IRC logs for #openttd on OFTC at 2013-02-26
⏴ go to previous day
00:36:45 <wakou2> Can someone point me to a simple guide to show me how to build a station with overflow. My trains wait at the last siganl on the entry rail. If I put signals on the platform entries, they jst 'stick a nose in' and bloc the whole line until THAT platform is free
01:01:37 <Eddi|zuHause> you need entry/exit signals for that, it won't work with path signals
01:05:47 *** Supercheese has joined #openttd
01:53:11 <wakou2> I have tried a few different variants with old stlye siganal, and have given myself a headache!
01:59:51 *** tracerpt has joined #openttd
03:38:49 *** Bad_Brett has joined #openttd
04:23:00 *** RavingManiac has joined #openttd
05:03:01 *** LordPixaII has joined #openttd
05:03:38 *** RavingManiac has joined #openttd
05:56:17 *** Eddi|zuHause has joined #openttd
06:37:40 *** valhallasw has joined #openttd
06:39:08 *** bo-31027_ is now known as bo-31027
06:42:56 *** Devroush has joined #openttd
06:47:41 *** oskari89 has joined #openttd
07:14:01 *** Twofish has joined #openttd
07:20:55 *** andythenorth has joined #openttd
07:22:46 <andythenorth> not only but also
07:43:58 *** RavingManiac has joined #openttd
07:44:15 <Supercheese> Soooo, bridges-over-stations, how's that going? Any patch for testing?
07:48:10 *** bo-31027 has joined #openttd
07:50:48 <oskari89> Bridges over stations, sounds nice :)
07:59:33 *** andythenorth has joined #openttd
08:07:57 <andythenorth> are all of those terms true? :)
08:08:04 <andythenorth> or is it a colloquilsiasdiadasm
08:10:12 <andythenorth> how do I do that?
08:10:32 <andythenorth> I can probably hack at it and do it in 3 or 4 passes, but that smells
08:10:48 <andythenorth> seems like Eddi|zuHause would be able to tell me one neat recursive function for it?
08:19:22 *** andythenorth has joined #openttd
08:23:29 <Supercheese> "Don't worry about minute grf license details" he says
08:25:18 <Supercheese> anyway, good night
08:41:55 *** chester_ has joined #openttd
08:59:18 *** HerzogDeXtEr has joined #openttd
09:11:32 *** HerzogDeXtEr1 has joined #openttd
09:17:34 *** andythenorth has joined #openttd
09:21:14 *** chester_1 has joined #openttd
09:27:40 *** HerzogDeXtEr has joined #openttd
09:39:46 *** tracerpt has joined #openttd
10:03:00 *** ChanServ sets mode: +v tokai
10:07:59 <tracerpt> think im going back to bed
10:11:42 <Eddi|zuHause> so why is andythenorth never here?
10:15:07 <peter1138> well he was an hour ago
10:16:52 <Eddi|zuHause> yes, and yesterday. and both times he asked a question to me and was gone by the tiem i read it
10:25:45 *** andythenorth has joined #openttd
10:27:16 <Eddi|zuHause> anyway, the "shortest" result i can come up with for his question: "tmp = flatten(input); for i in range(0,len(tmp): output["%d..%d"%(tmp[i],tmp[i+1]-1)] = [j for j,r in enumerate(input) if tmp[i] in range(r*)]"
10:27:55 <Eddi|zuHause> where "flatten" is something left to implement that transforms a nested list into a single list, with sorting and erasing duplicates
10:29:07 <Eddi|zuHause> then you get a dict out of it with {"x..y": [0,1], "z..a" [1]} etc.
10:29:25 <Eddi|zuHause> or with a slight modification you can get a list of strings
10:29:48 <Eddi|zuHause> oh, there he is :)
10:30:46 <Eddi|zuHause> i guess it needs to be range(*r)
10:44:06 <andythenorth> input is similar to [[0, 9999], [1900, 1990]]
10:45:58 <Eddi|zuHause> you have a strange way of appending to a list :p
11:02:33 <andythenorth> I could probably have used a comprehension, but my brain is a mess
11:03:00 <andythenorth> I don't like this code
11:03:07 <andythenorth> otoh it probably works :P
11:05:09 <andythenorth> dict.keys() might have arbitrary order?
11:27:36 <Eddi|zuHause> but you could also use an actual list instead of abusing a dict :p
11:35:28 <andythenorth> I write terrible code when I'm sick :P
11:35:41 <andythenorth> and probably when I'm not sick :P
11:38:46 <andythenorth> Eddi|zuHause: feel free to refactor :)
12:32:40 *** RavingManiac has joined #openttd
12:33:58 *** RavingManiac_ has joined #openttd
12:43:38 *** goodger has joined #openttd
13:09:01 *** Bad_Brett has joined #openttd
13:12:23 *** RavingManiac has joined #openttd
13:16:34 *** Extrems has joined #openttd
13:45:27 *** Twofish has joined #openttd
13:53:04 *** RavingManiac has joined #openttd
14:06:29 <Eddi|zuHause> andythenorth try: reduce(set.union, input, set())
14:06:49 <Eddi|zuHause> where input is your list of list
14:07:36 <Eddi|zuHause> this gives you the list of all trigger dates
14:07:45 <Eddi|zuHause> with removed duplicates
14:10:48 <Eddi|zuHause> and then: for date in triggers: for counter. (start, end) in enumerate(input): if date in range(start, end): output[date].append(counter)
14:11:20 <Eddi|zuHause> misses an output[date]=[]
14:16:43 <andythenorth> bah, can't believe I forgot about enumerate :(
14:19:31 <Eddi|zuHause> the inner loop could be simplified by a list comprehension
14:20:31 <Eddi|zuHause> so: for date in triggers: output[date] = [counter for counter, (start, end) in enumerate(input) if date in range(start, end)]
14:20:49 <Eddi|zuHause> andythenorth: that should be a comma
14:21:37 <andythenorth> excellent, that's perfect :)
14:22:37 <andythenorth> Eddi|zuHause: ever play python code golf? o_O
14:22:49 <Eddi|zuHause> i've seen people play it
14:23:02 <Eddi|zuHause> but sometimes "short" != "elegant" :)
14:23:15 <andythenorth> there should be python code ballet :P
14:26:18 <Eddi|zuHause> one time in school in a test, my teacher was impressed because my solution was so much shorter and cleaner than his default solution that he used mine instead :)
14:27:04 <Eddi|zuHause> (that was in pascal, and the question was something like "implement a function that applies word wrap at width x")
14:27:47 <andythenorth> did it have split() and join() ? :P
14:28:19 <Eddi|zuHause> but next(' ') or something
14:39:27 <Eddi|zuHause> andythenorth: pascal did not have dynamic arrays/lists, only arrays with fixed length
14:45:58 <andythenorth> that must have been fun
14:46:08 <andythenorth> so to append, did you copy the old array into a new array?
14:53:23 <Eddi|zuHause> back then, if you wanted to "append" things, you used linked lists instead of arrays
14:55:17 <blathijs> Just like now, except that you don't need to implement the linked list yourself nowadays ;-)
14:57:15 <andythenorth> V453000 you should do an RV set :P
14:57:46 <andythenorth> nodwells caused FIRS
14:57:52 <andythenorth> I needed a cargo for them to transport
14:57:55 <andythenorth> so I invented ensp
14:58:02 <andythenorth> then I had to invent the rest of FIRS
14:58:13 <andythenorth> irony: I never added the nodwells to HEQS :P
14:59:53 <Eddi|zuHause> should have called it PIVO :p
15:00:53 <V453000> you can code them as trains?
15:01:14 <andythenorth> code everything as trains
15:01:21 <andythenorth> I should stop talking about it, start doing it
15:01:30 <Eddi|zuHause> code them as hovercrafts that go on snow instead of water :p
15:01:34 <andythenorth> then we could drop the other transport types
15:01:43 <andythenorth> V453000: you know your ships can go on sea?
15:02:08 <andythenorth> you just need to use the canals trick
15:02:23 <V453000> if you put water/canal objects around the rails, yes
15:03:05 <V453000> but diagonals are going to be a bit eww
15:05:32 <andythenorth> diagonals will smell
15:05:35 <andythenorth> we need diagonal canals
15:06:03 <V453000> I guess 1 tile cant have a diagonal object and a diagonal rail on the other half
15:06:11 <Eddi|zuHause> for the canal-tracks, you only need to put two diagonal tracks on the tile
15:06:17 <V453000> But what you can have is 2 diagonal water-rails on 1 tile
15:06:28 <V453000> yeah but they include buoys for the "rail"
15:07:42 <Eddi|zuHause> i suppose using fences won't work either
15:08:33 <V453000> looks like no fence is created when next co canal
15:08:35 <Eddi|zuHause> shouldn't the buoys have different colour on the left and right side?
15:08:51 <Eddi|zuHause> fences check who owns the neighbouring tile
15:09:14 <V453000> different colours, dont know :)
15:09:48 <V453000> felt like r/w and g/y makes the junctions somewhat orientable
15:13:27 <V453000> but well, the purpose isnt to make these things buildable on water anyway
15:14:10 <Eddi|zuHause> i have no clue what the purpose is anyway :p
15:14:56 <V453000> just wtf value :P also I am kind of known for hating ships, sooo ... :)
15:15:43 <V453000> this makes "ships" viable
15:16:48 <Eddi|zuHause> V453000: so you buy "bays" in the depot, and it constructs and appropriately sized ship out of the bay sprites? :)
15:17:17 <Eddi|zuHause> then use CETS turning magic to keep the sprites together :)
15:17:31 <V453000> it is just a double headed train with wagons
15:17:54 <V453000> 4 wagons in this case
15:19:07 <Eddi|zuHause> might want to start with shorter wagons :)
15:19:48 <V453000> not sure about that, maybe
15:21:12 *** ntoskrnl has joined #openttd
15:53:58 *** RavingManiac has joined #openttd
15:54:37 *** andythenorth has joined #openttd
15:59:49 <Eddi|zuHause> as is everybody else who joins here :p
16:01:47 <Eddi|zuHause> denying you're crazy supports the fact that you are crazy
16:02:28 <V453000> I aint crazy, I iz NUTS
16:11:43 *** Kurimus has joined #openttd
16:17:02 *** Prof_Frink has joined #openttd
16:25:54 *** Stimrol has joined #openttd
16:51:16 *** HerzogDeXtEr has joined #openttd
16:52:24 <andythenorth> if date != 9999: # shonky magic special casing for end dates
16:52:36 <andythenorth> that's how andythenorth codes :)
16:54:16 <Eddi|zuHause> why have an end date anyway, just use the default clause for infinite?
16:55:16 *** tracerpt has joined #openttd
17:06:24 <andythenorth> Eddi|zuHause: I'm sure there's a better way, I just can't see it :)
17:07:10 <Eddi|zuHause> andythenorth: anyway, the maximum year is 5000000
17:07:14 *** ZxBiohazardZx has joined #openttd
17:07:58 <andythenorth> so anyone playing with vehicles never expire will get default graphics after 9999
17:08:05 <andythenorth> there must be a better way than this
17:08:07 <Eddi|zuHause> so put in 5000001
17:08:19 <andythenorth> well it just smells all wrong
17:08:20 <jonty-comp> overflow the data type
17:08:25 <andythenorth> I should figure out how to do it
17:08:51 <V453000> andythenorth: note that I often test graphics at 19200 :D
17:09:04 <V453000> ... normally I start game at 1920
17:11:23 <andythenorth> reverse slice last item of list
17:19:00 <andythenorth> eliminated 1 of 2 silly max year checks
17:19:28 *** TheMask96 has joined #openttd
17:21:38 <andythenorth> and 2nd 9999 never shows up in the compiled grf, it's just a config file thing
17:50:00 <andythenorth> I have no way to support different kinds of smoke
17:50:28 <Eddi|zuHause> you mean like setting the smoke effect by cb36?
17:51:17 <Eddi|zuHause> maybe you want cb10?
17:53:00 <andythenorth> I mean that I have engineered my config file system with no thought of smoke :)
17:53:15 <andythenorth> ship graphics can vary by date / random
17:53:24 <andythenorth> including steam / diesel variants
17:53:31 <andythenorth> smoke is lame anyway currently :P
17:54:54 <andythenorth> ach, I'll just pack more values in to my nested lists, and unpack them :P
17:55:00 <andythenorth> all will be well
17:55:23 <Eddi|zuHause> just attach a smoke effect to the graphics?
17:55:56 <Eddi|zuHause> NGC3982: different base graphics set?
17:56:37 <NGC3982> I have no idea. I have never seen that before.
17:56:41 <Eddi|zuHause> NGC3982: so? just change the setting
17:56:52 <NGC3982> When was it changed, and by who? Me?
17:57:21 <NGC3982> By NewGRF or something?
17:57:50 <Eddi|zuHause> most likely by downloading it
17:58:13 <NGC3982> Oh, now i know. I downloaded * from Bananas yesterday.
17:58:20 <NGC3982> I guess baseset's turn on by default
17:58:42 <Eddi|zuHause> it picks a random one if you didn't specify a specific one yet
17:59:28 <Eddi|zuHause> by "random" i mean the "return 4;" kind of random :)
17:59:49 *** Alberth has joined #openttd
17:59:49 *** ChanServ sets mode: +o Alberth
18:00:13 <NGC3982> The "nothing can ever be random" discussion is filled with poop.
18:01:19 <Eddi|zuHause> it's better than the "nothing can ever be infinite" discussion at any day :p
18:02:56 <Alberth> not even the disccusion itself? :p
18:04:23 <Eddi|zuHause> i never really understood it, but apparently there are people that say "no (physical) object can be infinite, so you cannot talk about infinity in maths either"
18:06:14 <Alberth> ah, just like the idea of constructive mathematics, where you can only show things to exist by giving an constructive algorithm how to create instances
18:06:15 <NGC3982> The big problem is trying to use things like that in a normal discussion
18:06:46 <NGC3982> If i - on IRC, talking about trains - use the word "random", if of course mean random as in "It's really, really hard to know", and nothing else.
18:06:59 <NGC3982> And simply caring about something else is semantics, wrong or not.
18:07:02 <NGC3982> And i hate that so bad.
18:07:07 <Eddi|zuHause> Alberth: yes, similar to the "i reject the axiom of choice" debate, which is mostly religious anyway
18:07:23 <NGC3982> Eddi|zuHause: ..And pretentious.
18:07:51 <Alberth> NGC3982: it does depend on what you see as normal :) I have often discussions that use abstract mathematical concepts :)
18:07:57 <Eddi|zuHause> the difference is that without the axiom of choice, there's still plenty of maths you can do
18:08:04 <Eddi|zuHause> but if you reject infinity, ...
18:09:14 <Alberth> you can only do finite things, obviously :D
18:09:18 * NGC3982 fills his stomach with infinities of sharon fruit.
18:09:45 <Eddi|zuHause> yes, but you can express infinity in a finite number of axioms :)
18:11:21 <Eddi|zuHause> like "a subset of the natural numbers is called infinite, if there is no upper bound". then "all even numbers" is an infinite set. i said that with a finite number of words.
18:12:44 <Alberth> I was already convinced :)
18:13:20 *** oskari892 has joined #openttd
18:13:24 <Eddi|zuHause> yes, almost everybody is convinced, except a handful of zealots who then poison any discussion about cantor's set theory :p
18:15:41 *** Progman has joined #openttd
18:23:03 <andythenorth> Eddi|zuHause: I could encode the smoke effect in the graphics filename :P
18:23:17 <andythenorth> or I could encode it in the graphics themselves, using colour :o
18:23:30 <andythenorth> that's actually wholly plausible
18:23:54 <andythenorth> hmm, I had a thought for Alberth and now I've lost it :(
18:23:59 <Eddi|zuHause> make the first pixel in one of the magic pink colours :p
18:24:02 <michi_cc> Eddi|zuHause: So what are those people answering if you tell them that some infinities are more infinite than others? ;)
18:24:33 <Eddi|zuHause> michi_cc: yes, that is cantor's set theory :p
18:25:14 <andythenorth> Alberth: I did think that bottle could be used as the basis of a newgrf maker, where players edit via their browser, but that wasn't what I wanted to ask you / tell you :P
18:25:23 * andythenorth is not feeling so good today, brain addled
18:25:39 <Eddi|zuHause> michi_cc: but basically it boils down to: "why are you still discussing this? there is no infinity. period."
18:26:55 <andythenorth> Alberth: it wasn't this either, but... "Feature: remove all instances of a string from all lang files" ?
18:27:41 <Alberth> upload a new base language file without the string?
18:27:52 <michi_cc> Eddi|zuHause: I'm an engineer, not an mathematicing. I have no clue at all how all those axioms are called, we just use them and are done :)
18:27:54 <andythenorth> good idea, might be worth documenting :)
18:28:24 <Alberth> but yeah, I have not really considered what type of changes you might want to do on the strings
18:28:25 <andythenorth> Alberth: I did improve the string edit layout on Sunday night, and add error class handling
18:28:35 <andythenorth> I also solved, sufficiently, the passing template message
18:29:26 <andythenorth> it smells of a hack, but I think it's a valid hack
18:29:29 <Eddi|zuHause> michi_cc: well cantor's set theory starts with "is there a set of all sets?" and then "is the set of all sets element of itself?" resulting in "there can't be a set of all sets, so there must be a nested hierarchy of sets"
18:29:46 <andythenorth> also there are error classes available for success / warning / info / failure
18:30:27 <Eddi|zuHause> and from that on you get a hierarchy of infinities, by starting with the natural numbers as the base infinity, and the power-sets as the hierarchy
18:30:33 <andythenorth> herp, it's annoying to forget a thought :|
18:30:56 <Eddi|zuHause> and to decide whether there are more kinds of infinity than in this hierarchy, you need the axiom of choice
18:33:30 <Alberth> oh, if it's any good, it'll come back. Otherwise, it was not good enough anyway :)
18:33:38 *** goodger has joined #openttd
18:34:10 <Eddi|zuHause> that's the worst answer to a forgotten thought ever :p
18:34:24 <andythenorth> but it doesn't stop the annoyance
18:34:40 <Eddi|zuHause> it's like "i had a proof of fermat's last theorem, but i forgot it" - "then it wasn't any good" :p
18:34:48 <andythenorth> Alberth: I was wondering how close WT was to done
18:36:39 <Alberth> I have 3 points to do. 1) fix doing a translation (coded but not tested) 2) cleanup old changes (currently they are kept infinitely, if you believe such a thing exists) 3) some form of user admin
18:36:56 <Alberth> where the 3rd one is a bit of a problem
18:37:32 <Alberth> then you need some documentation for authors and translators
18:37:51 <Alberth> then it is time to look around what else is missing
18:38:24 <Alberth> probably stuff is missing from a practical use point of view
18:38:44 *** frosch123 has joined #openttd
18:45:21 <DorpsGek> Commit by translators :: r25049 /trunk/src/lang (lithuanian.txt unfinished/thai.txt) (2013-02-26 18:45:13 UTC)
18:45:22 <DorpsGek> -Update from WebTranslator v3.0:
18:45:23 <DorpsGek> lithuanian - 13 changes by Stabilitronas
18:45:24 <DorpsGek> thai - 12 changes by angelix
18:54:46 *** Supercheese has joined #openttd
18:55:30 *** oskari89 has joined #openttd
19:03:31 *** supermop has joined #openttd
19:06:12 *** andythenorth has joined #openttd
19:06:28 <andythenorth> Alberth: I'm thinking maybe we let it loose on other people?
19:06:40 <andythenorth> either they run it themselves...or we ask about devzone hosting
19:07:16 <andythenorth> assuming anyone is interested :)
19:10:04 <Alberth> yes, after 1 + 2, imho
19:12:15 <andythenorth> also the project overview screen sucks right now, layout wise :)
19:12:19 <andythenorth> I should fix that
19:14:52 <andythenorth> also the language view - might benefit from an explicit 'start' fixing button for each status group
19:15:42 <planetmaker> well, of course the devzone offers hosting :-)
19:15:55 <planetmaker> or rather: also using :D
19:16:21 <andythenorth> planetmaker: want to try setting up WT there? o_O
19:16:29 <andythenorth> it appears to have little or no deps
19:16:37 <andythenorth> but I'd run it in a vm, it might be insecure :)
19:17:55 <andythenorth> Alberth: the idea I lost earlier - the ability to mark some strings in base lang as 'do not translate'
19:18:10 <andythenorth> there may be other better ways to do that
19:18:19 <andythenorth> it's not a common case, but occasionally comes up
19:18:29 <Alberth> quickly, make a feature request :)
19:19:20 <Alberth> hmm, I had a discussion with y3xo about this, iirc he was not so happy adding such a feature to the file format, but I need to look that up
19:19:51 <andythenorth> not sure it's valid yet
19:20:27 <Alberth> there is a route to /fix/<proj_name>/<lname> to select a random string to fix
19:21:04 <Alberth> it does have fix priorities internally, but these are constant currently
19:22:10 <Alberth> isn't that a custom_tag thingie?
19:22:46 <Alberth> or do you use planetmakers type of custom_tags, entirely generated ?
19:24:47 <Eddi|zuHause> andythenorth: iirc the conclusion was to put that in custom_tags.txt
19:25:29 <Eddi|zuHause> andythenorth: but potentially a website could have a language parameter or something :)
19:25:39 <andythenorth> I don't mind moving strings
19:25:51 <andythenorth> I hadn't understood the purpose of custom tags
19:26:17 <Eddi|zuHause> andythenorth: the makefile autogenerates custom_tags.txt to include the version/revision number
19:33:24 <andythenorth> so what do we think the right thing to do is?
19:33:43 <andythenorth> I could just leave these strings in place, and let translators do whatever
19:33:47 <andythenorth> it would be a nice hack btw
19:34:03 <andythenorth> I don't review translations, so just put the url to goatse or porn or something
19:34:56 *** Flygon_ has joined #openttd
19:36:20 <Eddi|zuHause> no, you mean LCARS!
19:40:39 <oskari89> Light Almighty Rail Car Set?
19:40:55 <oskari89> That would be nice, generic cars on rails
19:41:08 <Belugas> Larks Tongues In Aspic
19:41:36 <Supercheese> No one got the Star Trek reference? :<
19:42:02 <ZxBiohazardZx> didnt want to mention
19:43:42 <peter1138> that's not really a reference
19:43:46 <andythenorth> Supercheese: too busy not talking about licensing to think about your Star Trek obsession
19:44:11 <peter1138> it's like saying "no, you mean DARTH VADER" and calling that a starwars reference...
19:44:26 <Supercheese> well, it is, isn't it?
19:44:45 <Supercheese> tad more subtle I suppose
19:44:55 <Eddi|zuHause> Supercheese: you expect people to shout out "YES I GOT THE REFERENCE"?
19:49:48 <Supercheese> When it doubt, put it on a shirt and sell it
19:51:01 <andythenorth> I GOT THE REFERENCE
19:51:54 <Eddi|zuHause> i really don't think there are a lot of people here who never heard of LCARS before
19:54:54 *** Flygon__ has joined #openttd
20:09:39 *** Flygon__ has joined #openttd
20:10:23 <andythenorth> wtf, I can actually use a valid definition list
20:10:28 <andythenorth> never done that before, in 15 years
20:11:23 <Alberth> valid definition of what?
20:12:00 <Alberth> btw shall we disable the reloader?
20:12:09 <andythenorth> module reloader yes
20:12:19 <andythenorth> definition is of string statuses
20:12:30 <andythenorth> I can use <dl><dt><dd> for the first time ever
20:15:04 <Alberth> ah, 'download' is like the download button at a language page of course
20:15:54 <andythenorth> we should store those definition strings as content somewhere
20:15:56 <andythenorth> and make them available
20:16:01 <andythenorth> anywhere in the app
20:17:02 <Alberth> why? they don't need to be translated, or are you talking about other definitions than custom_tags now?
20:17:40 <andythenorth> other definitions
20:17:48 <andythenorth> "Unknown: means the state of the translation was not decidable"
20:17:59 <andythenorth> that is reusable copy within the web translator
20:18:06 <andythenorth> could be used as help text in any page
20:18:18 <andythenorth> I'll probably just stuff it in a dict somewhere
20:18:28 <andythenorth> there is a bad pattern, called:
20:18:42 <andythenorth> "andy puts everything in utils.py because he doesn't know where it should go"
20:19:55 <Alberth> looks good as temporary storage :)
20:20:57 <Alberth> datetime stuff fits better near Stamp, but there are some other pieces to add there as well
20:22:53 <planetmaker> so... you need a VM for that stuff?
20:23:00 <andythenorth> I reckon it's safer
20:23:06 <andythenorth> I wouldn't run it on the bare OS
20:23:31 <andythenorth> it offers file upload to the filesystem, and could be full of vulnerabilities :)
20:23:39 <planetmaker> well. nothing on the devzone except the HV runs on "the bare OS" ;-)
20:24:05 <planetmaker> but yes... neither the DevZone VM might be best choice. So new one
20:24:24 <Alberth> planetmaker: it needs some form of user authentication first, imho
20:25:01 <planetmaker> Alberth, yes... likely sensible :-)
20:25:13 <planetmaker> so you don't need it now... but soon
20:27:06 *** swissfan91 has joined #openttd
20:31:38 *** Flygon_ has joined #openttd
20:31:59 <andythenorth> Alberth: can you pull, then tell me where to find the correct identifiers for statuses (keys for dict rendered in project overview)?
20:33:14 <Alberth> webtranslate.data.STATE_MAP
20:34:12 <Alberth> pull is going to be tricky atm :)
20:34:32 <andythenorth> ho, some of them have dashes
20:34:41 <andythenorth> it freaks me out that this is never a problem :)
20:34:57 <andythenorth> I got habituated to underscores everywhere in variable names etc
20:35:31 <andythenorth> for a key, am I better using UP_TO_DATE or up-to-date?
20:35:52 <CornishPasty> andythenorth: Use up_TO-datE
20:36:07 <CornishPasty> Make sure it's case-sensitive too
20:36:10 <andythenorth> mixed bulgarian-latvian-genoan notation
20:36:20 <CornishPasty> doesn't it need to be like
20:36:32 <Eddi|zuHause> and then have different cases mean different variables :p
20:36:38 <CornishPasty> Eddi|zuHause: Yup
20:36:42 <andythenorth> or have some mean objs
20:36:44 <andythenorth> others mean vars
20:36:50 <andythenorth> others are methods
20:36:53 <CornishPasty> booThatCouldBeAnIntUp-TO_DatE
20:37:11 <CornishPasty> boolUp_TO-_D-Ate = 6
20:37:33 <peter1138> true/false/yes/no/file_not_found
20:38:32 *** ZxBiohazardZx has left #openttd
20:38:42 <CornishPasty> peter1138: What about the edge case? file_was_found?
20:38:59 <CornishPasty> Which actually returns when the filesystem is corrupted
20:39:52 <Alberth> andythenorth: those are strings, not variables
20:40:30 <Eddi|zuHause> i'm always amazed at the totally constructive error message windows gives when there's a ':' in the filename :p
20:40:32 <andythenorth> ok, so I was reading the code that returns state to string_edit for exmple
20:40:35 *** zeknurn has joined #openttd
20:40:52 <andythenorth> seems like what I'm doing should be reused elsewhere
20:40:53 <Eddi|zuHause> you can neither open the file nor rename it then :p
20:41:25 <Eddi|zuHause> but it fails to explain why
20:41:36 *** Flygon_ has joined #openttd
20:41:37 <Alberth> Eddi|zuHause: you must truly love such a system :p
20:42:12 <andythenorth> Alberth: I'm going to do what I think is right, and hope you can help fix later ;)
20:42:16 <peter1138> file_might_be_there_but_good_luck_getting_to_it
20:43:29 <Eddi|zuHause> and because this error appears only every odd year, you search for a few minutes where the problem actually is, before you remember
20:48:57 <DorpsGek> Commit by rubidium :: r25050 trunk/Makefile.grf.in (2013-02-26 20:48:50 UTC)
20:48:58 <DorpsGek> -Fix: use the CC_BUILD compiler for preprocessing the extra GRF's source instead of a hardcoded gcc
20:48:59 <DorpsGek> -Fix: do not let gcc include files from the "standard C" include directories; newer gcc/libc seem to otherwise automatically include some header files at the top of the preprocessed nfo files which causes NFOrenum/GRFcodec to make invalid assumptions about the NFO version
20:49:14 <Rubidium> ... or in other words, effectively an one line fix
20:50:37 <Kjetil> the commit message is longer than the diff ? :P
20:50:59 <Eddi|zuHause> that's more common than you think :p
20:51:19 <frosch123> yay, we are approaching a symmetric revision number
20:51:52 <Rubidium> really? I'd reckon we already passed it 25 revisions ago
20:51:54 <Kjetil> I guess I should't speak to loudly. Commited changes to code comments earlier today
20:53:13 <Eddi|zuHause> frosch123: i really hoped i wouldn't need to save/load anything :/
20:53:18 <frosch123> 25252 will be fun :)
20:53:39 <Eddi|zuHause> but yes, i didn't consider base grfs
20:54:15 <frosch123> Eddi|zuHause: breaking savegames by disabling grfs is already fun, breaking savegames by changing the order of grfs even more
20:54:43 <Rubidium> frosch123: IMHO 25025 is more symmetric than 25052 ;)
20:54:47 <frosch123> i would expect grfs to use the random seed for quite troublesome stuff, so it should not change unexpectedly
20:54:49 <Eddi|zuHause> frosch123: i don't understand the difference between patch vars and global vars
20:55:30 <Kjetil> In the year 2525 if man is still alive
20:56:01 <Eddi|zuHause> Rubidium: if you have a 7-segment digital display, 25025 would be both even and odd symmetric :)
20:56:26 <Rubidium> Eddi|zuHause: exactly what I was thinking about
20:58:08 <Eddi|zuHause> frosch123: but wouldn't that mean you first have to copy it to a parameter to use it in an action2?
20:58:26 <Eddi|zuHause> how is that better then?
20:58:55 <frosch123> i have no idea what you want to do with that var. but other vars of that type are patchvars
20:59:12 <frosch123> and i think in nml it makes no difference
21:00:53 <frosch123> Eddi|zuHause: btw. ais ans gs have the option to request random values in their parameters
21:01:07 <frosch123> so, you can also make it an a14 feature, and store the values in the grfconfig
21:01:23 <frosch123> then a grf can take multiple random values just as it likes
21:01:27 <Eddi|zuHause> that sounds interesting
21:01:37 <frosch123> and i like the idea of unifying ais, gss and grfs
21:02:08 <Eddi|zuHause> i guess that means ditch this, and go back to the drawing board
21:04:04 <Eddi|zuHause> so a new setting type "random", which will not be displayed in the gui, but fill the associated parameter bits with random values?
21:11:41 <andythenorth> Alberth: can the base language have errors?
21:11:45 <andythenorth> or are all strings 'correct'
21:12:31 <andythenorth> I assume always correct
21:12:37 <Alberth> normally they are correct
21:12:53 <andythenorth> I can't think how they are 'uncorrect'
21:12:58 <andythenorth> what would you compare against?
21:13:31 <Alberth> you can have problems in the form of adding illegal strings, like "{BLECK" but they should get blocked from being uploaded
21:14:01 <Alberth> or stuff like "{P x y}"
21:14:16 <Alberth> ie no command to derive plural from
21:14:48 <Alberth> or doing that in a language without plurals :p
21:21:30 <frosch123> he, suddenly looks professional :o
21:21:31 <andythenorth> I suspect that using this to detect if a language has errors (or not) is very silly: if counts[1] == bcounts[1]:
21:21:46 <andythenorth> would be nice to have a flag for errors / not
21:22:16 <andythenorth> I just build on the shoulders of giants :)
21:22:29 <frosch123> i have no idea about that stuff, so i am probably impressed by a few lines of css :p
21:22:56 <andythenorth> ^ I just apply styles from there :P
21:22:58 <frosch123> though i wonder what happened the the srif project
21:23:00 <andythenorth> and maybe a bit of brain
21:23:10 <andythenorth> this web translator needs to be finished :)
21:23:18 <andythenorth> FIRS translations are becoming a PITA
21:25:10 <andythenorth> hmm table row colouring for success / not :P
21:25:14 <andythenorth> probly not needed
21:26:13 <Rubidium> andythenorth: a nice 'progress' bar per language could give a clue about the language's state
21:26:19 <andythenorth> that's intriguing
21:27:05 <Rubidium> or rather, a 90 degree rotated stacked bar chart
21:28:17 <frosch123> better make the table sortable :p
21:28:44 <frosch123> oh yeah, filter for > 50% or so to fake statistics :)
21:28:53 <andythenorth> sortable and filterable can be done
21:28:58 <andythenorth> fake statistics I delegae
21:30:05 <frosch123> anyway, don't express the completion state with 5 digits precision, or i'll flame you like i did tb back then :p
21:30:47 <Rubidium> anyhow, isn't "fake statistics" a tautology?
21:31:08 <frosch123> no, it's an emphasis
21:31:29 <frosch123> "the statistic is faked" would be a tautology
21:31:29 <Rubidium> like "real statistics" exist ;)
21:37:06 <Alberth> andythenorth: the order of the explanations is weird compared to the order of the columns ?
21:37:47 <andythenorth> now you want a pony?
21:38:03 <andythenorth> can you make python dicts return in a predictable order? o_O
21:39:10 <Eddi|zuHause> sorted(dict.values)
21:39:19 <andythenorth> yes, that's predictable :)
21:39:29 <andythenorth> I should ask better questions
21:41:10 <Eddi|zuHause> well you can provide an ordering function, to change the order
21:41:22 <peter1138> 8~9999999999999999999999999900000000000000000000000000000000000000000-=69o99999998i8888888888888887778888
21:41:33 <SpComb> how's that new keyboard going?
21:41:39 <Eddi|zuHause> is that a password? :p
21:41:50 <peter1138> 8 was physically sticking down
21:42:02 <peter1138> i striped it all and cleaned it up
21:42:08 <peter1138> plus it's nice & clean
21:42:15 <SpComb> need to yank it off and clean it out
21:42:33 <peter1138> i soaked the plastic parts for a couple of hours
21:42:53 <frosch123> oh, at my university there was a keyboard in some computer pool.... when you pressed the 5 on the numpad, all nine keys went down :s
21:42:55 <Eddi|zuHause> one day when i came into an office, the person said "my computer behaves woird when it starts up, it asks me something about harddisks and stuff"
21:43:03 <peter1138> nah, 8 is still a bit dodgy :S
21:43:09 <Eddi|zuHause> turned out the F12 key was stuck :p
21:43:18 <Eddi|zuHause> so it displayed the boot menu
21:44:40 <andythenorth> I really like this templating language btw
21:45:57 <andythenorth> it would be good for templating nml
21:48:30 <andythenorth> "Do not render untrusted templates! They may contain and execute harmful python code."
21:48:42 <andythenorth> not good for a web translator :)
21:51:52 <Eddi|zuHause> andythenorth: nah, it can't be worse than SQL injection :p
21:52:05 <andythenorth> probly about the same :)
21:52:24 <andythenorth> but I think we'd have to do something stupid
21:52:34 <andythenorth> like eval(input_from_untrusted_user)
21:53:42 <andythenorth> more powerful than python's Template module
21:53:57 <andythenorth> but doesn't leave xml crap everywhere like chameleon
21:55:01 <andythenorth> and it has a simple %include syntax
21:56:07 <Eddi|zuHause> and you find a new templating language every other week, so what?
21:56:49 <andythenorth> so nothing in my case
21:57:01 <andythenorth> I wouldn't recommend chameleon to most people though
21:57:13 <andythenorth> not for templating nml anyway :P
21:57:22 <andythenorth> I used it because I know it
21:58:08 <andythenorth> oh we do do we? :)
21:58:27 <andythenorth> I had to add an order-control list, which I hate as a pattern
21:58:40 <andythenorth> could have done one of those weird mult-dict constructs
21:58:43 <Alberth> I used it in newgrf/language_file
21:59:06 <Alberth> or are they used elsewhere too?
21:59:13 <andythenorth> I want to use them elsewhere
21:59:29 <andythenorth> you use STATE_MAP to provide mappings to strings?
21:59:39 *** valhallasw has joined #openttd
21:59:41 <andythenorth> I think that should be replaced or unified with what I've done
22:00:01 <Alberth> sounds like a good idea :)
22:00:14 <andythenorth> then I can use the Title attribute for display
22:00:22 <andythenorth> and I can set a tooltip which shows the description
22:00:26 <andythenorth> everywhere they are used
22:00:36 <andythenorth> and I can call .lower() on them if they are in a sentence :P
22:00:40 <Alberth> I changed string editing, it needs adding a message
22:00:50 <andythenorth> is it committed?
22:01:17 <Alberth> and you have a 'needs_fixing' flag in the project page :)
22:01:49 <andythenorth> so what message is needed?
22:03:07 <Alberth> at line 385 of pages/string_edit.py, you are thrown back onto the same string page as you still have errors in that string
22:04:03 <Alberth> line 304 is the other call of that function, which is when you start editing a string for the first time
22:04:42 <Alberth> you also had a comment on the language page, didn't you?
22:05:09 <andythenorth> I don't recall :)
22:05:18 <andythenorth> it was a long day :)
22:06:59 <Alberth> I think the actual current text should be added there
22:07:13 <andythenorth> where has my nasty hack with request.query['message'] = foo gone?
22:07:20 <andythenorth> did you find a better way?
22:08:08 <Alberth> where did you put that?
22:13:35 <andythenorth> maybe I imagined it
22:15:24 <andythenorth> forgot reload is off :)
22:16:12 <andythenorth> Alberth: so to add a message when returning a template
22:16:13 <andythenorth> request.query['message'] = 'blah'
22:16:16 <andythenorth> before the return
22:16:32 <andythenorth> I'll test it for string edit
22:17:12 <Alberth> Added 4 bugs for eints that need being done
22:20:22 <andythenorth> one more commit from me
22:22:16 <andythenorth> 22:22, so probably
22:22:18 <Eddi|zuHause> just... one... more... commit
22:22:37 <andythenorth> no more nice numbers today, so bedtime
22:22:50 <Eddi|zuHause> what about 23:45?
22:23:03 <andythenorth> but I'm not staying up for it :P
22:23:13 <andythenorth> Alberth: l385-6 show a quick hack for error
22:23:19 <andythenorth> I'm not sure it's a good method still
22:23:42 <andythenorth> I would rather pass it to a method somewhere, so we can replace easily in future if wrong
22:23:58 <andythenorth> or pass an error object around when calling a template
22:24:13 <andythenorth> error / message /s
22:24:43 <Alberth> it does look a bit fishy indeed
22:25:04 <andythenorth> especially if you read the bottle source, where request.query is ReadOnly :P
22:25:11 <andythenorth> assuming I understand it correctly
22:25:36 <andythenorth> create a message class, pass instances of it
22:25:44 <andythenorth> bit heavyweight, could use a dict, but...
22:25:50 <andythenorth> ...seems more robust
22:25:57 <andythenorth> problem for not-today ;)
22:26:18 <Alberth> Mark it with # XXX Needs a better solution ?
22:27:24 <andythenorth> might be here tomorrow, otherwise thursday
22:32:26 *** andythenorth has left #openttd
22:32:34 *** KritiK_ has joined #openttd
22:42:39 *** Rhamphoryncus has joined #openttd
22:46:07 *** Cybertinus has joined #openttd
23:21:03 *** glx is now known as Guest377
23:37:40 *** Rhamphoryncus has joined #openttd
23:38:14 *** oskari89 has joined #openttd
23:40:55 *** tracerpt has joined #openttd
continue to next day ⏵
