IRC logs for #openttd on OFTC at 2008-07-17
⏴ go to previous day
00:10:50 <ArmEagle> Is the location of the savegames, screenshots, etc directory hardcoded in the client?
00:11:12 <Yexo> no, they are saved in the directory where openttd finds openttd.cfg
00:11:44 <ArmEagle> ah, it will check the gamedir first. then 'myDocuments'?
00:12:01 <Yexo> not sure of the order in which they are checked
00:12:12 <Yexo> iirc gamedir is indeed first
00:21:26 <glx> it's all explained in the readme
00:25:47 <glx> the part about save dir and openttd.cfg is important
00:34:18 *** Eddi|zuHause3 has joined #openttd
01:03:29 <ArmEagle> hmm, now the forums really seem to be down :)
01:18:17 *** fmauNeko is now known as fmauNekAway
01:21:16 * Belugas_Gone wonders if a playme.mp3 would have more chances to be examined than a readme.txt
01:48:21 <DaleStan> Belugas_Gone: Define it as the default first soundtrack.
01:53:37 <Belugas_Gone> going to bed. will be out for 3 days. see you and enjoy the night
02:53:33 *** fly_wolfcn has joined #openttd
05:28:30 *** Ridayah_ has joined #openttd
05:41:03 *** Ridayah has joined #openttd
05:44:19 *** einKarl has joined #openttd
06:01:13 *** michael_ has joined #openttd
06:20:54 *** Sir-Bob has joined #openttd
07:02:42 *** GoneWacko has joined #openttd
07:58:54 *** Brianetta has joined #openttd
08:11:11 *** Wezz6400 has joined #openttd
08:25:58 *** Vikthor has joined #openttd
08:33:12 *** fmauNekAway is now known as fmauNeko
08:41:37 *** Sir-Bob has joined #openttd
08:53:13 <Brianetta> BTW, all: Brianetta's Standard Server is now, once again, a going concern
08:56:15 *** Doorslammer|BRSet has joined #openttd
09:01:47 *** Wezz6400 has joined #openttd
09:21:24 *** grumbel has joined #openttd
09:40:49 *** TiberiusTeng has joined #openttd
10:19:58 *** GoneWacko has joined #openttd
10:25:16 *** Celestar has joined #openttd
10:56:01 <Yorick> meh, I could even set up a new company as spectator by executing CMD_PLAYER_CTRL when I am ingame...right?
10:56:59 * Yorick smells functionality in the shape of a 'play on this server' button
11:04:36 *** einKarl has joined #openttd
11:05:02 <Vikthor> Yorick: That's very good idea
11:06:12 <Vikthor> This would save lot of bandwith, especially on bigger maps, when somebody comes to server to take a look and decides to play
11:07:44 <Yorick> Noldo: when a client wants a new company on joining, he sends a CMD_PLAYER_CTRL
11:08:04 <Yorick> it creates a new company and put a client in there
11:08:41 <Yorick> I figured the current way does work when you're already connected
11:08:51 <blathijs> Yorick: It might require some changes to the code, since there might be things that won't expect spectators to turn into players all of the sudden
11:09:07 <blathijs> Yorick: But in general, reusing an existing command would be an elegant way I guess
11:09:49 <Yorick> it works better than I would expect, because the spetatirs do now turn into players all of the sudden
11:10:26 <Yorick> NetworkUpdateClientInfo(ci->client_index); :)
11:12:55 <Noldo> Yorick: do you have a patch ready?
11:14:07 <Yorick> because I just noticed it when investigating the code for my "Player"-can'tplay patch
11:18:57 *** Progman has joined #openttd
11:19:42 <CIA-3> OpenTTD: rubidium * r13712 /trunk/src/ (6 files in 3 dirs): -Fix: enforce the length restrictions of company and president name in the commands too.
11:28:12 <Celestar> timidity needs more CPU cycles than a 200-train game with Yapp
11:29:10 <Celestar> and I dunno how to use libtimidity
11:30:02 <Rubidium> the only correct usage of (lib)timidity is no usage
11:30:23 <Celestar> Rubidium: suggestions on how to get music playback then? :P
11:30:34 <Celestar> hm I could convert the mids to mp3s :P
11:31:35 <Yorick> hmm, openttd fails to build static since 0.6.2-RC1 here
11:34:23 <Yorick> it goes requireing zlib1.dll and freetype6.dll
11:34:35 <Rubidium> I can't see a change since 0.6.1 that'd've caused that
11:35:12 <Yorick> probably I've just cped too much without clearing the config.cache stuff
11:38:24 <Rexxie> is there a windows mobile port of OTTD?
11:39:21 <Rubidium> if you mean ancient, then yes there is something
11:39:22 *** HenkeSwe has joined #openttd
11:39:24 <Rubidium> but it is not official
11:39:27 *** Klanticus has joined #openttd
11:39:32 <Rubidium> i.e. bugreports don't belong here
11:39:53 <Rexxie> lol, I didnt know if there was a port, you dont have to worry about bug reports ^^
11:39:55 <HenkeSwe> Uhm, I'm wondering where I can find good graphics for the game
11:40:52 <Rubidium> imo the best graphics are on your TTD CD ;)
11:41:43 <HenkeSwe> But i've been looking at some screenshots and
11:42:48 <HenkeSwe> The stations are "in" the industries
11:42:56 <HenkeSwe> I played the game about a year ago
11:43:06 <HenkeSwe> So I don't know if that is standard now or something
11:43:29 <Rubidium> not really, but search the manual (on the wiki) for newgrf
11:43:52 <HenkeSwe> OK, I guess I can find how to make the game swedish in ther also =)
11:45:11 <HenkeSwe> you really should make a huge pack for this =)
11:45:16 <HenkeSwe> Cant wait to start a new game =)
11:45:17 <dih> grfcrawler.tt-forums.net?
11:45:48 <HenkeSwe> Is there any tips you guys have, what things to install and generic stuff before I start a new game=
11:45:52 <dih> i was thinking of it Ammler :-P
11:46:56 <Yorick> openttdcoop.org/wiki/grf?
11:47:19 <Ammler> Yorick: that isn't a tip, which grfs to use
11:47:37 <Yorick> he asks what things to install, I think
11:47:38 <Ammler> it will result with "how can I include all 140 GRFs at once :P
11:47:43 *** ChanServ sets mode: +v tokai
11:48:10 <CIA-3> OpenTTD: rubidium * r13713 /trunk/src/network/ (core/config.h network_udp.cpp): -Fix: possible crash on creating a network packet.
11:48:18 <Ammler> the GRFPack should only recommand to server admins...
11:48:37 *** tokai|noir has joined #openttd
11:48:43 <planetmaker> [13:48] <Ammler> the GRFPack should only recommand to server admins... <- ain't that a bit strict?
11:49:00 <Yorick> when I heard about that grfpack, I tried to nclude all 140 grfs at once, yes :)
11:49:14 <planetmaker> As lazy as I am, I wouldn't know 20% of them w/o the pack...
11:49:17 <Ammler> well, there are also server admins, which aren't able to use it, indeed. :-P
11:50:28 <dih> as of r13713 OpenTTDLib will fail
11:50:57 <DorpsGek> dih: Commit by rubidium :: r13713 /trunk/src/network (core/config.h network_udp.cpp) (2008-07-17 11:47:57 UTC)
11:50:58 <DorpsGek> dih: -Fix: possible crash on creating a network packet.
11:51:10 <dih> sorry for the highlight R.
11:51:12 <planetmaker> let's hope the maintainer of OpenTTDLib is quick with catching up :P
11:51:26 <Yorick> NETWORK_COMPANY_INFO_VERSION = 5, ///< What version of company info is this? ARGH
11:52:15 <Yorick> dih: lets get tcp support into openttdlib then :)
11:53:16 <Yorick> rubidium: how do you mean, possible crash...with the current client limit, it should not be crashing
11:53:36 <Noldo> now you are on to something aren't you
11:53:58 <Yorick> he basically committed the udpfix
11:54:44 <dih> Noldo: that udp packet is just able to overflow
11:54:52 <Rubidium> I can crash any pre 13713 server without password with only the creation of 1 company
11:54:55 <dih> as all details are squished into one packet
11:55:30 <dih> Rubidium: add a howto to the wiki
11:55:59 <planetmaker> wiki/OpenTTD/how-to-be-a-bad-guy.html ?
11:56:07 <Yorick> no, the sabotagers guide
11:56:22 <Yorick> the sharedailing exploit is explained there
11:56:24 <Noldo> Rubidium: isn't that a nice way to get rid of not-uptodate servers
11:56:59 <dih> that way you can find out which servers are not maintained
11:58:30 <HenkeSwe> Can anyone tell me where to find newstations
11:58:38 <HenkeSwe> Its homepage gives me 404 :(
12:02:27 <CIA-3> OpenTTD: rubidium * r13714 /trunk/src/network/ (network_client.cpp network_server.cpp): -Fix: remove the unique_id from the message that a client has joined as it is only exposes the unique_id more than needed. Patch by dih.
12:51:14 <Brianetta> Hmm, think the openttd.org DNS records should have had their TTLs shortened for today
12:54:14 <dih> is it not working for you?
12:57:20 <Yorick> "Due to DNS changes, you may have trouble accessing the site for up to 24 hours after that, if your ISP caches the DNS records for too long."
13:00:33 <dih> Yorick: we do know how dns works
13:01:04 <Yorick> hmm, no servers with 0.6.2 yet
13:01:43 <dih> i need to fix OpenTTDLib tonight
13:01:53 <dih> else my autoupgrading nightly server will kill my website :-P
13:02:37 <Yorick> there should be a check that stops me from creating 8 passworded companies while I'm not a spectator
13:03:09 <dih> there should be a check before you enter any channels or games :-P
13:03:23 <Yorick> CMD_PLAYER_CTRL should not accept players that are not spectator
13:06:06 <Yorick> that allows me to click my 'start company' button 8 times ingame and flood the server
13:06:40 <Noldo> what are you waiting for then?
13:06:59 <Yorick> no servers with 0.6.2 yet
13:10:33 <Rubidium> Yorick: 0.6.2 servers will not come soon
13:20:11 <dih> Yorick: you can go around trashing servers from the list, starting from the bottom
13:21:01 <DorpsGek> dih: rortom was last seen in #openttd 6 days, 16 hours, 36 minutes, and 8 seconds ago: <rortom> gn8\
13:21:32 <DorpsGek> Sacro: Bjarni was last seen in #openttd 2 weeks, 0 days, 17 hours, 13 minutes, and 38 seconds ago: <Bjarni> I didn't have any
13:23:06 <peter1138> rortom is in the ror channel on oftc
13:23:20 <peter1138> or do i mean freenode?
13:23:52 <peter1138> or do i mean quakenet?
13:25:45 <Sacro> or do you mean slashnet
13:26:51 <Yorick> no, quakenet #ror is rise of republic
13:28:10 <Yorick> #ror on freenode is ruby on rails
13:29:11 <peter1138> i said the ror channel, not #ror
13:39:22 <Yorick> ah, at least he's linking to openttdcoop.org
13:39:47 <Yorick> irc://blueyonder.uk.quakenet.org/openttdserver.de
13:47:54 <CIA-3> OpenTTD: rubidium * r13715 /trunk/src/ (51 files in 5 dirs):
13:47:56 <CIA-3> OpenTTD: -Fix [FS#2129]: C-like strings had to be rebound each time they were printed, otherwise the text could change due to the few number of slots that could be used to bind.
13:47:56 <CIA-3> OpenTTD: -Codechange: remove all BindCString and related functions and replace it by RAW_STRING which prints the C-string raw pointer that is on the 'print stack'.
13:57:43 <Celestar> I'm getting false "Train is lost" messages apparently
14:00:52 <Rubidium> I've never seen false train lost messages
14:01:37 <Yexo> I have seen them for very long routes, but 'lost' in that case is discutable
14:01:43 <Yorick> I have never seen train is lost messages
14:02:03 <dih> they are talking about 'falst' ones... you know that
14:02:10 <Celestar> Yexo: "long" as in how long? It's about 150 tiles here from the one station to the other (with no junctions in between)
14:02:10 <Yorick> no true ones, at least
14:02:17 <Yexo> long as in 1000 tiles long :)
14:03:14 <Yexo> but iirc the message is not reset if the route is found after it couldn't found one, so if you removed one piece in the rail, vehicle did pathfinding, couldn't find path, you immediatly replace that piece of rail, vehicle can find path, you'll still get the message
14:03:37 <Celestar> Yexo: yes I'm aware of that, but this isn't the case here
14:04:15 <Rubidium> lost messages are only shown when a train didn't reach it's destination without a specific amount of days
14:05:27 <Rubidium> that got changed once YAPF was introduces
14:05:30 <Yexo> Rubidium: not true when using yapf: if yapf can't find a path it'll return that, and the message is shown immediatly
14:06:25 <Rubidium> Yexo: it was true, until YAPF got introduces
14:07:43 <Rubidium> could be that YAPP is returning that a path is impossible or so
14:09:05 <Celestar> sometimes, this drag-and-drop of orders can mess up stuff
14:10:47 *** badzero has joined #openttd
14:10:52 <Celestar> \o/ I just found the probably biggest bug in my network :P
14:11:06 <badzero> hallo darv ich hier deutsch schreiben?
14:12:05 <Yorick> nein, you can't, it's english only
14:12:55 <badzero> ja! also mal ein lob an die leute die das verwirklichen mit open ttd! echt cool! nur eine frage wie kann es sein wenn ich ein neus game anfange das ich keine bussbahnhof und lkw ladeplatz bauen kann?
14:13:02 <badzero> was muss ich da machen?
14:13:19 <badzero> die pc gägner können bauen nur ich nicht wasn da los`
14:13:24 <Sacro> have i suddenly forgotten how to read ENGLISH
14:14:00 <Sacro> badzero: auf english bitte
14:14:29 <Sacro> oder ich wille der mutter gegessen
14:14:52 * Sacro wonders how many mistakes that sentence has
14:14:56 <dih> Sacro: you are chatting complete and utter bs
14:15:05 <ln> "yes! So get a lob to the people the reality with open ttd! really cool! only a question how can it be when I start the game neus I do not bussbahnhof and trucks ladeplatz can build?"
14:15:24 <ln> Sacro: it's english, it was on google.
14:16:02 <dih> badzero: geh mal in #tycoon.de
14:18:04 <badzero> ich bin in #tycoon.de Sacro
14:18:33 <Yorick> badzero: good, now ask your question there, or ask it in english here
14:18:36 <Celestar> Sacro and other:s problem solves (=
14:18:38 <badzero> ok problem ist schon gelöst glaube ich :)
14:19:08 <Yorick> peter1138: the topic needs a small update, I think
14:20:25 <Yorick> "no non-english speaking"
14:22:41 <dih> c'mon - it was hillarious, one must at least admit that :-P
14:23:14 <Celestar> why do trains have a knack for congregating? :P
14:30:10 <Celestar> 241 trains and counting :D
14:34:26 <Celestar> peter1138: 24% CPU usage
14:34:32 <Celestar> (playing a 2560x1024)
14:35:03 <Celestar> with 1280x1024 I'm down to 17% (and a 20% lower clockspeed)
14:35:15 <Celestar> so still, the blitter takes by far most of the CPU
14:35:35 <Celestar> what interesting, when minized, CPU load grows
14:38:04 <Celestar> yapp/yapf seem not to use a lot
14:38:13 <Celestar> at least not with a measly 240 trains
14:39:09 <Celestar> zooming fully out gets this thing to about 80%
14:42:01 * Celestar wonders when the first ICE appears in DBSetXL
15:15:02 <CIA-3> OpenTTD: rubidium * r13716 /trunk/src/network/network_server.cpp: -Fix [FS#2144]: any player could construct new companies.
15:16:21 <CIA-3> OpenTTD: rubidium * r13717 /trunk/src/players.cpp: -Codechange: place the whole 'create new company' command in #ifdef ENABLE_NETWORK as it may only be executed when we are playing in a network game.
15:25:17 <Yorick> Rubidium: could you also include PLAYER_SPECTATOR in that 13716...argh...I'm stupid...
15:26:01 <dih> hehe - i've been waiting for something like that :-D
15:27:07 <Yorick> I knew submitting that bug was gonna cause me being unable to do that "Set up new company" button
15:34:53 <Eddi|zuHause3> <HenkeSwe> Can anyone tell me where to find newstations <- the website of MB is temporarily unavailable, you can probably find it in the openttdcoop grf pack though...
15:35:27 <Yorick> hmm, have the opntitle.dat grfs been fixed yet?
15:35:29 <Ammler> or at www.ttdpatch.net
15:36:14 <Eddi|zuHause3> Ammler: i thought of that, too, but i didn't know if that might just link to MB's page... and i was too lazy to check
15:37:02 <Ammler> I asked MB for new links for GRFPack 7.1
15:37:56 <Ammler> the other GRFs from him, I loaded up to his thread at TT-Forums...
15:41:50 *** frosch123 has joined #openttd
15:43:02 <Ammler> has someone yapp 8.2 installed?
15:43:24 <Rubidium> I never install OTTD
15:43:46 <Eddi|zuHause3> Ammler: no, i think i'm still with 4.something
15:43:49 <Ammler> Rubidium: because it is always there?
15:44:15 <Ammler> you meant "make install"
15:44:34 <ArmEagle> I should try yapp some time.. Together with being ready for 2x2 main tracks
15:44:49 <Ammler> I am still not able to remove reserved tracks.
15:45:27 <Yorick> we should get a "remove track reservation tool
15:45:53 <Ammler> but he said, he fixed remove tool with 8.2
15:45:58 <Yorick> I've had a few crashes
15:46:01 <Ammler> should now be possible too
15:46:14 <Rubidium> maybe you don't run the 8.2 binary
15:46:31 <Rubidium> like you compiled when you were running so the binary wasn't overwritten
15:46:44 <Rubidium> so the next time you still started an old binary
15:47:33 <Ammler> not installed, but compiled :P
15:49:40 <Eddi|zuHause3> <Rubidium> like you compiled when you were running so the binary wasn't overwritten <- make usually complains about that here...
15:49:54 <Eddi|zuHause3> that "cp" failed, after linking
15:51:26 <Yorick> but copying the lang files does not fail
15:51:43 <Yorick> means you can't run openttd after that point, until you make again
15:51:57 <Rubidium> unless nothing changed in the lang files
15:52:12 <Yorick> don't they get other ids then?
15:52:21 <Rubidium> actually, that happens only when english_uk.txt lang file changes
15:52:22 <Ammler> Rubidium: the bug is easy reproduceable
15:52:48 <Ammler> but before I post in tt-forums
15:57:32 <Ammler> else yapp seems perfect now :-)
15:58:26 <SmatZ> [17:49:39] <Eddi|zuHause3> <Rubidium> like you compiled when you were running so the binary wasn't overwritten <- make usually complains about that here... <== maybe 'rm' before 'cp' would be better
15:59:01 <Ammler> can't reproduce on a new map, so the bug might be in the save itself.
15:59:20 <Eddi|zuHause3> yeah, that'll probably only fail on windows
15:59:37 <Rubidium> on windows removing a file in use fails too
16:00:41 <SmatZ> [17:58:56] <Rubidium> or cp -f ;) <== you got me :-P
16:02:19 <SmatZ> if it won't overwrite the file ... I don't know how this works
16:02:26 <SmatZ> because if the binary file changed
16:02:48 <SmatZ> and it was read from disk, the program would crash
16:03:34 <Eddi|zuHause3> SmatZ: on linux, the file is not physically removed while it is running
16:03:54 <SmatZ> Eddi|zuHause3: yeah for 'rm', but how would 'cp -f' behave?
16:03:59 <Eddi|zuHause3> just the handle from the directory is removed, and replaced by the handle to the new file
16:04:06 <SmatZ> does it remove the old file, or just overwrite it?
16:04:45 <Eddi|zuHause3> well, i doubt it's actually overwriting... but you might have to check on that
16:05:06 <Rubidium> the man says: if an existing destination file cannot be opened, remove it and try again
16:05:23 <Rubidium> makes me suspect that it removes the file
16:05:24 <Yorick> $ cp -f objs/release/openttd.exe bin/openttd.exe
16:05:26 <Yorick> cp: cannot remove `bin/openttd.exe': Permission denied
16:05:49 <SmatZ> good on one hand, but bad on the other...
16:06:05 <SmatZ> but at least cp -f would work on non-windows systems
16:06:09 <Eddi|zuHause3> Yorick: that is exactly what was to be expected
16:09:03 <SmatZ> open(cesta, O_WRONLY | O_TRUNC) <- if this fails, 'rm' is done
16:10:01 <SmatZ> so it tries to overwrite it at first
16:10:12 <SmatZ> but it is not a problem for running program
16:13:30 *** Doorslammer|BRSet has quit IRC
16:15:35 <hylje> MLs done that way still become quite slow under load, no?
16:23:10 *** wolfy is now known as Wolfensteijn
16:23:30 <hylje> not in openttd, no!! :-)
16:23:59 <Eddi|zuHause3> that'll never get into openttd
16:24:17 <Eddi|zuHause3> because Rubidium has a non-optimal cornercase
16:24:23 <Wolfensteijn> the lack of PBS is one of the reasons why I haven't touched openttd (or any ttd) for a year now
16:24:31 <Eddi|zuHause3> that supercedes all improvements
16:25:48 <Rubidium> yup, all crashes are non-optiminal, especially when it doesn't involve crashing vehicles but the binary
16:27:14 * peter1138 remembers the number of train crashes with the old PBS
16:27:26 <Eddi|zuHause3> well... imho that is an issue that might as well be solved after hitting trunk... some issues are even better to be solved there because there are probably more testers
16:27:52 <Eddi|zuHause3> very humble, anyway :p
16:28:33 <Rubidium> it also involves a lot of reviewing and with 200+ kB of patch that takes very long
16:28:47 *** lobster_MB has joined #openttd
16:28:53 <Rubidium> especially because it is in the most complex places of OpenTTD
16:29:16 <Eddi|zuHause3> if it was easy, it would have been done 2 years ago :p
16:29:24 <Rubidium> peter1138's 'increase the max spritecount a little'-patch was much bigger, but much easier to review ;)
16:29:38 <peter1138> And even that had mistakes :p
16:30:22 <Ammler> maybe someone should test, if you can still play old style games with YAPP patch :-)
16:30:24 <peter1138> Well, this is annoying.
16:30:36 <peter1138> I found 4 sticks of 32MB SIMMs, and 2 sticks of flash.
16:31:15 <peter1138> Turns out either all 4 sticks are duff, or the board is duff as it does not like any of them, and the flash is only 8MB :o
16:31:37 <peter1138> (4 * 16MB SIMMs works :o)
16:31:39 <Eddi|zuHause3> SIMMs? are that those, who go on 486/Pentium boards?
16:32:09 <peter1138> It's an IDT R4700 CPU...
16:32:56 <Ammler> the ML on that image does not need bridges for hubs, every train does switch from left to right as needed, junctions are connected on both sides, it workes quite well, we have about 600 trains on that network...
16:33:52 <Rubidium> number of trains per network doesn't tell us a lot
16:34:08 <Rubidium> 600 trains on a 2kx2k map is 'boring', on a 128x128 map it is amazing
16:34:37 <Eddi|zuHause3> what size was the pile transport?
16:38:33 <CIA-3> OpenTTD: rubidium * r13718 /trunk/src/players.cpp: -Fix: possible desync on the autorenew settings 20+ game years (i.e. 4.5+ hours) after a company was started.
16:38:50 <Ammler> maybe the first nongrf game of coop since 2 years :-)
16:40:39 <Eddi|zuHause3> well, YAPP makes a totally different game style possible :)
16:40:43 <fjb> Ammler: Your grf game you once invited me too looked much nicer.
16:41:20 <fjb> Eddi|zuHause3: Single track lines are actuelly working with YAPP: :-)
16:42:04 <Ammler> fjb: yeah, the game style of members quite changed...
16:42:04 <Eddi|zuHause3> single track was also quite good with the old PBS
16:42:12 <Eddi|zuHause3> but twoway stations!!
16:42:32 <Ammler> we didn't make games like pile transport on the MZ long time :(
16:46:21 *** Anon2216 has joined #openttd
16:46:56 <Ammler> Rubidium: the yapp game delivers as much as pile transport.
16:53:21 <Yexo> and how much does it transport?
16:53:46 <fjb> Is each passenger and each ton of cargo a packet?
16:54:55 <dih> Rubidium is on a commit-spree....
16:58:00 <fjb> The C++ syntax is really ugly...
17:10:35 <fjb> Things like this: ((const CargoPacket*)obj)->index + 1
17:10:44 <Eddi|zuHause3> fjb: no, packages with same cargo and same creation time get merged
17:10:56 <fjb> Really hard to read when you don't work with C++ every day.
17:11:10 <fjb> Eddi|zuHause3: Thank you.
17:12:11 <Eddi|zuHause3> up to 64k or something ;)
17:12:21 <Eddi|zuHause3> (and of course same source station)
17:12:57 <fjb> So there aren't 8000 packets at each station...
17:13:36 <fjb> glx: C sytax is ugly too. But C++ makes it even worse.
17:14:13 <Eddi|zuHause3> fjb: the only bad thing about that is that you have to manually cast everything
17:14:56 <fjb> Casting is evel. Why di I have a type save language when I have to cast everything anyway?
17:15:41 <frosch123> fjb: I doubt you would be happier if everything would be templated
17:17:07 <SmatZ> mmm 'ln' is the one I should fake fun of :-P
17:17:13 <Rubidium> Ammler: boring game then ;) my YAPP game deliver 25% more than Pile on 128x128
17:17:15 <fjb> frosch123: Templating is the thing that makes C++ syntax worse compared to C. It is a syntactical nightmare. Ok, they didn't know about usability in the early 70s.
17:19:14 <Ammler> Rubidium: I saw you ships on that map :P
17:19:44 <Rubidium> yup, otherwise I could only reach ~90% of Pile
17:20:18 <Rubidium> just because there isn't enough place to build stations and especially tracks
17:20:47 <hylje> i thought the point was about using those sparse tracks as efficiently as possible?
17:21:19 <Ammler> one of the last nonyapf games
17:25:00 *** KillaloT has joined #openttd
17:28:39 *** fmauNeko is now known as fmauNekAway
17:29:44 <glx> but it's should be possible to do it in a .bat or a .cmd
17:30:22 <Yorick> I've been trying to get file contents read to variables for decades in cmd
17:34:54 <Rubidium> shouldn't that be possible with the shell that's in vista?
17:34:59 <Rubidium> or is that shell like winfs?
17:36:01 <Yorick> powershell isn't in vista
17:37:44 <glx> powershell looks good, but I never tried to install it
17:40:50 <Yorick> powershell is not installed everywhere
17:42:10 <Yorick> neither is bash, on windows systems
17:42:18 <Yorick> I could offcourse include it
17:42:42 <joachim> bash is like the brain... everyone should have it, but...
17:43:33 <glx> Yorick: when .bat or .cmd can't do what you want you can try .vbs :)
17:44:22 <Yorick> I think php could do exactly what I need
17:44:36 <glx> you need to install php for that :)
17:45:18 <Yorick> for the silly users that can not run that script when it's written in 'english'
17:45:42 <Mchl> yeah, most likely php could do that
17:45:42 <Eddi|zuHause3> oh... i remember the times when DOS had macros :p
17:46:10 <Mchl> not sure about zipping though
17:46:15 <Yorick> everyone should get themselves a script interpreter implanted..
17:46:25 <Yorick> Mchl: you could get that with 7za.exe
17:46:31 <Yorick> that's not my main problem
17:46:35 <Yorick> it's actually my least
17:46:53 <Eddi|zuHause3> you could override internal commands with those
17:47:09 <Eddi|zuHause3> like you could replace "dir" with "echo y | del *.*"
17:47:50 <Eddi|zuHause3> of course, both have the same postcondition... you know exactly which files are in the directory afterwards :p
17:48:07 <Yorick> wouldn't del *.* -y do?
17:48:23 <Eddi|zuHause3> i don't think DOS had that kind of switch
17:48:25 <Mchl> where's fun with that :P
17:49:13 <Eddi|zuHause3> i've used Novell DOS most of the time
17:49:15 <Yorick> "dir" -> "Do you really wish to delete the files? y" "Files deleted" -> "NOOO, undo, undo!"
17:49:41 <Mchl> that's why I don't recall having such macros :)
17:49:50 <Eddi|zuHause3> Yorick: actually, "del" was one of the few commands that did not actually print anything if they succeeded
17:50:11 <Eddi|zuHause3> Mchl: i'm sure i read about them in a book for MS-DOS 5.0
17:50:24 <Eddi|zuHause3> Yorick: del was a builtin command
17:50:34 <Eddi|zuHause3> so you had to del command.com
17:50:43 <Mchl> I did all sorts of batching in 6.22, but never such macros
17:50:57 <Yorick> why doesn't powershell come in a single mouth-size exacutable?
17:51:04 <glx> Yorick: del is not an exe
17:51:43 <Eddi|zuHause3> afaik the novell dos command.com wasn't a real .com file, but an .exe renamed to .com
17:52:05 <Mchl> not that it ever mattered for DOS
17:52:22 <Mchl> I had .bat renamed to .exe
17:52:45 <glx> it matters for priorities
17:52:58 <Eddi|zuHause3> .com is tried first, i believe
17:53:33 <Mchl> that;s what I believe too :)
17:54:21 <Eddi|zuHause3> no, .bat was the last one
17:55:26 <Eddi|zuHause3> the order was: macro, builtin, (com, exe, bat)[repeat for each entry in %PATH%]
17:56:52 <kyevan> What's the min spec for openttd on Mac?
17:56:56 <Eddi|zuHause3> where the last part could be overridden by specifying an extension
17:57:21 <Eddi|zuHause3> kyevan: a monitor
17:57:34 <kyevan> Eddi|zuHause3: Oh, I don't have one of those :(
17:57:47 <Rubidium> monitor doesn't matter
17:59:26 <Eddi|zuHause3> Rubidium: ok, but it is recommended for non-server-usage :p
18:02:05 <Mchl> MS-DOS had doskey app for defining macros
18:02:20 <Mchl> still not sure if they were higher priority than builin commands
18:02:35 <glx> I used doskey for command history
18:04:08 <Mchl> and for custom command prompt ;)
18:04:25 <Mchl> was it doskey or something else...
18:09:39 <ln> evening, i'm out of prison again.
18:12:40 <ln> not again, the first time.
18:19:43 <Mchl> should 'private' go before, or after 'abstract'?
18:20:09 <Mchl> I mean, which would you consider more important when reading code?
18:24:19 <ln> Mchl: there's no such rank as 'abstract'.
18:27:55 <Mchl> it's just that PHP manual has an examlpe, where 'abstract; goes in front of visibility
18:29:29 <Eddi|zuHause3> i think in pascal/delphi, "abstract" came after the function declaration
18:31:08 <frosch123> I hope you do not want to declare a function both private and abstract :p
18:32:00 <Eddi|zuHause3> imho, "private" is something that you should not need to repeat for every function... but i don't know PHP
18:33:06 <Eddi|zuHause3> frosch123: i hoped he'd kinda figure that one out on his own :p
18:33:44 <Mchl> yeah... protected not abstract :P
18:33:56 <Mchl> all methods are public by default in php
18:37:28 <Eddi|zuHause3> Mchl: "default" doesn't mean you can't override it for a whole block of functions
18:40:32 * Mchl goes back to documentation
18:41:32 <kyevan> OpenTTD is rather slow on a 400 mhz mac running 10.3 >_>
18:41:43 <kyevan> Then again, what isn't? <_<
18:42:37 <Noldo> kyevan: use smaller map?
18:43:30 <kyevan> Going muchsmaller than 256^2 gets weird >_>
18:43:55 *** fmauNekAway is now known as fmauNeko
18:44:18 <kyevan> Only if you don't use RVs, Mchl :P
18:44:41 <kyevan> Hmm. It also doesn't like the color depth changing under it :P
18:44:50 <Eddi|zuHause3> kyevan: typical performance killers are ships, NPF and ECS
18:45:22 <kyevan> Eddi: Or runing on a machine under spec for the OS it's running on? >_>
18:46:03 *** fmauNeko is now known as fmauNekAway
18:46:05 <Eddi|zuHause3> kyevan: that shouldn't do anything when the OS is actually idle
18:46:45 <kyevan> glx: Wouldn't matter in this case.
18:46:58 <kyevan> Unless Windows runs on PPC iMacs LO
18:49:22 *** ben_goodger has joined #openttd
18:50:10 *** Progman has joined #openttd
19:08:53 <Yexo> are there more industry / cargo grfs except for ecs and pikka's basic industries?
19:09:33 <Yorick> the tourist industry set?
19:10:10 <Yexo> didn't know of that one, thx
19:10:40 <Ammler> but that is "just" a experimental grf
19:12:05 <Ammler> canada set has also newindustries in it...
19:12:33 <frosch123> technically "alpine" also
19:13:26 <Yexo> canada set as in "canadian trains set"?
19:18:06 <Noldo> Yexo: testing your AI ?
19:18:16 <Ammler> seems no new industries, only having all industries in all climates...
19:19:01 <Yexo> but both ecs and pikka's industries are too complicated (mostly because som eindustries need two cargos or they'll stop accepting it)
19:20:31 <Ammler> Yexo: maybe easier for AI then for human :P
19:21:00 <Yexo> maybe, but it required quite drastic changes to my AI
19:21:16 <Ammler> ECS got parameter settings, which allows disabling stockpiling(closing etc.
19:21:53 <Yexo> I'll have too read the readme for ecs sometime for that :)
19:23:07 <Yexo> In the last version I improved station handling, so now industries can use some work
19:25:06 <Noldo> few road are quite yellow
19:25:27 <ArmEagle> it's just one big jam!
19:25:31 <Yexo> most of the vehicles were driving, only a few were actually standing still
19:26:38 <Yexo> unfortunatly the maximum number of road vehicles in openttd is 5000 (the patch settings doesn't go higher)
19:29:16 <Eddi|zuHause3> actually, the german tram set [unreleased] has cargo trams that look like small trains
19:36:05 <yexo> how does the stockpiling work? Is that something openttd knows about, or is it internal to the grf
19:36:31 *** fmauNekAway is now known as fmauNeko
19:38:56 *** Brianetta has joined #openttd
19:39:28 <Rubidium> more or less internal
19:39:51 <Eddi|zuHause3> yexo: the grf uses callbacks for that
19:40:03 <yexo> so there is no way an api could be created for the ai to now the amount of goods stockpiled
19:40:28 <Eddi|zuHause3> i'm not sure about that...
19:40:31 <Rubidium> the amount of goods stockpile can be determined fairly easily (IIRC)
19:40:44 <Rubidium> how much can be stockpiled is something completely different
19:41:36 <Eddi|zuHause3> last time i checked, PBI put status messages like "coal stockpile is getting full (750t)" in the industry window
19:42:13 <Eddi|zuHause3> yexo: you can definitely find out that the stockpile is full, when the station stops accepting the cargo :p
19:42:13 <frosch123> google cannot translate between english and noai
19:42:14 <yexo> Eddi|zuHause3: the grf can put whatever it want in the industry window, but as long as that is done by the grf and not by openttd it's useless for an ai
19:42:58 <Rubidium> as I said, only the amount stockpiled can be determined by OTTD, the maximum amount and such cannot be determined by OTTD
19:43:00 <Eddi|zuHause3> yexo: yes, but you can extend the grf spec for status replies like that
19:43:26 <Eddi|zuHause3> and then ask the grf authors to implement the AI callbacks
19:43:29 <Rubidium> (except actually stockpiling cargo till it's full)
19:43:39 <yexo> Rubidium: where do I need to start looking in the code?
19:44:31 <Eddi|zuHause3> i'd suggest reading the grf spec on industries, and then following the code for these specs
19:45:54 <frosch123> newgrf_callbacks.h is also a nice way to find something for grepping
19:49:38 <frosch123> except you won't find the production callback there
19:50:26 <DorpsGek> Ammler: yorick was last seen in #openttd 38 minutes and 48 seconds ago: <Yorick> :Y
20:02:06 <Ammler> I like to put the GRFPack outside of the web , which function do I need in php to make a script download.php, which loads the pack?
20:03:21 <Ammler> (no direct links to the GRFPack should be possible)
20:03:50 <glx> send the header then outputs the file
20:04:08 <ccfreak2k> No, don't send the header!
20:04:15 <ccfreak2k> That's exactly what they expect you to doQ!
20:05:35 <yexo> readfile will do the trick
20:05:54 <Ammler> fopen and fpassthrough sounds good
20:06:32 <yexo> that's the same as readfile, but readfile is easier ;)
20:06:41 <Ammler> yep, just read about...
20:07:03 <Eddi|zuHause3> real programmers use a butterfly...
20:07:23 <Rubidium> why is readfile easier?
20:08:38 <yexo> Rubidium: what is easier: readfile(filename); or fpassthr(fopen(filename, "r"));
20:08:38 <Ammler> Eddi|zuHause3: something with rewrite?
20:08:47 <Rubidium> still, there isn't much that you can do about hotlinking without breaking some browsers
20:09:11 *** glx|away has joined #openttd
20:09:11 *** ChanServ sets mode: +v glx|away
20:09:27 *** glx is now known as Guest1403
20:09:27 *** glx|away is now known as glx
20:10:19 <Rubidium> readfile/fpassthru seems to be 55% slower than feopen/feof/fread/echo
20:11:13 <Rubidium> according to one of the comments on the php documentation page of fpassthru
20:11:14 <SpComb> gah, benchmarking PHP I/O
20:12:24 <SpComb> Ammler: what is it that you want to do exactly? "outside of the web"?
20:12:56 <SpComb> do you just want to disable hotlinking to the .zip? Tell your web server to do that
20:12:57 <Ammler> I wan't to make the GRFPack not accessable from web
20:13:13 <SpComb> Ammler: so kill your http server and unplug your server's network connection, then
20:13:19 <CIA-3> OpenTTD: rubidium * r13719 /trunk/src/ (42 files in 3 dirs): -Codechange: rename IsValidPlayer to IsValidPlayerID in line with all other structs/classes that are in a pool.
20:13:22 <SpComb> preferrably, you could also power it down and remove the power cable as well
20:13:58 <SpComb> I presume you mean "accessible in a limited way"
20:14:01 <Eddi|zuHause3> SpComb: he wants to enforce that people link to the index page, which features a kind of readme and a download link
20:14:19 <SpComb> right, that kind of stuff is best done in your web server's configuration
20:15:32 <Eddi|zuHause3> all those hosters who disallow image hotlinks must have some wayys
20:15:33 <SpComb> probably not the best one, but something like that
20:16:20 <SpComb> preferrably something that would redirect the user to another URL if they try and hit a download URL without having the correct referrer
20:16:30 <Ammler> I will check that with our web gurus :-)
20:16:41 <SpComb> I would know how to do it with lighttpd, but I don't use Apache that much
20:17:32 <Ammler> well, what about browsers without referrer?
20:17:45 <SpComb> tough luck for them, I guess
20:18:18 <yexo> or allow openttdcoop.org and empty as referers
20:18:25 <yexo> only disallowing strange sites
20:19:09 <SpComb> if it's really important then you could also script something with e.g. cookies, but you'd really want to be able to avoid having to ship the file data around yourself, and leave that to your httpd
20:19:39 <SpComb> (lighttpd lets a script set an "X-Lighttpd-Sendfile" header that tells lighttpd to send the contents of a file)
20:20:20 <Ammler> I thought more of something like needed a key if referrer not openttdcoop.org
20:23:18 *** lobster_MB has joined #openttd
20:24:31 <Noldo> Ammler: why don't you just redirect to the download page if the referer doesn't match
20:26:01 <SpComb> Noldo: < Ammler> well, what about browsers without referrer?
20:27:26 <Ammler> same answer as SpComb :P
20:28:03 <Ammler> I could redirect those without right referer
20:28:21 <Ammler> and keep the IP, if they try a 2. time, I can allow it...
20:28:48 <Noldo> why are you doing this in the firstplace anyway?
20:29:25 <Ammler> people should download the pack from our page
20:29:30 <fjb> Did the airport noise calculation change?
20:29:37 <Ammler> for credit page and readme etc.
20:30:03 <Rubidium> don't forget the ads
20:30:24 <Ammler> but indeed, it shouldn't be that complicated...
20:30:39 <SpComb> have the download page create an URL that contains the hash of a secret and a timestamp/ip, and then validate that in the valid_hash
20:30:58 <SpComb> that's the more complicated solution for those cases where you don't have a referrer to go by
20:31:12 <SpComb> allowing empty referrers is also a valid solution, far simpler
20:31:21 <Ammler> Rubidium: don't think so, that those counts :P
20:31:41 <Noldo> why would you want to be nice to browsers why don't want to tell where they came?
20:33:08 <SpComb> and that's also true, not everyone wants to save-as the download from the website, breaking wget-from-your-server is pretty annoying
20:34:01 <Rubidium> what's more annoying is download from website, go with shell/tool to file, copy to the server *again* typing the full path
20:34:05 <Noldo> well if the point is to make people visit the download page then there is no way to be wget friendly
20:34:24 <Rubidium> instead of "copy link location", alt-tab, wget "paste"
20:34:36 <Ammler> well, I need to discuss it with my mates anyway...
20:34:41 <SpComb> well, if you really cared about that, then those magic-hash-token URLs would work with wget as well
20:35:42 <Ammler> Noldo: I thought about a key or something for "friends"
20:35:48 <Noldo> limiting the pack to grfs that don't require this kind of stuff would solve it too
20:35:55 <Ammler> like zuu's windowsupdater and such...
20:37:27 <Ammler> Pack without the distributable GRFs is quite empty
20:37:51 <Ammler> MBs GRF would be almost the only GRFs then
20:38:48 <SpComb> and if they were all distributeable then you could distribute them automatically from within OpenTTD...
20:39:38 <SpComb> yes, and then add a "can I distribute this?" flag to .grf files
20:39:57 <Ammler> Oskar should include a direct download link to the Crawler
20:40:00 <SpComb> or could you just add a license tag to .grfs, and then decide if you're allowed to distribute them based on that?
20:40:22 <SpComb> although that presumeably doesn't cover the "read the readme before using this .grf" thing
20:40:22 <Noldo> SpComb: there is no technical solution to licensing
20:40:32 <SpComb> Noldo: of course not, we're talking about DRM here
20:41:22 <Noldo> I see reference checking as a kind of DRM too
20:41:31 <Ammler> easiest would be a direct download link on the crawler, if someone like to support it, he just need to include the link...
20:42:48 <Ammler> and maybe a HTML viewer in ottd for the readme :P
20:43:37 <dih> an HTML viewer? you are not thinking of including that in the game are you?
20:44:27 <dih> sure, grab an opensource browser and build it into OpenTTD :-P
20:45:10 <Rubidium> and then release openttd even more often because of security errors in the browser?
20:45:10 <Ammler> well, I thought, it might be SDL has already something like that...
20:45:15 <dih> OpenTTD would get a pretty bulky thing
20:45:42 <dih> where 4 MB actually is game
20:46:06 <Ammler> you do not need script support etc.
20:46:45 <dih> you just need to perfectly render xhtml
20:46:51 <Ammler> just some basic formats like bbcode or so
20:47:00 <dih> oh - and then perhaps include css support, because everybody will want that
20:47:13 <dih> and then the width of the div is not the same as in browser x
20:47:35 <Ammler> there is only one browser then :P
20:47:41 <dih> you can see where that would be heading can you? :-P
20:48:10 <Prof_Frink> rewrite the gui in Qt and use webkit
20:48:12 <Noldo> for everyrhing else there is emacs
20:48:22 <Ammler> yeah, I would going to use ottd also for browsing the internet :-)
20:48:42 <dih> Prof_Frink, probably then resulting in even less os's being supported
20:49:16 <SpComb> there's plenty of browser-widgets in libraries, depends on your dependencys, I guess
20:49:46 <Ammler> and last, a text only viewer would be enough too
20:49:54 <SpComb> for viewing HTML source?
20:50:08 <dih> and then 'auto-download' 50 grf's and let the game display all readme files just to satisfy the authors
20:50:36 <dih> if you leave it the way it is, it's up to the user
20:50:41 <SpComb> dih: and a big fat "Next »" button
20:50:50 <Ammler> dih: how many grfs do you know with readmes?
20:51:06 <dih> and you can only click the button after scrolling right to the end
20:51:47 <dih> point being: it should in no way be OpenTTD's duty to make sure the readmy has been read
20:51:56 <dih> as long as there is no distribution, that is done :_P
20:52:04 <dih> it's up to the client to do that
20:52:13 <dih> and that is fine, is it not?
20:52:20 <Ammler> that would support the authors wishes even better...
20:53:26 <Eddi|zuHause3> hm... this is the second time i had rpm segfault on me...
20:54:01 <Prof_Frink> You know it makes sense.
20:54:30 <dih> but that sounds too close to Eddepp
20:54:41 <dih> and we dont want that now, do we :-P
20:55:21 <SpComb> load the remotely for use, and then just "cache" it locally... that's not distribution, is it? :P
20:55:49 <SpComb> perhaps require the user to download it separately if they want to use the newgrf to start a new game/server
20:55:51 <dih> give more network control to grf's
20:56:02 <Eddi|zuHause3> by german law, "distribution" needs the transfer of a physical object
20:56:03 <dih> then write a grf that reads and sends new packet types
20:56:08 <SpComb> but download it if they are just acting as a client
20:56:10 <dih> and implement the system with a grf
20:56:27 <Eddi|zuHause3> downloading and copying does not constitute "distribution"
20:57:40 <SpComb> perhaps NewGRF authors would be OK with automatically downloading it for network game clients, but not have those be valid for starting new games/servers?
20:57:56 <DaleStan> And how would you enforce that?
20:58:29 <SpComb> how would you enforce the NewGRF author's rights and license conditions with the current state of affairs?
20:59:17 <Rubidium> by not providing the 'tools' to actually do so?
20:59:21 <SpComb> not so sure if "enforce" is the right term to use, because that's impossible, even without an autodownload system
20:59:28 <SpComb> dunno, someone else could provide them
20:59:37 <Ammler> SpComb: you are at #tycoon too
20:59:50 <Ammler> so ask next time you see oskar there...
20:59:53 <DaleStan> No, enforce "not valid for starting new games".
21:00:30 <Ammler> adding new fields to GRFCrawler...
21:00:41 <SpComb> DaleStan: well... it's a question of intent, and what the NewGRF authors think of that
21:00:54 <Ammler> like direct download link
21:01:07 <dih> i just got a Bus Error with 13718
21:01:13 <SpComb> so e.g. you could store them without the GRF name, just the GRFID
21:01:46 <Ammler> I thought about renaming all GRFs in the pack to that :-)
21:02:03 <SpComb> that wouldn't stop someone from writing a tool that just extracts those for normal use, but someone could also write something like that _right now_, and the newgrf authors would be powerless to stop it
21:02:19 <SpComb> unless they sued/C&D'd whoever was running the server for it
21:02:47 <Prof_Frink> dih: That's a *stupid* number for a bus.
21:03:10 <DaleStan> Someone could write something like what? There aren't currently any protections to evade.
21:03:47 <dih> Program received signal EXC_BAD_ACCESS, Could not access memory.
21:03:47 <dih> Reason: KERN_PROTECTION_FAILURE at address: 0x0000004e
21:03:47 <dih> 0x00148040 in Utf8Decode ()
21:03:55 <dih> when i try to view details on the DMU
21:04:59 <SpComb> DaleStan: well, e.g. NR-TTD contains a component that downloads a bunch of .zip files and extracts them, which is, to some degree, similar to what we're discussing now
21:05:24 <Brianetta> Bah, had to invite myself
21:05:32 <Brianetta> [22:05] --- You have been invited to #coop.members by ChanServ (kinetic.oftc.net)
21:06:08 <SpComb> although it downloads them from them from whereever the GRF author hosts the file normally over HTTP, and not from e.g. some random user's server
21:06:16 <Progman> Brianetta: you're lucky, you got an account at nickserv....
21:07:01 <Brianetta> Progman: Well, luck had less to do with it than registering...
21:07:19 <SpComb> someone could maintain a database that maps GRFIDs to HTTP URLs, and then some client-side module that maps the server's GRFs to HTTP URLs through that, and downloads them
21:08:46 *** dragonhorseboy has joined #openttd
21:09:20 <Ammler> would you as server admin trust that url?
21:09:52 <SpComb> Ammler: well, dunno, you as a server admin could operate your own DB or something
21:10:19 <Ammler> that's also a reason for the pack
21:10:31 <SpComb> and then your clients would need to trust you... remote code execution via grf download...
21:10:36 <Ammler> because they mostly remove old versions
21:11:04 <Ammler> well, It makes sense to trust the server you like to join, doesn't? :-P
21:11:16 <SpComb> (if you download a .grf for GRFID=482394 from server A, and then server B tells you to load that GRFID, do you download it again or use the one you have?)
21:11:27 <SpComb> although perhaps providing the md5sum would solve that issue as well
21:11:42 <Ammler> [23:07] <Ammler> GRFID and md5sum
21:11:57 <Eddi|zuHause3> <Prof_Frink> dih: That's a *stupid* number for a bus. <- why? trains have that kind of numbers all the time
21:12:22 <SpComb> but yes, implementing such a HTTP-redirect server on the OpenTTD server side, and a HTTP client on the client side probably wouldn't violate any licenses
21:12:25 <Eddi|zuHause3> and busses could easily be used as train replacement for certain situations ;)
21:12:58 <Eddi|zuHause3> germans have the word "Schienenersatzverkehr" for that ;)
21:13:01 <dragonhorseboy> the question would be..could this also work with older-than-current grfs of the same grfid?
21:13:26 <SpComb> unless the GRF authors continue to provide the old versions for download
21:13:36 <dragonhorseboy> then there's not much point if its buggy for the typical server that doesn't have same of everything
21:13:41 <SpComb> which they have to do in any case... directly or via the ottdc grfpack
21:14:21 <Eddi|zuHause3> i don't think that download feature is worth the trouble
21:15:19 *** Klanticus has joined #openttd
21:15:19 <peter1138> I could implement it just to annoy OzTransLtd further
21:15:20 <Ammler> Eddi|zuHause3: thanks to "our" pack, but ask those you see our grflist and first grep the internet for all GRFs
21:15:49 <SpComb> the current solution is to make a .zip of the NewGRFs you use in your server, and then (illegally) distribute it to your friends - or alternatively, tell them to download the ottdc grfpack
21:16:04 <Eddi|zuHause3> you can search for grfids in grfcrawler
21:16:08 <Ammler> then they join #openttdcoop and tel they have now found all grfs but can't still join.
21:16:16 <CIA-3> OpenTTD: rubidium * r13720 /trunk/src/vehicle_gui.cpp: -Fix (r13715): string wasn't correctly passed to draw function.
21:16:17 <SpComb> (grfcrawler has an XML API, btw)
21:16:25 <Eddi|zuHause3> SpComb: make a .tar, and they won't even need to extract it ;)
21:16:27 <fjb> peter1138: :-) You are evil.
21:17:23 *** fmauNeko is now known as fmauNekAway
21:18:25 *** fmauNekAway is now known as fmauNeko
21:18:33 <SmatZ> jen ta prace s tim balenim ma hodnotu!
21:19:24 <Ammler> SpComb: if we would like to use GRFCrawler, it needs to change to a wiki style
21:19:38 <Ammler> so everybody can add and edit entries.
21:20:09 <SpComb> haven't seen eis_os on #tycoon for a month
21:21:01 <Ammler> and wiki style also would need versioning... :-)
21:21:23 <SpComb> Ammler: well, feel free to write your own... :)
21:21:53 <Ammler> yeah, but first we need to know, that eis_os doesn't wan't to
21:22:46 <peter1138> He doesn't, because he asked the GRF authors and got no positive responses.
21:23:16 <Ammler> asked about "open" the crawler?
21:24:08 <Ammler> I have something similar like cia.vc in mind
21:24:32 <Ammler> active authors could claim exclusive rights for their entries.
21:24:36 <Eddi|zuHause3> SmatZ: after running it through google translator, i'm still not sure if that was cybersex or not :p
21:25:12 <peter1138> Right, but authors can't be bothered with all that.
21:25:30 <peter1138> They're perfectly content with just posting in a version in a forum thread.
21:25:30 *** GoneWacko has joined #openttd
21:25:43 <Ammler> then they do not need to make it exclusive and we would like to help
21:25:53 <Eddi|zuHause3> so we actually need a spider crawling through the forums
21:25:59 <SmatZ> Eddi|zuHause3: hehe, not :)
21:26:26 *** Wezz6400 has joined #openttd
21:27:12 <Eddi|zuHause3> i'm leaning towards yes, because i have read on bash.org how that "wrong window" thing works :p
21:28:39 <Ammler> http://grfcrawler.tt-forums.net/index.php?do=search&type=grfidlist&q=6D620900,%2044440A01,%2045520102,%2045520400,%2050430301,%2043415463,%205443E104,%2058534453,%204D413032,%2054560202,%20504E0002,%2052570103,%2052571201,%2044440401,%204A430002,%2045530400,%20524B0301,%204D656F17,%2045520500,%206D620004,%206D620601,%2043415261,%2053470102,%205257131E,%205257131D,%204D413034,%2052571302,%2041424631,%20659E5728,%2052455200,%2052453C00,%204C410002,%2
21:28:39 <Ammler> 052455300,%2074677462,%2056430001,%2044444040,%20554B5400,%20504E0001,%2054570200
21:29:03 <Ammler> around 60% is in the crawler
21:29:04 <SpComb> meh, reading through the most recent "NewGRF Web Service" topic, I honestly can't see any stark facts that would make such a system impossible
21:29:48 <peter1138> There's no technical reason it can't be done.
21:30:13 <SpComb> and I beleive that it's possible to work out acceptable solutions to the social/political/legal issues
21:30:30 <SpComb> the car analogy that I like to use here is that people *die* every day in motor traffic, yet people *still* drive around in cars each day
21:30:33 <Ammler> 1. it would need a repo and then some authors which participate
21:30:35 <peter1138> Even so, I'm going to bed.
21:30:37 <SpComb> a system doesn't have to be perfect
21:31:17 <SpComb> and in the end, I'm sure most NewGRF authors just want people to actually play/use their GRFs...
21:31:41 <Ammler> until now, nobody complained about our pack
21:32:12 <Eddi|zuHause3> apart from "i can't load all grfs at once" :p
21:32:21 <yexo> SpComb: the main reason I hear every time: Who can upload new grfs to the service? If you only allow grf authors, that list (who can upload) has to be maintained. If you allow everyone to upload, what about people uploading grfs from authros that don't allow that?
21:32:40 <glx> Ammler: you pack theorically only contains latest version of any grf
21:33:05 <SpComb> yexo: indeed, and it's probably impossible to figure out a perfect solution to that issue... but that doesn't stop you from implementing something that works mostly correclty
21:33:18 <SpComb> YouTube was mentioned in that thread, and that example is still relevant here
21:33:36 <Ammler> glx: yes, older versions aren't accessable anymore
21:33:46 <Ammler> except the legacy packs
21:33:59 <yexo> of course you can implemenbt something that works mostly, but if by doing so you make some grf authors angry (who therefore make their grfs incompatible with openttd), I won't be happy with the system
21:34:39 <SpComb> if you implement the system reasonably, then any NewGRF author that goes into counterattack mode is pretty unreasonable
21:34:51 <SpComb> and I'm under the impression that very few people play TTDPatch anymore
21:35:00 <Eddi|zuHause3> i still think an automatic spider kind of thing would be better than a wiki style repo
21:35:19 <Eddi|zuHause3> so the authors don't have to do double and triple accounting
21:35:42 <Ammler> Eddi|zuHause3: I thought the GRFcrawler worked that way some years ago...
21:36:17 <SpComb> but yes, I understand that it's clearly a horribly bad idea for the OpenTTD devs to include something in trunk that copies the .grf data from the server to the clients
21:36:22 <SpComb> it would have to work somewhat differently
21:37:00 <SpComb> because I can understand why GRF authors would be upset by that
21:37:02 <Eddi|zuHause3> SpComb: once there was an idea to provide grfcrawler links, but that kind of needs a browser dependency
21:38:09 <Prof_Frink> SpComb: So have the server provide an URL for the grf, and the client wget it
21:38:38 <ArmEagle> hmm, any chance it's now the wiki that's moving?
21:38:47 *** _Lakie_ has joined #openttd
21:39:03 <SpComb> Prof_Frink: yes, I suggested that half an hour ago
21:39:20 <Prof_Frink> SpComb: Great minds
21:39:34 <SpComb> and I also noted that e.g. Nr-TTD contains a module that does just that
21:39:50 <SpComb> (well, wgets a list of .zips/.grfs automatically)
21:40:05 *** _Lakie_ is now known as Lakie`
21:40:27 <Ammler> has not the TTDPatch configurator also something like that?
21:41:38 <SpComb> and I never heard a single complaint from a NewGRF author about it, even though I'm sure several were aware of it
21:43:13 <SpComb> wolf makes a lot of good points in that thread
21:45:29 <Eddi|zuHause3> i have never heard of a "nr-ttd" before
21:46:32 <SpComb> Eddi|zuHause3: have you been active in the TTDPatch section of the forums during the past two years?
21:47:15 <Eddi|zuHause3> i have only sporadically visited the forum
21:47:33 <SpComb> it's not related to OpenTTD in any way, so it's natural that you've never heard of it
21:49:23 <SpComb> perhaps what one needs is input from people who are actually NewGRF authors, not just a bunch of game devs speculating about stuff
21:50:55 <SpComb> a) require the GRFID and md5sum to be able to download the file b) a bit more difficult... would require something in the OpenTTD GUI...
21:51:19 <SpComb> well, not really. the ottdc_grfpack doesn't show anything in the GUI either, so nevermind that
21:51:59 <SpComb> if you use the same interpretation of "credit is given, with a link to my website" as the ottdc_grfpack, implementing that shouldn't be too difficult
21:53:07 <Eddi|zuHause3> "credit" is easy, just pack a readme and the grf in a .tar file
21:53:12 <Ammler> what I got from the authors is, that they like, if you read the their readmes and guides to configure the newgrfs proper
21:53:42 <Ammler> so you do not have ugly things on the map because of missconfiguration
21:53:46 <SpComb> except network game clients can't configure the NewGRFs used by the server in any case
21:53:49 <Ammler> and blame the GRF for that.
21:53:52 <SpComb> that's only relevant when creating a new game
21:54:48 <Eddi|zuHause3> so, provide a "read once" flag somewhere, and then show the readme (textfile) in openttd when adding it to the grflist for the first time
21:54:59 <Ammler> yes, and because a joining client can't confiugre it, id doesn't matter if they can download it in a pack.
21:55:08 <SpComb> Eddi|zuHause3: meh, it's not like the ottdc grfpack implements that kind of behaviour, does it?
21:55:31 *** Lakie is now known as Lakie_
21:55:34 *** Lakie` is now known as Lakie
21:55:43 <SpComb> if it was implemented it would need to apply to the ottdc grfpack as well
21:55:56 *** Lakie_ is now known as Lakie`
21:56:07 <SpComb> (and then PikkaBird's "ps, you may want to update av8 to 1.31" request could also be handled by letting PikkaBird update the newgrf himself)
21:56:27 <Eddi|zuHause3> yes, it would apply to any grf, the readme is shown unless the grfid is already stored in the .cfg as "already read"
21:57:31 <Eddi|zuHause3> but this is actually separate from any kind of download feature
21:57:48 <Ammler> SpComb: as I said, most (or all) authors are happy to have the grf in the pack
21:59:04 <SpComb> Ammler: when building the grfpack, was that tt-forums thread the only method that you used to get permission from the authors?
21:59:28 <SpComb> or did you contact induvidual authors directly and ask them?
21:59:36 <Ammler> we also mailed with about 2-3 authors.
22:00:39 <Ammler> as we thought about the pack, we thought about to contact every author
22:01:11 <Ammler> but then, it was difficult to determine who is resposible for the GRF etc.
22:01:22 <Ammler> so I made that thread.
22:01:33 <Ammler> and Brianetta helped me ::-)
22:02:00 <SpComb> so the implication is that most of the .grfs in the grfpack are there because the authors themselves came forward and gave you permission?
22:02:42 <Ammler> the grfs are in because nobody forbids me to do it...
22:02:57 <Ammler> I guess, foobar is one of the only
22:03:19 <Ammler> I asked because he strictly forbid distributing in his readme.
22:03:22 <SpComb> hmm... are most of those GRFs released with an explicit license that you follow?
22:03:48 <Ammler> do not ask such question...
22:04:39 <SpComb> well, I think the topic merits further discussion, including discussion with NewGRF authors, once we have some kind of concept
22:05:25 <Ammler> you can count of help from us, we can't play without grfs anymore...
22:06:34 <SpComb> Ammler: have you ever had a case where a NewGRF author explicitly forbade you from distributing their GRFs?
22:06:54 <Ammler> I once asked eis_os for his 2 grfs :-)
22:07:24 <Ammler> (he coded his grfs to not run with ottd)
22:07:50 <Ammler> sometimes, authors ask us to include their grfs...
22:08:03 <Ammler> but nobody did forbid us to :-)
22:08:48 <SpComb> solving the technical implementation is indeed not an issue, once we manage to work out the social/political/legal ones
22:09:00 *** kolobos has joined #openttd
22:09:08 <Ammler> technically, it is the versioning...
22:09:41 <SpComb> the OpenTTD server already provides the md5sum, doesn't that solve any versioning issues?
22:09:50 <SpComb> or what do you mean with them?
22:09:59 <Ammler> I mean you need to keep different versions availabe...
22:10:51 <Ammler> that's why you can't use just a link
22:11:55 <SpComb> the clients can download, store, and use the grfs based on the md5sum - the repo can keep several versions available - although it does need some kind of indication/warning so as to not make outdated/legacy GRFs an issue
22:12:22 <SpComb> but the NewGRF author being able to add updated versions right away themselves should counter some of the negative aspects of that issue
22:12:44 <Ammler> outdated GRFs should only be availabe for a running serverm, not a new one :-)
22:13:10 <SpComb> there's a distinction between using a GRF as a network client, and using a GRF as a singleplayer/network server
22:13:38 <Ammler> and you should be able to "help" the authors by keeping the REPO up2date
22:13:52 <SpComb> not a cryptographically secure disctinction, though
22:13:59 <Ammler> some authors even don't care about at all,
22:14:15 <Ammler> I can imagine, George would be one of those...
22:14:57 <Ammler> he dosn't care, if we use his GRFs, but he does not wan't more work with it.
22:15:18 <Ammler> so we need to maintain the repo for him...
22:16:49 <SpComb> the way I see it, there would have to be a central database containg GRF metadata (GrfID, md5sum, author, website, etc. - not the same as GRFCrawler, though), and if this were to make it into OpenTTD trunk, then the official client would refuse to automatically download the GRF unless it was in this database
22:17:41 <Rubidium> what is 'running' server and what is 'new' server?
22:18:08 <SpComb> Rubidium: network client, network server
22:18:13 <Ammler> loaded map is running, starting map is new
22:18:43 <Rubidium> looks like ottdcoop will not be able to run a game version of two years back as the 'current' newgrfs wouldn't be compatible with that version
22:18:59 <Rubidium> (read as if in 4 years time or so)
22:19:43 <Ammler> well, if the server wants to use the the repo support...
22:19:46 <Rubidium> but starting a new map doesn't fetch newgrfs from a server I reckon
22:20:00 <Ammler> we could still load the game and save it with new grfs.
22:20:16 <Rubidium> kolobos: complain at your ISP for not having a properly configured DNS server
22:20:34 <SpComb> (and note that when I say "the official client would refuse to automatically download the GRF unless it was in this database", I don't imply that "if the GRF is in this database, then everyone is free to do whatever they want with it" - these are all partial ideas...)
22:21:06 <Ammler> SpComb: one important thing is, you can't exc
22:21:28 <Ammler> expect grf authors add the grfs self to the database
22:21:55 <Ammler> but not because the do not like it, they just are too lazy or gone or what ever...
22:23:03 <SpComb> I'm sure if you can get the explicit agreement of some subset of NewGRF authors then you should be pretty safe making some kind of assumptions about the majority of the rest
22:23:57 <SpComb> with some kind of sane process in place to correct mistakes arising from that, most authors shouldn't have much to complain about
22:24:36 <Ammler> if we could provide them with nice statistics
22:24:44 <Ammler> they would really be happy
22:24:54 <Ammler> how much is my GRF used etc...
22:25:03 <SpComb> Ammler: out of interest... what's the reasoning behind "you may only download the ottdc grfpack directly from us"? Not having old, broken versions of it laying around?
22:25:34 <Ammler> and knowing who is using it
22:26:14 <Ammler> if someone would complain, we can then remove his grf
22:26:19 <SpComb> if this system was conveniently integrated directly into OpenTTD, I'm sure most people would be reluctant to download outdated GRFs from anywhere else :)
22:26:27 <Ammler> and should be sure, it isn't availabe anymore...
22:28:44 <SpComb> this would be one of the inherent properties of a centralized GRF database
22:30:12 <Ammler> hmm, you could also mark the grfs then with something like "stable", "experimental" etc....
22:31:03 <Ammler> or other experiences...
22:31:20 <dragonhorseboy> ammler... "stable" v1 or "stable" more-vehicles-to-use v2?
22:31:40 <dragonhorseboy> thats why I prefer if people would just leave any grfs they made up .. not taking one down unless it was buggy
22:31:51 <SpComb> (also useful to note that GRFCrawler shares the users database with tt-forums, which can also be used for this)
22:32:22 <Ammler> maybe not only user database
22:32:53 <Ammler> except ID/md5sum/file itself :-)
22:34:55 <Ammler> dragonhorseboy: you trash the old prodcut to sell the new... :-)
22:35:26 <Eddi|zuHause3> dragonhorseboy: 99.99% of all software is buggy
22:36:49 <SpComb> so essentially, this project would, in effect, combine GRFCrawler and the ottdc_grfpack, and then add some support into the client to make it all more convenient
22:38:25 <Ammler> GRFCrawler would only be usefull, if it does open a liitle bit
22:38:53 <Ammler> if you are able to fix links from other authors, if they are absent for months...
22:39:42 <SpComb> yes, I don't think that the current GRFCrawler code would be very useful
22:39:57 <Ammler> as I said, about 40% of the pack grfs aren't available over Crawler...
22:40:48 <SpComb> yes, that's probably true, Rubidium (or someone else?) also stated the same in the thread
22:41:05 <dragonhorseboy> ammler...I don't think so (re trashing)
22:41:08 <SpComb> eis_os has been pretty absent recently, and GRFCrawler dates from the TTDPatch era
22:41:45 <Ammler> didn't he just update it some months ago
22:41:47 <dragonhorseboy> what about canstation when it has a CLEARLY stated id incompactibility between two versions .. how would someone else download an older author-created gamesave if the stations can't appear right due to the newer grfs?
22:42:34 <Rubidium> I'd say that uhm... the GRF ID should've been changed?
22:43:22 <dragonhorseboy> nope..its still same last I checked
22:43:35 <dragonhorseboy> as such is with many other grfs
22:44:09 <Ammler> yeah, ISR has same problem currentyl
22:44:25 <Ammler> he forgot to change ID from 0.6 to 0.7, afaik
22:44:51 <dragonhorseboy> or jcindust....ukrs....etc
22:44:57 <dragonhorseboy> too many to list
22:45:56 <SpComb> make it convenient enough and most people will have a reason to keep their servers compatible with it, and thus the newest versions of the GRFs...
22:47:32 <Ammler> well, I ususally do also symlink the server data to the web
22:48:05 <Ammler> so if the server got old and has some outdated GRFs, you can get them that way :-)
22:48:49 <Ammler> since you don't have original files there anymore, shouldn't be a big problem...
22:52:18 <SpComb> also noteworthy to compare this centralized-metadata-db with flag-in-the-.grf
22:52:44 <SpComb> if you flip the bit in the .grf, and then give a copy to someone else, then they can redistribute at will, and there's no way to recall it
22:53:30 <SpComb> if the official game client always checks some central db before automatically downloading a grf, then you'd need to use a custom patched version of the OpenTTD client, which is an entirely different issue
22:54:10 *** dragonhorseboy has left #openttd
22:54:39 <Ammler> openttd has tar support
22:55:01 <Ammler> if someone likes to add readme etc, just put that all together in a tar...
22:55:25 <SpComb> perhaps better to unpack it into a dir, because most people won't go looking for a README in a plain .tar
22:56:11 <SpComb> OpenTTD may support reading the .grf from a .tar, but the user may not support reading the README from a .tar :)
22:56:36 <Ammler> thats why we do not distribute the grfpack as tar
22:57:30 <Eddi|zuHause3> i suggested displaying in openttd... if people don't look in a tar, they also don't look in a dir that was automatically created
22:58:17 <SpComb> they will if they're trying to figure out what NewGRFs to use when creating a new game, and thus trying to figure out what a NewGRF does
22:58:26 <SpComb> at least that's how I behave
22:59:43 <Ammler> SpComb: so you have ALL grfs availabe in your GUI
22:59:59 <Eddi|zuHause3> the "i want to add all grfs" people kinda contradict that theory, SpComb
23:00:03 <Ammler> and if you choose one GRF which isn0t on your hd, you download it...
23:00:16 <SpComb> many GRFs are mutally exclusive...
23:00:40 <SpComb> Ammler: dunno, the behaviour of this when creating a new game is a separate issue
23:01:04 <Eddi|zuHause3> SpComb: sure, you know that, i know that, but $random_noob doesn't know that
23:01:12 <SpComb> if I were to implement this right now, it would be purely for network clients, and the downloaded NewGRFs would be stored separately from other GRFs
23:01:44 <SpComb> it wouldn't stop people from copying those GRFs over, but that doesn't matter, it's just a question of intentions
23:02:00 <SpComb> then, later, you could try and figure out what GRFs could also be made available when creating a new game
23:02:02 <Ammler> as GRFID.md4sum.grf :-)
23:02:05 <SpComb> (only the newest versions, etc)
23:02:33 <SpComb> ${OPENTTD_DIR}/cache/GRFID.md4sum.grf
23:02:59 <Ammler> as GRFID.md4sum.grf.gz
23:03:27 <SpComb> this would encourage people running servers to mostly use NewGRFs that are in the db, and this would also encourage people to encourage GRF authors to have their GRFs in the db
23:04:22 <Ammler> but you need some GRFs to start with :-)
23:04:49 <SpComb> not an issue, you can use dummy NewGRFs when implementing, testing and then demonstrating the prototype
23:05:04 <SpComb> or just ask some NewGRF author to let you use their GRFs as an example
23:05:36 <Ammler> forum grfs like USSet, JPSet
23:05:38 * SpComb isn't very far from starting to implement a prototype
23:07:10 * Ammler isn't very far from falling to sleep...
23:07:24 <SpComb> yes, that's the sad truth :(
23:07:42 * SpComb is busy this weekend as well, in a place as far from the internet as anything
23:07:45 <Ammler> good night and keep some of the chat :-)
23:43:55 <CIA-3> OpenTTD: smatz * r13721 /trunk/src/blitter/32bpp_base.hpp: -Codechange: one less multiplication for colour mixing in 32bpp blitters (GeekToo)
23:46:03 <CIA-3> OpenTTD: truebrain * r13722 /branches/noai/ (6 files in 2 dirs): [NoAI] -Add: AICargoList_IndustryNNN (Yexo)
continue to next day ⏵
