IRC logs for #openttd on OFTC at 2009-09-14
⏴ go to previous day
00:06:22 <Audigex> it's very easy to drop support
00:06:28 <Audigex> but it doesnt mean the devs actually want to
00:06:41 <Audigex> given the choice, I think they'd have it running on my microwave
00:06:50 <Audigex> but there are only so many hours a day
00:07:05 <Audigex> as we all know, programmers are lazy
00:07:21 <Audigex> i'm sure the devs would rather spend an hour working with the other builds than an hour toiling with mac
00:20:10 <glx> we just need someone with a mac, some knowledge and some time :)
00:20:21 <glx> shouldn't be that hard ;)
00:43:26 <PeterT> can I make a suggestion for the next dev that comes online?
00:44:28 <PeterT> for the console: could you add the feature to use multiple commands at once?
00:44:51 <PeterT> rcon <pw> "kick 123 && reset_company 4"
00:45:33 <Sacro> glx: i have a mac, some knowledge and some time
00:55:39 *** KenjiE20|LT has joined #openttd
01:11:04 <welshdragon> anybody know a thing about distrubution cntres in openttd?
01:11:54 <welshdragon> i have trains that call there, but the seem to drop off/pick up the same amount of cargo - i.e. it stays on the same train
01:12:10 <welshdragon> i have them transferring and taking cargo
01:14:22 <Eddi|zuHause> well, that obviously won't work
01:15:44 <welshdragon> so, no sggestion?
01:16:27 <Eddi|zuHause> 1) use transfer&leave empty
01:16:41 <Eddi|zuHause> 2) use either cargodest or cargodist
01:17:07 <welshdragon> 2) is hard as i am playing Infrastructure Sharing patch
01:17:27 <Eddi|zuHause> well, that is your problem :p
01:17:30 <welshdragon> i'll do Transfer & Leave Empty :)
01:17:36 <welshdragon> Eddi|zuHause: i know :D
01:22:32 <Eddi|zuHause> your a is broken
01:23:07 <z-MaTRiX> no its only devil language
01:23:50 <Eddi|zuHause> so the devil speaks spanish?
01:25:06 <Eddi|zuHause> what's a dead beef cafe anyway?
01:27:25 *** z-MaTRiX has joined #openttd
01:39:40 *** roboboy has joined #openttd
02:11:47 *** HerzogDeXtEr1 has joined #openttd
04:23:12 <planetmaker> DaleStan: just read it.
04:23:14 <planetmaker> I didn't install the "keep dates within those bounds" workaround yet.
04:25:03 <planetmaker> DaleStan: I'll try to work out in more detail and locate what the problem is. But it might take a bit of time as I'm mostly away from any ttd-related things the next 4 weeks.
04:40:17 <DaleStan> planetmaker: I still have no clue. I even deleted all my data files, just in case something strange was living in there. But if my Holmesian deduction (Eliminate the impossible, and whatever remains, ...) is correct, @@LINT OFF should fix the problem.
04:40:58 <planetmaker> DaleStan: I switched to the new MacOS. Might be related. I'll give it a shot on my old system
04:41:24 <DaleStan> Not that I think that's anything resembling "Good Idea", but it'll get the code into a state where GRFCodec will accept it.
04:41:25 <planetmaker> the local makefile, though, has only an if for my mingw
04:41:58 <planetmaker> @@LINT OFF in the NFO file?
04:43:55 <DaleStan> Yep. Or -l- (lowercase L) on the command line.
04:44:54 <planetmaker> yes. without I get the error again.
04:48:24 <DaleStan> Well, I'm pretty sure I know where it's failing then. NFORenum converts every property flagged as a date into a date while checking action 0s.
04:49:06 <planetmaker> yes, I'd assume so.
04:49:19 <DaleStan> It's *supposed* to be guarded so out-of-range dates are shifted into range before being converted, but that apparently isn't working for you.
04:50:28 <planetmaker> I honestly need to check that with my 10.4 install. The boost version used there is the same, but the compiler a different one (and some other system intricacies surey, too)
04:51:43 <planetmaker> as this system (partly) calls itself 64 bit as opposed to 32 bit in the 10.4 macos version
04:56:06 <planetmaker> brb. reboot for testing
05:16:22 <planetmaker> DaleStan: beats me. on macos 10.4 I don't get that error. Seems that the boost routine somehow fails on 10.6
05:16:53 <planetmaker> the one on my new HD is a verbatim copy of that dir
05:24:50 <DaleStan> ... Actually, there needs to be an (...) after the catch.
05:33:59 <planetmaker> hm, the patch does not apply for me...
05:34:09 <planetmaker> Died reading sprite 189make: [sprites/2cctrainset.nfo] Error 1 (ignored)
05:34:29 <planetmaker> the context was slightly different... I added it there manually
05:37:13 <Rubidium> why are there so many people telling incorrect stuff on the forum?
05:38:48 <planetmaker> best case: trying to be helpful, but honest mistake :-)
05:39:02 <planetmaker> worst case: getting a know-all profile without basis
05:47:57 <DaleStan> ... OHSH... Nevermind. I know what's wrong.
05:54:55 <planetmaker> yes, that fixes it.
05:55:57 <planetmaker> I didn't test yet (again) on 10.4. But it was working there before.
05:57:11 <DaleStan> Well, now I know what's wrong. Now I just get to figure out the correct way to make it right.
06:00:27 *** Azrael- has joined #openttd
06:00:29 <DaleStan> Rubidium: What year does OpenTTD stop updating the date?
06:02:34 <DaleStan> Ah. Good. Before 32 bit signed overflow of the day count.
06:03:42 <planetmaker> src/date_type.h:#define MAX_YEAR 5000000
06:10:30 *** PhoenixII has joined #openttd
06:13:48 *** TrueBrain_ has joined #openttd
06:15:15 *** Beklugas has joined #openttd
06:15:53 *** stuffcor1se has joined #openttd
06:17:06 *** deghosty has joined #openttd
06:20:00 *** Phoenix_the_II has quit IRC
06:22:58 *** Cybertinus has joined #openttd
06:23:45 *** De_Ghosty has joined #openttd
06:42:28 *** Progman has joined #openttd
07:25:20 <Rubidium> yes, it stops before the int32 overflow so one can (in theory at least) add BC dates
07:48:36 *** bubastis has joined #openttd
08:10:39 *** TrueBrain_ is now known as TrueBrain
08:41:10 <dihedral> is it possible for me, on a linux system, to watch a file and if the file is deleted or created, run some code??
08:41:17 <dihedral> i need to watch /dev/ttyUSB0
08:42:02 <Noldo> if [ -f /usr/local/bin/irssi ];then
08:42:12 <Noldo> works on bash and tests if the file exists
08:42:35 <Noldo> there is propably one for devices too
08:42:58 <dihedral> not just test a file once
08:43:53 <Noldo> udev handles things happening when devices are connected and stuff
08:44:34 <Noldo> I think it does the device file creating too
08:47:46 <Noldo> hmm there is also something called File Alteration Monitor, but udev is propably better for device related things
08:50:13 <SmatZ> also some kernel *notify
08:51:04 <SmatZ> but those are library functions
08:52:50 <SmatZ> hmm but as the name says, the file has to exist (it's bound to inode number)
08:58:53 *** Polygon has joined #openttd
09:04:31 *** Terkhen has joined #openttd
09:11:56 *** tux_mark_5 has joined #openttd
09:24:34 *** Chris_Booth has joined #openttd
09:38:36 <dihedral> good morning Terkhen
09:47:57 *** fonsinchen has joined #openttd
10:08:52 *** Belugas has joined #openttd
10:08:53 *** ChanServ sets mode: +o Belugas
10:14:45 *** ChanServ sets mode: +v tokai
10:18:15 <petern> SmatZ, you can watch a directory, i think
10:24:37 *** KenjiE20 has joined #openttd
10:41:22 *** Vikthor has joined #openttd
10:49:39 *** [1]Mark has joined #openttd
10:49:43 *** Mark is now known as Guest2472
10:49:43 *** [1]Mark is now known as Mark
11:15:29 *** Biolunar has joined #openttd
11:56:35 *** Coco-Banana-Man has joined #openttd
11:59:55 *** Doorslammer has joined #openttd
12:13:06 *** DaleStan has joined #openttd
12:23:12 <CIA-4> OpenTTD: rubidium * r17534 /trunk/src/ (25 files): -Codechange: unify the naming of callback masks/flags
12:33:17 *** zachanima has joined #openttd
13:02:51 *** ChanServ sets mode: +v tokai
13:06:49 *** Cybertinus has joined #openttd
13:13:54 <Belugas> [06:46] * Quits: PeterT (~Peter@c-76-19-168-104.hsd1.ma.comcast.net) (Quit: I'm off) <--- indeed, way off
13:21:45 <Xaroth> oooOOoo only 2500 more revisions to go
13:27:40 <Vikthor> Xaroth: r20034? What's so special about that?
13:31:37 <dihedral> Vikthor, arn't you a clever little boy!!
13:33:42 * Belugas sends out a calculator to Xaroth
13:38:09 <dihedral> dont you just enjoy having nitpickers around??
13:38:37 <dihedral> esp. when it's so obvious what some one is saying!!
13:39:59 <petern> r20034 is r1000111000 in binary, it's clear now
13:40:12 <petern> r20034 is r1000111000 in ... ternary?, it's clear now
13:41:06 *** th1ngwath is now known as thingwath
13:43:41 *** lobster has joined #openttd
13:48:26 *** zachanim1 has joined #openttd
13:59:41 *** oskari89 has joined #openttd
14:16:12 *** tux_mark_5 has joined #openttd
14:17:32 *** ChanServ sets mode: +v tokai
14:28:57 *** Dred_furst has joined #openttd
14:31:10 *** Coco-Banana-Man has quit IRC
14:31:36 *** Coco-Banana-Man has joined #openttd
14:47:41 *** Progman has joined #openttd
15:12:59 *** zachanima has joined #openttd
15:15:08 *** Coco-Banana-Man is now known as ociporT|naM-ananaB-ocoC
15:20:36 *** Dreamxtreme has joined #openttd
15:23:06 * Belugas stomac is announcing lunch hour 40 minutes in advance
15:27:32 *** th1ngwath has joined #openttd
15:32:03 <muxy> thx. have you seen my comment on flyspray about master server ?
15:32:42 <yorick> good idea, you go make it
15:48:10 *** ChanServ sets mode: +v tokai
15:49:07 *** thingwath has joined #openttd
15:52:27 *** Terkhen has joined #openttd
15:53:32 <CIA-4> OpenTTD: rubidium * r17535 /trunk/src/ (crashlog.cpp saveload/saveload.cpp saveload/saveload.h): -Fix: compiling without networking failed; also unhackify the crash logger a bit.
15:59:16 *** ChanServ sets mode: +v tokai
16:13:05 *** Chris_Booth has joined #openttd
16:27:15 *** ChanServ sets mode: +v tokai
16:30:07 *** Terkhen has joined #openttd
16:33:10 *** |Jeroen| has joined #openttd
16:35:52 *** HerzogDeXtEr has joined #openttd
16:36:06 *** Azrael- has joined #openttd
16:43:56 *** Singaporekid has joined #openttd
16:54:10 *** frosch123 has joined #openttd
17:03:34 *** Doorslammer has joined #openttd
17:06:08 *** TheMask96 has joined #openttd
17:11:36 *** DaleStan has joined #openttd
17:32:38 *** michi_cc has joined #openttd
17:32:38 *** ChanServ sets mode: +v michi_cc
17:33:51 *** Phoenix_the_II has joined #openttd
17:45:27 <CIA-4> OpenTTD: translators * r17536 /trunk/src/lang/ (4 files in 2 dirs):
17:45:27 <CIA-4> OpenTTD: -Update from WebTranslator v3.0:
17:45:27 <CIA-4> OpenTTD: english_US - 4 changes by agenthh
17:45:27 <CIA-4> OpenTTD: hungarian - 3 changes by alyr
17:45:27 <CIA-4> OpenTTD: russian - 64 changes by Lone_Wolf
17:45:28 <CIA-4> OpenTTD: vietnamese - 23 changes by nglekhoi
17:51:49 *** michi_cc_ has joined #openttd
17:52:17 *** michi_cc_ is now known as michi_cc
18:06:48 *** George3 has joined #openttd
18:10:34 *** Doorslammer has joined #openttd
18:30:35 *** ChanServ sets mode: +v tokai
18:31:12 *** ChanServ sets mode: +v tokai
18:41:06 *** Grelouk has joined #openttd
18:45:41 *** andythenorth has joined #openttd
18:53:34 *** ociporT|naM-ananaB-ocoC is now known as Coco-Banana-Man
19:08:30 *** Audigex has joined #openttd
19:30:26 <CIA-4> OpenTTD: smatz * r17537 /trunk/src/ (crashlog.cpp gamelog.cpp): -Fix: don't crash while printing gamelog in crash handler if we crashed before while gamelog action was active
19:31:51 <frosch123> everyone should say that three times clear, loud and quick, please
19:33:09 <Prof_Frink> *that**that**that*
19:35:11 <frosch123> looks like small children pointing at each other
19:35:53 <Doorslammer> Hah, story of the BRSet
19:37:12 <yorick> Prof_Frink started it!
19:38:36 <Zuu> Guess not since he has an underscore in the end of his nickname.
19:42:40 <Prof_Frink> yorick: Go on then, finish the quote
19:43:34 <Prof_Frink> "What? You started it! "No we didn't!" "Yes you did, you invaded Poland!"
19:44:10 <yorick> ok, Prof_Frink: you invaded Poland!
19:44:45 *** HerzogDeXtEr has joined #openttd
19:44:53 *** Chris_Booth has joined #openttd
19:44:54 <Doorslammer> Or should I say, Comrade Frinkowski
19:53:42 <SmatZ> anyone here using gcc 4.5 trunk?
19:54:22 <Zuu> I was just thinking about buging you about a project page for PAXLink.
19:54:44 <Zuu> But I saw you need a few details, so I'll pm them to you.
19:55:20 <Yexo> only the identifier is important, is "ai-paxlink" ok?
19:55:25 <Yexo> you can change the rest later
19:55:38 <Zuu> My username is same as usual.
19:56:02 <Zuu> I guess the user database is not too big to scan through visually either.
19:56:55 <Zuu> No I need to re-learn how to use svn/cvs again. :-p
19:56:55 <Yexo> no link to that can be found, so you might want to add that in the description field
19:57:09 <Zuu> Haven't used that for like 5-6 years or so. (as developer)
19:57:15 <frosch123> [21:55] <SmatZ> < two weeks old <- but you have been coding for ottd longer than 9 months
19:59:53 *** Doorslammer has left #openttd
20:00:07 *** stuffcorpse has joined #openttd
20:15:43 <CIA-4> OpenTTD: frosch * r17538 /trunk/src/ (44 files in 3 dirs): -Codechange: Rename STR_CONTENT_FILTER_(OSKTITLE|TOOLTIP) for more general usage.
20:24:30 <CIA-4> OpenTTD: frosch * r17539 /trunk/src/network/network_gui.cpp: -Codechange: Replace a magic number.
20:28:31 <Zuu> By the way (tell me if I should head to the docs instead of asking you) is it an import or an export I want to do? Or just a commit?
20:29:16 <Yexo> or multiple commits if you want to include old versions
20:29:54 <Zuu> Tortise seams to not allow me to do that, but I gues if I just checkout r0 I can then commit.
20:30:15 <Yexo> yes, you need a working directory first
20:30:36 <Zuu> Commiting old versions might actually be a good idea.
20:30:51 *** Dreamxtreme was kicked by Belugas (THIS IS Epic fail!)
20:31:18 <Zuu> I should have all the v1-13 tars laying around and they are also on tt-forums.
20:31:30 <welshdragon> why kick dreamextreme Belugas?
20:31:39 <welshdragon> i see no recent comment by him
20:31:49 <Belugas> 'cause i can't kick him in #tycoon
20:32:15 <Belugas> [16:29] <Redirect_Left> Whats OpenTTD default port?
20:32:16 <Belugas> [16:30] <Belugas> OpenTTD uses TCP and UDP port 3979 for server <-> client communication and UDP port 3978 for masterserver (advertise) communication (outbound)
20:32:16 <Belugas> [16:30] <Dreamxtreme> ottd not using UPNP is epic fail
20:32:22 <Belugas> stupid user who knows it all
20:33:17 *** Redirect_Left has joined #openttd
20:33:23 <welshdragon> was only moaning as he can't start a server
20:33:39 *** Dreamxtreme has joined #openttd
20:34:15 <welshdragon> #TCOON INVADES #OPENTTD :O
20:34:29 <Redirect_Left> #tycoon is more win though.
20:36:15 <jonty-comp> that's why #openttd has more people in
20:36:16 * Zuu is happily unaware of what is going on in #tycoon
20:36:28 <jonty-comp> Zuu: that's definitely a good thing
20:37:31 <SmatZ> welcome here, gentlemen
20:38:10 * Dreamxtreme goes to the very lowest level of IRC
20:40:19 <CIA-4> OpenTTD: smatz * r17540 /trunk/src/town_gui.cpp: -Codechange: move functionality of PlaceProc_Town() to FoundTownWindow::OnPlaceObject()
20:42:18 * Dreamxtreme looks at the source code of the nightly
20:45:12 <andythenorth> if #tycoon style of discussion is coming in here, I'm off
20:45:16 <andythenorth> but that may be no bad thing
20:45:30 <Dreamxtreme> how can i support ottd ?
20:45:59 <Belugas> dunno. it all depends on your capacities
20:46:20 <Belugas> since i don't have a clue abouty those, i cannot tell
20:47:27 <Prof_Frink> If only there was an alternative to #tycoon...
20:47:42 <Dreamxtreme> what are you looking for
20:47:50 <Dreamxtreme> i can do most things lol
20:47:59 <andythenorth> most people can't
20:48:03 <andythenorth> join an elite band!
20:48:10 <jonty-comp> I tried, but I have no hope in hell
20:48:33 <jonty-comp> also, someone should port the NFO spec to the openttd wiki
20:48:41 <jonty-comp> the ttdpatch tikiwiki thing annoys me for some reason
20:48:47 <jonty-comp> perhaps because, well, it isn't mediawiki
20:49:31 <andythenorth> I have a standing bet with two people that they can't write an nfo compiler that is (a) sane (b) finished (c) implements everything nfo can do
20:49:41 <andythenorth> do I mean compiler?
20:50:11 <andythenorth> all these computer words, they baffle me
20:50:15 <andythenorth> strange computer world
20:50:18 <Yexo> andythenorth: "nfo compiler"? so something like grfcodec?
20:50:40 <jonty-comp> I thought grfcodec fulfils all those requirements :p
20:50:45 <andythenorth> yes, except leaving the nfo part out completely...
20:50:49 *** Nite_Owl has joined #openttd
20:50:54 <andythenorth> i.e. write in language that is not batshit crazy
20:51:03 <jonty-comp> if I could write C++
20:51:12 <jonty-comp> I would make it so you made GRFs using MXL
20:51:26 <jonty-comp> it would be so much easier, but rather difficult to plan out :o
20:51:30 <Yexo> oh, it's a bet they can't write a random-high-level-language to nfo compiler
20:52:36 <Yexo> jonty-comp: how exactly would writing it in XML make it easier? To be able to do everything that nfo already can, it'd easily become such a mess that it'd be worse than nfo
20:52:49 <andythenorth> XML is a dead route, they're thinking something like python -> valid newgrf
20:52:51 <jonty-comp> at least it would make a slight amount of sense
20:52:51 <Yexo> ^^ note that I don't think nfo is bad, it just has a steep learning curve
20:52:57 <andythenorth> or haskell -> valid newgrf
20:53:03 <andythenorth> or brainfuck -> valid newgrf
20:53:16 <jonty-comp> nfo is like the assembly language of game modding
20:53:24 <jonty-comp> everyone else uses Lua nowadays :(
20:53:43 <Dreamxtreme> ok i think i got this
20:53:52 <asilv> nfo was surprisingly easy to learn for me, I expected something much more difficult from other people's descriptions
20:54:39 <Terkhen> I'm still trying to understand callbacks :P
20:55:03 <asilv> mostly grf coding is just time consuming not difficult once you learn the basics
20:55:52 <jonty-comp> I gave up, I just get annoyed at it and then I can't write it anyway
20:56:01 <jonty-comp> not that I ever particularly made the effort
20:56:07 <jonty-comp> there are enough coders around
20:57:10 <Terkhen> after some time coding the numbers start making sense
20:57:16 <Belugas> but not enough GOOD coders
20:57:47 <Belugas> and i'm all but a waste right now... so i'll run home and sleep until tomorrow
20:58:49 <andythenorth> If anyone can implement brainfuck -> valid newgrf (that's worth playing with, not just a one-trick programming pony), I'll give prizes
20:59:58 <CIA-4> OpenTTD: frosch * r17541 /trunk/src/ (core/smallvec_type.hpp lang/english.txt newgrf_gui.cpp): -Feature: Filtering in Add-NewGRF dialog.
21:00:09 <jonty-comp> andythenorth: lolcode -> valid newgrf!
21:00:19 <jonty-comp> Belugas: well OK, there are already enough bad coders
21:00:22 <jonty-comp> and I would be a bad coder
21:00:35 <jonty-comp> well, I am already a bad coder in HTML and PHP and VBA and C#
21:01:19 <_ln> of which HTML is not a programming language
21:01:24 <Eddi|zuHause> andythenorth: i'm sure one could extend my NDL compiler with a brainfuck parser
21:01:59 <_ln> Xaroth: do not try to confuse us.
21:02:03 <andythenorth> Eddi|zuHause: but can anyone write valid brainfuck to do something useful?
21:02:09 <Nite_Owl> Oooo frosch: Filtering to what extent
21:02:18 <Eddi|zuHause> andythenorth: that i don't know ;)
21:02:18 *** Chris_Booth has joined #openttd
21:02:28 <Xaroth> _ln: they are interpreted languages, not compiled programming languages
21:03:06 <Eddi|zuHause> andythenorth: it's pretty trivial to encode hexadecimal, decimal or binary numbers in brainfuck
21:03:26 <Nite_Owl> Oooo frosch123: Filtering to what extent <name corrected for IRC>
21:03:40 <andythenorth> well if you think that's too easy...
21:04:06 <andythenorth> (not turing complete)
21:04:07 <frosch123> Nite_Owl: good from bad of course
21:04:19 <_ln> Xaroth: that's not a definition of programming language.
21:04:21 <jonty-comp> _ln/Xaroth: I never said they were programming languages, but I certainly write code in them :P
21:05:27 <Nite_Owl> I will check it out once the compiling is done tomorrow
21:06:21 * Dreamxtreme creates some new NFO's
21:06:35 <dihedral> jonty-comp, .... sure - as is PHP* (*at run time)
21:06:57 <Nite_Owl> No, no - the daily public compiling
21:07:08 <dihedral> daily pubic compiling?
21:07:29 *** TheMask96 has joined #openttd
21:08:03 <Nite_Owl> Whatever such language you prefer
21:08:27 <Nite_Owl> I am being too obtuse
21:09:11 <Nite_Owl> or obvious in not knowing the proper terminology
21:09:23 <Dreamxtreme> its easy to pick up and learn
21:09:38 <dihedral> that is something different
21:10:26 <_ln> Dreamxtreme: it's easy to pick up and learn wrong
21:11:33 <dihedral> perl can be easy to learn
21:11:57 <Dreamxtreme> firebug is the best tool for php
21:17:03 <Terkhen> anyone knows why UpdateNearestTownForRoadTiles needs _generating_world to be true when invalidating cached data?
21:18:06 <Eddi|zuHause> is _generating_world true in the scenario editor?
21:18:55 <Eddi|zuHause> Terkhen: most likely an optimisation because towns can't change during a game
21:19:49 <Eddi|zuHause> sounds like an expensive operation anyway
21:20:59 <Eddi|zuHause> in a simple approach, for each tile it is O(n^2) to find the nearest town, and there might be O(n^2) tiles to check, so worst case is O(n^4)
21:21:13 *** Audigex is now known as Audi[away]
21:21:26 <Eddi|zuHause> where n is the length of the edge of the map
21:22:02 <_ln> isn't there a separate notation for the worst-case and average-case and so on
21:23:09 <Eddi|zuHause> i know worst case is O and best case Omega, then there is Theta for when both classes are the same
21:23:25 <Terkhen> it seems like I won't be able to just remove the limitation
21:23:25 <Eddi|zuHause> (which is different from "average")
21:24:04 <Eddi|zuHause> Terkhen: it's the simple approach, an optimisation could be O(m*n^2) where m is the number of towns
21:24:07 <Yexo> Eddi|zuHause: for each tiles it's O(M), where M = total number of towns
21:24:39 <Eddi|zuHause> it's very likely that m << n^2 ;)
21:25:51 <Eddi|zuHause> but the original O(n^2) figure can be reduced when one assumes a fairly even distribution of towns, then it's O(k^2) where k is the largest distance between neighbouring towns
21:26:24 <Terkhen> mmm... I remember doing some optimization in a case similar to this one, but I don't remember what exactly
21:26:48 <SmatZ> [23:17:00] <Terkhen> anyone knows why UpdateNearestTownForRoadTiles needs _generating_world to be true when invalidating cached data? <== so cache isn't updated after each built road tile
21:27:14 <Terkhen> it seems like I found another prerrequisite for the found a town patch :P
21:27:34 <Eddi|zuHause> there are more cases like this, like nearest town from station (for station names), or industries around a station (for catchment area cargos)
21:28:44 <SmatZ> Eddi|zuHause: station name isn't generated from "nearest" town, but from "nearest" woth when station is built
21:28:56 <SmatZ> it doesn't change when town is founded / station is moved
21:29:16 <Eddi|zuHause> SmatZ: yes, but it's the same problem of finding the town in the first place
21:29:34 <Terkhen> okay, I'll give the road building code a look :)
21:30:30 <SmatZ> [23:26:45] <SmatZ> [23:17:00] <Terkhen> anyone knows why UpdateNearestTownForRoadTiles needs _generating_world to be true when invalidating cached data? <== so cache isn't updated after each built road tile
21:30:59 *** DaleStan_ has joined #openttd
21:31:00 *** DaleStan is now known as Guest2542
21:31:00 *** DaleStan_ is now known as DaleStan
21:31:07 <SmatZ> not sure it's correct, better verify my statement ;)
21:31:19 <SmatZ> it's usefule when "generating many random towns"
21:31:42 <SmatZ> but it's not set when building "long road"
21:31:52 <_ln> if there's a traveling salesman who wants to visit all the cities on a map using the shortest route, is that a problem?
21:32:13 <SmatZ> _generating_world is there used so town generates more houses or so...
21:32:22 <SmatZ> some parts of town building code depend on that
21:32:36 <SmatZ> it doesn't have much to do with the cache
21:33:15 <Yexo> _ln: of course not, just block all routes except one
21:33:16 <SmatZ> _ln: my behaviour is nondeterministic
21:33:46 <_ln> is there a newgrf with a salesman?
21:34:12 *** HackaLittleBit has joined #openttd
21:34:20 <Terkhen> SmatZ: my current hack is to turn _generating_world to true, execute UpdateNearestTownForRoadTiles, set it to false again and then execute DoCreateTown... the town is created as before, but the houses aren't fully built from the start
21:34:27 <Terkhen> I haven't seen any other differences
21:34:50 <Zuu> Hmm, the company takeover seams to give vehicles (and possible other things too) to the wrong company. I just played as AI company number 3 in my game (using cheats) and when I answered yes to take over TownCars then Rodje got the buses instead of Chopper :-)
21:35:08 <Zuu> And for some reason a new TownCars did not start either.
21:35:31 <Yexo> Zuu: you got a savegame from just before you got that question?
21:36:07 <Zuu> Yexo: No, but I had a similar issue before when I played as AI 1 (paxlink) and the player got all the vehicles.
21:36:19 <Zuu> So Im sure I can make a savegame for you.
21:36:38 <Yexo> not needed if you have to create one
21:37:42 <SmatZ> [23:34:17] <Terkhen> SmatZ: my current hack is to turn _generating_world to true, execute UpdateNearestTownForRoadTiles, set it to false again and then execute DoCreateTown... the town is created as before, but the houses aren't fully built from the start <== interesting idea
21:38:17 <Terkhen> it looks nice, but I still have to check that it doesn't introduce any problems
21:39:18 <Terkhen> I found it by accident anyways
21:39:59 <SmatZ> maybe using some "bool _leave_houses_unfinished" would be better ;)
21:40:18 <SmatZ> anyway, I am now merging and a bit changing your patch ;)
21:40:26 <SmatZ> there are some strange things, like
21:40:34 <SmatZ> this->RandomTownName();
21:40:38 <SmatZ> static void RandomTownName()
21:41:04 <SmatZ> also, storing grf_id and such as static variables of FoundTownWindow
21:41:11 <Terkhen> mmm... yes, I had to change it to static
21:41:12 <SmatZ> I will do some changes in that :)
21:41:42 <Terkhen> I probably made a mess with static variables as I always do
21:42:03 <SmatZ> as more variables were static
21:42:19 <Rubidium> IIRC not all houses are finished; just build a town in the sceneditor and see it grow in inhabitants for a while without getting extra houses (it's like 5-10% though)
21:42:20 <SmatZ> guess it comes from before-oop-in-GUI times ;)
21:42:34 <Terkhen> let me check why I had to change them to static
21:45:04 <Zuu> Hmm, can't reproduce it in my new game and didn't save the old one..
21:45:23 <Zuu> Though I have not tried to do things in the order as I did them in order to simplify it a bit.
21:45:34 <Terkhen> SmatZ: I think I had to convert them to static because I had to call RandomTownName in a static function, but I don't remember which one
21:46:16 <Zuu> But I'll try re-do what i did first time.
21:47:23 <Terkhen> Rubidium: okay, I'm going to check that... I am currently trying to implement a check that gives an error when trying to found a town when towns can't build roads and there is no roads around (to prevent towns with 0 population)
21:51:52 <Zuu> Still I failed to reproduce it. So I'll stop for now, but if it happens again I'll try to record it.
21:52:09 <Eddi|zuHause> hm... has the limitation that the center tile must be a road been lifted?
21:52:25 <Yexo> Zuu: lots of related code looks sane, so no obvious problems there
21:53:54 <Zuu> In both cases when it happened it was the company above in the UI-list (drop down lists of mani menu) of companies that got the buses.
21:58:49 *** bubastis has joined #openttd
22:01:36 * Zuu dreams of a script that goes trough .nut files and collect all function signatures from the implementation part of a class so that he can make sure all functions are declared as static
22:02:51 <Zuu> Unless it falls down from heven I should probably dust of my limited ruby skills and make that script.
22:03:01 <Yexo> maybe the nail code can be of help
22:03:11 <Yexo> if I recall correctly the parsing part was mostly done
22:04:05 <Zuu> Okay. It only need to read the function abc::def(var, var2) line and then output a list of all functions.
22:05:15 <Zuu> A better script would compare it against the available static declarations and either insert new static functions or inform me which are missing. But that extra part has not as high value as the first part.
22:07:46 <Zuu> My Builder class has a subset of its functions that are static and are used as static functions and then some users of the class create an instance and use that. This creates nasty bugs when a static function calls a non-static function which only shows if the static function was called from outside as a static function but not when it is called using an instance of the class. :-)
22:08:22 *** Audi[away] is now known as Audigex
22:08:51 <Zuu> The error that OpenTTD spills out is "index not found functionXYZ" where functionXYZ is a non-static function called from a static function.
22:09:29 <Zuu> I've fixed a few, but I need to make a propper fix or more bugs will be found/created. :-)
22:09:44 <Rubidium> Audigex, now you're back: announcing you away-ness (or lack thereof) by changing your nick or automated messages is unwanted in this channel
22:10:02 <CIA-4> OpenTTD: yexo * r17542 /trunk/src/ (company_cmd.cpp economy.cpp):
22:10:02 <CIA-4> OpenTTD: -Fix: don't access variables in the company struct after it has been deleted
22:10:02 <CIA-4> OpenTTD: -Cleanup: remove some never-used code
22:10:58 <Audigex> i've always just assumed it to be common irc practice
22:11:04 <Zuu> Using /away might be allowed (I don't know) since most clients filter that out anyways.
22:11:26 <Zuu> At least most clients I use, but then I only use x-chat :-p
22:11:46 <Yexo> X-chat doesn't ignore them, the username is drawn in grey instead of in black
22:12:43 <Zuu> Yea, but it doesn't show it in the chat log as some clients does.
22:12:49 <Yexo> Zuu: does a simple grep do what you want? Like "grep 'function' *.nut"?
22:13:39 <MyCatVerbs> Zuu: /away is very much allowed and recommended.
22:13:54 *** ChanServ sets mode: +v tokai
22:14:10 <MyCatVerbs> Away status doesn't get spammed at anyone; instead, one picks up on it when PRIVMSGing or WHOISing. Much better. ^_^
22:14:24 <Zuu> Yexo: Yea should do or beter "grep 'function ClassName::' class_name.nut
22:14:51 <Zuu> Just need to install Cygwin but that shouldn't harm to much to have it installed hehe
22:15:45 <Zuu> The question when you make a new windows install / buy a new computer is not if but when you end up installing Cygwin.
22:16:02 *** TinoDidriksen has joined #openttd
22:17:32 <Rubidium> but... but... I've only installed cygwin once :)
22:19:09 <Rubidium> (and Windows more than once)
22:19:39 <Dreamxtreme> anyone got Win 3.1 or old windows i want to load and play on vmware
22:20:30 <Dreamxtreme> oops wrong place lol
22:21:25 <Eddi|zuHause> i installed win 3.11 in DosBox once
22:21:51 <Eddi|zuHause> needed a little tweaking for 256 colours and sound
22:22:14 <FauxFaux> I have copies of the disks. ¬_¬
22:22:37 <Eddi|zuHause> i have an original (OEM) CD with win 3.11
22:23:05 * Rubidium remebers the 20 or so floppies for Windows 95
22:23:17 <Rubidium> and then the extra 8 or so for Internet Explorer
22:23:22 <Eddi|zuHause> ooooh, yes. i had win 95 beta on disks
22:23:37 <Eddi|zuHause> and it was sooooo unstable
22:24:09 <Eddi|zuHause> i must have some win 3.1 disks somewhere (those were 7)
22:24:46 <Rubidium> back in the good old days that you could boot and start Windows (3.11) from a single floppy :)
22:25:18 <Nite_Owl> I still have a Windows 95 CD - never saw it on disks
22:25:25 <Eddi|zuHause> i once got so annoyed by the compressed win95 boot disk, that i used an over-formatted disk and unpacked it
22:25:50 <Eddi|zuHause> i once installed win 95 from cd onto a computer without cd or network
22:26:19 <Eddi|zuHause> and the .cab files on the CD were 1.7MB, so they wouldn't fit on normal disks
22:26:29 <Eddi|zuHause> and windows refused to format to 1.7MB
22:26:56 <Eddi|zuHause> but from old DOS days i still had this handy tool "VGACopy"
22:27:07 <Eddi|zuHause> which could format 1.7MB floppies
22:27:52 <Eddi|zuHause> so i had two of these floppies, and switched them around between a computer with CD drive, and the one installing windows
22:28:00 <Eddi|zuHause> one .cab file at a time
22:33:27 *** Coco-Banana-Man has quit IRC
22:44:30 <HackaLittleBit> Eddy and don't forget the delightfull fight with IRQ, espacially with sound cards :)
22:46:39 *** ChanServ sets mode: +v tokai
22:48:46 <CIA-4> OpenTTD: rubidium * r17543 /trunk/src/sound/allegro_s.cpp: -Fix [FS#3200]: allegro's sound output was worse than excepted
22:51:21 <Nite_Owl> Need to feed - Later all
23:09:26 *** fonsinchen has joined #openttd
23:25:28 *** fonsinchen1 has joined #openttd
23:25:45 <HackaLittleBit> good night all
23:26:51 *** HackaLittleBit has quit IRC
23:33:14 *** Eddi|zuHause has joined #openttd
continue to next day ⏵