IRC logs for #openttd on OFTC at 2012-10-14
⏴ go to previous day
00:43:05 *** tokai|noir has joined #openttd
00:43:05 *** ChanServ sets mode: +v tokai|noir
01:19:42 *** Knogle^AFK has joined #openttd
03:36:58 *** hmmwhatsthisdo has quit IRC
04:56:17 *** Eddi|zuHause2 has joined #openttd
06:08:50 *** andythenorth has joined #openttd
06:09:54 *** sla_ro|master has joined #openttd
06:35:54 *** andythenorth has joined #openttd
07:07:11 <andythenorth> Terkhen planetmaker ¿ so if I read the nml correctly....to get the correct climate-aware ground appearance, FIRS has a stack of ground sprite + 7 childsprites (with complex conditions)
07:07:28 <andythenorth> does that sound like the correct understanding?
07:14:00 *** mahmoud has joined #openttd
07:25:15 *** Progman has joined #openttd
07:35:46 *** Alberth has joined #openttd
07:35:47 *** ChanServ sets mode: +o Alberth
07:40:08 *** sla_ro|master has joined #openttd
07:46:00 *** Devroush has joined #openttd
08:28:19 * andythenorth isn't convinced about magically concatenating identifiers
08:28:33 <andythenorth> not everywhere anyway
08:28:51 <andythenorth> in some cases seems better to write them out by hand and use find / replace if they change
08:32:04 <planetmaker> andythenorth, yes, that's how I wrote that template: draw on ground and conditionally the correct one over it
08:32:16 <andythenorth> it's pretty neat
08:40:37 <Terkhen> andythenorth: I don't remember, can you point me to the file which contains that code?
08:40:57 <andythenorth> Terkhen: it's ok, planetmaker answered :)
08:41:27 <andythenorth> sprites/nml/templates/spritelayouts_groundaware.pnml
08:41:33 <andythenorth> GROUND_AWARE_SPRITELAYOUT
08:43:23 <Terkhen> aah, yes, I remember it now :)
08:44:30 <Terkhen> I remembered something quite more complex, which probably was our original solution before that one :P
08:45:01 <andythenorth> this one looks pretty robust
08:45:13 * andythenorth wishes chameleon would spit a better error for a misclosed tag :P
08:47:48 * andythenorth got tripped up again by "a single-entry tuple is a string in python" :P
08:48:12 <andythenorth> always happens :P
08:48:24 <andythenorth> trying iterating over ('building_sprite')
08:48:32 <andythenorth> you get each char
08:48:57 <andythenorth> that's been confusing me every few months for about 7 years :P
08:49:03 <Alberth> sure, a parenthesized expression is just its value
08:49:23 <Alberth> a single-valued tuple is ('building_sprite',)
08:49:36 <Alberth> ie the , is what makes it a tuple
08:50:32 * Alberth forgot the syntax of a zero-valued tuple :(
08:51:31 <Alberth> planetmaker: FYI: I consider nml_langcheck.py 'done' for now.
08:51:56 * andythenorth now has tiles and layouts building with python for the grain mill
08:51:59 <andythenorth> not spritesets yet
08:52:07 <andythenorth> but maybe spritesets should be written out manually anyway
08:52:33 <Alberth> or derived from some name convention or so :)
08:53:11 <andythenorth> generating them reduces manual work to add snow version
08:53:54 <andythenorth> no idea how I'm going to handle non-standard tiles yet (e.g. animated)
08:54:33 <Alberth> every tile is animated? :)
08:56:23 <andythenorth> Alberth: ho, every tile is animated probly works :P
08:57:41 <andythenorth> if I make the spriteset a class, then I'll have the number of realsprites and frames in scope
08:57:47 <andythenorth> so I can just write them into the layout
08:58:06 <andythenorth> it's like an OO foam party here
08:58:24 <andythenorth> should have done this a year ago, instead of moaning about CPP :P
08:58:57 <Alberth> moaning is generally unproductive :)
08:59:39 <andythenorth> it's just how I get things done
08:59:43 <andythenorth> moaning is like breathing :P
09:09:36 <planetmaker> "done" is good news, Alberth :-)
09:12:20 <Alberth> that remains to be seen, now the things I didn't do correctly will surface :p
09:17:17 * andythenorth has not yet looked at the slope aware layout
09:18:20 <Alberth> one thing you may want to consider is a better nml primitive to do these things
09:21:59 <planetmaker> andythenorth, slope aware is simple, too... it's just calling the proper sprite as defined by the slope
09:22:20 <planetmaker> sprite: spritesetname(slope_to_sprite);
09:22:36 <planetmaker> or similar, need to lookup the exact nml function for conversion of slope to sprite
09:23:11 <andythenorth> the template will tell me
09:23:41 <andythenorth> it's all pretty obvious in the source, or compiled nml
09:23:53 <andythenorth> so far this has been easy, all I need is motivation + not to be holding a child ;)
09:24:13 <planetmaker> Is there a reason to re-write the templates?
09:26:08 <planetmaker> from what I see there it's only a slight syntactic variation on the existing template?
09:27:12 <andythenorth> the resulting nml should be ~1:1 same
09:27:19 <Alberth> andythenorth: you're extending nml for better handling of layout. Since this is a generic problem, chances are that others can benefit from your approach too. The better place for this extension is thus in nml itself.
09:28:04 <andythenorth> Alberth: write 1 first though to learn...
09:28:22 <Alberth> that's the usual approach :)
09:28:34 <andythenorth> also last time it was discussed, nml templating was 'parked, awaiting ideas / time'
09:29:02 <Alberth> templates and prmitives are not the same
09:29:18 * andythenorth is prepared to learn...
09:29:48 <Alberth> templates is about giving arbitrary constructs a name so you can re-use them (which is what is being done with CPP and m4 currently)
09:30:07 <Alberth> a primitive is an elementary language construct
09:30:25 <Alberth> ie the list of 'statements' supported by nml
09:30:37 <planetmaker> splitting the layout definitions into three files?
09:30:49 <andythenorth> alberth so primitives != first class objects or so?
09:31:15 <andythenorth> planetmaker: one of those is generated
09:31:26 <planetmaker> generated... and in the repo?
09:31:35 <andythenorth> thought I removed it
09:31:56 <planetmaker> it's not tip commit I linked
09:32:12 *** argoneus has joined #openttd
09:32:25 <andythenorth> yeah it's gone in tip ;)
09:32:50 * planetmaker finds the result A LOT more complicated
09:33:02 <Alberth> andythenorth: note sure, some primitives are first class objects, whether they all are would depend on the design of the primitive, I'd say
09:33:54 <Alberth> planetmaker: complexity is not an objective notion :p
09:33:59 <planetmaker> but possibly I don't yet understand it fully
09:35:02 <andythenorth> planetmaker: no you're correct, it is more complicated (complex?). Each industry becomes a first class python object, with a template, which is definitely more complex, but also makes many things easier
09:35:13 <andythenorth> it's the same as FISH and BANDIT tbh
09:35:37 <planetmaker> what becomes easier?
09:36:08 <andythenorth> templating, and economies
09:37:22 <andythenorth> it can always be reverted :P
09:38:24 <andythenorth> Alberth: so what primitives might be useful in this case?
09:39:45 <Alberth> I don't know, what are you missing?
09:40:10 <andythenorth> I don't know either :)
09:40:22 <planetmaker> solving an undefined problem?
09:40:27 <andythenorth> every time I work with nml, I want first class objects in scope
09:40:37 <andythenorth> I don't think nml should be providing that though
09:41:40 <andythenorth> so in FISH and BANDIT I have vehicle objects, in FIRS there are industry, tile, layout and spriteset objects
09:41:45 *** Biolunar has joined #openttd
09:42:03 <andythenorth> having objects means that templates can be simple, because they handle minimal logic
09:43:46 <planetmaker> andythenorth, would it then not make sense to provide these first-class objects in NML? similar like you can define them in c++ but not in c?
09:44:21 <andythenorth> I worry it imposes programming on non-programmers I guess
09:44:37 <planetmaker> writing a newgrf *is* programming, no?
09:44:47 <planetmaker> and... if they're not obligatory, people can just ignore it
09:47:34 <andythenorth> ^ there is code in there that will be moved out to a common module
09:47:49 <andythenorth> but it shows how I define the layouts, tiles, etc
09:50:09 <planetmaker> the top part I assume, the module definitions will be moved
09:50:54 <planetmaker> the top part I assume, the module definitions will be moved
09:50:57 <planetmaker> how does this template industries better, though? By deriving an industry objects from a parent object which defines the common things?
09:51:35 <andythenorth> I wonder if I just prefer the syntax
09:51:44 <andythenorth> I don't know why this is logically better
09:52:58 <andythenorth> I know that I wrote a CPP version of BANDIT, and got a kicking on this channel for doing things wrong, and it was no fun either
09:54:13 <planetmaker> I simply cannot comment on this code :-P
09:54:17 <andythenorth> when I try and do CPP things I keep coming up against the limitations of a macro language, when I want a programming language to hand I think
09:55:29 <planetmaker> yes, sure, it's much more limited. It's not a programming language nor a full-fledged macro one
09:56:22 <V453000> I really dont want to see full firs code :D
09:57:55 <andythenorth> V453000: firs code is what it is, it works, it handles a lot of complicated cases
09:58:01 <andythenorth> but very few people want to work on it :(
09:58:08 <andythenorth> and I kind of can't
09:58:40 <V453000> perhaps you will be able to in the future again :)
09:58:41 <planetmaker> why can't you? The cpp templates too complicated?
09:59:35 <andythenorth> it's just an ugly format CPP, mixing variadic macros with lots of parameters and ALLCAPS identified *into* nml which uses a similar format for params, and has ALLCAPS constants
09:59:37 <Alberth> I'd remove the 'layouts.append' magic
09:59:46 <andythenorth> Alberth: thought you might say that :)
09:59:53 <andythenorth> one rev does it explicitly by hand
10:00:00 <andythenorth> I thought I'd be clever and register them :P
10:00:28 <andythenorth> what would you do there?
10:00:28 <Alberth> I found that clever code tends to bite :)
10:00:44 <Alberth> do you need the layout_4 variable?
10:00:59 <andythenorth> "clever" is why the windmill tile has to be commented out :P
10:01:13 * planetmaker will stop the whining and just let andy do what he sees fit ;-)
10:01:14 *** valhallasw has joined #openttd
10:01:19 <andythenorth> layout_4 is the unfinished stuff
10:01:43 <Alberth> k, otherwise you could make a list directly
10:03:05 <Alberth> you could do 'layouts.append(layout_x)' under the definition, or at the end construct a list explicitly
10:03:13 <Alberth> the former is easier to get right imho
10:03:59 <andythenorth> I had the latter, it looked like a maintenance headache
10:04:03 <andythenorth> I'll do the former
10:05:39 *** Knogle^AFK is now known as Knogle
10:10:25 <andythenorth> Alberth: does my formatting of parameters (long lists when creating instances of classes etc) make you wince ?
10:10:59 <andythenorth> e.g. when creating tile instances and layout instances
10:11:46 <andythenorth> looks like jQuery :(
10:11:55 <V453000> im getting flashbacks to java which I hated at school
10:12:02 <V453000> not like I know any other programming at all though
10:14:04 *** Hyronymus has joined #openttd
10:16:22 <Alberth> why is 'tiles' a tuple in Layout?
10:19:58 <andythenorth> Alberth: do you mean 'why tuple not list' or 'why is it in layout'' ?
10:20:26 <andythenorth> could be a list of tuples
10:20:29 <andythenorth> is that better form?
10:21:13 <Alberth> I see tuples as things with a fixed length, which is not the case here
10:21:32 <Alberth> but maybe I am wrong :)
10:21:54 <andythenorth> it makes more sense when a tuple has a predictable length, e.g. (x, y, filename) or such
10:31:34 <andythenorth> Alberth: now handle non regular shapes ;)
10:31:41 <andythenorth> and things with gaps in
10:32:20 <Alberth> oh, you leave out a triplet?
10:35:41 <Alberth> that looks very much the same :p
10:36:34 <andythenorth> some tiles in the grid are empty
10:36:49 <andythenorth> the - does that of course in your version
10:38:00 <andythenorth> tbh, when coding an industry there is a lot counting from n tile (usually by point at it with my finger with the game on the screen) :P
10:38:09 <andythenorth> having the offsets from N tile is very handy at that point
10:38:16 <andythenorth> although I do appreciate the suggestion ;)
10:39:01 <andythenorth> industry layout is defined relative to the north-most tile in the layout
10:39:20 <andythenorth> the 0, 1 are offsets in south-east / south-west directions
10:39:50 <andythenorth> and I'm very used to reading columns of offsets
10:40:02 <andythenorth> I can kind of parse them now without thinking too much
10:40:18 <Alberth> you want a 2d layout :p
10:40:51 * andythenorth envisages some horrible matrix :)
10:41:07 <andythenorth> maybe I should draw them in a png
10:41:11 <andythenorth> and have PIL figure it out
10:41:33 <andythenorth> FIRS, "now with visual layout editor" :P
10:43:59 <Alberth> in this case it gives names to the images in the .png file
10:44:37 * andythenorth hadn't considered that option :P
10:51:45 * Alberth retracts the reference
10:53:54 <andythenorth> honestly, xml might have a place in FIRS in future :P
10:54:09 <andythenorth> although also...perhaps not ;)
11:00:23 <Alberth> I only use xml because I am too lazy to write a proper parser :)
11:03:41 *** LordAro has joined #openttd
11:05:33 <LordAro> planetmaker/Rubidium: apparently my last message to you was cut off: "thanks for your help!"
11:13:55 <LordAro> "23:02:31 < Kjetil> why doesn't the file specification you mention use stdint-types ?" <-- it probably does, but according to the spec i'm following, it does, so that's what i was using :L
11:21:13 <andythenorth> Alberth: currently I'm passing explicit ids (strings) to classes
11:21:21 <andythenorth> this is the best way to do it with python, right?
11:22:42 <Alberth> python prefers explicit (and so do I)
11:23:07 <Alberth> assuming you need the id somewhere of course :p
11:24:27 <andythenorth> so I've templated the simple cases for spriteset, tiles, layouts
11:24:32 <andythenorth> now I have to figure animation :P
11:27:00 <Alberth> LordAro: hi, one quick comment dirFile and hFile have inconsistent code style, it seems
11:28:39 <LordAro> i realise tis, i've basically just hacked it all together so it's functional, other than that, who knows :)
11:30:58 <andythenorth> what are the performance implications if I make every tile animated?
11:31:10 <andythenorth> most will only have one frame
11:32:25 <Alberth> make a special case in the code generator?
11:32:41 <andythenorth> not animated, animated
11:32:59 <andythenorth> or just don't enable animation on the tile
11:33:05 * andythenorth has some digging in code to do :P
11:36:02 <andythenorth> wrt zextent, how accurately does it need to be provided?
11:36:20 <andythenorth> if there are multiple building sprites in a layout, can they all use same zextent?
11:47:56 <andythenorth> if a building changes size (animated for example), zextent needs to change too?
11:48:45 <Alberth> I have no idea at all about zextent
11:52:06 <andythenorth> I never had any idea in nfo FIRS either
11:53:23 <andythenorth> I'll make an assumption and deal with consequences later :P
11:56:56 *** frosch123 has joined #openttd
11:57:08 <andythenorth> bonour Yexo and quak
11:57:34 <Yexo> andythenorth: as with all other variables that deal with bounding boxes: if the bounding box becomes too small it can cause glitches, I think it should never extend over tile borders but the smaller the better for performance
11:57:42 <frosch123> afternoon Alberth, andythenorth, Yexo :)
11:57:54 <Yexo> theoretically at least, I'm not sure it means much as long as you use reasonable values
11:58:14 <andythenorth> we have values already for FIRS, I'm just trying to figure how specific it should be
11:58:22 <andythenorth> if I can define it per spriteset (to largest sprite in set)
11:58:26 <andythenorth> or if it needs to be more fine-grained
11:58:49 <andythenorth> seems the tile layout only has one definition for it, ignoring animated frames that might change size :P
11:58:55 * andythenorth does per spriteset
11:59:56 <andythenorth> why is the default result for indsutry layouts a tile ID?
12:13:01 *** fonsinchen has joined #openttd
12:14:22 <fonsinchen> It seems to have problems creating some directories needed for publishing and then nothing is published.
12:14:55 <fonsinchen> For some reason it still marks "publishing" green.
12:15:09 <fonsinchen> Build #52 succeeded for some magic reason.
12:15:24 <fonsinchen> I wrote an email about that.
12:15:48 <fonsinchen> to info@openttd.org . People were telling me that would reach the right person ...
12:16:00 <TrueBrain> it does; but it is not always me ;)
12:16:14 <fonsinchen> I mean #52 seemed to succeed, but didn't actually publish.
12:16:28 <TrueBrain> the publish seems to fail on upload to openttdcoop
12:18:08 <fonsinchen> What is that publish.sh doing there? It seems it fails on creating a directory /home/openttd . Is that by ssh on openttdcoop?
12:18:56 <TrueBrain> the coop server made some changes in the way they want to receive files
12:19:00 <TrueBrain> for what ever reason
12:19:05 <TrueBrain> and that is giving issues
12:19:19 <fonsinchen> can I do anything about that?
12:20:49 <TrueBrain> started a new build
12:20:54 <TrueBrain> sadly, it has to do everything again ..
12:22:24 <TrueBrain> ah, I get why it goes wrong
12:22:42 <TrueBrain> how do you set the port for scp ...
12:23:26 <Ammler> the ssh port of bundles.openttdcoop.org changed to 10222, that should be about all changes
12:23:43 <TrueBrain> I am well aware, thank you :P
12:23:47 <TrueBrain> but non-default ports are a bitch
12:24:07 <TrueBrain> there is no universal syntax
12:24:13 <TrueBrain> like host:port:/dir or what-ever
12:24:23 <Yexo> to avoid any confusion: I relayed the problems above to Ammler. TrueBrain found the problem just before Ammler entered here
12:24:26 <TrueBrain> you have to add -p -P .....
12:26:39 <TrueBrain> Ammler: I am not a big fan of hiding configuration in system-wide files .. makes it rather hard to move anything if it is ever needed
12:28:03 <V453000> Pikka and people around him working on UKRS2: I really like where the set is progressing :) well done
12:28:45 <andythenorth> nml is a beautiful thing
12:29:15 <Ammler> pikka does nml too now?
12:29:53 <Ammler> ro do you just make generic comments about how nice the world is? :-P
12:30:04 <andythenorth> I am in the nml docs
12:30:24 <andythenorth> conveniently spritesets are zero-indexed
12:30:28 <andythenorth> which makes for nicer code
12:30:53 <andythenorth> pikka is firmly attached to nfo
12:31:03 <V453000> :) if it works for him
12:31:06 <Ammler> he might have is own secret framework
12:31:10 *** Chris_Booth has joined #openttd
12:31:18 <andythenorth> he has a build script
12:31:27 <andythenorth> it's a batch file or something iirc
12:31:40 <andythenorth> what happens in nml if I use animation_frame in a spriteset reference, but tile has no animation?
12:31:49 <andythenorth> [0 would be useful]
12:32:21 <Yexo> animation_frame is 0 by default until it's changed by openttd due to animation properties or by an animation callback
12:32:32 <Ammler> TrueBrain: would ipv6 with default port be better?
12:32:50 <Yexo> so it's possible to set animation_frame to 10 and after that disable animation, you'll get 10 as value in that case
12:33:04 <Ammler> still looking for a reason to finally setup ipv6 :-P
12:33:27 <TrueBrain> it should be fixed this way, so I dont care anymore :P
12:33:34 <andythenorth> Yexo: ¿ so I could control construction stages that way too (theoretical question, as none are drawn, nor do I plan to)
12:33:54 <andythenorth> this is very neat
12:41:38 <TrueBrain> right, it should be published now
12:46:28 <TrueBrain> fixed the warning too
12:46:35 <TrueBrain> fonsinchen: can you confirm all went okay now?
12:48:18 <andythenorth> hmm, turning animation on or off per spriteset? [andythenorth is thinking out loud]
12:48:43 <fonsinchen> looks right now. The new build shows up on openttdcoop.
12:57:34 <andythenorth> Yexo: I assume floating point maths isn't supported with animation_rate?
12:57:40 <andythenorth> e.g. 0.5 * animation_rate?
12:57:44 <andythenorth> animation_frame /s
12:57:57 <Yexo> floating point is nowhere supported by newgrfs
12:58:10 <Yexo> animation_rate / 2 will work though, it just rounds to an integer
12:59:08 <andythenorth> maybe I should rename my python animation_rate variable to avoid confusion
12:59:56 <andythenorth> ah no, it's animation_speed in nml, tis ok
13:19:39 *** KnogleAFK has joined #openttd
13:26:22 *** Eddi|zuHause2 is now known as Eddi|zuHause
13:26:37 <Eddi|zuHause> * andythenorth just remembered that his dream featured Eddi|zuHause <-- how could it, if you don't even know what i look like?
13:26:57 <andythenorth> it wasn't a hugely weird dream btw
13:27:05 <andythenorth> I was travelling in Europe, we bumped into each other
13:27:12 <andythenorth> we went on our way
13:29:12 <frosch123> does "british guy travelling on the continent" count as nightmare?
13:29:46 <Rubidium> Eddi|zuHause: maybe some photos of r20k spilled?
13:29:51 *** andythenorth has joined #openttd
13:29:56 <Alberth> perhaps for the britsh guy :)
13:30:28 *** andythenorth is now known as Guest1690
13:30:28 *** andythenorth_ has joined #openttd
13:30:28 *** andythenorth_ is now known as andythenorth
13:39:50 <Eddi|zuHause> so then... what did i look like?
13:42:38 <Rubidium> back then you had a beard
13:42:54 <Eddi|zuHause> i meant in his dreams
13:43:03 <andythenorth> can't remember :P
13:45:30 <Alberth> aka, an ordinary bloke :)
13:45:34 <Rubidium> although technically... I've never seen Eddi|zuHause
13:46:05 <Eddi|zuHause> nobody of you did
13:46:12 <Eddi|zuHause> not even the ones that picked me up
13:46:44 <andythenorth> Yexo: all spritesets in a layout must contain same number of sprites....any builtin workarounds to that? (before I template to solve it)
13:54:15 <Yexo> use empty sprites to fill up spritesets with not enough sprites
13:54:34 <andythenorth> so just add lines with [] ?
13:55:33 *** andythenorth is now known as Guest1691
13:55:33 *** andythenorth_ has joined #openttd
13:55:34 *** andythenorth_ is now known as andythenorth
14:10:07 *** KnogleAFK is now known as Knogle
14:32:33 * andythenorth wises python could time travel
14:32:48 <andythenorth> 'get the value from a property on an object which I have not created yet but will'
14:32:58 <andythenorth> like a forward reference
14:43:22 *** ChanServ sets mode: +v Alberth
14:43:22 *** ChanServ sets mode: +v planetmaker
14:43:22 *** ChanServ sets mode: +v peter1138
14:43:22 *** ChanServ sets mode: +o SmatZ
14:43:22 *** ChanServ changes topic to "1.2.2 | Website: *.openttd.org (translator: translator, server list: servers, wiki: wiki, patches & bug-reports: bugs, revision log: vcs, release info: finger) | Don't ask to ask, just ask | 'Latest' is not a valid version, 'Most recent' neither | English only | #openttd.dev for dev-talk | #openttd.notice for commit notices"
14:43:23 *** ChanServ sets mode: +o Terkhen
14:43:23 *** ChanServ sets mode: +o Yexo
14:43:23 *** ChanServ sets mode: +o blathijs
14:47:56 *** andythenorth is now known as Guest1697
14:47:56 *** andythenorth_ has joined #openttd
14:47:56 *** andythenorth_ is now known as andythenorth
14:50:26 <DorpsGek> Commit by frosch :: r24592 trunk/src/date.cpp (2012-10-14 14:50:20 UTC)
14:50:27 <DorpsGek> -Codechange [FS#5241]: Set up the new date completely before calling various daily or monthly processings. (dihedral)
14:55:11 <andythenorth> can I mix spritesets and sprite numbers in one layout?
14:56:47 <andythenorth> must be my fault then :)
15:17:36 *** Industrial has left #openttd
15:18:14 <DorpsGek> Commit by yexo :: r24593 /trunk/src (strings.cpp strings_func.h) (2012-10-14 15:18:09 UTC)
15:18:16 <DorpsGek> -Fix [FS#5333]: crash when a gamescript provided too many parameters to a GSText object
15:19:48 <LordAro> i literally just looked at the changelog, now i have to look again..
15:20:02 <LordAro> honestly, y u devs no work around my schedule?
15:21:33 <frosch123> what? you are implying we work at all :p
15:22:04 * LordAro concedes point to frosch123
15:26:06 *** andythenorth has joined #openttd
15:28:41 <DorpsGek> andythenorth: alberth was last seen in #openttd 1 hour, 43 minutes, and 11 seconds ago: <Alberth> aka, an ordinary bloke :)
15:39:21 *** andythenorth is now known as Guest1701
15:39:22 *** andythenorth has joined #openttd
15:41:19 *** andythenorth has joined #openttd
16:03:17 <LordAro> (without clicking on link) it's happening now?
16:05:02 <__ln__> yes, but he's only at 40000 feet atm
16:05:20 <__ln__> going to take a while until he reaches 120000
16:05:35 *** sla_ro|master has joined #openttd
16:09:10 <LordAro> still not sure about broadcasting it live - what are they going to do if something goes wrong?
16:10:00 <planetmaker> broadcast it live? :-)
16:10:38 <planetmaker> there's no point to watch something live where you know exactly what happens 100%
16:11:33 <LordAro> is a German lecturing an English person on their English skills again? (not sure)
16:26:30 <andythenorth> if the graphics chain switch for industry layouts (relative tile position) is templated....
16:26:40 <andythenorth> ...I can probably template the layout definitions as well right?
16:28:54 <planetmaker> LordAro, if you meant me: no, I didn't try. But surely they'll just broadcast anything happening live. And tbh, there's no real harm in that. We're not in US free-TV here. Luckily
16:29:41 <TyrHeimdal> IF something goes wrong after he jumps...there won't be anything to see
16:34:52 *** sla_ro|master has joined #openttd
17:03:37 <andythenorth> Yexo planetmaker: FIRS grain mill has 6 layouts, but the graphics switch only handles 4 of them
17:03:47 <andythenorth> and nothing looks broken in game (it has been this way for months at least)
17:03:51 <andythenorth> which is intriguing
17:05:03 <andythenorth> 2 layouts are repeated, probably to minimise savegame breakage
17:06:07 *** DanMacK has joined #openttd
17:17:43 *** Progman has joined #openttd
17:28:12 *** DanMacK has joined #openttd
17:37:51 *** Alberth has joined #openttd
17:37:52 *** ChanServ sets mode: +o Alberth
17:41:47 *** DanMacK has joined #openttd
17:45:31 <DorpsGek> Commit by translators :: r24594 /trunk/src/lang (3 files) (2012-10-14 17:45:21 UTC)
17:45:32 <DorpsGek> -Update from WebTranslator v3.0:
17:45:33 <DorpsGek> croatian - 7 changes by VoyagerOne
17:45:34 <DorpsGek> korean - 7 changes by telk5093
17:45:35 <DorpsGek> russian - 5 changes by Lone_Wolf
18:36:46 *** Chris_Booth has joined #openttd
18:46:27 *** andythenorth has joined #openttd
19:02:14 <LordAro> __ln__: fairly impressive, no?
19:06:32 <peter1138> WHY IS MY LOAD AVERAGE SO HIGH
19:06:56 <andythenorth> BECAUSE YOUR PLAYING TOO MANY AVIS
19:07:31 <andythenorth> or you've got the red bull baumgartner thing open
19:07:47 <andythenorth> made my laptop hot :P
19:08:18 * andythenorth breaks FIRS savegames again
19:11:13 <Prof_Frink> peter1138: Because you're hugin'ing huge panoramas?
19:14:12 <peter1138> it's since i switched from vservers to lxc
19:15:06 <DorpsGek> Commit by frosch :: r24595 trunk/src/group_gui.cpp (2012-10-14 19:15:00 UTC)
19:15:07 <DorpsGek> -Change [FS#5168]: Make the group GUI drag&drop highlighting behave more intelligible. (based on work by Juanjo)
19:21:43 <andythenorth> Yexo: can a tilelayout and a switch share same identifier in nml?
19:21:57 <andythenorth> [before I go concatenating cruft onto them]
19:31:06 <Eddi|zuHause> andythenorth: unlikely
19:31:42 *** peter1138 has joined #openttd
19:31:42 *** ChanServ sets mode: +o peter1138
19:32:14 <andythenorth> Eddi|zuHause: ta
19:38:00 * andythenorth experiments anyway
19:41:45 *** HerzogDeXtEr1 has joined #openttd
19:42:47 <andythenorth> Eddi|zuHause: you were correct ;)
19:44:27 <andythenorth> layouts property doesn't complain about last item having a trailing comma
19:44:31 <andythenorth> I thought it would :P
19:44:42 *** argoneus_ has joined #openttd
19:48:07 <andythenorth> Eddi|zuHause: FIRS economies looks a bit more plausible to implement now
19:48:51 <Eddi|zuHause> andythenorth: lists often allow a trailing comma
19:49:20 <Eddi|zuHause> they do in C, and they do in Python
19:50:05 <__ln__> C doesn't allow trailing commas afaik
19:53:11 <andythenorth> Eddi|zuHause: you could join me in 'templating FIRS with python' land :)
19:53:16 <andythenorth> you might not like it :P
19:53:47 <Eddi|zuHause> __ln__: the openttd code is full of those
20:12:08 *** hmmwhatsthisdo has joined #openttd
20:12:44 <__ln__> Eddi|zuHause: i'm basing my opinion on the fact that with GCC's -pedantic flag those aren't allowed.
20:13:38 <andythenorth> someone should teach my templated code to drop unnecessary linebreaks
20:13:43 <andythenorth> in the generated output
20:13:48 * andythenorth can't be bothered :P
20:15:12 <Yexo> <andythenorth> layouts property doesn't complain about last item having a trailing comma <- a feature to make scripting easier
20:15:34 <andythenorth> I can drop the comma with the templater
20:15:41 <andythenorth> but if it's not needed, I wont'
20:18:36 * andythenorth forsees a line of code forthcoming "industry.get_code()"
20:18:40 <andythenorth> that can't be good
20:19:00 <NGC3982> I have been to the Nose Lake Train Museum this weekend.
20:19:08 <NGC3982> It was a bit of a weird.
20:19:25 <NGC3982> The guys hosting it was getting drunk with the electricians, having a blast in the cinema wagon.
20:34:07 * andythenorth runs out of things to template
20:41:59 <dada_> hm, that's weird. I noticed some of my trains were 100 years old. turns out they could not find the route to the depot (even though it's very close by and properly connected to the tracks). when I changed one path signal to a presignal they suddenly did find the depot.
20:45:06 <Rubidium> path signals are by default red and thus have a much higher pathfinder penalty than normal signals. Coupled with the fact that trains will only go to a depot when they are at most a certain amount of pathfinder penalty away from the depot
20:45:18 <Rubidium> this will cause the effect you're seeing
20:46:29 <Rubidium> there might be a setting for the amount of pathfinder penalty. Another solution is using service orders
20:48:18 *** Supercheese has joined #openttd
20:55:55 <Eddi|zuHause> also note that the trains cannot change the reservation when deciding to go to the depot
20:56:58 *** andythenorth has left #openttd
21:00:37 *** supermop has joined #openttd
22:08:06 <Supercheese> ach, nimis tardus fui...
continue to next day ⏵