IRC logs for #openttd on OFTC at 2016-12-09
⏴ go to previous day
00:17:53 *** HerzogDeXtEr1 has joined #openttd
00:19:51 *** smoke_fumus has joined #openttd
00:23:07 *** Biolunar has joined #openttd
00:23:07 <supermop> they just renamed the airport in my home town after him
00:27:24 *** chomwitt has joined #openttd
02:43:17 *** TrueBrain has joined #openttd
04:04:37 *** ChanServ sets mode: +v tokai
04:05:03 *** Fatmice has joined #openttd
04:25:52 *** ConductorCat has joined #openttd
05:34:25 *** Supercheese has joined #openttd
08:26:51 *** BluesInTheNet has joined #openttd
09:26:00 *** Supercheese has left #openttd
09:36:54 *** Biolunar has joined #openttd
11:01:29 *** TheMask96 has joined #openttd
11:14:52 *** andythenorth has joined #openttd
12:40:13 *** Alberth has joined #openttd
12:40:14 *** ChanServ sets mode: +o Alberth
12:43:19 * andythenorth reading nml source to try and figure out the string literal problem
12:43:46 *** FR^2 is now known as Guest1113
12:50:12 <Alberth> reads to me like it expects a literal like "andy", or a string type string(...)
12:52:15 <andythenorth> chokes on TTD_STR_CARGO_SINGULAR_COAL
12:52:32 <andythenorth> which isn’t a literal?
12:53:26 <Alberth> no, a literal is something with " thingies
12:53:45 <Alberth> string(TTD_STR_CARGO_SINGULAR_COAL) would work I think
12:54:35 <Alberth> ^ I am pretty sure I copied that knowledge from one of your projects :p
12:56:36 <Alberth> yep, from fish: grep string fish.nml | grep ,
12:58:39 <andythenorth> grf strings work
12:58:45 <andythenorth> but the problem is TTD strings
12:59:45 <andythenorth> they have constants defined
12:59:53 <andythenorth> and they’re expected to be literals
12:59:55 <Alberth> string(1234) doesn't do anything useful (with a more useful number probably)
13:00:01 <andythenorth> but can’t use a literal inside string()
13:01:00 <andythenorth> I am just double-checking, but I expect to get ‘not found’ if I do string(TTD_STRING)
13:01:03 <andythenorth> yeah, not found :)
13:01:31 <andythenorth> interesting problem
13:02:34 <Alberth> I assume it's correct there :)
13:04:29 <andythenorth> but that means TTD strings can’t be used as substrings
13:05:50 <andythenorth> if I don’t use the default TTD cargo strings (where they exist), translators will have to duplicate existing translations
13:06:02 *** chomwitt has joined #openttd
13:06:21 <andythenorth> I already provide them in English for the docs, which can’t access TTD strings, so it’s no extra work for me
13:06:28 <andythenorth> but for translators, might be annoying
13:10:51 <Alberth> what happens if you use the number inside the 'string()' ?
13:11:10 <Alberth> also, do you use it as parameter, or as main string?
13:12:51 <Alberth> there is no regression test either, so it may have been tested last a looong time ago, if at all
13:13:45 *** welshdragon has joined #openttd
13:15:55 *** welshdragon has joined #openttd
13:24:58 *** welshdragon has joined #openttd
13:26:41 <argoneus> good morning train friends
13:30:22 <andythenorth> Alberth: using the string number fails as ‘First parameter of string() must be an identifier'
13:30:30 <andythenorth> which is expected I think :)
13:31:26 *** welshdragon has joined #openttd
13:37:58 <Alberth> string(STR_JUST_STRING, 1234) where STR_JUST_STRING : {STRING}
13:39:43 <Alberth> this would suggest you can just use the constant name instead of string(...)
13:39:53 <andythenorth> that was my initial expectation
13:40:56 <Alberth> 1234 can be a TTD_STR_... literal, I would expect
13:40:59 <andythenorth> the number doesn’t work as a parameter
13:41:06 <andythenorth> Expected a (literal) string
13:41:12 * andythenorth checks the generated nml
13:41:20 <andythenorth> it’s a very not-simple test case :(
13:41:57 <Alberth> it starts to sound like a bug, imho
13:42:38 * andythenorth would suspect more likely in FIRS, right now :P
13:42:43 <andythenorth> I need to make a simple test case
13:42:44 <Alberth> no, you'd get a "parse error" then :)
13:43:05 <Alberth> nml has a 'regression' directory with tests
13:44:02 <Alberth> no TTD_ text in there
13:46:53 <andythenorth> so I made a simpler test case which fails also
13:46:54 <andythenorth> string(STR_EXTRA_TEXT_SECONDARY, TTD_STR_CARGO_SINGULAR_COAL, string(STR_EXTRA_ALUMINUM_PLANT))
13:47:09 <Eddi|zuHause> i would expect you can only use TTD_STR_* by pushing it to the text stack
13:47:24 <Eddi|zuHause> not using the string(..., ...) method
13:47:34 <andythenorth> that is plausible, and might give a solution
13:55:28 *** maciozo has joined #openttd
13:57:25 *** welshdragon has joined #openttd
14:02:41 *** welshdragon has joined #openttd
14:06:38 *** welshdragon has joined #openttd
14:08:05 *** welshdragon has joined #openttd
14:20:30 *** welshdragon has joined #openttd
14:34:11 *** welshdragon has joined #openttd
14:37:53 *** welshdragon has joined #openttd
14:39:57 *** Guest1113 is now known as FR^2
14:53:05 *** umgeher_ has joined #openttd
15:14:46 *** welshdragon has joined #openttd
15:18:07 *** welshdragon has joined #openttd
15:21:49 *** sim-al2 has joined #openttd
16:20:39 <andythenorth> ach now I have to look up text stack syntax for nml :P
16:25:36 *** maciozo has joined #openttd
16:48:20 *** BluesInTheNet has joined #openttd
17:01:54 *** TheMask96 has joined #openttd
17:23:47 <Wolf01> Mid week holidays disturb me a lot
17:25:08 <Alberth> lack of mid-week holidays is also disturibing, every now and then :)
17:49:25 *** Progman has joined #openttd
18:00:15 *** andythenorth has joined #openttd
18:06:12 *** welshdragon has joined #openttd
18:18:06 *** andythenorth has joined #openttd
18:38:51 *** sla_ro|master has joined #openttd
18:51:49 *** welshdragon has joined #openttd
18:53:10 *** frosch123 has joined #openttd
19:09:47 *** JacobD88 has joined #openttd
19:14:07 *** Wormnest has joined #openttd
19:41:09 *** welshdragon has joined #openttd
19:45:44 <DorpsGek> Commit by translators :: r27684 trunk/src/lang/italian.txt (2016-12-09 19:45:36 +0100 )
19:45:46 <DorpsGek> italian: 1 change by lorenzodv
19:59:52 *** m1cr0man has joined #openttd
20:13:40 *** andythenorth has joined #openttd
20:18:11 <supermop> i still haven't picked a malware appliance yet
20:19:26 <frosch123> supermop: do you need a list of boat owners?
20:20:00 <supermop> my dad is lobbying for me to get a surface book, but 2700USD and still only 16gb ram is tough for my budget
20:20:00 <andythenorth> supermop: ‘many options, decide late’ is usually a good strategy
20:20:09 <andythenorth> or you just buy a mac, like me, because eh, choice is boring
20:20:11 <supermop> frosch123: sure why not
20:20:20 <supermop> andythenorth: that what the wife does
20:20:31 <andythenorth> it’s a much simpler form of consumerism
20:20:44 <andythenorth> particularly for phones, about which I have zero interest
20:20:58 <andythenorth> so I just buy ‘phone’ which = cheapest iphone with 32GB
20:21:23 <supermop> getting anything set up with autocad etc by week before christmas would be nice though, allowing me to work from 'home' in the week before new years
20:21:37 <supermop> andythenorth: pixel with 128 is nice
20:21:52 <supermop> where 'home' is possibly asia
20:22:43 * andythenorth prefers no research, no thinking, no angst about ‘what kind of person am I?’ or ‘what’s the most performance per £?’ :P
20:24:52 <supermop> over the past year ive used a home computer, when mine was still working, for 1)playing go, 2) reading tt-forums, 3) designing a few pieces of junk to 3d print
20:25:05 *** FLHerne has joined #openttd
20:25:08 <supermop> so really dont need much of a beast
20:29:06 <supermop> maybe ill just spend computer budget on gifts for wife
20:29:21 <supermop> and hope that she then buys me a computer
20:29:26 <andythenorth> buy wife new mac, use wife’s old mac
20:29:55 <andythenorth> install windows if you need it
20:32:26 <andythenorth> how do I build nmlc binary? o_O
20:32:31 <andythenorth> I thought it was just ‘make'
20:33:05 <andythenorth> I need an nmlc I can install in python
20:33:51 <frosch123> i thought you would just run /home/andy/nml/nmlc firs.nml
20:34:19 <andythenorth> yes I should do that
20:34:49 <andythenorth> but I still need to build that nmlc?
20:34:59 * andythenorth has serious Friday brain
20:35:18 <frosch123> or did you move to windows?
20:35:44 <andythenorth> oh that nmlc is just an entry point
20:35:48 <andythenorth> I should have opened it and looked :P
20:37:05 <andythenorth> I am being surrounded by brio train track construction right now
20:38:31 <frosch123> someone likes trains?
20:41:37 <andythenorth> I have to dig some more
20:41:38 <andythenorth> nmlc ERROR: "../lang/english.lng", line 5: Undefined command "TITLE"
20:41:49 <andythenorth> firs nml can’t be built without all the args
20:41:57 <andythenorth> which is why I was trying to install it :)
20:42:08 <andythenorth> (it=nmlc in my python bin dir)
20:42:22 * andythenorth will figure it out
20:43:44 <andythenorth> ach I need to give the path to custom_tags.txt
20:44:44 <andythenorth> ok, building now
20:46:45 * andythenorth must parent for a bit :)
20:50:47 <frosch123> hmm, is the string() stuff even meant for parameters?
20:51:00 <frosch123> it looks more like static concatenation of string
20:52:55 <frosch123> looking at extra_text_primary.pynml: the textstack stuff is done via STORE_TEMP, not via string
21:05:04 <frosch123> the previous construction from strings was only done for plain rail tiles, likely was weird for translations, and does not work for road/tram types :)
21:09:24 <frosch123> trashing 22 translations per language :)
21:10:14 <frosch123> hmm, does trashing translations actually work in eints?
21:10:27 <frosch123> i think it just keeps them as invalid
21:57:07 <andythenorth> it’s not a great user story for translators afaik
21:57:23 <andythenorth> the FIRS change I want to make will improve translations for the long term
21:57:37 <andythenorth> but in the short term, translations will be very broken, until translators fix them
21:58:14 <andythenorth> I can do this with text stack if I need to
21:58:36 <andythenorth> it’s just usually easier to start by concatenating strings, especially because they’re static
22:00:07 <Alberth> there is a minimum number of old translations that eints keeps, and a time out, iirc
22:00:15 <frosch123> i wonder how to trick eints
22:00:30 <frosch123> i could commit the script-fixed translations first
22:00:36 <frosch123> and then commit the baselanguage change
22:01:06 <frosch123> maybe that results in valid but outdated strings, so people could still look at them
22:01:11 <Alberth> reduce the numbers, restart, eints throws stuff in the bin, stop eints, revert numbers, start eints again
22:01:43 <Alberth> no idea what values you use, maybe it's ok already :)
22:02:56 <Alberth> oh, base strings change of course, it trashes deleted base strings afaik, no way to ever see them again
22:03:34 <Alberth> ie, it's all attached to the string name
22:04:20 <Alberth> upload language should know, I guess
22:07:50 <andythenorth> dunno whether to try switching to text stack
22:07:58 <andythenorth> or just put the cargo strings I need in the lang file
22:08:19 <frosch123> how many strings does firs need?
22:08:31 <frosch123> in the compiled output
22:10:40 <andythenorth> total string count? Or just for this one use case?
22:11:36 <frosch123> hmm, firs does not compile
22:12:37 <Alberth> it seems, after modifying the base language, it walks through all translations, and deletes all string names that are no longer available (pages/upload_language.py, lines 174
22:12:44 <frosch123> i guess you already pushed stuff :)
22:13:08 <frosch123> nmlc info: D0xx strings: 84/1024 <- ok, looks like there is no problem currently
22:13:25 <frosch123> i was worried that composing all the combinations would result in many strings
22:13:32 <andythenorth> DCxx strings: 236/256
22:13:40 <andythenorth> but that’s...ok?
22:13:49 <frosch123> those are cargo names :)
22:14:12 <andythenorth> that’s….possibly a problem soon
22:14:23 <frosch123> maybe also industry names, not sure
22:14:27 <andythenorth> especially if I duplicate the OpenTTD default strings
22:14:51 <andythenorth> I use 19 default cargo names
22:15:37 <andythenorth> also I pushed ‘not broken compile’ :|
22:15:40 <frosch123> Alberth: are you talking to me or to andy?
22:15:49 <frosch123> we kind of had two string discussions :)
22:18:17 <andythenorth> I don’t understand the result of using nml ‘string(ID, substring)’ w.r.t string count
22:18:38 <andythenorth> does each resulting string use up a string ID?
22:18:56 <frosch123> if you use STR_A and STR_B then string(STR_A, STR_B) adds a third string
22:19:11 <frosch123> it does not actually use the textstack, it statically composes a new one
22:19:14 <andythenorth> I was hoping there were magic bytes for ‘insert another string’
22:19:24 <andythenorth> I might switch to the text stack for this
22:19:35 <andythenorth> does the text stack have any performance costs?
22:19:53 <frosch123> none that you can measure
22:20:05 <frosch123> rendering the actual font will always take longer
22:20:09 <andythenorth> it will increase compile time marginally, there will be a lot more nml switches
22:22:59 *** Ethereal_Whisper has quit IRC
22:24:53 *** HerzogDeXtEr has joined #openttd
22:26:44 <DorpsGek> Commit by frosch :: r27685 /trunk/src/lang (56 files in 2 dirs) (2016-12-09 22:26:37 +0100 )
22:26:45 <DorpsGek> -Change: Preemptively invalidate translation changes of next commit for eints.
22:27:27 <DorpsGek> Commit by frosch :: r27686 /trunk/src (7 files in 2 dirs) (2016-12-09 22:27:22 +0100 )
22:27:28 <DorpsGek> -Change: List railtype of rail tiles explicitly in the tile info window.
22:27:38 <frosch123> let's see whether that tricked eints
22:28:58 <frosch123> i may get scolded by the farm for one commit, but the eints result is fine :p
22:35:05 * andythenorth must find nml text stack docs
22:36:33 <frosch123> pretty sure there are none :)
22:38:52 <andythenorth> so just store temp, and {STRING} (or with formatting)
22:39:17 *** Ethereal_Whisper has joined #openttd
22:39:32 <andythenorth> oh and I might have to stuff the dwords :P
22:39:37 <andythenorth> just like old times
23:01:31 <frosch123> road vehicle property 05 ?
23:01:48 <frosch123> would be the same number of for trains
23:03:21 <frosch123> did someone already check whether road compatibility works the same as for railtypes?
23:03:39 <frosch123> is railtype compatibility transitive?
23:04:29 <frosch123> so, if off-road vehicles could run on town roads, and normal vehicles could run on highway, would that imply that off-road vehicles could run on highways?
23:08:15 <Rubidium> but isn't normal road like the universal railtype?
23:08:47 <frosch123> venice road is pretty much incompatible to everything
23:09:15 <frosch123> anyway, looks like railtype poweredness is not transitive, so the same applies to roads
23:09:47 <andythenorth> type compatibility is always nuts :D
23:10:00 <andythenorth> the only way I could make sense of it in a grf was…
23:10:12 <andythenorth> …types aren’t compatible, except for el-[type]
23:11:23 *** Defaultti has joined #openttd
23:13:09 <Wolf01> Anything that can run on normal road could also run on any other road-like type, the inverse is not entirely true
23:13:39 <Wolf01> A truck should be able to run on dirt road too
23:14:30 <Wolf01> The problem might be if you define a tram type to be like a road
23:14:38 <frosch123> there was a story about some garbage truck driver, who get stuck in the middle of some forest
23:15:54 <frosch123> apparently he had an rendezvous during working hours in his "company car" and took a detour :)
23:17:33 *** sim-al2 has joined #openttd
23:27:07 <Eddi|zuHause> anyone ever wonder why germans sleep "out" and english sleep "in"?
23:28:09 <Eddi|zuHause> the name for when you start your day late
23:28:52 <Eddi|zuHause> that's when you didn't intend to...
23:29:25 <Eddi|zuHause> that's... always :p
23:30:24 <Wolf01> I had the habit to really wake up only after drinking the coffee at work, at least 1 hour after starting from home
23:31:44 <Eddi|zuHause> that's probably the opposite of "sleeping in"
23:51:09 <Wolf01> Eddi|zuHause, do you think urban games will release another patch for TF or they will just make TF2 with the fixes?
23:51:30 <goodger> they already did that
23:51:57 <Eddi|zuHause> i'm sure there will be fixes for about a year after release
23:52:13 <Eddi|zuHause> and then they're probably working on the next thing
23:53:03 <Wolf01> The last patch was really needed, but I still have those annoying performance drop for no reason at all :(
23:54:40 <Eddi|zuHause> it's a bit early to lose hope, i think...
23:54:51 <Eddi|zuHause> these types of fixes take some time
continue to next day ⏵
