Back to Home / #openttd / 2015 / 07 / Prev Day | Next Day
#openttd IRC Logs for 2015-07-27

---Logopened Mon Jul 27 00:00:26 2015
00:38-!-Pikka [] has joined #openttd
00:56-!-Eddi|zuHause [] has quit []
00:56-!-Eddi|zuHause [] has joined #openttd
01:07-!-Supercheese [] has left #openttd []
01:07-!-Supercheese [] has joined #openttd
01:11-!-Alberth [~alberth@2001:981:c6c5:1:be5f:f4ff:feac:e11] has joined #openttd
01:11-!-mode/#openttd [+o Alberth] by ChanServ
01:22-!-Supercheese [] has quit [Quit: Valete omnes]
01:31-!-andythenorth [] has joined #openttd
01:32<andythenorth>also bye
01:32-!-andythenorth [] has quit []
01:33-!-sla_ro|master [slamaster@] has joined #openttd
01:55<@Alberth>hi bye too, andy
01:57-!-Supercheese [] has joined #openttd
03:00-!-andythenorth [] has joined #openttd
03:28-!-roidal [] has joined #openttd
03:42-!-andythenorth [] has quit [Quit: andythenorth]
03:46-!-Cybertinus [] has quit [Remote host closed the connection]
03:46-!-Cybertinus [] has joined #openttd
03:58-!-andythenorth [] has joined #openttd
04:26-!-wicope [] has joined #openttd
04:39<@planetmaker>hi hi
04:39<@planetmaker>someone should tell Zuu that DevZone can host git repos just fine ;)
04:41<@planetmaker>just that building NewGRF projects from git repos is not supported - but only due to lack of support from the build script which someone[TM] could also fix in the form of providing patches
04:46<@Alberth>send a PM ?
04:46<@planetmaker>:) or wait till I meet him here
04:46<@Alberth>at least he replied to the forum, although I haven't seen him here today
04:47<@Alberth>I am sure he'll turn up somewhen :)
04:47<@planetmaker>as said in the other channel: I really like the idea of one (or maybe two, one git on hg) repo to build test builds from for us
04:47<@planetmaker>and on DevZone it's easy for us to configure it to whatever needs we might have :)
04:47<@Alberth>yeah, we don't need it often, but it's nice to have
04:48<@Alberth>I should really play a game with his patch
04:48<@planetmaker>gs2newgrf talk one, yes?
04:48<@planetmaker>it's intriguing me, too :)
04:50<@Alberth>hmm, moving both git and hg at the same time is confusing :)
04:50<@planetmaker>yes :)
04:50<@planetmaker>what's git's equivalent for hg manifest?
04:51<@Alberth>oh, nice command :)
04:51-!-andythenorth [] has quit [Quit: andythenorth]
04:52<@Alberth>(yeah, I didn't know it :) )
04:52<@planetmaker>seems git ls-files or so
04:52-!-andythenorth [] has joined #openttd
04:52<blathijs>planetmaker: ls-files lists files in the index / working copy only I think
04:52-!-andythenorth [] has quit []
04:53<@planetmaker>blathijs, so I can't specify a revision? That default it only lists working copy is totally fine
04:53<@planetmaker>and that's what hg manifest does, too
04:53<@Alberth> ls-tree, by the looks of it
04:53<blathijs>planetmaker: Yeah, you can't specify a revision AFAICS
04:53<@planetmaker>yup, that's where I found it, Alberth ;)
04:54<@planetmaker>hm, yours is different :)
04:54<@Alberth>many websites have more than one page :p
04:55<@planetmaker>now that you tell me... a whole new world opens :)
04:55<@Alberth>yw :)
05:23-!-JezK [] has quit [Remote host closed the connection]
05:23-!-JezK [~jez@2407:7800:400:107f:3db5:daca:8457:e66a] has joined #openttd
05:40-!-zeknurn [] has quit [Read error: Connection reset by peer]
05:43-!-zeknurn [] has joined #openttd
05:54-!-wicope [] has quit [Remote host closed the connection]
06:06<@Alberth>it is?
06:15-!-Hiddenfunstuff [] has joined #openttd
06:17-!-JezK [~jez@2407:7800:400:107f:3db5:daca:8457:e66a] has quit [Remote host closed the connection]
06:20<argoneus>is 15 the hardcoded amount of max companies?
06:22<@planetmaker>yes. But it's a limit in many parts. Thus changing the constant will not do the trick. It needs many deep code changes
06:22<@planetmaker>that includes NewGRF specs, command handling etc
06:23<argoneus>I see
06:23<argoneus>um, forgive my ignorance
06:23<argoneus>but what does the number of companies have to do with specs?
06:24<argoneus>is it that the number is a 4 bit number or something?
06:24<@Alberth>it is
06:24<@Alberth>0000 to 1111
06:25<@Alberth>so some fields are 4 bit for the company number
06:25<@Alberth>more companies means you have to move data, or find room elsewhere, which may not always exist
06:27<argoneus>this game really is on a tight budget huh
06:27<argoneus>as in
06:27<argoneus>you don't want to raise many requirements
06:30<@Alberth>it's not simple adding stuff
06:30<@Alberth>you have to find all occurrences, see how you can handle the additional data
06:31<@Alberth>and that increase works through everywhere
06:31<@Alberth>maps get bigger save files get bigger
06:32<@Alberth>if you add a byte to each tile, at 4k x 4k, that 16MB
06:33<@Alberth>we may need a different map layout to fix, but that means you need a new data structure that is still bloody fast to access
06:34<@Alberth>it's not that we don't want it, but such undertakings take loads and loads of time
06:35<@Alberth>so you need someone dedicated to the problem enough to do it
06:36<@Alberth>and you may find there is no good solution
06:37<@Alberth>suggesting it takes 5 minutes, implementing something like this is soon a few years, as nobody works full-time on the project
06:38<@Alberth>assuming you don't mind having desync and crash issues lying around for a few years
06:39<@Alberth>and all that for a handful of MP players
06:41-!-andythenorth [] has joined #openttd
06:42*andythenorth wonders how to patch build script for git :P
06:42<andythenorth>maybe not right now, but in 2 weeks, after holidays
06:53<@Alberth>what's there to patch?
07:05-!-Nathan1852_ [] has joined #openttd
07:22-!-Wolf01 [] has joined #openttd
07:22<Wolf01>hi hi
07:30-!-Supercheese [] has quit [Read error: Connection reset by peer]
07:31-!-Supercheese [] has joined #openttd
07:49-!-tokai [] has joined #openttd
07:49-!-mode/#openttd [+v tokai] by ChanServ
07:56-!-tokai|noir [] has quit [Ping timeout: 480 seconds]
07:58-!-tycoondemon [] has joined #openttd
08:15<@planetmaker>Alberth, basically it needs patching (or at least checking) the build scripts by the devzone and the (standard) Makefile for NewGRFs:
08:17<@planetmaker>that should be it, I think. Most of the work is actually already done. Mostly it's untested, though
08:20<@planetmaker>also @ andythenorth ^
08:21<@planetmaker>I structured it basically in a similar way to openttd's configure script which checks for the VCS and gets the needed information accordingly. But leaves the rest untouched
08:25<@Alberth>ah right
08:25<@Alberth>a little more generic than I expected :)
08:27<@planetmaker>the first two are fairly generic, yes; but it's simple stuff
08:27<@planetmaker>and the latter three are taylored towards newgrfs, of course and it's basically the same thing again :)
08:28<argoneus>has anyone here had successful passenger only games?
08:28<@planetmaker>does mail count as passenger? :D
08:28<argoneus>I guess
08:28<argoneus>I'm not sure how I want to connect the cities
08:28<argoneus>like if I want one line that goes through all of them in a loop
08:29<argoneus>or just between pairs
08:29<@Alberth>try both? :)
08:29<@Alberth>with cdist, a big loop is going to be a mess
08:30<argoneus>yeah I use assymetric cdist
08:30<argoneus>I thought like having a train+bus in each city
08:30<argoneus>then a bunch of buses in the cities
08:31<argoneus>thats what im gonna do I guess
08:32<Eddi|zuHause>with cdist you want a mesh, not a loop
08:32<@Alberth>the problem is that pax want to travel to neighbouring towns
08:32<@Alberth>to ALL neighbouring towns
08:32<argoneus>ill just do it like real life
08:32<Eddi|zuHause>argoneus: as many direct connections as possible
08:32<argoneus>have a bus network in every city
08:32<argoneus>and pair cities together with trains
08:32<argoneus>that works, right
08:33<Eddi|zuHause>argoneus: straight lines with many cities may help reduce loading/unloading times
08:33-!-wicope [] has joined #openttd
08:34<Eddi|zuHause>if you have A<->B and B<->C, people going A<->C have to unload and load again at B, if instead you have A<->B<->C, they can stay in the train
08:35<Eddi|zuHause>that is especially helpful with train sets that have long distance trains with low loading speed
08:35<argoneus>hmm, im using NUTS
08:57-!-zeknurn [] has quit [Read error: Connection reset by peer]
09:01-!-andythenorth [] has quit [Quit: andythenorth]
09:01-!-zeknurn [] has joined #openttd
09:05-!-Suicyder [~Suicyder@] has joined #openttd
09:21<argoneus>Eddi|zuHause: but doesn't that mean I have one big loop?
09:22<Eddi|zuHause>argoneus: no
09:22<argoneus>or by stay in the train you mean stay in the station
09:22<Eddi|zuHause>argoneus: you have a network of straight lines
09:23<argoneus>Eddi|zuHause: so train from city A goes to city B, unloads passengers and goes back
09:23<argoneus>another train comes from city C to city B to the same station, and loads the people?
09:23<Eddi|zuHause>argoneus: you only combine A<->B and B<->C if there is no shorter direct route A<->C
09:25<Eddi|zuHause>argoneus: this is the ideal network:
09:26<argoneus>oh like that
09:26<Eddi|zuHause>(except that the map is not euclidean, so this "ideal" method doesn't quite work)
09:34<argoneus>note to self
09:34<argoneus>don't make PAX in 1930
09:34<argoneus>the cities are too small to make profit
09:36<Eddi|zuHause>that is totally the opposite of my experience...
09:36<SpComb>you gotta build those railway stations and claim your right-of-ways near the center of the towns before they grow too big :)
09:36-!-tycoondemon2 [] has joined #openttd
09:37<Eddi|zuHause>go through the largest cities, find two that are close (there pretty much always are), build a tram in each of them, and connect them by train
09:37<Eddi|zuHause>you don't have to be too close to the town center with cdist
09:38<Eddi|zuHause>just make sure the tram line between station and town center has excess capacity
09:38<Eddi|zuHause>because usage will skyrocket
09:38-!-smoke_fumus [~smoke_fum@] has joined #openttd
09:41-!-tycoondemon [] has quit [Ping timeout: 480 seconds]
10:06-!-andythenorth [] has joined #openttd
10:08<argoneus>cargodist broke for one of my friends
10:08<argoneus>it says 2200 crates of goods to reykjavik valley
10:09<argoneus>he has a train refitted for goods to load them and unload at reykjavik valley, but it's stuck loading
10:09<argoneus>it doesn't want to load the goods
10:10-!-andythenorth is now known as Guest1576
10:10-!-andythenorth [~Andy@] has joined #openttd
10:11<@Alberth>cdist needs a little time to adapt to new routes
10:12-!-supermop [] has joined #openttd
10:12<Eddi|zuHause>argoneus: try sending the train empty one roundtrip
10:12<argoneus>yup he fixed it
10:12<argoneus>he sent it to depot and back
10:12<argoneus>and now it loads
10:12-!-glx [] has joined #openttd
10:12-!-mode/#openttd [+v glx] by ChanServ
10:13<Eddi|zuHause>sometimes cdist has weird effects when things take too long
10:14-!-Guest1576 [] has quit [Ping timeout: 480 seconds]
10:18-!-andythenorth [~Andy@] has quit [Read error: Connection reset by peer]
10:33-!-supermop [] has quit [Read error: Connection reset by peer]
10:34-!-supermop [] has joined #openttd
10:36<argoneus>is there a proper use for entry/exit signals? I figured that most of my use cases can be replaced by a single path signal
10:36<argoneus>specifically in station design
10:36<argoneus>I know there is but I don't remember
10:39<@planetmaker>there mostly is no use for bock signals when you use path signals
10:39<@planetmaker>for none of the block signals, including entry-, combo- and exit-signals
10:40<argoneus>so block signals are only good for the tricks like throttling traffic?
10:41<@Alberth>the only time I went back to block signals is when I found that path signals wait for moving trains
10:41<@planetmaker>dunno what you mean with 'throtteling traffic'. But yes, you can use them for funky effects like building a calculator with them
10:42<argoneus>Alberth: wait for moving trains?
10:42<@Alberth>which at high rates of incoming traffic is quite nasty :p
10:42<@Alberth>when a train is leaving, it waits
10:43<@Alberth>instead of picking another platform or so
10:43<argoneus>I thought trains reserve a path before they leave
10:43<@Alberth>they do
10:43<@Alberth>I mean the next train waits for the train already there
10:43<argoneus>doesn't it pick a free platform?
10:44<argoneus>I thought that was the selling point
10:44<@Alberth>normally not a problem, except when you really cannot afford to block the entry at all
10:44<@Alberth>it does, but after a very short while
10:44<@Alberth>like I said, it's only important when you're dealing with very high rates of traffic
10:45<argoneus>so um
10:45<argoneus>is something like this bad?
10:45<@Alberth>really don't worry about it
10:46<@Alberth>should be fine, except I'd remove the block signals directly after the junction
10:47<@Alberth>you want sufficient space after the junction for a train to stop
10:47<argoneus>oh right
10:47<argoneus>it could stop at the crossroad
10:47<argoneus>and that's bad
10:47<@Alberth>for performance it doesn't matter, as path signals don't care about multiple trains in the block
10:48<argoneus>but idk
10:48<argoneus>is it bad I'm sad path signals were added? :D
10:48<argoneus>block signals just take so much more thought and they look cool
10:49<argoneus>but they're inferior
10:49<@Alberth>nobody forces you to use path signals
10:49<argoneus>well I want to have optimized everything
10:49<@Alberth>premature optimization is the root of much evil :p
10:50<@Alberth>you're likely to waste loads of time at non-relevant parts
10:50<argoneus>that's the fun part though
10:51<@Alberth>hmm, if you make the split after the first path signal longer, you could add a signal just before the junction
10:51<@Alberth>and I'd remove the connection between the 2nd and 3rd platform
10:52<@Alberth>let me find an example
10:57<@Alberth> with less clutter
10:58<@Alberth>each platform has a an entry and exit shared with another platform
10:59<argoneus>that looks cool
11:01<@Alberth> this is a bit closer to what you are doing, a cross before each pair of platforms
11:02<@Alberth>obviously, the 2nd train of a pair cannot move while the 1st is moving and vice-versa
11:02<@Alberth>if your loading times are long enough, that's not a problem
11:06-!-Nathan1852__ [] has joined #openttd
11:06<argoneus>thanks :3
11:08-!-Nathan1852__ [] has quit [Read error: Connection reset by peer]
11:08-!-Nathan1852 [] has joined #openttd
11:08<@Alberth> don't be too sure, it may get messy :p
11:08<@Alberth>partly also because of other tracks nearby, and industries and other stations that are in the way :p
11:09-!-Pikka [] has quit [Quit: Leaving]
11:13-!-Nathan1852_ [] has quit [Ping timeout: 480 seconds]
11:22-!-Ttech [] has quit [Quit: Este é o fim.]
11:25-!-HerzogDeXtEr [] has joined #openttd
11:26-!-Ttech [] has joined #openttd
11:34-!-supermop [] has quit [Ping timeout: 480 seconds]
11:50-!-TheMask96 [] has quit [Ping timeout: 480 seconds]
11:52-!-wicope [] has quit [Remote host closed the connection]
11:56-!-TheMask96 [] has joined #openttd
11:58-!-smoke_fumus [~smoke_fum@] has quit [Quit: KVIrc 4.2.0 Equilibrium]
12:48-!-frosch123 [] has joined #openttd
13:12-!-jottyfan [] has joined #openttd
13:17-!-Progman [] has joined #openttd
13:52-!-liq3 [] has quit []
13:52-!-Pensacola [] has joined #openttd
13:55-!-gelignite [] has joined #openttd
13:55-!-gelignite_ [] has joined #openttd
13:55-!-gelignite_ [] has quit [Remote host closed the connection]
14:04-!-roidal [] has quit [Quit: WeeChat 1.2]
14:26-!-Progman [] has quit [Remote host closed the connection]
14:36-!-DanMacK [] has joined #openttd
14:37<DanMacK>Hey all
14:37<DanMacK>@seen andythenorth
14:37<@DorpsGek>DanMacK: andythenorth was last seen in #openttd 7 hours, 54 minutes, and 29 seconds ago: <andythenorth> maybe not right now, but in 2 weeks, after holidays
14:39-!-DanMacK [] has quit []
14:39<argoneus>is there a quick way to show current amount of players on a website?
14:39<argoneus>without implementing the whole admin port thing
14:43<Eddi|zuHause>anyone find it amazing that dan comes here, does his usual thing, and gets the reply "i'm back in 2 weeks"? :p
14:44<Eddi|zuHause>argoneus: sure, just use the interface the game uses to get server info
14:44<Eddi|zuHause>which is one single UDP package
14:44<argoneus>oki, thanks
14:45<argoneus>one more thing
14:45<argoneus>do you think it's a bad idea to poll for this every time someone loads the website? like maybe have it on a cron and store it in redis or whatver?
14:45<argoneus>I'm not an experienced webdev (._.)
14:45<Eddi|zuHause>UDP packages are very lightweight
14:46<Eddi|zuHause>the game sends hundreds of them out every time you open the server page
14:46<argoneus>yeah but if I put it on a website
14:46<argoneus>then people could use that to ddos the server kinda
14:46<argoneus>then again
14:47<argoneus>they might as well start the game, find it in the server list and ddos that
14:47<Eddi|zuHause>you'll have quite a hard time DDOS-ing anything through UDP
14:47<argoneus>you can still flood the line
14:48<argoneus>and make requests take longer
14:48<argoneus>or throw them away
14:48<Eddi|zuHause>UDP packages are thrown away all the time
14:48<Eddi|zuHause>which is why the game sends out like 3 equvalent ones in the hope at least one gets through
14:49<argoneus>is there any doc for this basic server info pack(et|age?)
14:49<argoneus>or do I need to dig in source
14:49<argoneus>I can't seem to find it on the wiki
14:50<Eddi|zuHause>i'm fairly sure there is an explanation for this somewhere
14:50<argoneus>on wiki or in doxygen?
14:50<Eddi|zuHause>i'd try doxygen
14:52<@Alberth>sure that'd be more efficient than just doing the admin port thingie?
14:53<argoneus>Alberth: isn't it a bad idea to query the admin port every time someone loads the website?
14:53<argoneus>then again I always worry about how to make things properly and in the end I don't even have a working prototype because I get bored
14:57<frosch123>argoneus: <- i think you are looking for that
15:02<@Alberth>argoneus: I am just very amazed about "admin port is too much work, let's do custom source code checking and hacking to get an inferior partial solution instead"
15:03<argoneus>well, but
15:03<argoneus>well, you're right
15:03<@Alberth>argoneus: given that there exist servers today that use the admin port, I'd say it can't be that bad :)
15:03<argoneus>but do I want to cache the information
15:03<argoneus>or query for it evertime someone loads the website
15:03<argoneus>I'm not sure how heavy it is
15:04<argoneus>it won't matter in my case but exercise is exercise
15:04<@Alberth>I said something about premature optimization already didn't I? :)
15:04<argoneus>yeah but
15:04<@Alberth>no way to know in advance, just try it, and optimize when needed
15:05-!-shadowalker [~dark@2a03:b0c0:0:1010::102:d001] has quit [Quit: WeeChat 1.0.1]
15:05<Eddi|zuHause>how many million views per minute do you think your website gets?
15:06<argoneus>like 0.000005 million
15:06<Eddi|zuHause>that is about the order of magnitude i think that "performance" should even enter the consideration
15:06<argoneus>it's just that
15:06<argoneus>you hear these nightmare stories
15:07<argoneus>how people made some random shit project, and it blew up
15:07<argoneus>and then they couldn't handle the traffic and it fucked up
15:07<argoneus>then again if you get a facebook esque that blows up I assume fixing your infrastructure is the least of your problems with all the money you have
15:07<@Rubidium>it's not like the OpenTTD website is optimised, but it survived slashdot fairly easily
15:07<Eddi|zuHause>sure. but for every such project that blew up, a million projects didn't blow up.
15:07<frosch123>just redirect people to pikkas fundraiser, if the load is too big
15:07<Eddi|zuHause>it's just how it works...
15:07<argoneus>so basically
15:08<argoneus>for random projects, functionality > scalability?
15:08<Eddi|zuHause>__ln__: didn't we go through this last year already?
15:08<argoneus>I just keep underestimating today's computers
15:09<__ln__>Eddi|zuHause: there was no physical evidence of a submarine back then; now they've found one.
15:09<Eddi|zuHause>__ln__: i'm sure it's totally unrelated
15:10<@Rubidium>argoneus: good practice is to write simple to understand code and optimize the hot spots. Arguably you can get more performance gain from tuning the web server / php / python than querying the server
15:10<@Rubidium>if you're that concerned about performance for a connecting client, then just generate static HTML every X seconds
15:10<+glx>and your wrapper around admin port can do all the necessary caching
15:11<@Alberth>argoneus: scalability is a non-issue until you actually have functionality
15:11<+glx>(if caching is really needed)
15:12<argoneus>I don't even know how to cache :<
15:12<Eddi|zuHause>argoneus: the point is, you don't have to know. just prepare your code to have a cache added later, when the need arises
15:13<argoneus>yeah I don't even know how to do that, I don't do webdev
15:13<Eddi|zuHause>argoneus: there are definitely standard solutions for caching
15:15-!-HerzogDeXtEr [] has quit [Quit: Leaving.]
15:28-!-Progman [] has joined #openttd
15:44<NGC3982>I started this EVE thing
15:44<NGC3982>And i haven't used the rest of the internet for the last year.
15:44-!-Alberth [~alberth@2001:981:c6c5:1:be5f:f4ff:feac:e11] has left #openttd []
15:49-!-Pensacola [] has quit [Remote host closed the connection]
15:53-!-Wormnest [] has joined #openttd
15:57<Supercheese>that is precisely why I checked out EVE and said, "Nope."
15:57<Supercheese>would be far too consuming
16:25-!-jottyfan [] has quit [Quit: Konversation terminated!]
16:29-!-sla_ro|master [slamaster@] has quit []
16:35-!-frosch123 [] has quit [Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn]
16:36<@Terkhen>good night
17:05-!-Wormnest [] has quit [Quit: Leaving]
17:08-!-gelignite [] has quit [Quit:]
17:39<Wolf01> A train cannot turn it's direction on the rail by itself or magic game mechanic. This is different to games like OpenTTD!
17:40-!-DDR [] has quit [Ping timeout: 480 seconds]
17:40<Wolf01>need to build a double headed train
17:49-!-Progman [] has quit [Remote host closed the connection]
18:05-!-Suicyder [~Suicyder@] has quit [Quit: HydraIRC -> <- The alternative IRC client]
18:13<peter1138>what game?
18:20<Wolf01>the manual driving, shunting and rolling stock connection is wonderful *_*
18:21<Wolf01>too bad I have other things to do, like killing those zerg
18:39<Nemoder>the factorio trains are pretty cool except by the time you've built enough stations for them to be useful you already have all the resources you need to finish the game
18:48-!-Nathan1852_ [] has joined #openttd
18:55-!-Nathan1852 [] has quit [Ping timeout: 480 seconds]
19:05-!-Biolunar [] has joined #openttd
19:09-!-Bluelight [] has joined #openttd
19:15-!-Bluelight [] has quit [Quit: ChatZilla [Firefox 39.0/20150630154324]]
19:24-!-Hiddenfunstuff [] has quit [Quit: HydraIRC -> <- Go on, try it!]
19:24-!-Wolf01 [] has quit [Quit: Once again the world is quick to bury me.]
19:26-!-Nathan1852_ [] has quit [Ping timeout: 480 seconds]
19:42-!-Biolunar [] has quit [Ping timeout: 480 seconds]
20:39-!-JezK [~jez@2407:7800:400:107f:3db5:daca:8457:e66a] has joined #openttd
22:29-!-glx [] has quit [Quit: Bye]
23:27-!-liq3 [] has joined #openttd
23:38<Flygon> This is the part where we implement Nutella as an industry
23:56<Supercheese>Yuck, Nutella is gross
23:56<Supercheese>anything hazelnut is
---Logclosed Tue Jul 28 00:00:27 2015