IRC logs for #openttd on OFTC at 2013-08-14
⏴ go to previous day
00:15:38 *** DarkAce-Z has joined #openttd
01:05:05 *** SamanthaD_ has joined #openttd
01:11:36 *** TinoDidriksen has joined #openttd
01:19:32 *** montalvo has joined #openttd
01:19:32 *** Arkabzol has joined #openttd
01:31:09 *** Supercheese has joined #openttd
02:24:53 *** Dark-Ace-Z has joined #openttd
02:29:04 *** Dark-Ace-Z is now known as DarkAceZ
02:32:20 *** roboboy has joined #openttd
03:00:21 *** Belugas has joined #openttd
03:00:21 *** ChanServ sets mode: +o Belugas
03:18:26 *** Belugas has joined #openttd
03:18:26 *** ChanServ sets mode: +o Belugas
04:56:16 *** Eddi|zuHause has joined #openttd
05:28:53 <maddy_> I have been thinking about the signals thing we talked about yesterday: I will forget the idea for scripts, but maybe do something similar to 'programmable signals', only not use that as a base but start fresh
05:30:52 <maddy_> the supported features and specifically the GUI needs some thought to be sure
05:30:58 <SamanthaD_> is this so that prios won't have silly track elements?
05:31:22 <maddy_> yeah, at the simplest level, but I think I will support more stuff later on
05:32:15 <SamanthaD_> what I would like are signals with penalties going through them for the pathfinder
05:32:52 <maddy_> we have that already, passing a pbs signal from the backside?
05:32:57 <SamanthaD_> (make trains prefer the "center lane" on multi-track tracks)
05:33:18 <SamanthaD_> yeah, I knew about that
05:33:36 <SamanthaD_> it's just that I have this track structure I've been playing with that involves alternating signals and cross-over tracks
05:33:55 <maddy_> it actually works out pretty good for the penalty, because you can change the penalty value manually (in my game a value of 500 worked good for my purpose, instead of the default 1500)
05:34:53 <SamanthaD_> I dunno... I might go for a "priority signal" but more than that might be feature cruft...
05:35:04 <SamanthaD_> but that's just IMHO
05:36:49 <maddy_> the most simple version I am thinking of, is that you could choose a set of signals which the programmable signal depends on, and then select a boolean operator for it (AND, OR) which is used to process the inputs
05:37:22 <maddy_> it could be called a logic signal
05:38:06 <SamanthaD_> definitely need thouht about the GUI for that though...
05:38:39 <maddy_> yeah, that's the hard part, how to do it well
05:39:23 <SamanthaD_> maybe we could just go with a consistent implementation? All signals are now controlled via Lisp code!
05:41:42 <maddy_> for more advanced features, I would really like to support signals which act based on the incoming train, so you could do splits based on cargo type, train length and the like
05:42:25 <SamanthaD_> that's easy: just use a waypoint
05:50:52 <maddy_> right, I didn't think of that, good point, so that functionality is not required
05:51:10 <SamanthaD_> I use waypoints like that in almost every game!
05:52:36 <V453000> not even waypoints are necessary to do that :)
05:53:19 <maddy_> for train length? yeah I saw some logic constructs for it
05:55:28 <V453000> e.g. if both trains go somewhere a tiny little bit different, you can navigate them differently too, even though they would normally take the same path for most of the trip
05:56:02 <V453000> we call it a pathfinder trap
05:56:17 <V453000> you basically simulate that somewhere is a path, but the trains cant access it when they approach it
05:56:56 <SamanthaD_> I usually use them in order to direct different trains to different platforms
05:57:06 <SamanthaD_> like... on small stations I might have one platform for loading and one for unloading
05:57:15 <SamanthaD_> or one for livestock and one for grain...
05:57:57 <V453000> well yeah but that is the same as if you made 2 separate stations which is probably more convenient in general
05:58:05 <maddy_> I read about the pathfinder trap, but didn't fully understand it, sounds interesting though
05:58:45 <SamanthaD_> I've never heard about this trap thing...
05:58:53 <SamanthaD_> sounds like a bug we need to squash ;)
06:01:34 <SamanthaD_> you guys have fun, eh?
06:03:26 <maddy_> V453000: what do you think about my idea for 'logic signals' ?
06:07:18 <V453000> unless it has gui visibility like cargo link it is completely worthless
06:07:47 <V453000> you can do everything by now already
06:08:22 <V453000> the signals would only do it invisibly - which if using a couple of signals would become absolutely chaotic if there is no graph thing to show which thing links to what, graphically
06:08:54 <V453000> it does not technically add anything new with it either, but at least is somewhat usable
06:11:12 <maddy_> yeah, the graph thing would really be a key I think, to show the links between the signals
06:11:28 <V453000> yes, rails do that now, which is a lot better
06:11:58 <V453000> but even with it I do not see a single reason for such feature
06:12:25 <V453000> unless you can think of something which will open new possibilities
06:13:05 <maddy_> the main reason is that you could do more 'clean' track layouts, without 'unused' priority tracks cluttering your track layouts, and of course not taking up space
06:15:31 <V453000> because laying down tracks is better in every regard
06:15:44 <V453000> it is visible, better maintainable, more adjustable
06:15:58 <V453000> and mainly improvesp layers understanding of pre-signals
06:19:56 <maddy_> well, I disagree with every regard, but it has lots of good things too, I agree
07:04:25 *** sla_ro|master has joined #openttd
07:30:21 *** valhallasw has joined #openttd
07:48:47 <maddy_> what do you think about my idea for logic signals, featuring only hard-coded choices, as opposed to scripts...it should be more manageable
07:51:17 <dihedral> give me some examples, i did not ready the backlog :-P
07:51:22 *** Devroush has joined #openttd
07:53:03 *** Pensacola has joined #openttd
07:53:20 <maddy_> very simple: ability to use AND/OR/NOT boolean operators and link to other signals as inputs
07:55:07 <dihedral> link only to the previous signal's as input
07:58:02 <maddy_> the key would be ability to show the links between the signals visually, so it's easy to see where the inputs come from
07:58:20 <dihedral> if you work with the pre signals, then you already have your OR type
07:59:03 <dihedral> and you already have the state forwarding / receiving
08:01:07 <maddy_> I will use the existing patch for programmable signals as a guide, but I will probably start fresh
08:03:07 <dihedral> i would rather suggest using the code for the existing pre signal types
08:14:46 *** LordAro has joined #openttd
08:33:33 *** killertester has joined #openttd
08:44:36 <peter1138> getting sidetracked there i see
08:45:50 *** oskari89 has joined #openttd
09:25:08 *** Devroush367 has joined #openttd
09:34:16 *** zeknurn has joined #openttd
09:40:38 *** MrShell has joined #openttd
09:48:13 <Eddi|zuHause> maddy_: the main problem with "signals as inputs" is that it won't work with path signals, there you would need (reservation of) arbitrary tile as input
09:55:21 <Eddi|zuHause> hm... so WHY THE HELL does my code not work?
09:56:09 <Eddi|zuHause> i get some stuff from the database, update it, and then the update is not there when i read it again :/
10:00:44 <dihedral> master / slave database setup?
10:05:55 <Eddi|zuHause> mismatch between editor's codepage and the #encoding-line
10:07:29 <Eddi|zuHause> yes, of course u"ä" does not equal u"ä"
10:08:58 <maddy_> Eddi|zuHause: yes, it would probably not work with path signals, my initial implementation would support only block signals I think
10:09:52 <Eddi|zuHause> maddy_: that's kind of the problem with all these patches. they try to cater some niche which fails to adress the main issue...
10:11:35 <planetmaker> Eddi|zuHause, what's the main issue? :-)
10:12:03 <Eddi|zuHause> whatever involves path signals, which are the most common signals of all :)
10:12:07 <planetmaker> (IMHO for programmable signals it's a sane UI beyond textual display, visual feedback on the map)
10:12:13 <Eddi|zuHause> like... "signals on bridges" (WTF!)
10:13:32 <maddy_> well I have to start somewhere, who knows maybe it can be improved later
10:14:34 <peter1138> Eddi|zuHause, "most common signals of all" < not on V453000's networks :D
10:15:59 <maddy_> planetmaker: any ideas for the UI / visual display on map are welcome btw
10:16:13 <Eddi|zuHause> well that is no surprise that V453000 does the exact opposite of what i would do :p
10:23:59 *** Ristovski has joined #openttd
10:30:55 <Eddi|zuHause> maddy_: concerning "simple" rules: what i always missed was something like "keep this signal red for 10 days, then allow it to go green for 30 days [repeat]"
10:38:37 *** KenjiE20 has joined #openttd
10:38:48 <LordAro> oh hey, the forums are fast again
10:39:34 <maddy_> Eddi|zuHause: good idea
10:40:24 <dihedral> Eddi|zuHause, wtf? what is that for??
10:40:33 <Eddi|zuHause> maddy_: a method to synchronize the signals like vehicle timetables would help there :)
10:40:45 <dihedral> you have mad ideas Eddi
10:41:11 <Eddi|zuHause> dihedral: mixed cargo/passenger line won't allow cargo trains to enter the line short before a passenger train is scheduled
10:41:31 <dihedral> Eddi|zuHause, what will happen if you computer denotes the suffix of your nick, and for some odd reason you need to take your home computer to work?
10:42:00 <Eddi|zuHause> dihedral: we'll figure that out when it happens :p
10:42:20 <dihedral> Eddi|zuHause, you can simply solve that by using a 'priority' type setup and making cargo trains a tile longer than passenger trains :-P
10:42:45 <dihedral> then you can loop a singal state forward to the passengar lane so that when a cargo train comes the passenger lane is always red :-P
10:42:51 <Eddi|zuHause> dihedral: can't put "priorities" on path signals
10:43:09 <dihedral> build it with tracks
10:43:15 <dihedral> oh - path signals ...
10:43:28 <Eddi|zuHause> dihedral: not enough space
10:43:42 <dihedral> that's your problem :-D
10:44:40 <Eddi|zuHause> yes, i often have a different set of problems and solutions than openttdcoop :p
10:44:42 <V453000> nobody says you need to have path signals everywhere :)
10:45:14 <Eddi|zuHause> V453000: but they do help when space is an issue :)
10:46:04 <V453000> if you need signal A to do whatever (e.g. priority), you dont need or want path signal there anyway
10:46:37 <Eddi|zuHause> V453000: if you can put the "priority" and "load balancer" stuff into the programmable part, instead of extra tracks, there is no reason not to use path signals
10:47:09 <V453000> how is it related to path signals at all
10:47:32 <Eddi|zuHause> it shoudl just work with them, or i won't use the patch
10:48:31 <V453000> if that worked, it would add something new indeed
10:48:37 <V453000> otherwise it is worthless as I said
10:51:30 <V453000> I wouldnt say that path signals are simple at all though
10:51:52 <V453000> every beginner has easier time understanding what is a signal block than where are safe waiting post
10:52:36 <dihedral> i think and|or|not signals for pre signal type could be interesting
10:52:48 <dihedral> where OR already exists
10:53:04 <peter1138> signal blocks, yes, pre-signals, no
10:53:42 <peter1138> but then true, most people don't give a fuck about safe waiting points and use path signals as block signal replacements
10:54:07 <V453000> I did not say pre-signals on purpose :)
10:54:10 <Eddi|zuHause> every beginner doesn't understand signals at all :p
10:54:27 <V453000> well then, why are path signals (especially the 2way ones) default? :d
10:54:40 <V453000> if basic block signals are most intuitive to new players
10:55:02 <Eddi|zuHause> that was an unproven statement by you
10:55:36 <V453000> yes I only see and teach players, has no value
10:57:34 <Eddi|zuHause> well, then implement a feature that hides presignals from the signal gui, for beginners
10:57:58 <Eddi|zuHause> then the block signal could be default, with path signals as an option
10:58:09 <V453000> how does that make any sense
10:58:11 <Eddi|zuHause> for when they learned what ha signal is
10:58:30 <Eddi|zuHause> because nobody(tm) needs presignals :)
10:58:56 <V453000> only retarded people dont
10:59:24 <Eddi|zuHause> if it were up to me, i would have ripped them out of the game and every tutorial out there
11:00:05 <Eddi|zuHause> you could re-implement their functionality with programmable signals
11:01:02 <V453000> simple logic of pre-signals does not kill the fun, retarded programming of signals would.
11:01:24 <Eddi|zuHause> then you need the ability to define "signal templates"
11:01:31 <Eddi|zuHause> to easily construct junctions
11:01:48 <Eddi|zuHause> without defining each signal individually
11:02:14 <V453000> which would end up as pre-signals ... difference is?
11:02:42 <Eddi|zuHause> difference is... a) more flexibility, b) less old clutter
11:04:06 <V453000> still, where would you save such a template
11:04:25 <Eddi|zuHause> in the config file, next to newgrf presets
11:04:36 <V453000> so everybody would need to re-define pre-signals?
11:04:38 <V453000> that doesnt make sense
11:05:11 <Eddi|zuHause> or where you store your copy-paste junctions
11:05:28 <V453000> anybody uses copypaste?
11:05:51 <Eddi|zuHause> anybody uses presignals?
11:07:24 <V453000> go visit actual servers, not forums filled with idiots who talk more than play
11:07:51 <Eddi|zuHause> anybody uses servers?
11:12:46 <maddy_> does there have to be a one-solution-fits-all type thing though? like say path vs block signals: arguably you could do some things with either type, depending on the players preferences
11:14:06 <V453000> I think the only thing to add could be a PBS signal which is red when presignals ahead are red
11:14:15 <V453000> you could do anything with that too
11:25:19 *** HerzogDeXtEr has joined #openttd
11:32:01 *** Devroush has joined #openttd
11:37:44 <dihedral> peter1138, what's wrong with using the existing pre-signal setup and extending that? they already have forwarding and receiving previous signals state, it's easy to add and change their reaction based on forwarded states
11:40:45 <peter1138> dihedral, i wasn't responding to you :)
11:41:00 <V453000> there isnt any problem with it, that is their problem :)
11:48:03 <maddy_> I'll see what I can do, if anything, being so beginner at openttd even as a player, not to mention as a coder
11:50:09 <planetmaker> coding a game is playing the meta-game ;-)
11:54:24 <Eddi|zuHause> sooo... UI concept: the signal GUI is changed, to allow a "insert new signal type here" button. then a "programmable signals" gui opens up, you can select an appearance of the signal out of the current (8?) signal type graphics, a button whether it acts as a block or path signal, and a behaviour setup where instead of "listen to signal at <X>" you enter "listen to {all|nearest} signal of type <X>" ["... and make <decision"] which
11:54:25 <Eddi|zuHause> automatically gets filled in when placing the signal/changing the junction
11:55:40 <Eddi|zuHause> player can manually override the selected signal(s) after the signal is placed, with the "regular" programmable signal GUI
11:57:05 <Eddi|zuHause> maybe newgrf signals can provide more than 8 signal graphics
11:57:22 <peter1138> what was wrong with the already implemented programmable signals anyway?
11:57:44 <Eddi|zuHause> likely "doesn'T work with path signals" :p
11:57:59 <maddy_> peter1138: nothing as such, I will plan to use it as a guide, but I like to write stuff from the scratch, to learn the openttd code better
11:58:00 <peter1138> there's more to signal state than checking the state of other signals
11:58:52 <Eddi|zuHause> the main thing to learn from that patch is probably "how and where was this stuff stored?"
11:58:53 <peter1138> whether it's possible to "hold this slow train a second while that faster train goes through" sanely is another matter :D
11:59:30 <peter1138> of course, that's "worthless" to players like V453000, but not to everyone
11:59:30 <Eddi|zuHause> peter1138: i don't think the signal should be able to detect the type of train that is waiting
11:59:45 <Eddi|zuHause> performance? complexity?
11:59:56 <peter1138> think of it as an automated man in a signal box or control centre
12:00:16 <peter1138> of course it's not the signal doing the decision really, but even so
12:00:16 <Eddi|zuHause> a turk signal! :p
12:00:35 <peter1138> pfft performance pfft complexity
12:00:38 <maddy_> peter1138: someone suggested that different trains could be already directed to specific tracks by using waypoints, which to some degree defeats the purpose of detecting any info about the incoming train
12:01:35 <Eddi|zuHause> maybe the other way around would be "enough", impose a pathfinder penalty depending on which train requested a path
12:01:51 <Eddi|zuHause> then the signal does not have to actively check for a train waiting
12:02:03 <Eddi|zuHause> and it can influence trains from further away already
12:02:26 <Eddi|zuHause> so you can guide slow trains towards the "slow trains wait here" signal
12:02:43 <Eddi|zuHause> and let the fast train pass on the "fast lane"
12:11:49 <dihedral> <maddy_> I'll see what I can do, if anything, being so beginner at openttd even as a player, not to mention as a coder <- i have not played in a very long time myself
12:12:00 *** Aristide has joined #openttd
12:13:54 <maddy_> dihedral: right, I have played on and off for years, with long breaks in between but I always return to this great game
12:15:23 <planetmaker> dihedral, bad bad bad ;-)
12:15:44 <planetmaker> I guess you stopped playing when a certain nightly server ceased to run :-P
12:33:22 *** NeuhNeuh has joined #openttd
13:00:56 <__ln__> does anyone have an idea what would be a good way to compress something like 15..30 bytes of data?
13:05:25 <Wolf01> but might not be your case
13:06:22 <Eddi|zuHause> for compressing it always matters WHAT kind of data
13:07:12 <Eddi|zuHause> there is no algorithm that can compress every stream of 30 bytes by 1 bit
13:08:38 <Eddi|zuHause> because there is one less bitstream of 1 to 8*30-1 bits than bitstreams of 8*30 bits.
13:10:07 <__ln__> i am well aware of the inherent limitations
13:10:14 <Eddi|zuHause> it's like the game where there's one less chairs than players
13:11:21 <Wolf01> you could turn up the chairs and let 4 people sit for each chair... ehm
13:11:27 <__ln__> i don't expect to compress any given input by even 1 bit, but i'm looking into the possibility of compressing some inputs by more than insignificant amount of bits.
13:11:51 <Wolf01> you could try to apply a Huffmann algorythm
13:12:39 <krinn> with such few bits, any compression system will higher the result on failure
13:12:53 *** Tom_Soft has joined #openttd
13:13:14 <__ln__> krinn: indeed, that's why probably the very first bit should indicate whether the data is compressed at all.
13:14:37 <Eddi|zuHause> well there are only a handful of generic compression algorithms, so when you don't give us any details about the structure of the data, just pick one of those
13:14:53 <Eddi|zuHause> huffman code is good start, or an arithmetic code
13:16:02 <Eddi|zuHause> or just pipe it through gzip
13:16:21 <__ln__> i know many algorithms or at least their practical implementations will generate so much headers and stuff that for very small inputs the compressed size will basically be larger than the source.
13:17:25 <Eddi|zuHause> yes. so what is the actual usecase where shaving a few bits off a 30 byte stream would outweigh the effort and latency of compressing?
13:18:37 <krinn> and generally for stream people prefer having more bits than less :)
13:19:09 <__ln__> the use case is transferring a few dozen bytes of binary data on the phone, one person reading it out loud, and the other one writing it down on the other end.
13:20:35 <Wolf01> it works better than havind 2 people doing that :P
13:20:55 <Eddi|zuHause> acoustic coupling :p
13:21:14 <Eddi|zuHause> the phone system is digital nowadays, you know :p
13:21:17 <__ln__> Wolf01: perhaps, but if one person doesn't have a modem, it doesn't work very well.
13:21:54 <__ln__> Wolf01: besides, i challenge you to find me a price for an acoustic modem that you can use with a cell phone :)
13:21:57 <Eddi|zuHause> __ln__: whatever you're trying to do, you should probably rethink your approach :p
13:22:02 <krinn> __ln__ i think the best method would be compressing the whole message and then send the compress message 30/30bytes
13:22:16 <Wolf01> the cellphone already has a modem
13:22:45 <__ln__> not necessarily, and certainly not an acoustic one
13:22:49 <Wolf01> or you won't be able to talk to another person
13:22:52 <Eddi|zuHause> the cellphone has also SMS which can transfer 30 bytes easily :p
13:23:50 <__ln__> Eddi|zuHause: i don't trust that people can write 30 characters of arbitrary A-Z0-9 without typos :/
13:24:10 <Eddi|zuHause> __ln__: add an error code :)
13:24:28 <__ln__> however, generating a QR code would be a possibility, and reading it with the phone's camera if it has one.
13:26:38 <krinn> you may count 1 and 0 and send the value as 3 (3 bit set to 1) -4 (4 bits set to 0).... if you have a good range of pattern of 0|1 you get good result and an easy system for the other guy to uncode it
13:28:25 <peter1138> bit-wise RLE is probably going to use more bits representing the counters than the data
13:28:42 <krinn> in the worst case you will have to said it 240 times :)
13:28:56 <Wolf01> or the run-length compression, but works well only if you have multiple consecutive 0 or 1, like 3-0 and 2-1, 2-0, 2-1 => 000110011
13:30:11 <Wolf01> and since you might need to do that in binary, you'll end up with more bits than the original :P
13:31:09 <Wolf01> this remembers an old game I found
13:32:13 *** Aristide has joined #openttd
13:32:26 <__ln__> ... and after that you learned the number 3 in school?
13:33:35 <planetmaker> hehe, nice sequence, Wolf01 :-)
13:33:46 <planetmaker> reminds me of "An eternal golden braid" by Hofstadter
13:35:02 <Wolf01> I started madness again?
13:35:13 * LordAro wonders how easy it would be to script...
13:35:14 <planetmaker> did it ever cease to exist, Wolf01 ? ;-)
13:35:18 <V453000> is that numbers of beers per round?
13:35:40 <V453000> if not, it is a bad game
13:37:35 <planetmaker> 13211311123113112211
13:39:01 <LordAro> 11131221133112132113212221
13:39:40 <LordAro> probability of a mistake... increasing...
13:39:45 <planetmaker> 3113112221232112111312211312113211
13:39:56 <planetmaker> don't they increase?
13:40:50 <planetmaker> anyway, I'm sure we wrote a programme for that sequence in school like 20-odd years ago or so :-P
13:41:32 <Wolf01> I think I did it in Q-Basic or so
13:41:45 <planetmaker> either Pascal or Prolog. Dunno anymore :-)
13:44:31 <krinn> tooked me 4 years to know it was name after the mathematician :P
13:45:18 <V453000> no it is named after the pressure units
13:48:27 *** valhallasw has joined #openttd
13:50:35 *** sla_ro|master has joined #openttd
13:50:52 <LordAro> 1321132131111213122112311311222113111221131221
13:54:59 <maddy_> UK english is the default language?
13:56:40 <LordAro> 3121312132111213122112311311222113111221131221
13:56:59 <LordAro> i think i managed to write a '1' instead of a '2' :L
13:57:09 <krinn> anyone have a look at the ai railpf ? it looks totally buggy
13:57:40 <Wolf01> and this time the error is at the very beginning :P
13:59:03 <Eddi|zuHause> krinn: it's probably Zuu's fault :p
13:59:45 <krinn> seriously, to a point i'm thinking it's me that is wrong
13:59:46 <Wolf01> lol, 312131 -> 11111111
14:01:14 <Eddi|zuHause> hm, actually i don't think i've seen this sequence before...
14:02:05 *** Aristide has joined #openttd
14:02:14 <LordAro> nobody tell him, let him work it out >:)
14:14:32 <krinn> lol Eddi|zuHause must use woobly cars there
14:15:09 <Eddi|zuHause> click on the double arrow :)
14:16:11 <Eddi|zuHause> krinn: yeah, i never understood the wobbly lines on english roads...
14:16:20 <krinn> could have done it on a pic with less trash
14:18:46 <Eddi|zuHause> the pic isn't the point...
14:20:01 <dihedral> <planetmaker> I guess you stopped playing when a certain nightly server ceased to run :-P <- actually long before that :-D
14:22:20 <krinn> there's a double >> to move
14:23:29 <Eddi|zuHause> allegedly it doesn't work if you use the "new" interface to google maps
14:24:01 <peter1138> "Get the new Google Maps. Try it now" < I assume I'm not using the "new" interface
14:25:02 <Wolf01> I can't see a >> also in the classic interface
14:26:21 <krinn> Eddi|zuHause, told you to avoid joking with people using lynx
14:28:45 <peter1138> still using dodgy browser detection code :S
14:32:56 *** ntoskrnl11 has joined #openttd
14:35:24 <maddy_> what's the difference of DoCommand and DoCommandP?
14:36:02 <Eddi|zuHause> maddy_: one is sent over the network for synchronisation and the other assumes that it's already run synchronously on every client
14:36:53 <Eddi|zuHause> so the initial user action must be synchronised, but all follow-up things deeper down the chain don't have to be
14:38:41 <peter1138> or rather, they are already running on all systems
15:12:19 <Eddi|zuHause> @calc 832.7*30065
15:12:19 <DorpsGek> Eddi|zuHause: 25035125.5
15:23:35 <LordAro> gets very long very quickly :)
15:35:26 <LordAro> interesting, i seem to have generated a 100 line, 50MB file :)
15:35:51 <LordAro> gedit is having trouble opening it
15:36:21 <Wolf01> last time I froze my computer after ~200 lines
15:37:10 <Eddi|zuHause> "this was an april's fools joke from 2009"
15:38:02 <maddy_> gedit doesn't seem very efficient with large files, I've noticed it too
15:38:27 <LordAro> hmm, geany didn't do much better
15:39:25 <Eddi|zuHause> with the "overabundance of RAM", programs have lost the ability to handle partial files
15:40:11 <Eddi|zuHause> i vaguely remember cursing a lot the last time i searched for a hex editor that can open a whole partition
15:41:45 <LordAro> no wait, geany handled it fine
15:41:55 <LordAro> the file wasn't formatted correctly
15:42:05 <Wolf01> and developers lost the ability to write programs which could run in a TI-83 calculator, despite doing the same work
15:42:41 <Eddi|zuHause> it could have been python, and much shorter and easier to run...
15:44:31 <Eddi|zuHause> "for (i = it; i < str.length() && str[it] == str[i]; i++);" <-- what horrible code style is this
15:45:43 <LordAro> i did have it on separate lines, but i decided to make it smaller :L
15:45:54 <TinoDidriksen> For shame. Should be ++i instead.
15:46:04 <Eddi|zuHause> smaller != more elegant
15:46:12 <TinoDidriksen> Readability is much more important than fewer lines.
16:00:30 *** andythenorth has joined #openttd
16:01:30 *** mindlesstux has joined #openttd
16:15:23 <Eddi|zuHause> "Microsoft revives the 'Ping of Death' for IPv6"
16:16:51 <Eddi|zuHause> ohh what fun we had with those in the 90s :p
16:17:10 <MNIM> It's not april first, is it?
16:18:00 <Eddi|zuHause> no, latest patchday :)
16:18:24 <peter1138> best not to apply it just yet thme
16:29:23 <MNIM> you'd think microsoft would know people have better things to do than updating every time they start up their PC.
16:29:59 <MNIM> I still don't get why they can't just make updates download and install during runtime, like a real OS.
16:33:23 <Eddi|zuHause> they do know that, hence why they started shipping out patches only once a month
16:36:16 <MNIM> So now I have a doubly outdated OS, and every once in a while I have to wait an hour instead of ten minutes to update?
16:36:26 *** Progman has joined #openttd
16:38:26 *** dumbdumbgj has joined #openttd
16:39:33 <dumbdumbgj> Are you looking for a big smelly elephant with green balls? come to WWW.CENTEX.NET - WE also have black tranny and mushroom sauce... thats WWW.CENTEX.NET
16:40:07 *** dumbdumbgj has left #openttd
16:46:38 *** valhallasw has joined #openttd
16:48:39 <maddy_> ok I got as far as getting my patch started: added a new signal type and added it to signal gui
16:51:52 <Pinkbeast> Aren't there a limited number of bits for signal types in the map array?
16:53:38 <maddy_> probably, but I only added one new type (7 now, instead of 6 previously)
16:54:22 *** Midnightmyth has joined #openttd
16:54:25 <Pinkbeast> maddy_: Ah, very well.
17:01:27 <V453000> LOL that is awesome :D
17:03:04 *** oskari89 has joined #openttd
17:10:25 <andythenorth> so I have a request for bigger layouts for FIRS plantation-type industries
17:10:31 * andythenorth wonders how big?
17:11:20 <andythenorth> current biggest is 5x5
17:12:04 <Xaroth|Work> could be fun though
17:12:25 <andythenorth> 64x64, no gaps :P
17:12:35 <andythenorth> you'd have to manually build it in scenarios :P
17:13:18 <Xaroth|Work> openttdcoop could have a ball with a few of those, trying to make the most efficient station evar :P
17:14:11 <V453000> nobody said it would produce proportionally as much :P
17:14:38 <andythenorth> FIRS compiles so slow
17:14:43 <andythenorth> it inhibits me working on it :P
17:21:43 <andythenorth> faster faster faster :(
17:22:06 <V453000> how long does it take?
17:22:11 <V453000> nuts takes around 15 minutes I think
17:27:36 <dihedral> wow - joan takes 1min 44sec (last build), grapes 26 secs, berries (all optional plugins) 31 secs
17:29:00 <Xaroth|Work> that's java for you
17:29:26 <andythenorth> nml industry spritelayouts scale horribly afaict
17:30:24 <Xaroth|Work> libottdadmin2 forced recompile takes.. <1s :P
17:30:44 <dihedral> Xaroth, at least i will not have to tell "everybody" how to make the environment in order to run it
17:31:09 <Xaroth|Work> make the environment?
17:31:38 <dihedral> setup the environment
17:31:50 <dihedral> ap+ needed expect, tcl expect bindings and what not
17:32:03 <dihedral> and it would only run on linux :-)
17:32:19 <Xaroth|Work> now you just have to explain them how to install java :P
17:33:08 <dihedral> and say java -jar <grapes.jar> <configfile>
17:39:54 *** frosch123 has joined #openttd
17:43:16 <andythenorth> these big layouts look stupid
17:44:53 <dihedral> andythenorth, i don't think they look stupid
17:45:07 <andythenorth> maybe it just needs to balance better
17:45:18 <andythenorth> seems very repetitive to me
17:45:46 <andythenorth> I guess plantations are repetitive :)
17:46:53 <DorpsGek> Commit by translators :: r25722 /trunk/src/lang (5 files) (2013-08-14 17:46:41 UTC)
17:46:54 <DorpsGek> -Update from WebTranslator v3.0:
17:46:55 <DorpsGek> croatian - 1 changes by VoyagerOne
17:46:56 <DorpsGek> indonesian - 48 changes by UseYourIllusion, Yoursnotmine
17:46:57 <DorpsGek> lithuanian - 2 changes by Stabilitronas
17:46:58 <DorpsGek> swedish - 17 changes by Joel_A
17:46:59 <DorpsGek> tamil - 12 changes by aswn
17:47:44 *** oskari892 has joined #openttd
17:48:09 <peter1138> so anyone using btrfs?
17:48:43 <peter1138> seems to be capable of migrating from ext3/4 to btrfs without a needing a backup...
17:49:21 <andythenorth> is it a new british trains newgrf?
17:54:41 <andythenorth> is this game done yet?
17:54:49 *** valhallasw has joined #openttd
18:10:32 *** valhalla1w has joined #openttd
18:19:03 <dihedral> peter1138, good luck
18:19:18 <dihedral> iirc btrfs is still beta
18:23:37 <andythenorth> anyone want to draw cargo icons?
18:23:40 <andythenorth> how wrong can it go?
18:23:44 <andythenorth> need coffee and suchlike
18:24:41 <krinn> btrfs is flaw by design, and such thing for a fs, should just be its dead
18:24:49 <maddy_> what kind of datatype could I use for something dictionary/hashmap type thing, where I have a key and a corresponding value?
18:24:50 <planetmaker> andythenorth, did you look at NUTS for coffee icons?
18:25:01 <andythenorth> maybe V453000 will draw me some icons?
18:25:11 <planetmaker> I *think* it has already coffee support. maybe I err
18:25:18 <Eddi|zuHause> maddy_: one of the standard types?
18:25:44 <V453000> I am about to draw coffee this or next week actually
18:25:47 <maddy_> Eddi|zuHause: yeah, where/what are they?
18:25:56 <V453000> will be some box-ish things on flatbeds, so could make those as icons
18:26:07 <Eddi|zuHause> how should i know... i don't write c++ :p
18:26:18 <andythenorth> V453000 need Copper, Sugar, Diamonds, Rubber, Coffee
18:26:24 <andythenorth> but too lazy too draw them :)
18:27:01 <andythenorth> I spose copper is just iron ore, but pink :P
18:27:04 <V453000> diamonds are on flatbeds already, you could grab those
18:27:14 <andythenorth> sugar is white grain
18:27:20 <V453000> maybe sugar/copper/rubber can be grabbed from hoppers
18:27:29 <andythenorth> coffee comes in a cup? :P
18:27:45 <V453000> coffee will be in boxes on flatbeds
18:27:53 <krinn> andythenorth, i'm not sure oracle won't sue you, they are kinda mad
18:27:57 <Eddi|zuHause> coffee should be bulk
18:28:04 <V453000> go see nuts/gfx and look for "hopper" files
18:29:28 <andythenorth> maybe too big for icons :)
18:29:39 <andythenorth> they are 10x10 px
18:29:44 <V453000> maglevhopper might be k
18:30:19 <krinn> at that size andythenorth i'm afraid even a cow looks like copper
18:30:59 <Eddi|zuHause> you'd be amazed what some people can do with a mere hundred pixels
18:31:33 <krinn> it's not what people can do, it's what people can see :)
18:31:50 <Eddi|zuHause> people can see a LOT of things :)
18:33:33 <krinn> well, i can nearly see nothing on the cargo png, the blue background doesn't help too, but when zooming you can see he manage to really do things
18:34:14 <krinn> lmao the pax is one dude & a woman, but at normal size... it's like waste
18:45:13 *** Pereba_ has joined #openttd
18:50:14 *** Pereba_ is now known as Pereba
18:54:17 *** Aristide has joined #openttd
19:03:55 *** andythenorth has joined #openttd
19:18:49 <andythenorth> time to make tea
19:23:42 <andythenorth> planetmaker: there was a param 'test_industry' for FIRS
19:23:47 <andythenorth> for fast dev compiles
19:23:52 <andythenorth> not sure it works with new makefile? :)
19:24:03 *** Bad_Brett has joined #openttd
19:24:09 <andythenorth> or I've forgotten how to use it
19:26:30 <Bad_Brett> C:\Program\OpenTTD\data>nmlc --grf TrainTest.grf TrainTest.nml
19:26:30 <Bad_Brett> nmlc ERROR: nmlc: An internal error has occurred:
19:26:30 <Bad_Brett> nmlc-version: 0.3.0.r2063:9e5e21dc5219
19:26:30 <Bad_Brett> Error: (IOError) "cannot identify image file".
19:26:30 <Bad_Brett> Command: ['nmlc', '--grf', 'TrainTest.grf', 'TrainTest.nml']
19:26:31 <Bad_Brett> Location: File "C:\Python27\lib\site-packages\PIL\Image.py", line 1980, in open
19:27:29 <andythenorth> looks like a PIL error :)
19:27:29 <Bad_Brett> it seems that this error should occur if you try to use pcx files - which I don't do
19:27:44 <planetmaker> pcx files should work, though
19:27:54 <planetmaker> afaik. Unless pil doesn't support them ;-)
19:28:57 <Xaroth|Work> depends on how PIL is compiled
19:29:06 <Xaroth|Work> it supports it if you have the system libraries for it
19:29:11 <Bad_Brett> the strange thing is that it worked a few hours ago
19:29:20 <Bad_Brett> maybe some image file has been corrupted
19:29:32 <Bad_Brett> in the cache maybe?
19:30:23 <Xaroth|Work> can you try removing PIL
19:30:26 <Xaroth|Work> then installing Pillow
19:30:43 <Xaroth|Work> (through pip works best)
19:31:24 <Bad_Brett> i'll try that if clearing the cache won't do the trick
19:32:41 <planetmaker> Xaroth, I first indeed would guess a corrupted image file
19:33:56 <Xaroth|Work> planetmaker: yeh, most likely
19:34:06 <Xaroth|Work> but it could also be some obscure issue with PIL that Pillow has solved
19:34:49 <Bad_Brett> since i could compile the same file earlier today, it seems doesn't seem very likely though
19:35:28 *** Progman has joined #openttd
19:35:33 <planetmaker> Xaroth, very unlikely... and... pillow with nmlc is supported... but I can't say that it's well tested
19:35:58 <Xaroth|Work> Pillow is backwards compatible with PIL
19:36:09 <Xaroth|Work> and PIL should be counted as deprecated tbh
19:37:38 <Bad_Brett> so the search begins... 8192 png's
19:37:39 <planetmaker> the version which caused the report is the pillow support ;-)
19:38:01 <Bad_Brett> perhaps it's quicker to create fresh copies :)
19:38:16 <planetmaker> Bad_Brett, just look for create / modification time?
19:38:23 <planetmaker> since it last worked
19:38:52 <Bad_Brett> that may indeed do the trick
19:39:00 <Xaroth|Work> how is that the other side of the coin, pm?
19:40:51 <planetmaker> (slightly) different call paths
19:41:15 <planetmaker> those are different. changing that broke (some) PIL installs
19:41:43 <Xaroth|Work> easy_install can cock up sometimes, yes
19:42:17 <andythenorth> we should just write a buildout for whole thing :P
19:42:25 <andythenorth> but buildout is a strange beast
19:43:15 <Xaroth|Work> pip + requirements.txt should often be enough
19:44:47 <andythenorth> I would (and do) virtualenv in that case :)
19:45:39 <wakou2> Might be interesting for some of you...
19:46:06 <Xaroth|Work> andythenorth: if pip works and you have an up-to-date requirements.txt , setting it up in venv should be a walk in the park
19:46:33 <andythenorth> I could be convinced easily :)
19:46:45 <andythenorth> I have no experience with pip and requirements.txt
19:46:56 <Xaroth|Work> pip is the better variant of easy_install
19:47:03 <andythenorth> I have experienced multiple other python packaging systems - and their problems
19:47:26 <andythenorth> currently installing nml is a faff
19:47:42 <andythenorth> it took me several days of attempts the first time
19:47:52 <Xaroth|Work> and requirements.txt is just a text file with 1 package (plus optional version specifier) per line
19:48:01 <planetmaker> preferred method: hg checkout and put that in the path :D
19:48:04 <Xaroth|Work> so you can tell pip to install all requirements in 1 go
19:49:22 <andythenorth> add it to nml? ;)
19:49:23 <Xaroth|Work> and, at the same time, your setup.py file can read requirements.txt to get the proper values for setuptools/distutils
19:49:35 *** Bad_Brett has joined #openttd
19:50:22 <andythenorth> which has a number of issues
19:51:40 <planetmaker> I think anyone is welcome to submit patch(es) to improve the situation
19:52:00 <planetmaker> tbh, I didn't yet care much about the installation issue(s)
19:52:17 <andythenorth> I am trying to remember the actual issues I had, rather than the issues caused by bad / conflicting advice on irc :)
19:52:21 <andythenorth> both happened :P
19:54:20 <andythenorth> problems I recall
19:54:32 <andythenorth> - there is a fair chance that user has a broken setuptools :P
19:54:45 <andythenorth> - some versions of PIL are just borked and don't install
19:54:51 <andythenorth> (PIL is known bad, hence pillow)
19:55:02 <andythenorth> - if you are on OS X, all bets are off
19:55:14 <Xaroth|Work> OSX is known to fuck packages up
19:55:28 <Bad_Brett> when everything else fails: restart your computer :)
19:55:29 <Xaroth|Work> python, perl, heck they even managed to mess up PHP PEAR :|
19:55:34 <andythenorth> if you are given advice about screwing with your path for specific python packages, this is PROBABLY wrong
19:55:56 <andythenorth> - get a clean version of python 2.6 or 2.7
19:56:01 <andythenorth> - never get python from macports
19:56:05 <andythenorth> - never get anything from macports
19:56:13 <planetmaker> well... works for me :D
19:56:35 <andythenorth> - installing python modules system wide is a cluster fuck if you work on multiple projects
19:56:50 <Xaroth|Work> use virtualenv :)
19:56:54 <andythenorth> - always use a virtualenv, even if people tell you not to
19:57:14 <Xaroth|Work> unless you're extremely stubborn, then good luck fixing things yourself if you cock up :P
19:57:38 <andythenorth> what I learned (for OS X)
19:57:56 <planetmaker> anyway... back to creating new DevZone :D
19:57:58 <andythenorth> - alias the pythons correctly
19:58:13 <andythenorth> - create a virtual env for your newgrf projects
19:58:20 <andythenorth> - install the nmlc deps in the virtualenv
19:58:26 <andythenorth> - pray that setuptools doesn't screw you
19:58:33 <andythenorth> - pray that PIL doesn't screw you
19:58:57 <andythenorth> and then all is well
19:59:17 <andythenorth> planetmaker: *new* DevZone? o_O
20:01:31 *** frodus_dev has joined #openttd
20:04:17 *** Aristide has joined #openttd
20:12:12 *** Aristide has joined #openttd
20:19:24 <andythenorth> I'm not very inspired
20:20:52 <gynter> Can you add a crane maybe?
20:21:10 <andythenorth> I removed the cranes
20:21:21 <Xaroth|Work> should add tiles or something so people who want can expand it a tad (like the diagonal shores above and below it)
20:21:34 <andythenorth> just use stations
20:21:41 <andythenorth> or someone made FIRS object tiles I think
20:23:43 <Wolf01> 1113122113121113222123211211222123211231131122211211131221131112311332211213211321322113311211131112132123222113111221131221 LordAro, could you check this please?
20:24:59 <LordAro> Wolf01, which iteration?
20:25:38 <Wolf01> I was explaining this to a friend, and I let my hands continue...
20:25:50 <LordAro> 11131221131211131231121113112221121321132132211331222113112211 <-- that's 14 :L
20:26:09 <LordAro> 311311222113111231131112132112311321322112111312211312111322212311322113212221 <-- that's 15
20:26:21 <LordAro> 132113213221133112132113311211131221121321131211132221123113112221131112311332111213211322211312113211 <-- 16
20:27:09 <Wolf01> or I missed some numbers
20:28:02 <LordAro> 11131221131211132221232112111312212321123113112221121113122113111231133221121321132132211331121321231231121113122113322113111221131221 <-- 17
20:31:16 *** HerzogDeXtEr1 has joined #openttd
20:38:29 <andythenorth> does anyone want a truck set?
20:38:42 <andythenorth> I keep _nearly_ starting one
20:39:47 <andythenorth> what does it do?
20:40:13 <Bad_Brett> something that hasn't been done before?
20:41:09 <Xaroth|Work> A space one would be cool as well
20:44:36 <Bad_Brett> i must say, that Mars One project is quite exciting
20:47:34 <andythenorth> I like imagination for made up vehicles :)
20:47:39 <andythenorth> need V453000 for that
20:50:22 <andythenorth> I feel like this trading post should have cobblestone dock
20:50:30 <andythenorth> but then the in-game dock won't match :)
20:51:48 <Bad_Brett> unless you add a matching dock... but that's not possible at the moment, right?
20:52:49 *** valhallasw has joined #openttd
20:52:49 <Bad_Brett> at least not with nml
20:54:59 <andythenorth> ach, maybe I try that
20:55:17 <Bad_Brett> it looks quite good if you ask me
20:56:24 <andythenorth> it needs another building or something
20:56:30 <andythenorth> but it will do for now
20:56:43 <andythenorth> at least there are no longer 4 'port' industries all look same
20:59:31 <Bad_Brett> ugh... using extra turning angles is my stupidest idea so far... i'm about to go insane
21:01:06 <Bad_Brett> you can say that again
21:01:34 <Xaroth|Work> too much something for too little something
21:03:40 <Bad_Brett> i would like to be able to use longer vehicles, and i hate when they split on the middle
21:04:30 <Bad_Brett> i think i have a solid approach, however, the code is getting so complex that it drives me insane :)
21:05:14 <andythenorth> sometimes you're fighting the game
21:05:19 <andythenorth> it is what it is :)
21:05:24 <andythenorth> don't let me discourage you though
21:07:29 <Bad_Brett> everything is eddi's fault... it was him who showed me CETS
21:08:17 <planetmaker> Bad_Brett, it's important to have goals. But also important to get things done. In smaller steps than re-inventing the world :-)
21:08:28 <planetmaker> (even when that can be your goal)
21:09:14 <Bad_Brett> yeah i know... however, sometimes doing the right preperations may save time in the end
21:10:54 <andythenorth> I need some idea for Bulk Terminal port
21:11:04 <andythenorth> problem is, cargos might change per economy
21:11:25 <Rubidium> that's no problem in a stuff economy
21:11:47 <andythenorth> I should make a stuff economy easter egg
21:12:29 <planetmaker> is there two kind of stuff or only one?
21:13:03 <andythenorth> red, green and blue stuff
21:13:16 <andythenorth> this trading post is no good :(
21:13:22 * andythenorth has looked at it again
21:13:28 <andythenorth> the life of a newgrf author is hard :P
21:14:26 <Bad_Brett> by the way, how was the tt-forums get-together?
21:15:56 * andythenorth might have to draw something, instead of copy+paste existing sprites :/
21:18:47 *** tokai|noir has joined #openttd
21:18:47 *** ChanServ sets mode: +v tokai|noir
21:20:53 <andythenorth> 17k downloads of FISH 2 alpha :(
21:20:55 <andythenorth> and it's not good
21:22:19 <Xaroth|Work> right, cleaned up openttd-admin.py ... stripped out most of it, now to re-write it O_O
21:33:58 <Bad_Brett> would someone be interested in an accordion grf?
21:43:24 <andythenorth> it will do for now, as an idea
21:46:00 <Bad_Brett> LordAro, what's the status on your AI?
21:46:14 <LordAro> umm, working, i think
21:46:26 <LordAro> i haven't looked at it in about 2 years :/
21:48:19 <Bad_Brett> is it only for road vehicles?
21:53:56 <Bad_Brett> i remember using it a lot a few years ago
21:55:23 <LordAro> tbh, now that i understand a bit (quite a lot) more about coding, i'd probably scrap the whole thing and start again :L
22:00:59 <Bad_Brett> one thing i remember last time i tried the different AI's, was that high running costs seemed to drive them out of business quickly... the ones that could handle trains never seemed to survive if I used the NARS grf
22:02:33 <Bad_Brett> my fear is that no AI will be able to handle the gold rush scenario either :(
22:20:40 *** zeknurn has joined #openttd
22:21:13 *** Aristide has joined #openttd
22:59:37 *** amiller has joined #openttd
23:20:30 <wakou2> Hi folks what grfs should I use if I want to start date at (say) 1900?
23:23:36 <krinn> wakou2: if it's to get vehicle, egrvts gave u horses
23:24:11 <wakou2> krinn: TY and early trains?
23:25:50 <krinn> wakou2, i don't know, trains appears ~ 1940 i think
23:27:00 <V453000> most newgrf train sets start in 1920
23:27:11 <V453000> 1920 could be UKRS2 or NARS I guess
23:27:26 <V453000> other train sets arent even noteworthy
23:28:48 <krinn> V453000, trying jedi tricks on his mind ?
23:29:19 <krinn> "this is not the newGRF you are seeking"
23:29:53 <V453000> 2cc set is bulltrash, everybody knows that
23:30:30 <krinn> so if so many knows it, maybe it's worth mention it no?
23:32:13 <V453000> im not sure if a newGRF being known as useless means it is noteworthy
23:33:04 <wakou2> & just getting zbase update from online content... 273Mb!! Is that right??
23:33:29 <krinn> "crappy state" is a personal thing, while popular, even bad for some or many, it should have something worth if everyone pick it up
23:33:58 <wakou2> Ok... (drums fingers) :)
23:34:17 <V453000> well since everyone picks up everything (to at least try it), I do not see how your sentence has any valid logic either
23:35:48 <krinn> wakou2 has just prove that theory is wrong :)
23:36:34 <V453000> k now I dont understand at all
23:37:40 <krinn> since everyone picks up everything -> if everybody was doing that, wakou2 would have done it and don't ask what newGRF to use
23:42:28 <V453000> ofc because he is just plain lazy and doesnt even bother trying things
23:42:31 <V453000> I dont even consider those people
23:43:54 <krinn> another new theory base on facts ?
23:44:43 <V453000> lazy people have no place in this game
23:45:04 <V453000> they will just get bored sooner or later anyway so why bother
23:45:39 <krinn> i dunno, maybe because you're the bored one here, assuming everyone is lazy just because asking something
23:48:33 <V453000> so you consider asking in some IRC channel a better option than trying to download the train set and check yourself when trains come out
23:49:12 <krinn> considering how many trainset are there, it's just enough yes, but he was asking vehicle, not specially trainset
23:50:16 <V453000> I still think trying, discovering, and learning on your own is a lot more productive
continue to next day ⏵
