Back to Home / #openttd / 2013 / 07 / Prev Day | Next Day
#openttd IRC Logs for 2013-07-19

---Logopened Fri Jul 19 00:00:15 2013
00:07-!-DabuYu [DoubleYou@] has quit []
00:41-!-roboboy [] has quit [Ping timeout: 480 seconds]
00:56-!-Eddi|zuHause [] has quit []
00:56-!-Eddi|zuHause [] has joined #openttd
01:30-!-Prof_Frink [~proffrink@] has quit [Remote host closed the connection]
02:04-!-sla_ro|master [slamaster@] has joined #openttd
02:22-!-tycoondemon [] has quit [Ping timeout: 480 seconds]
02:35-!-roboboy [] has joined #openttd
02:39-!-tycoondemon [] has joined #openttd
02:42-!-montalvo [] has quit [Quit: Textual IRC Client:]
02:51-!-Devroush [] has joined #openttd
02:59-!-tycoondemon [] has quit []
03:00-!-tycoondemon [] has joined #openttd
03:15<dihedral>good morning
03:16<__ln__>it's too early to say
03:19<Xaroth|Work>I'd have to agree with __ln__ for now
03:20<Supercheese>Some of us haven't even had half an hour of Friday yet ;)
03:21<roboboy>for me it's 5:20 PM
03:22<Supercheese>must be on the other side of the date line
03:22<Supercheese>well, going backwards anyway
03:22-!-Jomann [] has joined #openttd
03:22<roboboy>well it's Friday here
03:29-!-appe [] has joined #openttd
03:29-!-appe is now known as NGC3982
03:36-!-Zuu [] has joined #openttd
03:45-!-pjpe [~oftc-webi@] has quit [Quit: Page closed]
03:50<NGC3982>Jesus. The LEGO provided for James May's Toy Stories had a sales worth of £900k.
03:50<NGC3982>I hope it advertised itself well.
04:00<peter1139>postfix/pickup[25537]: warning: 59976C0A4B9B: message has been queued for 1345 days
04:00<peter1139>that's pretty old :p
04:03-!-permagreen [~donovan@] has joined #openttd
04:13-!-roboboy [] has quit [Ping timeout: 480 seconds]
04:13<Supercheese>good night
04:13-!-Supercheese [~Password4@] has quit [Quit: ChatZilla [Firefox 22.0/20130618035212]]
04:40-!-Ristovski [] has joined #openttd
05:00-!-zooks [] has joined #openttd
05:11-!-Midnightmyth [] has joined #openttd
05:11-!-roboboy [] has joined #openttd
05:20-!-zeknurn [] has quit [Read error: Connection reset by peer]
05:20-!-zeknurn [] has joined #openttd
05:32-!-pugi [] has joined #openttd
05:33-!-Alberth [] has joined #openttd
05:33-!-mode/#openttd [+o Alberth] by ChanServ
05:40-!-zooks [] has quit [Ping timeout: 480 seconds]
05:50-!-MatrixCL [] has joined #openttd
05:55<Zuu>Hmm, NAI could get SCP support for allowing AIs to query the stockpile count.
05:58<Zuu>Hmm, my manager page should give a good overview over GSs with SCP support :-)
05:58-!-oskari89 [] has joined #openttd
06:00<peter1139>To securely copy files?
06:00<Xaroth|Work>but, if they support SCP, do they also support SSH?
06:01<@Alberth>scp means something else here, but I forgot what it is
06:02<Zuu>Xaroth|Work: Nope, its Script Communication Protocol, which is quite unsecure as communication happens through signs on tile 0,0.
06:02<Zuu>Its used to allow GSes and AIs to communicate.
06:02<Zuu>Eg. for a GS to expose goals etc. to AIs.
06:03<Zuu>Its a user-land library that sort of abuses the signs :-)
06:03<@Alberth>shouldn't that be in the story book, in a human-readable form?
06:04<Zuu>How does that help AIs?
06:04<@Alberth>they read it from the story book?
06:04<Zuu>using SCP, NoCarGoal expose the 3 goal cargos so that AIs can focus on these and give you a challenge
06:04<@Alberth>is there a definition of what you can express?
06:04<Zuu>AIs are unaware of the story book
06:05<Zuu>Using SCP you can express things not defined by the Script API
06:05<@Alberth>using signs at the world doesn't sound like the right solution to me
06:05-!-zooks [] has joined #openttd
06:06<Xaroth|Work>sounds like it needs a patch
06:06<Zuu>It was the best solution available when scp was written. Later we could add API extensions that adds a hidden communication channel that SCP can use on newer OpenTTD versions and fall back to the sign method for 1.3 and older.
06:06<@Alberth>Xaroth|Work: more like it needs a proper solution :p
06:07<@Alberth>why hidden? that undermines the idea of AIs
06:08<Zuu>I though that was your problem, that the players can see the signs at tile 0,0 if they look over there.
06:08<Xaroth|Work>Alberth: that too
06:09<Xaroth|Work>Zuu: i think the problem more is the abuse of signs for something that should be put in programmatically...
06:09<Zuu>Writing an AI that parses the strings in the StoryBook doesn't sound like the rigth solution to me.
06:10<@Alberth>no it's not, you write the technical information in scp format in the book, so ais can read it. the book expands the information to human readable form
06:11<@Alberth>at least, that would a solution more in my line of thought
06:11<Zuu>Wouldn't that need a central definition of all things that GSs can expose to AIs?
06:12<@Alberth>like scp?
06:12<Zuu>scp don't require that
06:12<Zuu>scp is a communication protocol. It allow clients to send comands with arguments over a communication channel.
06:12<@Alberth>so how do I write a GS that tells any AI to transport coal?
06:12<Zuu>Its up to the GS to offer a specification of which commands that it offers.
06:13-!-SamanthaD [] has quit [Ping timeout: 480 seconds]
06:13<@Alberth>so require that to provide a human readable expansion as well?
06:13<Xaroth|Work>AI should never, ever, be using data that a human cannot obtain
06:13<Xaroth|Work>that's called cheating :P
06:15<Zuu>Alberth: If you want to write an AI with support for NoCarGoal, you read the readme of NoCarGoal which specify which incomming command that it accepts and what commands that it will send back to the AI.
06:15<Zuu> <-- readme
06:15<@Alberth>I want to write an AI that handles any GS. Does that mean I have to figure out how to talk to every GS out there?
06:16<@Alberth>sounds like a nice exponential explosion
06:16<Zuu>Yes. At the time we wrote the library we figured that it would be hard to add generic support for every GS variant that anyone would think off.
06:17<Zuu>However, it is still open for conventions where all GSs use the same command ontop of scp to exchange some information.
06:17<@Alberth>indeed, I see that proposal as cheating
06:17<@Alberth>I cannot read bytes
06:17<Xaroth|Work>I agree with Alberth
06:18<Zuu>But you can read text information in news, story book etc.
06:18<Zuu>Its up to the GS to only provide that same information to AIs and not more.
06:18<@Alberth>it's good to have a protocol separate from contents
06:19<@Alberth>how do I know the AI gets the same information I do?
06:19<Zuu>That requires that you create a protocol that support every information exchange that any GS ever will like to use.
06:20<@Alberth>for common goals, I know that other human players read the same information I get
06:20<@Alberth>no , it needs a language, a set of expressions common to understand what is wanted
06:20-!-gynter [] has joined #openttd
06:21<@Alberth>true it may not be complete, but that can be changed as needed
06:21<@Alberth>just like eg the NewGRF definition
06:21<Zuu>I do see your point, and that it would be nice to have a protocol where GSs tell OpenTTD that I want goal X and OpenTTD provide the strings (including translation) and AI API for that. However, GS authors will be creative and come up with ideas not possible to do this way as long as we allow manual town growth trottle etc.
06:21<@Alberth>or OpenTTD windows
06:22<Zuu>What SCP offers is a way for scripts that goes beyond the fully supported golas in OpenTTD to still allow AIs to have the same access to information as players do. It do also allow a GS author to give more information, but that is unfortunately hard to avoid. All you can do is to read the source code then of the GS.
06:24<@Alberth>Define a language, and be free in expanding it with new expressions
06:24<@Alberth>I doubt you'll get past 100 commands or so
06:24<@Alberth>AIs are limited in smartness
06:26<@Alberth>the big advantage is that you get a common language between GSes and AIs, which imh is worth more than being able to express anything you like
06:26<@Alberth>since the latter is hardly useful, unless you make specific GS/AI combinations
06:29<Zuu>What we did forsee was that when scp would get adopted by more GSes, there would be an AI library that offecs an AI client with a simple API to talk to all GSs with SCP support (and a published API).
06:29<@Alberth>yep that's likely to happen
06:30<Zuu>However, the interest to include scp in GSs (and AIs) have so far been quite low.
06:31<Zuu>The only scripts that I know support it so far is NoCarGoal and CluelessPlus.
06:31<@Alberth>it's a chicken and egg problem
06:31<@Alberth>and it's not in the regular API is perhaps also a point
06:32<Zuu>Indeed, to discover it you need to read about it in the forum or see it mentioned in the readme of a GS that supports it.
06:32<@Alberth>eg I didn't know it existed. Tbh, I also don't look for it
06:33<gynter>nmlc have issues using full paths?
06:34<@Alberth>gynter: I don't think so, what does it say?
06:34<Zuu>The actual implementation was made by krinn. I did mostly provide feedback to him, wrote some wiki documentation and tested it.
06:34<gynter>Alberth: Nothing, just hangs. If i make paths to relative it's fine
06:34<@Alberth>gynter: interesting, could you please file a bug report?
06:35<@Alberth>preferably with a small example, although
06:35<@Alberth>paths are very system specific, often :)
06:35<gynter>I might know whats the problem, sec
06:37<gynter>Yep, nmlc didn't like "ö" in path
06:37<gynter>thou older nmlc worked fine, I just updated to latest
06:38<@Alberth>Zuu: another point is perhaps that from a GS or AI author point of view, it's mostly useless now. I have to write both a GS and a AI, and define my own language between them, which sounds like too much work if you only want to do one of them
06:38<@Alberth>gynter: I changed path handling so paths are handled the same way at both windows and unix
06:38-!-ntoskrnl [] has joined #openttd
06:39<gynter>hmm never mind, I lied
06:39<gynter>still same problem
06:39<gynter>but older version (can't recall which one ofc) was ok
06:39<gynter>something r15xx
06:39-!-Brumi [] has joined #openttd
06:40<@Alberth>older version broke at users writing windows paths in nml that completely failed to work at unix systems
06:40<Zuu>Alberth: The idea is that a GS author just add the support at his end and an AI adds just the support at that end. It shouldn't need that you are the author of both ends to use scp. However, for GS authors it is useful to bug-test the scp interface using a test AI.
06:40<@Alberth>due to differences in interpreting case
06:41<@Alberth>Zuu: how? I don't have a language to speak or listen
06:42<Zuu>The GS first need to offer scp support before an AI can talk to it. The GS will declare the set of commands that it can exchange with AIs. Eg. the 'language'.
06:42<Zuu>The GS author will then document this in his readme or where he/she see it fits.
06:43<gynter>strace shows mremap(0x7f6405258000, 146919424, 165281792, MREMAP_MAYMOVE) = 0x7f6405258000 and thats it
06:44<Zuu>For example if an AI send the command 'CurrentGoal' with arg0=company ID to NoCarGoal, then you will get the current goals for that company. The readme specifies exactly in which order data will be returned (in the data array which scp pass to the AI)
06:44<@Alberth>gynter: that has little meaning, as it's deep inside the Python interpreter
06:45<Zuu>Sending Setting('play_years') to NoCarGoal GS will return number of years that the game covers.
06:46<Zuu>(the name of the commands is exchanged in a init procedure. byond that commands are translated to IDs by scp to compact the data transported to the other side)
06:46<@Alberth>gynter: path handling is in nml/ line 212 or so (def find_file(path): ) maybe you can add some debug prints there?
06:49<@Alberth>Zuu: For example if an AI send the command 'CurrentGoal' with arg0=company ID to NoCarGoal <-- ah nice, it can cheat to read all goals of all companies?
06:50<Zuu>That is available for players in trunk
06:50<Zuu>and the goal is the same for all players in NoCarGoal
06:51<@Alberth>so you don't need to have a company at all :)
06:51<Zuu>Its only there if the goals will ever be different in any future version of NoCarGoal (which is unlikely)
06:51<@Alberth>but ok, if players can read it too, that's fine
06:52<Zuu>Anyway, a GS currently can do more evil things than providing AIs with additional informaiton. You need to trunst the GS you play.
06:53<@Alberth>my point is perhaps that as AI author, I am not interested in a specific GS
06:53<@Alberth>I don't know what my users want to use
06:54<Zuu>Well, once GSs provide information that AIs can read, it is possible to create a library that provides more generic support. Eg. indicate what cargos it should focus on. However, the nature of GSs is that they can add very specific goals that are hard to handle with general rules.
07:00-!-ntoskrnl [] has quit [Ping timeout: 480 seconds]
07:05<gynter>Alberth: it seems to me, that for some reason it'll stay on the line 228 while loop. Gona add some more debug code.
07:05<@Alberth>k, thanks
07:10<Zuu>Alberth: However, even though I've argued for the reasons why scp was created, I have also had ideas that are more on the lines of your idea. As a start for that, my idea was that towns can set a GSGoal as requirement for growth. Only when the goal is marked as completed, the town will grow. Togeather with your idea of being able to specify the conditions to accomplish the goal using a rule set, it could bring both more translated goals and bette
07:10<Zuu>r AI support.
07:12*Alberth has to think about towns settng a goal
07:16<Zuu>The town goals today, are perhaps not really true goals, but more conditions that needs to be met before the town grows.
07:18<@Alberth>Yeah, I agree with that. And while it's somewhat unusual, it seems like a nice idea to me
07:18<Zuu>But we got about 4-5 GSs that adds such rules. Some uses the OpenTTD API with town effects which is something AIs can access. However, others have just a text in the Town Window for players with the goal and manually trottles the town growth.
07:19<@Alberth>o/ planetmaker
07:19<@Alberth>town authorities to the power! :)
07:19<@planetmaker>Zuu, from the little of the backlog I just saw: is the script communication protocol not obsolete, if the scripts could communicate e.g. via json objects?
07:19<@planetmaker>like admin port and GS?
07:20<Xaroth|Work>that's not the sollution tho
07:20<@Alberth>did you also read my objection against secret communication channels for AIs?
07:21<@planetmaker>like not being able to do what players can't?
07:22<Zuu>planetmaker: The network layer of the protocol could become obsolete if the API provides an alternative communication channel that is better than encoding data into signs (limited to 31 chars and only allow visible characters - eg not even the full lower byte). However the SCP Lib could still exist and just transparently use the new channel when available.
07:22<@planetmaker>while that's true, I do believe writing a syntactic and semantic parser for all languages to read story book seems a bit too high expecatation for AI authors
07:22<gynter>Alberth: the problem is with os.path.split('/')
07:22<gynter>This always returns ('/', '') therefor path will never be ''
07:22<@planetmaker>if it was something else, I missed it
07:23<@planetmaker>Zuu, sure :-)
07:23<@Alberth>gynter: for absolute paths?
07:24<@Alberth>right, makes sense indeed
07:25<Xaroth|Work>planetmaker: instead, metadata could be added to the story book containing the data that the parser would otherwise extract
07:25<Xaroth|Work>as long as it contains the same information as what players see
07:25<Zuu>I think what Alberth proposes is a GSCondition object or similar where you can add eg [destination = town 52], [cargo_list = {1, 3}], [end_date=1960-01-01] to create a goal condition which AIs can intreprete using the AI API and we can auto-generate a human readable string to show in the goal list.
07:25<Xaroth|Work>or that
07:26<Zuu>The auto generated goal string will then be translatable via the web translator of OpenTTD and offer better language support than GSes will ever do.
07:26<@Alberth>gynter: hmm, perhaps move the leading / to "drive" ? that's supposed to handle prefixes like c:, but this one seems like a good prefix as well
07:27<gynter>Alberth: the easiest solution would be to make the... you type faster :P
07:27<Zuu>It will then cover 80 % of what SCP offers and will in practice cause SCP to be more or less obsolete except for AI+GS combos, though in that case you easier write a sign communication yourself without help of SCP.
07:27<gynter>Alberth: yes, that would be good solution imho
07:28<@Alberth>gynter: I wrote the code, so I can skip the "what does this code do" step :)
07:29<@planetmaker>^ for me that's usually a function of delta time ;-)
07:37<@Alberth>that happens when my code is more clever than I remember :)
07:38<Xaroth|Work>I'm glad you're not working on skynet then...
07:40<@Alberth>gynter: this may be a better solution?
07:41-!-TheMask96 [] has quit [Ping timeout: 480 seconds]
07:41<@Alberth>Xaroth|Work: perhaps that's why I don't write web apps?... :)
07:42<@Alberth>bummer, you can write // in the middle of paths :(
07:43<gynter>Alberth: i'd use if not filepart :)
07:43<@planetmaker>yes... and tbh, that can be helpful under certain occasions
07:44<gynter>Alberth: os.path.normpath fixes that afaik
07:44<@Alberth>gynter: I prefer to be explicit in comparisons :)
07:45<@Alberth>and good point about normpath
07:47<@Alberth>planetmaker: opinion on that paste?
07:47-!-TheMask96 [] has joined #openttd
07:48<@Alberth>gynter: can you check whether it fixes your problem?
07:49<gynter>sec, but the pythonic way is to use not instead of empty comparison :)
07:53<@Alberth>I have been bitten too many times by Python for not being specific in what type I expect to have
07:53<gynter>this code is not really linux friendly there :D
07:54<gynter>you don't need read permission to access the directory, only execute, gona make few changes there
07:55<@Alberth>it's surprising how many applications die when you remove read access of $HOME :)
07:56<gynter>yea, shit loads :D
08:00<gynter> raise ScriptError("Path \"%s\" is not unique (case conversion gave %d solutions)" % (os.path.join(path, comp), len(matches)))
08:00<gynter>really :) ?
08:01<gynter>/home/User and /home/user are different paths :)
08:01<gynter>ofc one should be incredibelly dumb to do something like that, but still
08:02<@Alberth>at windows, you may get more solutions
08:03<@Alberth>although typically, people add paths in the nml source that don't actually exist at unix systems
08:05<@Alberth>the goal was to handle the latter kind of problems
08:20-!-ntoskrnl [] has joined #openttd
08:26-!-Brumi [] has quit []
08:29<@planetmaker>Alberth, I didn't follow too closely. What problem exactly does it fix?
08:30-!-amiller [] has quit [Ping timeout: 480 seconds]
08:30<@Alberth>not getting into a livelock on the / path (which happens while breaking down an absolute path)
08:32-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
08:35<@planetmaker>I didn't yet test it, but looks like it can do the trick
08:35<@planetmaker>so please, if it fixes it, go for it
08:37<@Alberth>gynter is still working at more fixes :)
08:38<@planetmaker>:-) Let'im roll, I think :-D
08:42-!-MatrixCL [] has joined #openttd
08:43-!-oskari89 [] has quit []
08:47-!-Midnightmyth [] has quit [Read error: Connection reset by peer]
08:48<@planetmaker>hehe, nice analysis, Alberth on the target group of mobile tt
08:48<@planetmaker>(and en-passant scaling down of the importance of the majority of people posting there :D)
08:50<peter1139>yes makes sense
08:52-!-Brumi [] has joined #openttd
08:53-!-Invulner [] has joined #openttd
08:55<Invulner>Hello, can someone recommend me a nice newgrf with locomotives only? I want to use regular wagons but with more than 1 locomotive on early sub-arctic map
08:56<Invulner>or at least a newgrf without many same-cargo wagons
08:58-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
08:58-!-zooks [] has quit [Read error: Operation timed out]
08:59<V453000>same-cargo wagons? :o
08:59<@planetmaker>I wondered about the meaning of that, too
08:59<Invulner>like 10 types of passenger carriages
09:00<V453000>:D ah, yes, I love that too
09:00<V453000>well all I can say is NUTS
09:00<V453000>norwegian train set does not have wagons, but it is not too suitable for "early" games
09:00<@planetmaker>yeah, maybe that's worth a try
09:01<Invulner>alright, i'll check those out, thank you!
09:01<@Alberth>V453000: would "regular wagons" work?
09:01<V453000>work with what?
09:01<@Alberth>nuts engines
09:01<V453000>no of course not
09:02<V453000>but 1 wagon for every cargo is not confusing :P
09:02<@Alberth>but Invulner wants to have lots of wagons, apparently
09:03<Invulner>i wanna have lots of locomotives, but few carriages
09:03<V453000>"without many" :P
09:03<V453000>then NUTS is perfect, it has lots of useful trains Invulner
09:03<Invulner>yeah i've just downloaded it, seems good
09:04<V453000>universal wagons for freight cargoes, and express wagon for all of passengers/mail/gold/valuables/diamonds
09:04<Invulner>could've done without wetrails though
09:04<@Alberth>Invulner: it's called NUTS for a reason :p
09:04-!-MatrixCL [] has joined #openttd
09:04<V453000>well you dont have to use them :)
09:05-!-Prof_Frink [~proffrink@] has joined #openttd
09:05<Invulner>ok thanks again, seems to be just what i needed
09:06<V453000>if you want to read how to use it effectively,
09:07<peter1139>i like the big graph
09:07<V453000>mainly the Wiki part
09:08<peter1139>takes an element of surprise out though ;p
09:08<V453000>graph is nice but outdated :d
09:08<V453000>current is 5.5
09:08<peter1139>too much work to remake? :p
09:08<V453000>no I just didnt think of it lately :)
09:08<V453000>might do soonish
09:10<V453000>about surprise, I fully agree, but people who play it for nth time (n is not 1 :) ), they like to have a table to know what comes when
09:17-!-LordAro [] has joined #openttd
09:17*LordAro waves
09:17<@Alberth>LordAro has arrived
09:17<@Alberth>hi hi
09:18<Xaroth|Work>dun dun dunnnnnnnn
09:18<LordAro>hai Alberth & planetmaker :)
09:18<LordAro>& Xaroth|Work :P
09:30-!-ntoskrnl [] has quit [Ping timeout: 480 seconds]
09:32-!-DarkAce-Z [~BillyMays@] has joined #openttd
09:35-!-SpComb^_ [] has joined #openttd
09:36-!-DarkAceZ [~BillyMays@] has quit [Ping timeout: 480 seconds]
09:37-!-SpComb [] has quit [Ping timeout: 480 seconds]
09:41-!-amiller [] has joined #openttd
09:42-!-HerzogDeXtEr [] has joined #openttd
10:05-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
10:06-!-Elukka [] has joined #openttd
10:09-!-MatrixCL [] has joined #openttd
10:17-!-Tom_Soft [] has joined #openttd
10:17-!-permagreen [~donovan@] has quit [Ping timeout: 480 seconds]
10:17-!-Tom_Soft [] has quit []
10:27-!-Invulner [] has quit [Quit: Leaving]
10:31<@Belugas>sir Alberth!
10:42-!-Brumi [] has quit []
10:44<gynter>sry it took so long, have daily job alos :P
10:47-!-mindlesstux [] has quit [Ping timeout: 480 seconds]
10:49-!-Devroush [] has quit [Ping timeout: 480 seconds]
10:50<@Alberth>that looks useful, could you make an issue for it, or shall I do that?
10:50<gynter>Please do that, i'm too lazy to register myself :P
10:53<gynter>hmm, getting some assert errors when increasing town name possiblility
10:53<gynter>10 -> 100
10:53<gynter>10 -> 90 is fine
10:55<gynter>if self.num_bits is None:
10:55<gynter>this makes me cringe :D
10:57-!-Lord_Aro [] has joined #openttd
10:57-!-LordAro is now known as Guest384
10:57-!-Lord_Aro is now known as LordAro
10:58<TWerkhoven>what do i put under [server_bind_addresses] to make adminport listen beyond localhost, just the local lan ip?
10:59<@planetmaker>uhm... don't assign any value unless you have a server with two network addresses
11:00<@planetmaker>and the admin port simply listens on the port which it is assigned to
11:00<@Alberth>gynter: created
11:00<@planetmaker>same IP as the client needs to assign
11:02<gynter>somehow i have a set of town names autobroken into a 256 part chunk
11:02<@planetmaker>no larger chunks supported afaik
11:03<gynter>255 should be the largest
11:03-!-Guest384 [] has quit [Ping timeout: 480 seconds]
11:03<@planetmaker>oh, an off-by-one :-)
11:04<gynter>The problem is that if I increase the probability for Tallinn in from 10 to 100 then it breaks.
11:05-!-Djohaal [~Djohaal@] has joined #openttd
11:06<peter1139>i didn't think town names had the concept of probabilities
11:07<@planetmaker>peter1139, they have for... ages :-)
11:09<peter1139>here's me thinking the name generator was meant to making up names from parts, not just picking from a big list
11:09<peter1139>*make up
11:10<@planetmaker>it was, yes :-)
11:10<@planetmaker>yet the big lists is the FAAR more common way to use it
11:10<gynter>Well letting it to randomly combine IRL city names doesn't really make sense :)
11:11<peter1139>randomly placing IRL city names on a random map doesn't make much sense either :p
11:12<@planetmaker>gynter, the real art imho is to make a random townname generator which matches your language and region
11:12<@planetmaker>kinda like localized default names
11:12<gynter>That's the point :)
11:12<@planetmaker>no, a simple list doesn't
11:14<@planetmaker> ;-)
11:15<gynter>I know how to do this, but that's not the intent of my newgrf
11:17<gynter>Thou this would be cool also :)
11:17-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
11:17<gynter>could get like few million possible town names :P
11:17<@planetmaker>exactly. creating more variety
11:18<gynter>well 4k is good too :P
11:20-!-beomouth [~james@] has joined #openttd
11:29-!-mindlesstux [] has joined #openttd
11:31-!-roboboy [] has quit [Read error: Operation timed out]
11:40-!-TheMask96 [] has quit [Ping timeout: 480 seconds]
11:43-!-TheMask96 [] has joined #openttd
11:47-!-Progman [] has joined #openttd
11:59-!-Alice3 [] has joined #openttd
12:10-!-amiller [] has quit [Ping timeout: 480 seconds]
12:14-!-MatrixCL [] has joined #openttd
12:19-!-Progman [] has quit [Remote host closed the connection]
12:51-!-frosch123 [] has joined #openttd
12:55-!-Brumi [] has joined #openttd
13:08-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
13:13-!-MatrixCL [] has joined #openttd
13:18-!-RavingManiac [~RavingMan@] has joined #openttd
13:22-!-glx [] has joined #openttd
13:22-!-mode/#openttd [+v glx] by ChanServ
13:25-!-amiller [] has joined #openttd
13:30-!-andythenorth [~Andy@] has joined #openttd
13:31<@Alberth>hi hi
13:34-!-andythenorth [~Andy@] has quit [Quit: andythenorth]
13:54-!-permagreen [] has joined #openttd
13:54-!-amiller [] has quit [Ping timeout: 480 seconds]
14:04-!-beomouth [~james@] has quit [Ping timeout: 480 seconds]
14:05-!-RavingManiac [~RavingMan@] has quit [Ping timeout: 480 seconds]
14:07-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
14:10-!-MatrixCL [] has joined #openttd
14:12-!-abchirk_ [] has joined #openttd
14:12-!-beomouth [~james@] has joined #openttd
14:14-!-Jomann [] has quit [Read error: Operation timed out]
14:15-!-amiller [] has joined #openttd
14:18-!-MatrixCL [] has quit [Ping timeout: 480 seconds]
14:28-!-Pol [] has joined #openttd
14:46-!-Brumi [] has quit []
14:53-!-andythenorth [] has joined #openttd
14:53<andythenorth>so cargodist....
14:53<andythenorth>what do I do? o_O
14:54<Zuu>Stop using full load and use timetables instead
14:54<@Alberth>run away?
14:55-!-montalvo [~montalvo@] has joined #openttd
14:55<Zuu>Provide only point-to-point connections and no shared stations => same behaviour as before :-)
14:56<@Alberth>just disable cargodist is a lot simpler then:)
14:57<Zuu>not enabling it in the first place is even easier :-)
14:57<Zuu>andythenorth: There is a forum thread in general forum on how to adopt the playing style to cargodist.
14:57<@Alberth>I have been doing a few short starts with cargodist to find out how to do map-wide cargo distribution
14:58<@Alberth>the easiest seems to be to make a big network with constant trains running around, visiting those industries that are relevant for their wagons
14:58<Zuu>If you disable town growth, that helps not getting overcrowded trains. :-)
14:59<@Alberth>I hardly do pax transport :)
15:07<andythenorth>biab :)
15:09-!-montalvo [~montalvo@] has quit [Quit: Textual IRC Client:]
15:10-!-gelignite [] has joined #openttd
15:11-!-andythenorth [] has quit [Quit: andythenorth]
15:19-!-Alberth [] has left #openttd []
15:28-!-Pol [] has quit [Remote host closed the connection]
15:41-!-KritiK [] has joined #openttd
15:44-!-SamanthaD [] has joined #openttd
15:55-!-Extrems1 [borgs@] has quit [Ping timeout: 480 seconds]
15:57-!-andythenorth [] has joined #openttd
16:19-!-Extrems [borgs@] has joined #openttd
16:22<andythenorth>can someone finish my FISH docs, ta?
16:23<andythenorth>too hot to code well :P
16:23<andythenorth>it's a little bit FIRS-ish right now :P
16:24<andythenorth>also I need to render some properties here
16:37<andythenorth>how much can I trust a dict to return in same order when iterated?
16:48<@planetmaker>o/ andythenorth
16:49<@planetmaker>did you see the patch I posted?
16:49<@planetmaker>not really for the readme, but... meh
16:49<andythenorth>or firs?
16:49<@planetmaker>you kind broke shipped bundles with r1159
16:49<@planetmaker>see bug tracker
16:49<andythenorth>I'll apply now
16:50<@planetmaker>I quite don't get the reason for 1159 anyway
16:50<@planetmaker>what was the reason behind it?
16:51<andythenorth>previous commit broke bundle build, I didn't want bundle server nagging me (and it was way past bedtime) ;)
16:51<andythenorth>I would have done it right...but tired
16:52<@planetmaker>I don't understand that... you didn't touch docs generation there?
16:53<@planetmaker>oh, you did
16:53-!-oskari89 [] has joined #openttd
16:54<@planetmaker>I guess it's broken again :D
16:54<andythenorth>we'll find out in a minute
16:54<@planetmaker>it then needs some magic like the other targets
16:55<andythenorth>same issue got solved in FIRS, but different makefile structure iirc
16:55*andythenorth looks
16:55<@planetmaker>in any case, DOC_FILES should not be altered really. That defines what gets shipped
16:56<andythenorth>FIRS has Makefile_doc
16:56<@planetmaker>the GENERATE_DOC variable may be pointed to a new target, though in
16:56<andythenorth>should maybe point to src/
16:56<@planetmaker>FIRS likely uses the old one? Or does it use the new and just has still dead code there?
16:57<andythenorth>could be dead code
16:57<@planetmaker>possibly, yes
16:57<@planetmaker>ok, let's see
16:57<andythenorth>either dead code, or FIRS renders docs twice :P
16:58<@planetmaker>that's for another day :-)
16:58<@planetmaker>docs are rendered not by, yes, but only src/
17:00<andythenorth> renders docs too
17:00<andythenorth>that could be adjusted
17:00<andythenorth>it's current state, but not essential
17:00<andythenorth>also should strictly be in /scripts :)
17:01-!-y2000rtc [] has joined #openttd
17:01<@planetmaker>a readme.txt is not generated any longer, no?
17:03<andythenorth>not right now
17:03<andythenorth>it will be restored
17:03<andythenorth>in a bit of a 'work-in-progress' moment right now
17:03<@planetmaker>+GENERATE_DOC := fish-nml
17:03<@planetmaker> in might be needed to make it build then
17:04<@planetmaker>I wonder what DevZone waits for :D
17:04<@planetmaker>without the + of course
17:05<@planetmaker>hm, might even not be needed. But would be proper
17:08<@planetmaker>with respect where you place the build scripts: doesn't really matter where they reside
17:08<@planetmaker></monologue> :-P
17:11<andythenorth>I know :)
17:11<andythenorth>advice from others was that it is cleaner to put build script with rest of makefile stuff
17:11<andythenorth>works either way though
17:11<@planetmaker>it has some point, yes
17:12<@planetmaker>basically, the scripts dir is empty of makefile stuff meanwhile ;-)
17:12<@planetmaker>it's the place for the translation status, though
17:12<@planetmaker>and possibly the gimp script. which you don't need, though
17:12<@planetmaker>Makefile is now easy: Makefile, Makefile.config and ;-)
17:13<@planetmaker>was alberth's complaint that it's too scattered. And he was right
17:13<@planetmaker>I didn't find stuff myself :D
17:14-!-Pereba [~UserNick@] has joined #openttd
17:18<andythenorth>any legacy stuff to delete? :)
17:18<andythenorth>FIRS might have some ;P
17:18<@planetmaker>dunno FIRS now. But... FISH does
17:19<NGC3982>I was about to throw the keyboard on the couch. Instead i tossed it right on the floor.
17:19<@planetmaker>andythenorth, scripts/Makefile*
17:19-!-DarkAce-Z is now known as DarkAceZ
17:19<@planetmaker>hm. no :-)
17:20<andythenorth>thought so
17:20<andythenorth>do you want to do the honours?
17:23<@planetmaker>there, pushed
17:24<andythenorth>great :)
17:25<@planetmaker>firs still uses an older Makefile... maybe I can update it
17:32-!-kais58__6 [] has quit [Read error: Operation timed out]
17:33-!-kais58__6 [] has joined #openttd
17:34<andythenorth>that's better
17:35<andythenorth>easier to read that than check properties in game
17:36<andythenorth>helps debug
17:40-!-amiller [] has quit [Ping timeout: 480 seconds]
17:44-!-kais58__6 is now known as kais58|AFK
17:51-!-frosch123 [] has quit [Quit: be yourself -- except: if you have the opportunity to be a unicorn, then be a unicorn]
17:59-!-y2000rtc [] has quit []
18:01-!-oskari89 [] has quit []
18:03-!-beomouth [~james@] has quit [Remote host closed the connection]
18:07-!-tokai|noir [] has joined #openttd
18:07-!-mode/#openttd [+v tokai|noir] by ChanServ
18:07-!-LordAro [] has quit [Ping timeout: 480 seconds]
18:12-!-tokai|mdlx [] has quit [Ping timeout: 480 seconds]
18:15-!-sla_ro|master [slamaster@] has quit []
18:24-!-andythenorth [] has quit [Quit: andythenorth]
18:29-!-cyph3r [] has joined #openttd
18:48-!-Supercheese [~Password4@] has joined #openttd
19:06-!-Ristovski [] has quit [Quit: Leaving]
19:43-!-Alice3 [] has quit []
19:50-!-EyeMWing [] has quit [Ping timeout: 480 seconds]
20:11-!-pugi [] has quit []
20:17-!-Elukka [] has quit [Ping timeout: 480 seconds]
20:17-!-KritiK [] has quit [Quit: Leaving]
20:22-!-Zuu [] has quit [Ping timeout: 480 seconds]
20:22-!-Pereba [~UserNick@] has quit [Quit: AdiIRC is updating to v1.9.1 Beta Build (200713) 64 Bit]
20:25-!-SamanthaD [] has quit [Read error: Operation timed out]
20:33-!-MatrixCL [] has joined #openttd
20:33-!-MatrixCL [] has quit []
20:35-!-HerzogDeXtEr1 [~Flex@] has joined #openttd
20:41-!-HerzogDeXtEr [] has quit [Ping timeout: 480 seconds]
20:42-!-SamanthaD [] has joined #openttd
20:47-!-gelignite [] has quit [Quit:]
21:10-!-wakou2 [] has joined #openttd
21:10-!-cyph3r [] has quit [Quit: Miranda IM! Smaller, Faster, Easier.]
21:38-!-tswett [] has joined #openttd
21:39<tswett>Ahoy. Is it still possible to reach the difficulty screen in 1.3.1?
21:49-!-roboboy [] has joined #openttd
21:50<Supercheese>I think that was integrated into the advanced options panel
21:51<Supercheese>Things like interest rate, max loan, subsidy multiplier are now under Economy
21:53-!-glx [] has quit [Quit: Bye]
22:16-!-wakou2 [] has quit [Quit: Konversation terminated!]
22:17-!-roboboy [] has quit [Ping timeout: 480 seconds]
22:30-!-HerzogDeXtEr1 [~Flex@] has quit [Read error: Connection reset by peer]
22:31<tswett>*nod* Thanks.
22:31<tswett>All right, gotta go. See you guys.
22:31-!-tswett [] has quit [Quit: Page closed]
23:15-!-amiller [] has joined #openttd
---Logclosed Sat Jul 20 00:00:17 2013