IRC logs for #openttd on OFTC at 2008-12-18
⏴ go to previous day
00:19:46 <Eddi|zuHause> @calc 2**32/3600/24
00:19:46 <DorpsGek> Eddi|zuHause: 49710.2696296
00:31:27 *** Eddi|zuHause has joined #openttd
01:19:46 *** Dr_Jekyll has joined #openttd
01:39:12 *** sigmund_ has joined #openttd
02:31:27 * fjb almost never needs 4-way junctions.
02:33:07 <benjamingoodger> a three-way version? :)
02:37:17 <fjb> How do I go from south west to north east?
02:38:40 <jerker> up a slope, over a bridge, left turn, right turn, through a tunnel?
02:38:44 <benjamingoodger> fjb: straight over, through the bottom-left bridge, merge, then through the tunnel
03:39:44 *** Eddi|zuHause has joined #openttd
04:07:48 *** elmex_ is now known as elmex
04:31:52 *** Frobozz has joined #openttd
07:38:53 *** ttd12345 has joined #openttd
07:39:12 <ttd12345> Your TRG1R.GRF file is corrupted or missing! You can find TRG1R.GRF on your transport tycoon deluxe CD-ROM
07:39:21 <ttd12345> i am getting above error
07:39:29 <ttd12345> do u konw anythign about it
07:39:31 <Forked> then you probably need the trg1r.grf file..
07:39:46 <ttd12345> but i m not getting it anywhere
07:40:28 <Forked> and put it in (assuming windows here..) the "my documents\openttd\data\" folder
07:41:02 <ttd12345> also there are same errors for TRGIR.GRF,TRGCR.GRF ,TRGHR.GRF,TRGTR.GRF,sample.cat
07:41:15 <Forked> yes, you need those files from the original transport tycoon deluxe game to play openttd
07:41:38 <ttd12345> but where i will get above files so that i can put them in "my documents\openttd\data\" folder
07:41:54 <benjamingoodger> from the original transport tycoon deluxe game
07:41:57 <ttd12345> is there any way u know
07:42:04 <Forked> buy the original game.. you can probably get it online for $5 + shipping or something
07:44:16 <dihedral> hihi - i am not saying anything publicly :-D
07:44:23 * dihedral has learned from the last instance
07:45:03 <dihedral> Forked, can you see the regards to the 'last instance'?
07:45:14 <Forked> I can, but I refuse to.
07:45:27 <dihedral> shove it up your fuzzy rear end :-P
07:45:45 <Forked> dihedral: get enough coffee yet? ;)
07:45:52 <dihedral> no - just arrived at the office
07:46:03 <benjamingoodger> sounds like you've had too much TBH
07:46:22 <Forked> and already giggling? this will be interesting :)
07:46:30 <benjamingoodger> exactly my point
07:46:34 <dihedral> i have 'club-mate' here to drink too
07:46:58 <Forked> amazon.com lists a geforce 9400 GT 1GB card under TTD related accessories :)
07:47:10 <dihedral> 20mg caffeine on 100ml
07:47:21 * benjamingoodger had a lovely conversation on #gnucash about the most efficient way to caffeinate
07:47:27 <dihedral> Forked, yes - you can play OpenTTD on the card itself
07:47:58 <dihedral> leaves the cpu to be dedicated for windows vista
07:48:13 <benjamingoodger> it probably would have more processing power than the CPU of the original target computer for TTD...
07:50:05 <Forked> I'm thinking I should install linux again on the gaming computer..
07:51:42 <benjamingoodger> dihedral: windows vista needs a whole GPU to itself also...
07:51:43 <benjamingoodger> never a bad idea
07:51:45 <benjamingoodger> my gaming system has linux running on it atm, actually
07:52:26 <Forked> well my main concern is getting x3: terran conflict running under wine
07:52:37 <Forked> since it's not ported yet :\
07:53:04 <benjamingoodger> for this reason, our friendly IBM overlords have provided us with a reset switch on our PC-compatibles
07:53:46 <benjamingoodger> I can't see any problem with it
07:59:53 *** ttd12345 has joined #openttd
08:06:59 *** roboboy has joined #openttd
09:01:51 *** Zealotus has joined #openttd
09:28:38 *** DorpsGek has joined #openttd
09:28:38 *** ChanServ sets mode: +o DorpsGek
09:44:48 *** Gekz__ is now known as Gekz
09:52:26 *** HerzogDeXtEr has joined #openttd
09:59:03 *** Rexxars has joined #openttd
10:09:06 *** Zealotus has joined #openttd
10:37:10 *** eQualizer has joined #openttd
11:19:02 *** Singaporekid has joined #openttd
11:21:13 *** mucht_work has joined #openttd
11:21:48 *** ChanServ sets mode: +v tokai
11:21:53 *** jerker is now known as jerkerTaggartTranscontinental
11:24:14 *** Dred_furst has joined #openttd
11:28:09 *** jerkerTaggartTranscontinental is now known as jerkerTaggart
12:07:38 <Forked> so there was some thunder and lightning here earlier
12:07:41 <Forked> the whole building was shaking
12:07:52 <TrueBrain> too bad we missed that
12:08:01 * Forked gets nightmareish flashbacks from school
12:08:30 <Forked> in the fourth grade there was an assignment with a picture of two gold fishes.. and the question was "what does the fish say?" .. I wrote down "blup blup" and moved on..
12:08:43 <Forked> and was informed that no, that was not right .. the two fish were suppose to have an intelligent conversation about something
12:10:15 <TrueBrain> I say you are right
12:10:24 <TrueBrain> "blup blup" is most likely the most intelligent conversation two fish can have
12:11:38 <TrueBrain> SimCity for iPhone .. lol
12:12:15 <Rubidium> the major problem is that even how intelligent the conversation is, there is no transliteration from fishish to the latin alphabet except mapping everything on blup.
12:14:28 <Rubidium> the same happens with Smurf, but to a lesser extent; e.g.: lets smurf today
12:17:57 <Forked> I guess it was one of those "you were suppose to use your imagination, not use it as a shortcut" assignments
12:18:02 <Forked> it still bothers me though :\
12:21:31 <Rubidium> like my (disliked by teacher) tactics for German grammar; when you had a sentence like "D__ Fehler__ hatte d__ Fehler__ gemacht" and you had to fill in the __. Apparantly always making it plural and thus having to learn only 1/4th of the tables and not having to bother about the gender wasn't the intended solution.
12:22:43 <Forked> I also had a math thing with a class that was on a trip.. X number of students blablabla bus and train and whatnot..
12:22:54 <Forked> I assumed they were going home to.. so I just doubled my result ..
12:23:02 <Forked> appears they were suppose to stay at the destination forever
12:23:28 <CIA-1> OpenTTD: rubidium * r14689 /trunk/config.lib:
12:23:28 <CIA-1> OpenTTD: -Change: make configure die on commonly made user mistakes, like not having SDL
12:23:28 <CIA-1> OpenTTD: development files or zlib headers installed; you can still compile a dedicated
12:23:28 <CIA-1> OpenTTD: server or a binary without zlib, but you have to explicitly force it.
12:30:00 <TrueBrain> damn, I am still bored :p
12:30:51 <Forked> let me tell you all about my childhood ..
12:31:06 <Forked> not a very good joke, but at least it was a joke .. not something I was about to do!
12:31:51 <TrueBrain> my keys like 'backspace', 'home', .. etc no longer work ...
12:32:05 <TrueBrain> also, I need to launch 'hald' before running 'slim' or 'kdm', else things go wrong
12:32:07 * petern ponders the feasibility of this 'non-sea edges' patch
12:32:49 <Rubidium> that patch's size is exploding with the addition of even more IsValidTile checks *all* over the place
12:34:02 <petern> and thus everything becomes slightly slower
12:35:01 <Forked> slow is good, no? gives you time to react :\
12:35:14 <Rubidium> and we'll get months of building X at the edge crashes
12:35:17 <TrueBrain> I agree with Forked :p
12:35:45 <TrueBrain> Rubidium: that holds for any progress .. always a chance on crashes and problems
12:38:13 <Rubidium> true, but this patch breaks assumptions in basically ALL _cmd.cpp files, which is an awful lot of places
12:38:28 <Rubidium> and there isn't quite a way to grep for them
12:38:39 <TrueBrain> so it is a good thing, I say
12:38:45 <TrueBrain> as then finally all those 'assumptions' are removed
12:40:41 <Rubidium> TrueBrain: so effectively you want to remove the MP_VOID at the bottom of the map too?
12:41:02 <TrueBrain> but I agree, it shouldn't make things slower :)
12:42:20 *** Belugas has joined #openttd
12:42:20 *** ChanServ sets mode: +o Belugas
12:42:28 <Rubidium> that's why I proposed adding MP_VOID to the top edges too; then *all* assumptions about the map are still valid and not all the if (IsValidTile(xx)) needs to be done
12:43:01 <Rubidium> resulting in a much smaller diff, much less chance of bugs and a smaller (if existing at all) impact on speed
12:47:59 <petern> bastard, i clicked on that :(
12:48:10 <TrueBrain> doesn't go to anywhere
13:07:26 <Yexo> Rubidium: as I said before, adding MP_VOID to the top edges removes only a few of the IsValidTile checks, most of them are also needed for the bottom edges
13:08:41 <Yexo> main problem is that quite a lot of functions assume that they are never called with a tile near the edges (because that were water tiles)
13:10:32 <Rubidium> most functions wander only 1 tile away from their own tile, so IsValidTile isn't needed when doing switch (GetTileType(t - 1)) and friends
13:10:37 <Yexo> and adding MP_VOID tiles to the top edges will give problems with older savegames (those can have buoys/ships) at the top edge, so I doubt the patch would become any smaller
13:17:54 <Rubidium> so make freeform map edges a patch setting and disallow changing it in-game: setting disabled -> old behaviour (no free form edges ofcourse), setting enabled -> extra MP_VOID at top of map
13:21:42 <Rubidium> jerkerTaggart: that's just a worse version of the Ultimate 3-way
13:25:22 <Rubidium> someone entered "pimk my ride"
13:28:16 <jerkerTaggart> Rubidium; If so it may be removed, however, as far as I can see, it has the same speed on the slow path (5 wagons) and is more compact (covering 20 extra squares vs 38 for the Ultimate 3-way)
13:32:13 <Yexo> Rubidium: I just went over the patch, and if the north edge was made MP_VOID only 11 !! chunks with IsValidTile could be removed
13:32:34 <Forked> way to fuck up a perfectly fine car
13:43:32 <Yexo> jerkerTaggart: your junction could be made even smaller by moving the rightmost bridge one tile to the top-left
13:44:13 <jong> hi there, I would really like to play some games with cargodest, but on the Multiplayer list there is just one server (Jezekus's cargodest server), which needs to grf's: CS Stations Set - Buildings and CS Stations Set - Platforms & Stops v0.3 alfa. I cannot find these, even after a long search on the forums and the newgrf part of the website. Also, if there are other servers with cargodest on, I'm more than happy to join these.
13:45:17 <Ammler> jong: www.tycoonez.com
13:45:42 <jong> Ammler: thank you, I'll check it out
13:46:15 <Ammler> hint: use not english language :-)
13:48:27 <jong> great, I'll use my slavic language skillz ;)
13:57:17 <jong> well, the cargodest server will restart in 2 years
13:57:34 <jong> so, it'd be fun if more people started together with me
14:12:06 <Rubidium> Yexo: I count many more
14:13:07 <Yexo> there are a lot more instances of IsValidTile, but not many more that can be removed
14:14:21 <Yexo> on second thought, I didn't count those before IsCompatibleTrainStationTile and they can also be removed
14:15:09 <Ammler> jerkerTaggart: if you set a signal every 2. tile, you need to double b
14:16:02 <petern> signals every 2 * train length!
14:16:28 <jerkerTaggart> Yexo: I added a variant so that the similarity with the 4-way will be clearer. Good suggestion to move the bridge, will fix later.
14:17:03 <Rubidium> Ammler: please use English and not German
14:18:33 <Yexo> jerkerTaggart: one point I don't like is that trains going NW-SW block trains going SW-SE (no signal before or after that bridge)
14:18:50 <Ammler> Rubidium: that wasn't german either :P
14:19:13 <Rubidium> so 2. isn't something German?
14:19:41 <Rubidium> Ammler: no, the "2."
14:19:53 <Rubidium> it makes my brain go haywire on parsing the sentence
14:20:32 <Rubidium> "if you get a signal at every 2" doesn't make sense and neither does "tile, you need to double <whatever you tried to write>"
14:20:42 <Ammler> hmm, ah, it should be 2nd or something :-)
14:21:18 <Ammler> that was indeed german, sorry.
14:21:51 *** Singaporekid has joined #openttd
14:23:20 <Ammler> but still it doesn't matter, how big the signal gap is, but it should be the same over the network, imo.
14:23:37 <Yexo> Rubidium: does it make sence to only add MP_VOID tiles to the NW edge?
14:24:39 <Ammler> at least over the same line :-)
14:25:15 <Rubidium> Yexo: only as in really only NW, or NW, SE and SW?
14:25:34 <Yexo> NW, SE and SW, so not to NE
14:25:48 <Yexo> as NE will wrap to the MP_VOID tiles at SW
14:26:51 <Rubidium> true, but OffsetByDir may do -1, -1
14:26:53 <Yexo> another option might be to let GetTileType return MP_VOID for tile >= MapSize()
14:27:12 <Yexo> Rubidium: even that case will always be ok
15:01:55 *** stillunknown has joined #openttd
15:09:49 *** Progman has joined #openttd
15:19:23 <edeca> With new industries, what causes production to go crazy? I have a forest producing hundreds of tons a month with lorries picking up the wood, but an iron ore mine that seems to produce almost nothing even though it has 2 huge trains servicing it
15:20:22 <Yexo> most of the time it's % transported last month
15:21:32 <glx> and how grf author decided to react to different things
15:22:43 <Yexo> Rubidium: http://www.tt-forums.net/download/file.php?id=103241 <- new patch, I've removed a lot of the IsValidTile calls. Instead of creating a row of MP_VOID tiles I've modified GetTileType to return MP_VOID on tiles > MapSize(). That's a hack, but it works to show the size difference in the diffs. This one is 68KiB instead of 85KiB
15:24:01 <Rubidium> with MP_VOID at top you don't need to change the 0s to INVALID_TILE
15:25:07 <Yexo> I know, that's another 15KiB or so (see FS#2454)
15:26:06 <Yexo> if you think that's the way to go I'll implement it
15:26:31 *** Dred_furst has joined #openttd
15:34:40 <petern> 0 -> INVALID_TILE should be done anyway, separately...
15:35:35 *** frosch123 has joined #openttd
15:36:32 <petern> is it only stations that use 0?
15:37:01 <Yexo> I haven't found anything else
15:37:32 <Yexo> planes might be using it if they don't have a destination, not sure about that
15:46:19 *** Eddi|zuHause has joined #openttd
16:23:19 *** HerzogDeXtEr has joined #openttd
16:48:18 * LA had a terrible (read: interesting for him, but possibly not worth the effort) idea
16:48:33 <LA> to make a toyland -> christmas conversion :D
16:49:41 <frosch123> just replace the landscape with the arctic one, and add snow to the houses
16:50:00 <LA> to make things like little santas etc. :P
16:50:14 <LA> or atleast GUI conversion to something christmas like
16:52:18 <LA> but I guess I cba to do it.. and if even I am not, then anyone else is neither
16:55:59 <George__> Is anybody here who can explain me the dependences between animation and CPU usage?
16:56:38 <glx> CPU usage increase with animations
16:56:59 <George__> I've got a question from the user "someone playing TTDPatch or on low spec hardware can avoid loading those 12 sprites and any similar ones in other industries? "
16:57:01 <De_Ghosty> ottd use only like 1 threads?
16:57:04 <LA> he knows that stuff about newGLX
16:57:06 <De_Ghosty> it doesn't use 2 cores..
16:57:18 <De_Ghosty> draw should have it's own personal thread
16:57:35 <petern> feel free to code that
16:57:43 <petern> it's been discussed to death
16:57:45 *** mortal` has joined #openttd
16:57:53 <glx> De_Ghosty: drawing is not the CPU eater
16:58:15 * LA would listen to a dev, when one is speaking...
16:58:16 <De_Ghosty> in openttd coop it is
16:58:45 <LA> yapf and ships are cpu eater lol
16:59:19 <George__> So, I would like to know - would disabling animation stages would help (grf will return the same sprite for every animation stage)
17:00:06 <George__> Or providing a sprite that has less bounding boxes (1 instead of 10)
17:00:14 <Yexo> George__: users can already disable animation, so what is the point of emulating that via newgrf?
17:00:31 <petern> you'd have to remove the animation code
17:00:48 <petern> otherwise you'll incur the same animation penalty but without the benefit of the image changing
17:01:02 <George__> does stoping the animation would help?
17:01:04 <LA> parameter controlled action7 :P
17:01:36 <frosch123> [18:03] <George__> does stoping the animation would help? <- yes
17:02:56 <George__> I supposed that I should change it according to Global var 1B
17:03:16 <frosch123> despite of that I doubt animated tiles have a big impact on performance
17:03:54 <George__> frosch123: So, if I specify animation for tiles, but stop it, animation does not take cpu?
17:04:14 <petern> events are still processed
17:04:43 <petern> following varaction chains
17:05:00 <frosch123> and var 1B is not supported
17:05:38 <frosch123> and likely will never be. You have to use a grf parameter
17:07:43 <George__> frosch123: > despite of that I doubt animated tiles have a big impact on performance
17:07:45 <Yexo> out of curiosity, what is var 1B? (wiki.ttdpatch.net is down)
17:07:45 <George__> Does it mean low profit with lot of efforts?
17:08:09 <George__> Yexo: 1B B display options; bit 0=town names, 1=station names, 2=signs, 3=animation, 4=transparency, 5=full detail
17:08:46 <frosch123> Yexo: wiki2.ttdpatch.net
17:08:52 <Eddi|zuHause> Yexo: there used to be wiki2
17:09:04 <Yexo> that one still works, thx all :)
17:09:54 <frosch123> George__: start with setting the animation_speed property to very slow, and ask the user whether it had any impact
17:10:23 <frosch123> (also disable the animation speed callback :) )
17:12:12 <Eddi|zuHause> the last time i tried ECS it totally choked on a big map. like 90% of time was spent in resolveSpriteGroup
17:12:36 <glx> Eddi|zuHause: lot of callbacks ;)
17:13:23 <Eddi|zuHause> that's when i decided to use PBI instead ;)
17:16:51 <George__> Eddi|zuHause: I have Core2Duo 2.4 CPU and I do not have any problems. But if it is a big problem for users with weaker computers - I'm ready to spend time in removing unnecessary callbacks if someone will be a tester for that.
17:17:40 <Eddi|zuHause> i don't have that weak of a computer... like an athlon 2000+ or 2400+ or something
17:18:06 <Eddi|zuHause> since the second core does not help openttd at all...
17:18:12 <TrueBrain> how ever fast your computer becomes, relative CPU time never changes :)
17:18:23 <TrueBrain> if it spents 90% of the time in resolveSpriteGroup, it does that on all CPUs :p
17:18:27 <TrueBrain> (well, give or take a few % :p)
17:19:05 <George__> frosch123: what callbacks take the most CPU?
17:19:30 <petern> Eddi|zuHause, athlon 2000/2400+ is pretty weak compared to core2duo ;)
17:20:09 <Eddi|zuHause> yes, but it's not that low end, like people using 800MHz PCs :p
17:20:35 <Eddi|zuHause> (which i still expect openttd to run fine on)
17:20:57 <Eddi|zuHause> and it was a freshly started map... no vehicles or anything
17:21:24 <Eddi|zuHause> and it felt like playing TT on a 25MHz computer with a full map
17:22:13 <frosch123> George__: those which are called often :) E.g. when the animation counter or tileloop triggers :)
17:22:14 <petern> was that before certain things were improved?
17:22:22 <petern> (or was that generation only?)
17:22:30 <Eddi|zuHause> that was generation only afaik
17:22:40 <Eddi|zuHause> i'm not talking about that ;)
17:23:05 <frosch123> it was the "distance to water/land" variable, which is most likely used on generation :)
17:24:59 <frosch123> hmm, maybe the "animation speed" callback is the slowest of all
17:25:27 <George__> yes, but this variable is used only once a month :)
17:25:33 *** LA is now known as LordAzamath
17:26:33 *** [com]buster has joined #openttd
17:26:46 <SpComb> put the mouse cursor anywhere on the row, and it highlights... left click, and it opens... middle click, and nothing happens
17:26:56 <LordAzamath> especially if noone responds to task 2458
17:26:58 <George__> frosch123: CBs 20/27/142 ?
17:45:36 <George__> Ok, I'll remove callback 27 from everywhere. What is the next one?
17:47:41 *** |Jeroen| has joined #openttd
17:52:11 *** [com]buster has joined #openttd
17:53:15 *** Alberth has joined #openttd
17:55:20 <George__> Unfortunately, there were not much callbacks 27, only in agricultural vector :(
17:56:48 <xand> hmm, I have openttd running on two machines, both version 0.6.3... on one, when I tell it to make trees transparent from the Transparency Options window, they disappear completely, but on the other, they go transparent like buildings do; why the difference?
17:57:12 <Yexo> in the gui tab under patch settings
17:57:55 <Yexo> sorry, it's under "interface", the option called "invisible trees"
17:58:11 *** HerzogDeXtEr has joined #openttd
18:00:01 *** Yeggzzz is now known as Yeggstry
18:03:22 <George__> frosch123: Very interesting statistics. Could you provide instructions how to get it myself?
18:05:51 <George__> frosch123: :( I do not have tools for compilation. Could someone compile me a win32 version?
18:06:41 <Rubidium> mrfrenzy: buildottd is failing for months already
18:07:16 <mrfrenzy> ooh sorry didn't know
18:07:23 <mrfrenzy> it was quite useful when I last used it
18:09:04 *** ProfFrink has joined #openttd
18:09:32 *** ProfFrink is now known as Prof_Frink
18:11:58 *** mortal`` has joined #openttd
18:23:21 <Eddi|zuHause> just out of curiosity, if i had an action 4, could i use gender and plural codes? and would ttdp understand those?
18:23:56 <glx> action 4 follows ttdp format
18:24:32 <Eddi|zuHause> because it would be kinda useful if you could put a gender e.g. to industries
18:25:03 <frosch123> and as translators can change and abuse the genders as they like, your grf would be specific to certain revisions
18:25:47 <Eddi|zuHause> yes, but that is a rare special case, and you can check revisions in the grf
18:28:02 *** Der_mit_Kruecken_geht has joined #openttd
18:28:04 <Eddi|zuHause> anyway, the "new industry founded" message is the one that annoys me the most
18:28:54 <Rubidium> it shouldn't be too hard to make something that OTTD likes
18:29:12 *** stillunknown has joined #openttd
18:29:29 *** NukeBuster has joined #openttd
18:30:57 <Eddi|zuHause> my memory is fading somehow... how did TT handle genders of industries? [in the german translation]
18:31:44 <Eddi|zuHause> and how got ottd's system get introduced?
18:33:41 <glx> Eddi|zuHause: newgrf can specify their own message
18:34:03 <frosch123> only when failing :p
18:34:27 <glx> new industry text ID (1B)
18:35:40 <Eddi|zuHause> apparently, tron introduced the gender system
18:36:25 <Eddi|zuHause> glx: i don't expect most newgrf authors to use those
18:38:31 <glx> of course they need to define the custom new industry text for all languages (even if only one requires special treatment)
18:41:11 <Rubidium> it should be fairly trivial to cook up a few changes to the specs to make genders and plurals supported
18:41:26 <Eddi|zuHause> hm... no i misread the log, ludde introduced the gender system, and tron updated the german translation
18:42:09 <Rubidium> basically "porting" SCC_SETCASE, SCC_SWITCH_CASE, SCC_GENDER_LIST and SCC_PLURAL_LIST
18:42:35 <Eddi|zuHause> strgen must introduce a special symbol for case and stuff, you just need to add those to the action 4 strings
18:42:45 <glx> the only problem is how to be sure gender and cases order remains the same ?
18:43:00 <Rubidium> glx: you can't be sure of that
18:43:31 <Eddi|zuHause> you don't, it's the grf coder/translator's responsibility. just make sure errors are handled gracefully
18:43:53 <frosch123> Eddi|zuHause: Seems like TTD solved it by skipping a lot "der", "die", "das"
18:45:30 <Eddi|zuHause> systems like these would be a lot more standardised if english had genders :p
18:45:48 <LordAzamath> Estonian doesn't have genders
18:46:00 <LordAzamath> so nothing else shouldn't have either
18:46:02 <benjamingoodger> genders are Bad, Very Very Bad
18:46:21 <benjamingoodger> inflected persons are Bad, Very Very Bad
18:46:28 <LordAzamath> We are gone so far that even he/she is with one word..
18:46:31 <glx> Eddi|zuHause: english has genders
18:46:35 <frosch123> LordAzamath: Isn't that compensated by about 6 genders in all your neighboured countries?
18:47:01 <glx> but it doesn't really use them
18:47:02 <Eddi|zuHause> glx: yes, but they appear at exactly one point. in the 3rd person singular pronouns
18:47:19 <LordAzamath> hmm.. I thought that there are only 2 genders.. but perhaps the russians have the rest 4 too
18:47:36 <Eddi|zuHause> frosch123: no, you mean cases
18:47:52 <CIA-1> OpenTTD: translators * r14691 /trunk/src/lang/ (10 files in 2 dirs): (log message trimmed)
18:47:52 <CIA-1> OpenTTD: -Update: WebTranslator2 update to 2008-12-18 18:47:25
18:47:52 <CIA-1> OpenTTD: arabic_egypt - 132 fixed, 1 changed by khaloofah (133)
18:47:52 <CIA-1> OpenTTD: czech - 10 fixed, 25 changed by Hadez (35)
18:47:52 <CIA-1> OpenTTD: esperanto - 15 fixed by Athaba (15)
18:47:53 <CIA-1> OpenTTD: greek - 7 fixed by ouranogrammi (7)
18:47:55 <CIA-1> OpenTTD: indonesian - 88 changed by fanioz (88)
18:48:19 <benjamingoodger> LordAzamath: english does have a gender-neutral pronoun, "one", but it has acquired the status of snobbish idiom, and thereby "you" is preferred in informal speech
18:48:25 <benjamingoodger> which is annoying to most others
18:49:13 <Eddi|zuHause> gender-neutral != neutral gender ["it"]
18:50:12 <benjamingoodger> I mean gender-neutral as opposed to neutral gender.
18:50:42 <benjamingoodger> french: "on", esperanto: "oni", etc. god knows about the germanics
18:50:58 <Eddi|zuHause> it's worse in german, where you would say "man"
18:51:10 <Eddi|zuHause> which is awfully close to "Mann"
18:51:11 <benjamingoodger> how inconvenient
18:51:34 <Eddi|zuHause> "awfully" for the feminists ;)
18:51:57 <benjamingoodger> possibly just as in "male person". ¬.¬
18:52:34 <LordAzamath> Atleast we got 14 cases of a noun :P (or 28 if you add plural forms too)
18:52:34 <petern> you'll not see nothing like the mighty quinn
18:52:38 <benjamingoodger> anyway, time for me to go, I have a pub quiz to attend... thirty per cent of my estranged friends have returned to the county while their universities close over christmas
18:54:01 *** [alt]buster has joined #openttd
18:54:14 *** [alt]buster is now known as [com]buster
18:55:04 <Wolf01> who's the pathfinder guru?
19:07:56 <Rubidium> pick any of blathijs/michi_cc/KUDr I reckon
19:08:11 *** ChanServ sets mode: +v tokai
19:14:08 <petern> i seem to have made a directory that only openbve can see :o
19:16:37 <SpComb> if it has a handle open on it
19:17:04 *** [com]buster has joined #openttd
19:23:17 <Wolf01> blathijs, michi_cc : fix the vehicles pathfinding please! they always use a single roadstop and they make stupid tours around the stations to reach the nearest (by bird fly) stop instead of the first free one
19:24:08 <frosch123> that one is assinged to celestar :p
19:25:48 <TrueBrain> gimme his home address, I will sent him a letter!
19:26:28 <Wolf01> post trucks will get stuck!
19:38:44 *** [alt]buster has joined #openttd
19:38:53 *** [alt]buster is now known as [com]buster
20:00:39 *** [alt]buster has joined #openttd
20:00:42 *** [alt]buster is now known as [com]buster
20:02:39 *** FauxFaux has joined #openttd
20:06:25 <blathijs> Wolf01: I think the key there is to make the heuristic drop to 0 when the bird's distance to the station becomes smaller than the station's spread
20:06:43 <blathijs> Wolf01: Or rather, subtract the station's spread from the heuristic might be even better
20:07:08 <xand> is there a known issue where "Loading indicators" stop working after a while? it's set to "Own company" but they no longer appear
20:07:54 <frosch123> blathijs: the problem is, road vehicle queueing does not know about the difference of entering drivethrough stops from different directions
20:08:29 <frosch123> does does it know, that it is useless to drive through one roadstop to reach another one
20:08:44 <Wolf01> but I remember the last time we tried something like this, the performance loss was bigger than the enhancement
20:09:23 <blathijs> frosch123: Can you explain how drivetrhough roadstops are different in this regard? I've never used (nor seen) them
20:09:41 <Eddi|zuHause> what it needs is a list of entry points, reachable bays from that entry point, and exit points of a road station
20:09:53 <frosch123> road vehicle queueing sees that drivethrough stops have 2 slots for vehicles to load
20:10:04 <petern> celestar said he was going to fix them
20:10:14 <frosch123> but when the vehicles approach from one direction, they can only enter one of them
20:10:19 <blathijs> frosch123: Ah, but only one of them for each direction?
20:10:31 <glx> petern: and you said you were going to fix the newgrf warning for msvc ;)
20:10:31 <frosch123> so all vehicles are constantly sent to the stop, as one slot is still free
20:10:36 <Rubidium> petern: he also said that cargodest would be done in two weeks (when he started)
20:11:50 <Wolf01> maybe with "done" he meant "a playable patch"
20:13:38 <petern> would be handy if vehicles could get to the 'wrong bay' anyway
20:13:53 <petern> maybe just if near one-way road
20:14:33 <Eddi|zuHause> things like that _could_ be solved by newgrf roadstations :p
20:14:54 <Eddi|zuHause> define efficient big bus stations
20:15:10 <petern> i don't think richk got that far ;)
20:15:12 <Eddi|zuHause> tram turning loops with several parallel stops
20:15:24 <Eddi|zuHause> yeah, but i can dream ;)
20:16:11 <Eddi|zuHause> but those also must be considered when preparing multistop now
20:16:19 <Eddi|zuHause> otherwise it must be updated again
20:20:38 <blathijs> frosch123: Wolf01: Not sure about fixing the drivethrough roadstop thing, but I think this might be a useful patch
20:21:09 <blathijs> frosch123: Wolf01: It's completely untested (not even sure if it compiles), but gotta go again now :-)
20:21:38 <Wolf01> thanks, I might give it a look when I have more spare time :P
20:21:57 <frosch123> what is it supposed to do ?
20:22:50 <Wolf01> I might try to merge it with the drag&drop roadstops
20:23:43 *** [alt]buster has joined #openttd
20:24:07 *** [alt]buster is now known as [com]buster
20:27:21 *** DephNet[Paul] has joined #openttd
20:30:19 <Wolf01> Yexo, how's going with the freeform edges?
20:30:44 <Yexo> the patch works but needs more testing
20:31:19 <Yexo> I still need to add a setting to disable terraforming the edges and make the NE and NW tiles MP_VOID
20:31:41 <Yexo> but that shouldn't change anything visible (except that you'll lose one row/column of tiles)
20:32:18 <Wolf01> the current sea row/column
20:33:21 <Wolf01> so it's like all the map will be translated by one tile to make room for the southern edge
20:33:49 <Yexo> no, the main problem is that currently it's possible to have buoys on that row of sea tiles
20:34:09 <Yexo> the as long as there are any ships / buoys, you won't be able to enable it
20:34:15 <frosch123> Wolf01: currently you cannot build on the first column/row either
20:35:29 *** DephNet[Paul] has joined #openttd
20:36:29 <Wolf01> uhm, this suggests me that I will have some problems with ships, I usually make very long routes with buoys
20:38:00 <Yexo> Wolf01: only if you place your buoys on the north edges
20:38:17 <Yexo> and then only if you do that in a game were you have not yet enabled the freeform map edges feature
20:38:29 <Wolf01> no, I mean I will need to terraform, or make long canals
20:38:57 *** DephNet[Paul] has joined #openttd
20:40:30 *** DephNet[Paul] has joined #openttd
20:40:31 <Wolf01> or I can connect cities with trains/aircrafts/rvs... but cruise ships are so nice :P
20:43:07 *** DephNet[Paul] has joined #openttd
20:46:21 *** [alt]buster has joined #openttd
20:46:24 *** [alt]buster is now known as [com]buster
20:51:26 *** Vikthor has joined #openttd
20:53:42 <svippy> And why does the newest openttd rely on a lib I don't have?
20:54:07 <svippy> ./openttd: error while loading shared libraries: libicui18n.so.38: cannot open shared object file: No such file or directory
20:54:11 <Eddi|zuHause> the language stuff most probably
20:54:11 <Yexo> svippy: openttd-useful.zip was updated some time ago
20:54:36 <svippy> Eddi|zuHause hit the nail on the head.
20:54:51 <Eddi|zuHause> svippy: support for left to right languages, like persion or hebrew
20:55:13 <frosch123> haha, left to right, how uncommon :)
20:55:29 <svippy> Yeah, English is so crazy, frosch123.
20:55:34 <Prof_Frink> Eddi|zuHause: "suppart"?
20:55:34 <Eddi|zuHause> can't figure out the sed command for that now :p
20:56:26 *** DephNet[Paul] has joined #openttd
20:56:26 <Prof_Frink> (I was halfway through s|$|/| when you added the trailing delimiter)
20:56:50 <svippy> Is OpenTTD looking for translators of Artificial Languages as well?
20:57:29 <frosch123> ido, esperanto: yes
20:57:59 <svippy> Okay, more specific; fictional languages.
20:58:24 <Prof_Frink> Klingon? Pirate? lolcat? 1337?
20:58:24 <frosch123> maybe one or both of bokmal and nynorsk are artificial too :p
20:58:48 <svippy> Qwenya is more my deal.
20:59:27 * Prof_Frink ponders lolopenttd
21:00:34 <svippy> Make up a town naming scheme for that as well, Prof_Frink.
21:01:08 <Prof_Frink> i made you a airyplane, but i crashed it. 100 dead :(.
21:02:00 *** DephNet[Paul] has joined #openttd
21:04:26 *** DephNet[Paul] has joined #openttd
21:13:30 *** [alt]buster has joined #openttd
21:19:24 *** [alt]buster is now known as [com]buster
21:27:22 *** Terkhen has joined #openttd
21:36:07 *** [com]buster has joined #openttd
21:45:24 <SpComb> but building a bouy still just segfaults :(
21:46:40 <SpComb> apparently because there's no towns... because the genmap stuff got into an infinite loop there
21:48:32 <SpComb> yeah, seems you can't really play OpenTTD without any towns, too many NULL pointers all over the place :(
21:49:52 <glx> (required for industries too)
21:50:51 <SpComb> and I need something that's not sea
21:52:12 <Wolf01> use the freeform edges patch and place one single building in the center
21:52:52 <Wolf01> oh, I think I found an unwanted behavior... maybe it's wanted, but not by me :P
21:53:05 <Wolf01> the sea can flood the purchased land
21:54:24 <Eddi|zuHause> Wolf01: you cannot place towns too close to the edge of the map
21:54:57 <Wolf01> place them outside.. there's so many space over there
21:55:15 <blathijs> cool, the patch compiled :-)
21:56:46 <Eddi|zuHause> yeah, widescreen phones :p
21:57:17 <SpComb> svippy: heh, no, it's just an ion3 frame, window mode
21:57:22 <Wolf01> that should run really fast on pda
21:57:26 <SpComb> it's the map size that's the interesting bit
21:57:40 <svippy> What, can't OpenTTD handle small maps?
21:57:47 *** [alt]buster has joined #openttd
21:58:10 <SpComb> svippy: no, there's a lot of uint overflow-below-zero and other assumptions that the code makes
21:58:12 <Eddi|zuHause> at least map size is a power of two :p
21:58:13 <svippy> I hate Hindston, always disrespecting with their one road tile.
21:58:20 *** [alt]buster is now known as [com]buster
21:58:41 <SpComb> Eddi|zuHause: yeah, I started out by just lowering the minimum value for the map_x/map_y values to zero, and I'm now trying this with them both set to two :)
22:00:42 <Wolf01> ahah! you can make mini games which can be used as vista's sidebar widgets, you have a 9 tiles city you can see changing with years
22:01:12 *** ChanServ sets mode: +o Bjarni
22:01:18 <SpComb> the saveload stuff also assumes that the map size is a multiple of MAP_SL_BUF_SIZE or something
22:01:58 <SpComb> oh darn, you can't build stations on the edges of the map either :(
22:02:21 <DorpsGek> Bjarni: bjarni was last seen in #openttd 53 seconds ago: <Bjarni> hello people
22:03:34 <SpComb> dunno, but map_x/map_y are both 2
22:04:10 <Bjarni> does that mean that a 256x256 really is 255x255?
22:04:36 <Yexo> the tiles at the south border are MP_VOID (and thus invisible)
22:05:57 <Bjarni> looks like it's a bit hard to get a railroad though that town
22:06:29 <Prof_Frink> Lol, "pron"ningworth
22:07:31 <SpComb> Bjarni: you just build around the north edge, with stations in the middle :)
22:08:04 <SpComb> hmm... it's not generating any passengers, for some reason
22:08:29 <blathijs> Wolf01: Hmm, the problem I was trying to fix with the patch I pasted you, doesn't really seem to exist :-)
22:15:40 <blathijs> Wolf01: Ah, it's the multistop code that selects a specific tile for the road vehicle to go to, it's not the pathfinder that selects the roadstop at all
22:15:57 <blathijs> Wolf01: In that case, you should probably bug Celestar about your problem :-)
22:19:07 <glx> blathijs: he's already aware of the problem and said he will fix it
22:19:20 *** [alt]buster has joined #openttd
22:20:16 <Yexo> on the other hand the fs bug report is already from april and it's still not fixed
22:20:38 <Yexo> SpComb: nice :) what size map is that, 8x256?
22:21:29 <SpComb> I can't go below a 4x4, as it can't generate any towns then
22:23:49 *** [alt]buster is now known as [com]buster
22:25:58 *** Brianetta has joined #openttd
22:34:59 * SpComb wonders how the TILELOOP_* stuff works
22:36:30 <SpComb> nobody's bothered to comment any of these magical bitmask things :(
22:40:18 <Eddi|zuHause> if you commented magic it wouldn't be magic
22:42:16 *** rangaparmastan has joined #openttd
22:42:22 <SpComb> (why not use IsValidTile?)
22:42:45 *** Nite_Owl has joined #openttd
22:43:21 <Eddi|zuHause> because asserts are optimised out for release builds, so they indicate what you have to make sure before calling
22:43:51 <Eddi|zuHause> you don't have to call IsValidTile multiple times in a row, so it is not there
22:44:26 <SpComb> IsValidTile is completely inline'd, it would read a lot better than some cryptic assert on an "TILELOOP_ASSERTMASK" mask
22:45:06 <SpComb> as far as I can tell, `assert((tile & ~TILELOOP_ASSERTMASK) == 0)` means more or less the same thing as `assert(IsValidTile(tile))`
22:45:06 <Rubidium> but the assertmask doesn't check whether the thing is a valid tile
22:46:44 <Rubidium> SpComb: TILELOOP_ASSERTMASK becomes (bitwise) something like 0000111100001111 for a 256x256 map
22:47:06 <Rubidium> and 0001111000001111 for a 512x128 (or vice versa) map
22:48:36 <Rubidium> so TILELOOP_ASSERTMASK makes sure that the tile is somewhere within the most northern 16x16 tiles
22:50:44 <SpComb> clever, but somewhat hard to understand by looking at the code :/
22:51:21 <SpComb> and the `tile += 9` ensures that the tile loop eventually goes through all tiles?
22:52:38 <SpComb> and this evidently breaks completely when the map doesn't even have 16x16 tiles in the north corner
22:55:05 <SpComb> and TILELOOP_CHKMASK then masks those 0 bits in between the two clusters of 1 bits
22:58:22 *** Yeggstry is now known as Yeggzzz
23:01:28 *** [com]buster has joined #openttd
23:05:07 <SpComb> well, I lowered TILELOOP_BITS to 1, and now it works, but the algorithm is probably horribly broken now
23:31:26 *** ChanServ sets mode: +v tokai
23:43:30 *** [alt]buster has joined #openttd
23:43:33 *** [alt]buster is now known as [com]buster
23:46:21 *** Zealotus has joined #openttd
23:46:48 <SpComb> TGP is somehow broken on the coastline tiles...
23:47:35 <Eddi|zuHause> tgp artificially cuts off the coast line irregularly
23:49:04 <SpComb> the function comments talk about "only for x/y coordinates which are multiples of the minimum edge length"
continue to next day ⏵
