53 <fonsinchen> Whatever it was: I remember I couldn't build a working PAX route to fishing boat, because the fishing boat would only produce PAX if it had received some before.
21:21:18 <Eddi|zuHause> that really sounds like ECS
21:21:26 <fonsinchen> However, Cargodist routes PAX symmetrically so it would only send pax to the boat if it was expecting to send some back, too.
21:21:34 <Rubidium> it might be useful that if cargo is routed intelligently, that it tries to deliver the optimum ratios of input cargos to secondary industries. However, that requires quite a bit of work which would be useful for AIs as well
21:22:32 <Eddi|zuHause> fonsinchen: maybe add a random amount to any accepting destination for symmetric cargo?
21:22:38 *** Djohaal has joined #openttd
21:23:01 <Eddi|zuHause> it may break this catch 22
21:23:09 <fonsinchen> There is some allowance to send cargo to destinations which don't send anything back.
21:23:16 <Eddi|zuHause> even very small chance should suffice
21:24:04 <fonsinchen> But the "chunked" behaviour of ECS was the second problem. Sometimes some PAX would arrive at the boat and a short time later they wanted to go back
21:24:43 <fonsinchen> so Cargodist took some time to calculate routes for them and when the routes went into effect the passengers were gone
21:25:02 <fonsinchen> the routes were heavily fluctuating all the time.
21:26:18 <fonsinchen> I believe this behaviour probably is a pain with vanilla openttd, too, so I didn't put much work into changing cargodist for it.
21:26:45 <Eddi|zuHause> yes, afaik they're working in "shifts", so you bring a bunch in, they work for some time, and then all want to go back at the same time
21:26:54 <andythenorth> so output limits might be the solution to forcing players to switch to different industry type?
21:27:08 <andythenorth> but lower production ratio isn't?
21:27:25 <Eddi|zuHause> andythenorth: yes, i found the lower production ratio annoying
21:27:49 <Zuu> andythenorth: Why not just close down the industry if you want to phase it out?
21:29:07 <FLHerne> andythenorth: Both are :-)
21:29:35 <FLHerne> Not too much production rate difference though, please
21:29:56 <andythenorth> how much would you like?
21:30:31 <Eddi|zuHause> Zuu: but there has to be some incentive for the player to switch to the new industry before the old one gets phased out
21:31:00 <FLHerne> Perhaps 4 or 5:8 instead of 6:8 later?
21:31:26 <Eddi|zuHause> all industries should be able to reach 8:8 ratio
21:31:48 * andythenorth wonders if the solution is actually just to remove this industry
21:31:59 <andythenorth> sometimes editing is necessary
21:32:26 <Eddi|zuHause> Elukka: that's a rather small rocket :)
21:32:39 <FLHerne> andythenorth: Which industry?
21:32:56 <FLHerne> Just closing served industries is incredibly annoying :P
21:33:19 <FLHerne> Perhaps close all unserved ironworks? Can the grf tell?
21:33:31 <FLHerne> andythenorth: Because of the giant piles of cargo?
21:33:42 <andythenorth> what giant piles of cargo?
21:34:37 <FLHerne> The ones that would appear rapidly if you clsed a served industry. Except that wasn't what you were proposing :P
21:34:48 <FLHerne> Or at least I don't think so
21:34:53 <andythenorth> I'm considering it
21:35:06 <andythenorth> how are those piles of cargo different to the ones you get with an acceptance limit?
21:35:28 <andythenorth> the change is identical: the station simply stops accepting cargo
21:35:38 <andythenorth> why is that desirable in one context, but not another?
21:35:50 *** Superuser has joined #openttd
21:36:19 <FLHerne> andythenorth: Because in one case it applies predictably to all industries of that type, and can be considered when deciding how much cargo to send
21:36:44 <andythenorth> industries would close at a predictable date, let's say 1900
21:36:54 <andythenorth> with a few months randomness
21:37:16 <andythenorth> the effect on player is identical to when acceptance limit is reached: they have to build new routes
21:37:18 <FLHerne> andythenorth: That would be a bit odd if all industries of one type closed simultaneously :P
21:37:45 <andythenorth> to me the two methods have 1:1 the same implications
21:37:57 <andythenorth> just one of them confuses AIs less
21:38:16 <FLHerne> andythenorth: With acceptance limit, you can simply avoid building excess routes in the first place, rather than unexpectedly having to find the nearest industries that accept the relevant cargoes and build lines to them
21:38:40 <andythenorth> how do you avoid building excess?
21:38:45 <andythenorth> by building other routes?
21:38:58 <FLHerne> Indeed. But that's different
21:39:28 <andythenorth> building routes != building routes ?
21:39:41 <FLHerne> Would it be possible to have acceptance rates decrease to 0 over a decade or so? And then close them?
21:40:07 <FLHerne> Would represent gradual reduction in output as they got outcompeted by new tech :D
21:40:18 <FLHerne> And give the operator time to respond
21:40:50 <andythenorth> you'd need a news message about it?
21:41:18 <FLHerne> andythenorth: Ideally yes. Not possible yet?
21:42:00 <andythenorth> Zuu: can I inform the AI about a changing acceptance limit by having it read news messages?
21:42:01 <FLHerne> The difference is that with acceptance limits, you can centre a new subnetwork around a second industry after the first is overloaded
21:42:15 *** DanMacK has joined #openttd
21:42:33 <FLHerne> That's different from having to bodge your old network to serve several new industries after th old hub closes
21:42:53 <andythenorth> FLHerne: but if industry closes, you can centre a new network on a second industry
21:43:21 <Superuser> OpenTTD's OpenMSX should totally use some music from Simutrans
21:43:22 <FLHerne> andythenorth: But the old network then has to be either completely rebuilt or substantially reconfigured
21:43:31 <Superuser> rework it a bit to be in the same style and you are done.
21:44:04 <FLHerne> Slowly declining production at least allows the user to realign/replace one route at a time, rather than all at once
21:44:29 <andythenorth> FLHerne: so I could space closures over ~20 years for same effect
21:45:05 <Superuser> Just wondering, does turning off 'full detail' change anything at all?
21:45:11 <Superuser> I see no change whatsoever.
21:45:31 <Zuu> andythenorth: AIs (and GSs) don't read news. But they have an event system which is basically a stack of events which the AI/GS can try to pop events from to see if anything have been added there by OpenTTD.
21:45:46 <Zuu> From a quick look I can't find an event for acceptance changes.
21:45:50 <Superuser> oh, I see it now, lamp-posts go away, that's about it
21:48:32 <Zuu> Oh, and the events aren't saved in savegames so upon save+load all unread AI events are lost.
21:49:07 <FLHerne> Black is secondary, red is primary. If one secondary closes, I have to link all those primaries to a different secondary.
21:49:27 <FLHerne> Of course, in a real game the secondaries would be further apart
21:50:09 <andythenorth> I don't see how it's different to closure tbh
21:50:17 <andythenorth> mathematically, I think they reduce to same thing
21:50:20 <andythenorth> 'acceptance stopped'
21:50:20 <FLHerne> Also, if you close all industries of one type the user has to divert to several alternative secondaries, because no one secondary would accept all the same cargoes
21:50:29 *** mahmoud has joined #openttd
21:50:34 <FLHerne> andythenorth: I'm complaining about closure :P
21:50:53 <andythenorth> I still don't see the issue :)
21:51:05 <andythenorth> you end up building routes either way
21:51:08 <FLHerne> Slow closure is better than fast closure, because the user can change things piecemeal rather than all at once
21:51:58 <DanMacK> Slow would be more realistic
21:52:14 <FLHerne> DanMacK: Don't mention that, he'll do the opposite :P
21:53:11 <andythenorth> my favourite idea so far is output that falls as more cargo is delivered
21:53:22 <andythenorth> possibly tending to zero, or negative :P
21:54:15 <FLHerne> Do you mean output:input ratio?
21:54:18 <DanMacK> Oh that might be cool
21:54:31 <FLHerne> If you don't, that's a bit strange :P
21:56:33 <andythenorth> deliver > x amount, production falls
21:56:42 <andythenorth> deliver too much, production falls to 0
21:56:58 <andythenorth> 'industry is overflowing with iron ore, production halted'
21:57:09 <andythenorth> it would never restart though :P
21:58:02 <Zuu> That will really punish AIs that give higher scores to connections that feed existing nodes with more cargo.
21:58:34 <andythenorth> Eddi|zuHause: how about I just remove this Iron Works (merge it with Steel Mill, issue solved)
21:59:04 <FLHerne> andythenorth: Differentiated early industries are nice :-)
21:59:19 <Eddi|zuHause> andythenorth: i'd still like a shifting between old (small) economay and new (large) economy
21:59:41 <Eddi|zuHause> at least for some economies
21:59:41 <andythenorth> when do you start your games
22:00:01 <Eddi|zuHause> historically, 1920, but CETS aims towards 1870
22:00:25 <FLHerne> Anywhere between 1840 and 1980, depending on how much time I expect to have in the next week or two
22:00:25 <Superuser> are you talking about Simutrans or something?
22:00:38 <FLHerne> Superuser: Sheck the channel topic :P
22:00:49 <FLHerne> This is OTTD, not SImutrans :D
22:01:05 <FLHerne> Re OpenMSX, it *is* open-source :P
22:01:17 <Zuu> but well, if AIs can see what the industry does, then do whatever you like... or mark it as uncompatible with AIs.
22:01:44 <FLHerne> Catch is, it needs MIDI files. Not common anymore
22:01:52 <andythenorth> Eddi|zuHause: the only 'early' industry you'll really see is Iron Works in that case
22:02:00 <andythenorth> there are no others apart from Forge
22:02:17 <Superuser> Yeah, Simutrans is Midi
22:02:27 <Superuser> and most of them are fairly jazzy
22:02:30 <FLHerne> Superuser: What license?
22:02:39 <Superuser> should be Artistic License 2 if I'm not mistaken
22:02:57 * FLHerne hasn't paid attention to Simutrans much :P
22:03:02 <Superuser> as it's part of the Simutrans distribution, not tied to the game
22:03:03 <Eddi|zuHause> soo... i'm halfway through making a floorplan for pixa, but i got bored... what do i do now?
22:03:07 <Superuser> I know, it sucks compared to OpenTTD hahaha
22:03:21 <Superuser> it's actually awful, I hate it and am hopelessly lost when I play it
22:03:22 <FLHerne> What license is OpenMSX? (*really* haven't been paying attention :D)
22:03:29 <andythenorth> Eddi|zuHause: install more patience? :P
22:03:30 <Superuser> It's like the Dwarf Fortress of transportation games
22:04:04 <Superuser> except that Dwarf Fortress actually has a plan, whereas Simutrans just does the feature creep
22:05:21 <andythenorth> OpenTTD has no plan
22:05:36 <Superuser> yeah, but you don't exactly introduce new features
22:05:48 <Superuser> you had a definite 1.0 version
22:06:29 * Superuser has actually started playing since about v1.10
22:06:49 <FLHerne> Superuser: Game scripts? Infrastructure maintenance? Rivers? Stuff I don't remember about? :P
22:07:07 <Superuser> nothing that impacts the core gameplay much : )
22:07:15 <Superuser> the out of the box experience is still largely the same
22:07:36 <Superuser> and that's because it is very good, if it ain't unfun don't change it (phrase adapted)
22:07:58 <Eddi|zuHause> andythenorth: we do have a "2.0 vision" thread in the forum :p
22:08:11 <andythenorth> we even have a roadmap on the wiki
22:08:16 <andythenorth> or did someone delete it?
22:08:21 <andythenorth> it was a lovely work of fiction
22:08:39 <Superuser> anyway Simutrans is meh but the music is pretty cool, the OpenMSX team should definitely give it a listen and maybe adapt it
22:08:46 <Eddi|zuHause> i think someone made it readonly
22:08:56 <Superuser> or just stick it straight in, most tracks should be appropriate
22:08:58 <Eddi|zuHause> and since then probably nobody updated anything
22:09:07 <Superuser> (there are some 45 music tracks in Simutrans)
22:09:38 <Eddi|zuHause> Superuser: you can just make a simutrans music set, independent from openmsx
22:09:52 <FLHerne> Superuser: Perhaps make a separate 'Simutrans' music pack?
22:09:57 <Eddi|zuHause> just put the midi files in a .tar and add a .obm file
22:10:07 <FLHerne> |Gah! Stop stealing my ideas! :P
22:10:07 <Superuser> but... but... no-one will listen to it... :(
22:10:26 <FLHerne> Superuser: Put it on BaNaNaS :-)
22:10:33 <Eddi|zuHause> Superuser: put it in bananas with the appropriate license, and certainly people will get it
22:10:46 <Superuser> damn, gotta scour for the license then
22:10:47 <FLHerne> Eddi|zuHause: Beat you to it this time :D
22:10:52 <Superuser> but okay, challenge accepted :)
22:11:25 <andythenorth> Eddi|zuHause: tell me a number for output limit on Iron Works?
22:11:31 <Eddi|zuHause> FLHerne: just because you left out half of it :p
22:11:34 <FLHerne> Eddi|zuHause: How does the bananas requirement to be the author work with music packs
22:11:54 <FLHerne> Does the compiler count as an 'author'?
22:12:13 <FLHerne> Agreed, 500 sounds good :-)
22:12:24 <FLHerne> Perhaps have it even lower pre-1850ish?
22:12:51 <Eddi|zuHause> andythenorth: and didn't you have a lovely windmill graphics that could serve as an old-timey food processor?
22:13:43 <andythenorth> it's used by the grain mill
22:13:55 <Eddi|zuHause> FLHerne: if he wrote the .obm file, that counts :)
22:14:34 <Eddi|zuHause> andythenorth: yes, windmill: old food processor (with limit), grain mill: new food processor (without limit)
22:17:38 <andythenorth> Eddi|zuHause: would limits by date on all industries suit your aim?
22:18:57 <andythenorth> what about reduced primary production?
22:19:05 <andythenorth> lower production has same effect as limits?
22:20:05 <Eddi|zuHause> andythenorth: the idea was that the player has some incentive to switch deliveries to the new steel works. simple primary production limits won't achieve that
22:20:25 <Eddi|zuHause> but gradually increasing primary production with time sounds nice
22:20:33 <Eddi|zuHause> independent of that
22:20:33 *** sla_ro|master has joined #openttd
22:21:00 <andythenorth> I think I can solve the iron works / steel works issue by merging them tbh. I am interested in your case for early industry behaviour, which might be orthogonal to iron works issue
22:21:01 <Eddi|zuHause> andythenorth: anyway, this could be a "metamorphosing" economy
22:21:32 <andythenorth> I wish someone would bite the bullet and decide that GS can have control over newgrf industry
22:21:37 <Eddi|zuHause> so you offer one economy which always has the new industries and none of the old industries
22:21:37 <andythenorth> it would be more appropriate
22:22:05 <Eddi|zuHause> and another economy with the old industries at the start and adding in the new industries gradually
22:22:15 <andythenorth> that makes sense
22:22:22 <Eddi|zuHause> i don't think GS will help there
22:23:34 <andythenorth> GS control over closure, production rates etc would be appropriate
22:23:45 <andythenorth> it allows the newgrf to be more vanilla
22:26:03 <Eddi|zuHause> the problem is you cannot combine GS
22:26:09 <Eddi|zuHause> it's all or nothing
22:26:28 <FLHerne> Eddi|zuHause: They should change that, too :-)
22:26:49 <Eddi|zuHause> FLHerne: that'll cause all kinds of nightmares
22:27:59 <Eddi|zuHause> andythenorth: and the other problem is how do you make the GS so that it understands the loaded NewGRFs and finds out what to do? things like "old, phased out" industries can't work with a GS
22:32:10 <andythenorth> ok, so I have no real objection to output limit
22:32:18 <andythenorth> I think players will struggle to understand it
22:32:26 <andythenorth> it's better than input limit
22:32:38 <andythenorth> and low production ratios just don't really work
22:33:01 <andythenorth> I might merge the iron works with steel mill though, which mostly eliminates this issue
22:36:03 <Eddi|zuHause> don't throw the code away :)
22:37:10 <andythenorth> some economy like 'everything changes'
22:37:21 <andythenorth> keep the iron works for that
22:46:07 *** andythenorth has left #openttd
23:43:29 *** Superuser has joined #openttd
continue to next day ⏵
