IRC logs for #openttd on OFTC at 2007-09-20
⏴ go to previous day
00:02:05 <Eddi|zuHause> it's a schrödingbug ;)
00:02:16 *** Diabolic1Angel is now known as Diabolic-Angel
00:02:44 <glx> it doesn't exists when we want to see it
00:02:49 <Eddi|zuHause> it ceases to be a bug if you de-bug it ;)
00:04:20 <Ammler> the assert came with 2 different nightlies
00:04:41 <Ammler> so, its nothing reacently changed
00:05:06 <Eddi|zuHause> does it still happen with the unpatched binary?
00:06:08 <Ammler> but that would be more strange :)
00:17:49 <Eddi|zuHause> PS: other side of the hill: www.informatik.uni-halle.de/~krause/Ravenswald%20Transport,%2029.%20Sep%201925.png
00:20:44 <Amix> but would be nice with sort of info telling newest addons to the page etc
01:31:33 *** Eddi|zuHause2 has joined #openttd
02:09:07 *** Diabolic-Angel has quit IRC
02:28:41 *** De_Ghost has joined #openttd
04:59:48 *** gfldex|DC is now known as gfldex
06:16:38 *** Frostregen_ has joined #openttd
06:22:48 *** Frostregen_ is now known as Frostregen
06:26:45 *** Greyscale has joined #openttd
06:41:02 *** boekabart has joined #openttd
07:08:58 *** ChanServ sets mode: +v tokai
07:25:26 *** stillunknown has joined #openttd
07:27:02 *** Vikthor has joined #openttd
07:35:23 *** ChanServ sets mode: +v tokai
07:36:07 *** TinoM|Mobil has joined #openttd
07:40:07 *** caladan has joined #openttd
08:09:06 *** dihedral has joined #openttd
08:09:19 <dihedral> Rubidium: are you around?
08:10:03 <dihedral> or somebody else who could tell me something about the grf data sent in the udp packets? :-)
08:11:08 <boekabart> is someone to just say 'good morning' any help at all?
08:15:42 <dihedral> i have this integer for grf_id: 4014b52 but should have this: 524B0104
08:15:47 <dihedral> anybody with any ideas?
08:16:37 <dihedral> sorry - little correction: i have this integer: 67193682
08:17:18 <dihedral> it is sent as uint32
08:19:48 <Rubidium> ReportText('speed: ' + str(speed) + ', GPS: ' + gps_data[GPS_LAT] + ', ' + \
08:19:48 <Rubidium> gps_data[GPS_LONG] + ' with ' + gps_data[GPS_SAT] + ' satellites')
08:20:05 <Rubidium> but yes, you need to change the endianness and convert it to hex
08:20:31 <boekabart> Rubidium: looks interesting ; what are you working on? (gps)
08:20:34 * Rubidium has a GPS that does not do what it's supposed to do...
08:21:13 * boekabart points out that a gps is not a system meant to teleport you to any global position
08:22:52 <Rubidium> the GPS is supposed to get the location it is at, but it kinda fails to do so
08:23:08 <boekabart> go outside, it might help LOL
08:24:02 <dihedral> how do i then 'change the endianness' ?
08:25:20 <SpComb> hmm, I assumed it would support specifying byte order
08:28:35 <Rubidium> boekabart: most likely not
08:29:05 <Rubidium> cause it worked... then it stopped working... then it start working after <unknown>... then it stopped working again
08:29:51 <dihedral> Rubidium: perhaps you need some more <unknown>
08:32:13 <dihedral> if i just convert dec 2 hex i get what i posted by mistake: 4014b52
08:34:55 <Ammler> dihedral: now change from little endian and you have it
08:36:11 *** ChanServ sets mode: +v tokai
08:38:38 * dihedral is giving it a try :-)
08:44:09 <dihedral> how the f... do i do that?
08:49:40 *** ChanServ sets mode: +v tokai
09:04:00 <Rubidium> dihedral: there are like a gazillion places where they explain how you can "switch" between endiannesses
09:17:32 <dihedral> Rubidium: thanks - i shall have a look
09:34:48 *** Progman has joined #openttd
09:58:34 <dihedral> Rubidium: thanks - finally understood what i had to do :-)
09:58:55 <dihedral> Ammler: thanks to you too :-)
09:59:50 <dihedral> Rubidium: though i do wonder why everything is sent one way - and grf id another?
10:04:06 <boekabart> i'd say it's because everything can be 'the native endiness' internally
10:04:25 *** _minime_ has joined #openttd
10:04:30 <boekabart> except stuff in Savegames and Data packets - they have to be in 1 uniform way indepentent of endiness of the client
10:04:52 <boekabart> ...independent, too
10:05:22 <boekabart> ... so that a BE client can talk to a LE client without getting all confused
10:06:45 <dihedral> hence i am asking Rubidium, the the "i'd say" becomes a 'i know' :-)
10:10:13 <dihedral> i dont understand this endianness enough to grasp why things are one way and other another...
10:11:00 <Ammler> dihedral: hmm, wikipedia has a good article about it, also in "our" language.
10:12:15 <Ammler> most importend is that the address of little endian doesn't change, if you change i.e. from 1Byte to 4 Byte
10:13:11 <dihedral> does that mean most significant byte and least significant byte swap places?
10:14:05 <dihedral> "our" language.... i am bilingual :-)
10:14:23 <dihedral> my mother is from london :-)
10:14:27 <Amix> just wonder. in which release will tram be a option?
10:14:50 <Amix> and will different companies be able to use the same tracks?
10:16:43 *** Greyscale has joined #openttd
10:16:51 <dihedral> Amix: you know the nightly builds?
10:20:26 <N101> dihedral: do the nightly builds allow use of other companies tracks?
10:23:31 <Amix> but allmost no one got nightly
10:23:51 <Amix> and ive been told that trams will be bundled with 0.3.6
10:23:57 <dihedral> N101: they include trams
10:24:09 <dihedral> answers Amix' first question
10:25:21 <Amix> the tram solution in openttd is way better than the one in locomotion for sure
10:32:41 <Rubidium> Amix: who told you take trams will be in 0.5.4 and/or 0.3.6?
10:33:33 <boekabart> can I ask .. what is 0.3.6 ?
10:34:15 <dihedral> as the current stable release is 0.5.3
10:34:23 <dihedral> some time there was a 0.3
10:34:36 <Amix> people talks about trams
10:34:43 <Amix> then some talks about nightly
10:34:52 <Amix> others about 0.5.4 release
10:35:02 <dihedral> 0.6 being more like it
10:35:10 <boekabart> dihedral: ah, right.
10:35:14 <boekabart> sounds more logical :)
10:35:49 <Amix> what would be cool, if someone would make avenues
10:36:25 <Amix> give me coding skills then
10:40:41 *** Darkebie has joined #openttd
10:44:17 <Rubidium> well, I can assure you that 0.5.4 will not have trams. The person saying it will be in 0.5.4 must be slapped.
10:44:58 <Amix> are they to unstable, or ?
10:47:11 <Rubidium> way too many changes to go into a bugfix release
10:47:36 <Rubidium> furthermore the code the trams are based on are so vastly different that backporting isn't an option
10:49:04 <Rubidium> though all of that has been described on the wiki too
11:07:11 *** skidd13 has joined #openttd
11:20:49 *** Dephenom has joined #openttd
11:36:01 <Rubidium> gfldex: STOP with the overabundant joins and quits!
11:40:34 *** frosch123 has joined #openttd
11:41:49 <frosch123> < skidd13> (_patches.autoslope && !_is_old_ai_player && (IsValidPlayer(_current_player) || (_current_player == OWNER_NONE && _game_mode == GM_EDITOR)))
11:42:00 <frosch123> That will also enable autoslope, when placing industries
11:45:36 *** mikk36[EST] has joined #openttd
11:46:10 <skidd13> I just thought over a simple solution to start with.
11:46:58 <Ammler> Rubidium: your client can also ignore such notes
11:47:42 <Rubidium> Ammler: yes, but then I won't know when somebody I need to ask comes online either
11:57:27 <skidd13> frosch123:Where's your problem with the enabled autoslope for industries? It's enabled for the player placed industries in game?
12:02:04 <frosch123> Err, don't know. Perhaps I only tested the random-placed.
12:03:00 <frosch123> I am not that familiar with the editor. Is there a way to randomly generate industries?
12:03:46 <skidd13> yup The first element of the matrix-widget of the create industry window.
12:06:07 <frosch123> And should they also use autoslope?
12:07:25 <skidd13> I don't see any use for them. It's a larger check but I'm not so deep in the industries code to be able to decide.
12:10:15 <skidd13> Belugas is the person who cares for the industries code. Just ask him.
12:10:46 <frosch123> I was talking about a user-view.
12:11:18 <frosch123> The land leveling for industries was introduced by TGP, not by newindustries.
12:11:25 <_42_> Sacro, Bjarni (~Bjarni@0x50a4166d.virnxx14.adsl-dhcp.tele.dk) was last seen quitting #openttd.notice 12 hours 56 minutes ago (19.09. 23:15) stating "Quit: Leaving" after spending 7 hours 4 minutes there.
12:13:12 <boekabart> actually - they put a market near/under/around the rails, not the other way around
12:15:04 <skidd13> frosch123: As I said earlier I don't see the use of autoslope for industries. The leveling should be enough IMO (user view).
12:17:10 <skidd13> But for NewIndustries it might be usefull. As I said the one who can answer this question is Belugas.
12:18:29 <frosch123> I already discussed the autoslope aspect with newindustries with Belugas.
12:19:20 <frosch123> The point is: IMO autoslope on placing industries in editor is not wanted by user.
12:19:38 <frosch123> So it must be disabled.
12:19:57 <frosch123> Same holds for player-funded industries in the game, I guess.
12:21:13 <skidd13> I tried it a while (with the conditions) but was not able to autoslope by placing a industy.
12:21:28 <boekabart> frosch123: how about: enable ONLY if the player Manually places any building at a CHOSEN location
12:21:52 <boekabart> building being rail/road/station/industry
12:22:18 <boekabart> but not if the player 'funds' an industry that gets built at a random location
12:23:51 <frosch123> This is getting more complicated every time
12:24:03 <skidd13> That brings one thing in my mind. Why don't adding a control bit to the level land command.
12:24:38 <frosch123> skidd13: Generate a random map, place some town on a hill, and fund a industry at the hill side
12:25:46 <skidd13> Adding a controll bit would remove the whole check stuff and would require only a few changes in the GUI-call-functions
12:26:23 <boekabart> skidd13: I think I agree - only enable it _if_ it was a player-initiated build command
12:27:06 <skidd13> And it will be faster code, cause less checks :D
12:27:47 <skidd13> And the NoAI could be able to use it.
12:29:02 <frosch123> I have no OTTD on this machine, but if you want to try: In industry_cmd the command-flags are passed to TerraformTile_Industry
12:29:35 <frosch123> NoAI? Doesn't it behave just like a human player?
12:31:31 <skidd13> frosch123: if you want to try <-- It's OK.
12:31:31 <skidd13> NoAI <-- Depends on the programmer :D
12:36:37 *** Vikthor has joined #openttd
12:58:01 *** ChanServ sets mode: +o Bjarni
12:58:37 <Bjarni> Sacro: you want tips for MS train sim again?
12:59:00 <Bjarni> basically just do as I do and you will do fine
13:06:02 * Bjarni wonders why that movie makes him think of Hull
13:06:18 <Bjarni> somehow I'm not surprised
13:07:17 <Bjarni> Asia, Africa and South America produces a lot of stuff like this
13:08:06 <SpComb> someone's actually playing on bugmenot's server :P
13:16:25 <Bjarni> Sacro: btw what was it about your house leaking?
13:17:54 <SpComb> does OpenTTD have any kind of equivalent to TTDPatch's switches.xml? I.e. to let me support modifying the config file settings for different OpenTTD versions automatically?
13:23:00 *** eQualizer has joined #openttd
13:28:03 <SpComb> ugh... looks... semiparseable
13:28:14 *** Diabolic-Angel has joined #openttd
13:28:27 <Sacro> Bjarni: the hot water tank was leaking down via my light socket
13:28:35 <Sacro> which was making some very interesting sparky noises
13:30:23 <Eddi|zuHause2> SpComb: well, you could as well just compile it :p
13:30:31 <Bjarni> so how much is ruined? the entire floor?
13:31:02 <SpComb> I think I'll just run a couple of regexps against the file
13:31:16 <SpComb> and stab anyone who adds an extra newline or whatever somewhere
13:36:02 <Eddi|zuHause2> well, you could at least push it through the pre-processor, will probably simplify the layout a bit
13:42:50 *** Diabolic-Angel has quit IRC
13:43:24 <SpComb> but that's for later, first I need support for savegames
13:48:07 <SpComb> " -g [savegame] = Start new/save game immediately" --- the wording's slightly unclear, does it make it load up said savegame?
13:49:07 <glx> -g -> newgame, -g savegame -> load savegame, -g scenario -> load scenario
14:07:49 *** Deathmaker has joined #openttd
14:13:55 *** Greyscale has joined #openttd
14:24:58 *** Greyscale has joined #openttd
14:34:18 *** Hendikins has joined #openttd
14:36:05 *** Diabolic-Angel has joined #openttd
14:48:42 *** dihedral_ has joined #openttd
14:48:49 *** Hendy is now known as Hendikins
14:49:24 *** dihedral_ is now known as dihedral
14:49:43 <dihedral> just got my first newgrf query to work :-)
14:50:34 <Ammler> cool, so version 0.3 will be released soon :)
14:51:13 <dihedral> i shall query one of the coop games :-)
14:51:57 <Ammler> hmm, yes, the public server uses about some grfs
14:52:24 <Ammler> openttdcoop.ppcis.org:3980
14:53:09 <Ammler> from where do you get the infos about newgrfs? (Name)
14:53:12 <glx> and still no assert I guess ;)
14:54:01 <Ammler> glx: the assert was on an other server (german community)
14:54:13 <dihedral> you query and get the server_info, which holds grf id and md5sum
14:54:24 <dihedral> send that data back to the server and get names
14:54:45 <dihedral> but i only get 4 new grf details from the coop game
14:55:19 <Ammler> we use about 30 NewGRFs
14:56:35 <dihedral> i have the id's and md5's but when getting a UDP_SERVER_NEWGRFS it only includes 4
14:56:42 <dihedral> is Rubidium around by any chance?
14:58:52 <boekabart> dihedral: It has to do with max. packet size; it think you get more than 1 of those
15:00:51 <boekabart> after you get one, then you reply with another request in which you request more
15:01:58 <dihedral> i get more data in a UDP_SERVER_DETAIL_INFO packet
15:02:11 <dihedral> and i get more info in the simple server info packet
15:02:25 <boekabart> you send a PACKET_UDP_CLIENT_GET_NEWGRFS right?
15:02:43 <dihedral> yep - would not get the UDP_SERVER_NEWGRFS back other wise :-)
15:03:01 <dihedral> i have 4 grf files, that are correct
15:03:42 <dihedral> the rest is simply missing!
15:04:32 <dihedral> so it tells me that i am at least doing /something/ right
15:05:04 <dihedral> also checked the details against those at servers.openttd.org
15:05:30 <boekabart> and it doesn't have to do with line 208 (network_udp.cpp) num_grfs = p->Recv_uint8 (); ??
15:06:02 <dihedral> my packet has a 26 for the server to read there
15:06:29 <boekabart> if (packet_len > SEND_MTU - 4) { // 4 is 3 byte header + grf count in reply
15:06:32 *** Socarys has joined #openttd
15:06:54 <dihedral> all i send back to the server, is what i get in the info packet anyway
15:07:00 <dihedral> so it aint the packet size
15:07:06 <dihedral> perhaps on the response
15:07:15 <Socarys> hi, i have a problem with nightly r11129... in multiplayer after12month the server restarted
15:07:19 <boekabart> the pack you get is << 1460 bytes?
15:08:09 <dihedral> i dont even care to read more data from the file pointer
15:09:03 <dihedral> the packet is constructed correctly, as otherwise it would go through on the server side as an invalid packet
15:09:09 <dihedral> and be ignored, dropped
15:09:21 <dihedral> as i get a response from the server, something must be going through
15:10:48 <dihedral> as i get details for the grfid i send as second place and no details come back from the first grfid in the packet
15:10:56 <dihedral> something is not as i exprected it
15:11:09 <dihedral> but anyhow - i need to be getting ready
15:11:15 <dihedral> heading out for this evening
15:11:22 <dihedral> at least it's getting somewhere
15:11:39 <dihedral> i shall do some more work tomorrow
15:13:06 <SpComb> openttd doesn't seem to like paths as the argument to load
15:13:27 <SpComb> i.e. `save "7/1"` works fine, but `load "7/1.sav"` fails
15:14:12 *** Leviath has joined #openttd
15:15:39 <gfldex> that is has to be prevented on application level
15:15:52 <gfldex> is there a dialog for invalid file names?
15:16:42 *** Digitalfox has joined #openttd
15:16:58 <gfldex> and you should remove or rename that file befor you do any backups
15:17:16 <SpComb> dialog? blaa/foo.sav: No such file or directory.
15:17:29 <SpComb> and why before I do any backups?
15:19:01 <gfldex> well, technically it's a valid file name but it can get funky if you confront applications with it
15:19:47 <gfldex> and a (semi) bad filename in an archiv can drive you nuts if you use anything more complex then tar
15:23:03 <gfldex> zip it and send that save game to a windows person if you want to drive him crazy :)
15:23:55 <SpComb> seems I need to cd into the directory to use load
15:24:14 <SpComb> it builds a list of files, and then goes through them, comparing against the given filename
15:28:13 <gfldex> you have to work with inodes to get around that
15:28:53 <gfldex> so the function that gives you the list of file names for the load dialog has to hand out both the file name and the inode number
15:29:08 <gfldex> then you have to use the inode number to load the file and not the full path
15:29:19 <gfldex> but nobody is doing it that way
15:29:53 <gfldex> the only program that is working on inodes i know of is postfix
15:34:01 <gfldex> gnome terminal just crashed and showed me a send-a-crash-report dialog. for every single terminal one dialog
15:34:19 <gfldex> and it opened a browser window with help for gnome terminal :)
15:34:45 <gfldex> anyway, that / - problem is a bug because it's not portable
15:35:23 <gfldex> the application should tell the user not to use / \\ and any unprintable chars
15:36:04 <gfldex> i think posix forbids at least / and the linux standard base forbids it for sure
15:38:27 <gfldex> there is a SanizieFilename(char *filename) in fileio.cpp, seams not to get used that often
15:48:48 *** Wezz6400 has joined #openttd
15:52:25 <gfldex> SpComb: did you wrote a bug report?
15:53:10 *** Wezz6400_ has joined #openttd
15:55:19 *** |Jeroen| has joined #openttd
15:56:15 <SpComb> no, I just used `cd dir` and `cd ..`
16:00:55 <SpComb> right, now the backend supports savegames, just need to make the web frontend support them
16:03:46 *** Wezz6400_ is now known as Wezz6400
16:13:25 *** |Jeroen| has joined #openttd
16:33:58 *** Brianetta has joined #openttd
16:39:12 <Nite> hi - anyone know about "myottd"
16:39:24 <Nite> i cant reach teh website !?
16:45:10 <SpComb> there, now it should have support for savegames
16:45:20 <SpComb> one click to save, one click to load a different savegame
16:46:25 <Nite> is there newgrf support yet?
16:47:58 <Nite> are there other sites where you can host easily?
16:48:11 <Nite> ist there a way to multiply runningcosts?
16:48:31 <SpComb> what do you mean? Some different site where you can run your own dedicated server?
16:50:52 <Nite> i think the us set runingcosts should be 10(ten) times higher - any ideas?
16:51:08 <SpComb> is it some OpenTTD patch setting?
16:51:19 <SpComb> you'll have to wait for me to implement support for those
16:51:33 <Eddi|zuHause2> Nite: there's a difficulty setting for running costs
16:52:04 <SpComb> I'd just need to implement support for it
16:52:10 <Nite> yeah but the dificult setting only adjusts running cost very slightly
16:52:28 <SpComb> (assuming that's what he meant, maybe not)
16:53:04 <SpComb> may have understood wrong, I'm not going to modify OpenTTD to add some freighttrains patch, just implement support for configuring them in MyOTTD
17:00:47 *** stillunknown has joined #openttd
17:07:06 <SpComb> ok, now it saves the game before it shuts down, and loads it when it starts up, so the server reboots shouldn't kill your game anymore
17:09:32 <SpComb> and I think I've finally managed to get svn under control
17:13:31 <SpComb> how easy/hard would it be to extract the current/start date from a savegame file?
17:14:23 <SpComb> I can also encode that info in the filename using the console getdate command, but if the date's at a fixed offset in a reasonably simple format, it would be cleaner
17:17:56 <SpComb> also, no way to get the starting date of the current game via the console
17:21:15 *** stillunknown has joined #openttd
17:22:41 <Eddi|zuHause2> you could as well just start a server to load the game and query it...
17:24:07 <SpComb> well, I want to display the start and current dates in the savegames list
17:29:30 <Eddi|zuHause2> well, saveload.cpp might possibly explain more about the savegame format
17:41:04 *** Frostregen has joined #openttd
18:06:02 *** thgergo has joined #openttd
18:06:53 <Ammler> is there someone who did patch chrisIN?
18:07:55 <Ammler> I am not able to patch my SVN source with that.
18:08:48 <Ammler> sometimes, it addes contend more then once in one file
18:09:10 <Ammler> omg, (sry for my english)
18:13:15 <Ammler> I patched with >patch -p0 -i chrisIN.diff
18:34:54 <SmatZ> Ammler: I was able to apply chrisin patch
18:35:07 <SmatZ> didn't try to other revisions...
18:35:41 <SmatZ> I everytime use "patch -p 0 -l -F 666 < file"
18:36:13 *** Hendikins has joined #openttd
18:39:12 <Ammler> SmatZ: the problem is with new files
18:39:33 *** ChanServ sets mode: +v tokai
18:39:34 <Ammler> when I patch multiple times
18:41:03 <Ammler> hmm, and I don't understand "fuzz factor"
18:46:58 <Ammler> SpComb: uploading a save?
18:50:02 <SpComb> it also saves the game on shutdown and resumes it on startup, so me restarting the server won't kill it entirely
18:53:58 <Ammler> SpComb: do you only make the software or do you really like to host MyOTTD?
18:54:22 <SpComb> I'll be hosting MyOTTD in the future as well
18:54:55 <Ammler> did you think about the CPU usage?
18:55:45 <SpComb> it's probably not particularly high, but I'll obviously have to put limits on how many servers can exist
18:56:23 <Ammler> hmm, or that only servers are running, where someone like to connect
18:56:35 <SpComb> the source code's also available, of course, svn://svn.marttila.de/myottd
18:56:46 <Ammler> ok, paused server don't need cpu
18:57:30 <Ammler> and what are requirements?
18:57:57 <Ammler> SmatZ: my patch problem was only the new files
18:58:11 <Ammler> can I tell svn to delete files on revert?
18:58:28 <SpComb> python 2.4, postgres, twisted, pylons, sqlalchemy, psycopg, pypgsql, possibly some other things
18:58:43 <SpComb> nor is it packaged for distribution, so you'd have to install all of those yourself, and set up the db
18:59:15 <Ammler> hmm, its not a "common" setup
18:59:40 <SpComb> not really, no, it's python+postgres
19:02:21 *** Hendikins has joined #openttd
19:17:05 <Eddi|zuHause2> <Ammler> sometimes, it addes contend more then once in one file <- "patch" can not handle new files correctly, you have to delete them (patch -R) before applying the new patch
19:17:46 <Eddi|zuHause2> and "fuzz factor" means that it could add the lines, but the context lines did not match correctly
19:18:20 <Eddi|zuHause2> best is to apply such a patch to a clean checkout each time
19:20:03 <Ammler> Eddi|zuHause2: thanks, I did a new checkout
19:20:24 <Ammler> and patched only with patch -p0 < patch
19:21:00 <Ammler> (sry for my short lines, chatted to much over ottd)
19:21:35 <CIA-1> OpenTTD: rubidium * r11133 /trunk/src/sprite.h: -Fix (r11102) [FS#1245]: in some cases vehicles would be drawn over houses when they should be drawn behind the houses. Patch by frosch.
19:23:40 <Prof_Frink> Random suggestion #284927: Could you make 'summary' news items display like chat messages?
19:27:03 <Eddi|zuHause2> www.informatik.uni-halle.de/~krause/Ravenswald%20Transport,%2014.%20Mar%201926.png <- snowy towns are awesome :)
19:27:22 <Ailure> that would certainly be uesful
19:27:58 <Eddi|zuHause2> Prof_Frink: you mean ticker items?
19:28:34 <Ailure> I played with the idea of having a constantly scrolling ticker if there's a few news items behind
19:28:38 <Ailure> like the ticker in SC3000
19:28:47 <Ailure> but making it show as chat messages is probably less of a hassle
19:29:58 <Eddi|zuHause2> i hated the ticker in SC3000
19:30:03 <Prof_Frink> The trouble with the scolliness is it takes too long, so you tend to get a backlog of messages
19:30:36 <Ammler> Eddi|zuHause2: did you "hack" alpine.grf?
19:31:05 <Ammler> or have you a working newindustry release?=
19:31:06 <Eddi|zuHause2> no, but i do have a patched build ;)
19:31:32 <Eddi|zuHause2> well, "working" is slightly exaggerated
19:31:43 <Ailure> [21:29] <Eddi|zuHause2> i hated the ticker in SC3000
19:31:50 <Ailure> the ticker in TTD is way worse either way :P
19:32:07 <Eddi|zuHause2> yeah, but i tend to ignore that one anyway ;)
19:33:38 <Ammler> Eddi|zuHause2: with changing snowline?
19:33:58 <Ailure> seems like I hadn't updated my local trunk for over 500 revisions
19:34:00 <Eddi|zuHause2> yeah, that one works great ;)
19:34:20 <Ammler> is that in roadmap for 0.6?
19:35:49 <Ammler> snowline also affect industry?
19:37:09 <Ailure> as I understand it, if it's a crucial part of newindustries then yes it is
19:37:54 <Rubidium> snowlines crucial for newindustries?
19:38:32 <Rubidium> and for that matter, changing snowlines have been implemented ages ago
19:38:47 <Rubidium> I wouldn't even be surprised if it is in 0.5
19:39:15 <Ailure> hmm yes its' been in for awhile
19:39:27 <Ailure> well, it dosen't change during the game but it's probably possible to make it do that
19:39:53 <Eddi|zuHause2> once upon a time there was a patch setting to change the snowline height
19:40:12 <Ailure> you can't change the snowline during the game
19:40:17 <Ailure> but you can set before starting it at least
19:40:22 <Eddi|zuHause2> and yes, snowline also affects industires
19:41:19 <Eddi|zuHause2> but not all industries in alpine have a snowy version yet
19:42:48 <Ailure> don't some industry that appears both below and above snowline look the same anyway?
19:43:21 <Ailure> like the power station >_>
19:43:29 <Ailure> which I alway thought looked a bit weird without snow
19:44:12 <Ailure> then of the orginal industries
19:44:19 <Prof_Frink> I suspect that's related to the funny-coloured grass on large airports
19:44:21 <Ailure> only the "forest" one is snow-covered
19:44:57 <Ailure> I always thought that looked weird
19:45:14 <Nite> well about snowline - you CANB simply set it.
19:45:15 <Rubidium> ergo... you *can* change the snow line height during the game
19:45:17 <Prof_Frink> Ailure: It's because those tiles are reused
19:45:26 <Eddi|zuHause2> there are some additional snowy industries in alpine.grf
19:45:26 <Prof_Frink> From Transport Tycoon
19:46:26 <Ailure> the grass on the airport is the old-style grass
19:46:48 <Eddi|zuHause2> and the factories have grassy roofs also (same way as the town buildings
19:46:55 <Eddi|zuHause2> s/grassy/snowy/
19:47:06 <Ailure> it's using the old grass
19:47:13 <Ailure> damn I prefer the look of the stone in TT
19:47:25 <Ailure> and same for it's rough land
19:47:38 <Ailure> the grass is a bit bright though
19:48:31 <Ailure> TTO is near unplayable when you're used to the more advanced signals
19:49:13 <Prof_Frink> Ailure: The best bits of TTO that got dropped in TTD were the monorail graphics
19:49:38 <Ailure> looking at old savegames from back then is close to embrassing
19:49:55 <Ailure> then there wasn't even one-way signals
19:50:20 <Ailure> I never liked the way monorail or maglev was implemented in TTD
19:50:24 <Ailure> it was somewhat better in TTO
19:50:34 <Ailure> but I prefer something like how most trainsets does it
19:50:42 <Prof_Frink> Especially the monorail bridge
19:50:48 <Ailure> maglev is there, but it dosen't replace railway
19:51:12 <Ailure> the monorail bridge looked a bit plain
19:51:16 <Ailure> at least for longer distances
19:51:26 <Prof_Frink> Heh, in my TTO games, monorails replaced railway
19:51:44 <Prof_Frink> good ol' single-headed X2001s
19:52:07 <Ailure> Looks nothing like the real x2000 trains
19:52:12 <Ailure> which are here in Sweden
19:52:22 <Ailure> it's a regular railway train too
19:52:43 <Ailure> pretty modern looking though
19:53:46 <Ailure> reminds me about the talk about the economy rebalancing
19:54:05 <Ailure> I liked the idea of monorail being the cheapest but somewhat less flexible than regular railways :P
19:54:05 <Prof_Frink> Meh, looks like a grey TGV
19:54:24 <Ailure> it's built for the swedish railways too
19:54:32 <Ailure> which aren't very straight apparently :P
19:55:06 <Prof_Frink> Yeah, I'm reading its wiki page
19:55:44 <Ailure> I live next to a train station so I see them now and then
19:55:51 <Ailure> they have a... intresting sound on it's engine.
19:56:41 <Prof_Frink> Is it anything like the scream you get from a 125 flying past?
19:56:45 <Ailure> It's supposed to compete with aircraft transport, which is ironic as it does sound little bit like an aircraft
19:56:57 <Ailure> nah it's not a really that much awful of a noise
19:57:03 <Ailure> just unusual for a train
20:01:47 *** Greyscale has joined #openttd
20:05:31 *** gfldex_ has joined #openttd
20:43:59 <ln-> prison break 3x01 is there.
20:44:56 <Wolf01> we are waiting for 2x01, the next week
20:45:19 <glx> I just saw 2x03 and 2x04 here
20:45:38 <ln-> by "there", i meant the world wide web.
20:45:49 <Rubidium> ln-: it's been "there" for ages
20:45:59 <ln-> the finnish tv is at about 2x05 or so.
20:46:19 <ln-> Rubidium: it was only aired last sunday or something, is that "ages"?
20:47:31 <ln-> correction, aired this monday.
20:47:45 <Rubidium> you can even view stargate atlantis 4x01 and 4x02
20:48:01 <ln-> and by world wide web i mean the www, not p2p.
20:48:43 <Rubidium> oh, and Stargate Atlantis 4x01 is going to be aired Friday the 28th
20:52:04 <Rubidium> ergo... if you're "lucky" you can view episodes way before they're aired
20:54:08 <TrueBrain> Rubidium: lol, did you watch 401 and 402 that are 'available' currently? :)
20:54:11 <TrueBrain> is without background music
20:54:13 <TrueBrain> VERY cool to watch :p
20:54:18 <TrueBrain> you hear people saying: SHOT SHOT
20:55:25 <Rubidium> TrueBrain: that's more or less why I said viewed and not watched
20:55:37 <Rubidium> but yes, it's very interesting how it is without sound
20:55:38 <TrueBrain> I could 'watch' it for 12 minutes
20:55:41 <TrueBrain> then it really annoyed me :p
20:57:13 <Rubidium> yes, it's very annoying
20:57:39 <Rubidium> another thing that annoys me, but what is done "right" in those two is the fact that you can't hear anything in space
20:57:53 <Rubidium> but you *always* hear the spaceships passing
20:58:33 <TrueBrain> Rubidium: I love the most in the last of 3, that they moved an astroid with ships
20:58:42 <TrueBrain> far out in space they put on their fuel engines...
20:58:46 <TrueBrain> and SPEEDY they went
20:59:46 <Rubidium> now you've spoiled it for glx... you bastard
21:00:23 <glx> well they eventually will never show it
21:01:04 <glx> as they already done with other shows
21:01:10 <ln-> does anyone want me to spoil prison break season 2?
21:01:25 <TrueBrain> ln-: only if you want a ban? :)
21:04:53 <ln-> they all are out and live their lives happily ever after.
21:05:43 <Rubidium> ah well, if that's what ln- calls "happily" ;)
21:08:50 <Ammler> is it possible to switch off "exclusive rights"?
21:09:30 <Ammler> it the only possibilty to do it too
21:13:22 <SmatZ> no, there is some patch, but not in trunk... but you may disable it by modifying server
21:14:43 <gfldex_> is there a way to tell gprof to show a bit smaller numbers?
21:16:15 <SmatZ> what 'smaller numbers'?
21:18:02 <gfldex_> it tells me that functions need 0.00 m/s. I would like to have something like 0.0005 m/s in the output
21:18:35 <SmatZ> I don't think it is possible
21:18:56 <SmatZ> eg. if system kernel allows interrupt more often
21:19:40 <gfldex_> it takes the samples with a lot higher resolution
21:19:57 <gfldex_> you can tell if you stick the whole program into a loop of 1000
21:21:20 <Ammler> :) just realized, that the one who bought exklisve right drops do appalling after that
21:23:28 <SmatZ> gfldex_: the running program is interrupted 100 times a second, and everytime it logs what function is being run. Switching tasks takes some time, and the system timer may be able to generate interrupts only at 100Hz. So then it is not possible to have any higher resolution.
21:25:36 <gfldex_> i doubt they use an interrupt. my guess would be symbol table magic
21:25:36 *** glx|away has joined #openttd
21:25:36 *** ChanServ sets mode: +v glx|away
21:26:15 <Rubidium> Profiling also involves watching your program as it runs, and keeping a histogram of where the program counter happens to be every now and then. Typically the program counter is looked at around 100 times per second of run time, but the exact frequency may vary from system to system.
21:26:30 *** glx|away is now known as glx
21:27:37 <Rubidium> gfldex_: and "The run-time figures that gprof gives you are based on a sampling process, so they are subject to statistical inaccuracy. If a function runs only a small amount of time, so that on the average the sampling process ought to catch that function in the act only once, there is a pretty good chance it will actually find that function zero times, or twice."
21:29:41 *** gfldex_ is now known as gfldex
21:32:48 <gfldex> that would render gprof rather useless
21:35:15 <SmatZ> gfldex: gprof only reads the statistical data logged by program
21:35:38 <SmatZ> how do you mean 'useless'?
21:35:54 <SmatZ> how else would you measure runtime of functions, and number of their runs?
21:37:15 <TinoDidriksen> gprof should be used together with gcov to get best data. Hardly matters if something takes 3 seconds if it only happens once per execution.
21:37:51 <Rubidium> gprof should add a hardware component that gets triggered each time a function call is made, so it can do proper logging
21:38:03 <gfldex> it's good to find out what functions are called often but it cant really tell you if a function that's not called often in your test (that will differ from those of users) is awfully slow
21:38:09 <SmatZ> maybe I misunderstood the question then
21:39:01 <gfldex> you could use debug registers but that would limit the number of functions you watch to a few 100
21:39:20 <Rubidium> making the function that is called most often faster isn't usually the best way to make an application faster
21:39:55 <SmatZ> making the function that takes the most time faster is the best way to make whole program faster
21:40:08 *** ChanServ sets mode: +o orudge
21:40:10 <Rubidium> not true either (all the time)
21:40:48 <Rubidium> (as in: not true for all cases either)
21:40:55 <SmatZ> it depends how much can you make it faster, but you should look at functions that take the most time as the first
21:41:17 <SmatZ> when something takes only 1% of runtime, making it 10 times faster won't make the program faster...
21:41:49 <SmatZ> but function that takes 20% when is made 10 times faster, the program will be ~ 18% fater
21:42:20 <Rubidium> calling something much much less is also a very good candidate
21:43:54 <SmatZ> it is a way to make something faster... depends if you look at 'total time including called functions' or only 'time spent in function'
21:44:23 <Rubidium> I usually see people who are primarily focussed only on the thing on the top
21:44:39 <Rubidium> and only that function
21:45:25 <Eddi|zuHause2> hm, why does my express train not get a dining car?
21:46:18 <Eddi|zuHause2> the readme says i get a dining car if i have at least 8 wagons
21:47:17 <gfldex> is that the DB set Eddi?
21:48:03 <gfldex> no wonder then. if it comes to DB you better have a walk
21:48:45 <Eddi|zuHause2> it is 1927 and i have a BR 18 with 1 mail and 11 passenger wagons
21:49:32 <gfldex> wasnt that right in the middle of the depression?
21:50:31 <Eddi|zuHause2> no, that was way earlier
21:52:12 <gfldex> i found something called OProfile. needs a 2.4 kernel tho :(
21:52:55 <Eddi|zuHause2> but i want a dining car now!! *stomps on the ground*
21:55:13 <Nite> gues 8 wagons on some special loco at some special time
22:03:39 <Eddi|zuHause2> i don't have the patch to double check the behaviour...
22:08:56 <Eddi|zuHause2> [Whois] Ailure ist Ailure@194.47.44.201 (lol)
22:09:26 <Ailure> ....haha I had lol as full name
22:10:04 <Eddi|zuHause2> hm, i just cheated to 1955, and i get a dining car there...
22:10:31 <Nite> does this make you less hungry now?
22:10:54 <Eddi|zuHause2> no, because i have 1927, and that is an awful long time to wait :p
22:13:27 <Ailure> using telnet for IRC is nice too
22:13:49 <Nite> trilian really sucks for irc
22:14:15 <Amix> amirc and wookiechat for amigaos works ;)
22:46:44 <Eddi|zuHause2> www.informatik.uni-halle.de/~krause/Ravenswald%20Transport,%2028.%20Sep%201927.png <- my network is finally getting shape (even without PBS ;))
22:47:58 <Eddi|zuHause2> i could use a little faster bridges...
22:55:41 <gfldex> you play DB, your trains will be late anyway
22:56:37 <gfldex> according to your currency you play RB actually
23:18:13 <Eddi|zuHause2> there is no DB in 1927, it's DRG
23:19:20 <Eddi|zuHause2> Deutsche Reichsbahn Gesellschaft
23:20:26 <Eddi|zuHause2> (german reich railway corporation)
23:21:25 <Eddi|zuHause2> damn, it is difficult to build double track through the mountains...
23:22:06 <svippy> And OpenTTD lacks mountains and cliffs and valleys and bottomless pit holes.
23:22:14 <svippy> Uh, bottomless tar pits*
23:23:29 <svippy> You'd think you can have a game without bottomless tar pits?!
23:23:52 <Nite> locomotion has great landscape/cliffs valleys - BUT nothing else
23:29:25 <Eddi|zuHause2> ottd has great landscape... i occasionally miss the possibility to dig my rails on the lower end of a slope instead of adding a foundation
23:29:47 *** Greyscale has joined #openttd
23:30:20 <Amix> ottd lacks avenues for cities etc
23:30:46 <Amix> the tramsystem, is great
23:30:57 <Amix> they get stuck sometimes
23:31:03 *** Hendikins is now known as Hendikins|FlyingAgain
23:31:15 <Amix> its not possible to turn them arround etc
23:32:18 <SpComb> how does the number of connected clients affect the cpu useage of the openttd process? From what I can see, not very much (in terms of total CPU time, it's pretty clear which openttd process is the one with clients, though)
23:33:32 <Nite> more vehicles/routes mean more cpu more clients merely
23:34:12 *** ThePizzaKing has joined #openttd
23:41:55 <Eddi|zuHause2> i must say... huge map and daylength 4 is great, you finally get to build a big network using steam engines
23:42:34 *** Ammller has joined #openttd
23:43:14 <Eddi|zuHause2> daylength 32 is kinda overkill...
23:46:33 *** Brianetta has joined #openttd
continue to next day ⏵
