IRC logs for #openttd on OFTC at 2012-07-01
⏴ go to previous day
00:18:05 *** Stimrol has joined #openttd
00:27:22 *** kkimlabs has joined #openttd
01:30:10 *** kkimlabs has joined #openttd
02:52:19 *** kkimlabs has joined #openttd
03:16:49 *** kkimlabs has joined #openttd
03:57:24 *** plastics has joined #openttd
03:58:52 *** NGC3982 has joined #openttd
04:27:48 *** John_Snow has joined #openttd
04:37:27 *** tokai|mdlx has joined #openttd
04:56:16 *** Eddi|zuHause has joined #openttd
05:44:24 *** sla_ro|master has joined #openttd
06:11:48 *** sla_ro|master has joined #openttd
06:20:21 *** cmircea has joined #openttd
06:33:46 *** telanus has joined #openttd
06:48:24 *** kkimlabs_ has joined #openttd
07:02:39 *** Devroush has joined #openttd
07:22:20 *** andythenorth has joined #openttd
07:22:38 * andythenorth ponders doing trucks as a railtype
07:30:08 * andythenorth is now considering it seriously
07:30:23 <andythenorth> can I haz new train pathfinder please?
07:30:53 <andythenorth> it will require a 'proceed on sight of clear path ahead'
07:30:55 <Supercheese> 45 degree turns with trucks would indeed be verrrry cool
07:31:23 <andythenorth> i.e. reserving a path for the next n clear tiles, n to be modulated by stopping distance
07:31:33 <andythenorth> I am thinking mining trucks
07:31:48 <andythenorth> using rail would allow them to be properly large
07:32:34 <andythenorth> it would be annoying however to have to build signals on every tile, hence new pathfinder
07:46:09 *** TomyLobo has joined #openttd
08:03:04 *** Alberth has joined #openttd
08:03:04 *** ChanServ sets mode: +o Alberth
08:10:35 <TomyLobo> how do i remove a newgrf from a save?
08:16:46 <TomyLobo> [23:58:49] <Beul> andythenorth, removing a newgrf that ads a new rail type causes some interesting behavour but not even that managed to crash the game:p
08:16:49 <planetmaker> removing NewGRFs usually leaves the savegame in a broken state
08:17:12 <TomyLobo> well i dont want to play it with that newgrf anyway
08:17:26 <TomyLobo> (that newgrf killed passenger cars)
08:18:02 <planetmaker> When you enable the newgrf_developer setting you can do that indeed. But ... don't, if you are not developing newgrfs
08:18:08 <andythenorth> TomyLobo it's just the way it works
08:18:25 <andythenorth> nobody likes it this way, but we are where we are ;)
08:18:38 <TomyLobo> planetmaker i am prepared to shoot myself in the foot, dont worry :)
08:19:14 <planetmaker> TomyLobo: as soon as you enable developer tools your support is void. Don't come asking and wondering about crashes, oddities or strange behaviour
08:20:16 <planetmaker> and, btw, don't expect oddities to show immediately in your game then. It might even only show after several game years or so
08:20:31 <planetmaker> as such: one tempered savegame has lost warranty forever
08:21:08 <TomyLobo> hrm, it didnt bring back the passenger cars
08:21:36 <TomyLobo> it was worth the try :)
08:21:44 <planetmaker> I could have told you ;-)
08:21:57 <planetmaker> you cannot bring back vehicles by removing newgrfs ;-)
08:22:58 <planetmaker> start a new game?
08:23:03 <TomyLobo> they expired thanks to that newgrf
08:23:11 <planetmaker> choose your newgrfs carefully before starting a game
08:23:35 <TomyLobo> the newgrf doesnt list an introduction date, or i would just set it to that and see what happens
08:23:44 <andythenorth> try reset_engines in console
08:23:57 <planetmaker> check briefly ingame whether you got everything. Quit, modify newgrf list and test again. Until you find a combination which works for you
08:24:21 *** eQualizer|dada has quit IRC
08:24:54 <andythenorth> TomyLobo: this issue is known to suck
08:25:17 <andythenorth> issues come up like 'play for 70 years, then discover that there's no vehicles to transport cargo x'
08:25:53 * andythenorth favours recommended combinations of known-good grfs
08:25:59 <andythenorth> which might come with new banaaananaas
08:26:48 <TomyLobo> 8/32bpp Trains 2CC is that newgrf
08:26:58 * andythenorth goes back to converting FISH to nml
08:27:46 <TomyLobo> reset_engines worked, thanks :)
08:28:29 <andythenorth> converting to nml is....interesting
08:28:37 * andythenorth ponders scraping the nfo for current values
08:29:07 <andythenorth> what would be nice is....
08:29:24 <andythenorth> ...a tool that could take a grf, and output a csv of all the vehicle action 0 properties
08:29:43 <andythenorth> this would, incidentally be useful for other things
08:30:03 <andythenorth> for example, bundles server could run it to provide a manifest of vehicles in a grf automatically
08:30:07 <andythenorth> as could bananas
08:30:08 <Alberth> there is an nfo parser in the nml issue tracker :p
08:30:35 <andythenorth> JSON might be better than csv
08:30:49 <Alberth> a wiki with useful combinations perhaps?
08:31:34 <andythenorth> xml would fit with my xml-driven grf framework
08:31:40 * Alberth has been doing xml processing for 2 days :p
08:32:27 *** Biolunar has joined #openttd
08:33:37 <Rubidium> oh, come on... why XML?
08:33:52 <Rubidium> use SGML. Much more versatile
08:35:15 <Alberth> I know, I have the annotated standard here, and have even read most of it
08:36:09 <Alberth> but standard Python does not come with sgml
08:36:13 * andythenorth uses xml because that's what the templater uses :P
08:36:42 <andythenorth> it may sound like insanity, but the data is not xml
08:36:49 <andythenorth> but the template code is
08:39:16 <Alberth> php conquerored the world with it, so it gets copied everywhere, no matter how conceptually broken it is :)
08:41:04 <andythenorth> using xml-valid templating languages means that html dev tools don't explode
08:42:36 *** chlorine has joined #openttd
08:42:45 <planetmaker> andythenorth: you know grf2html? "a tool that could take a grf, and output a csv of all the vehicle action 0 properties"
08:43:01 <planetmaker> and all others :-)
08:47:13 <andythenorth> planetmaker: good point. maybe it could output json
08:51:39 <andythenorth> I suppose you people like your buy menu sprites in the right place
08:52:27 * andythenorth grumbles and invents new properties :P
08:54:53 <CIA-4> OpenTTD: rubidium * r24364 /trunk/src/lang/slovak.txt: -Fix-ish: Slovak compilation failure
08:56:39 *** TGYoshi has joined #openttd
08:56:59 <CIA-4> OpenTTD: rubidium * r24365 /trunk/ (Makefile.lang.in Makefile.setting.in config.lib): -Fix-ish: add C(XX)FLAGS_BUILD to all the executables that are compiled for the build enviroment
09:06:56 *** LordAro has joined #openttd
09:13:13 <andythenorth> why do offsets differ between nml and nfo? :P
09:14:43 <andythenorth> the x offs seems to be different by about 30pz
09:16:49 <andythenorth> it's the buy menu string leading spacing missing; nothing to see here :P
09:19:42 <LordAro> trouble is, it doesn't work :)
09:23:18 <Alberth> that is the usual trouble with code :)
09:25:10 <LordAro> i know where it is failing (this->filename seems to be valid _only_ when the file download is in progress) but i can't see anything else to use...
09:25:47 <andythenorth> offsetting text by n pixels in newgrf <- deprecated?
09:26:00 <andythenorth> use NBSP instead, and expect font size issues?
09:31:32 <andythenorth> Ammler / planetmaker I can't push as I have .pyc files in my repo, any suggestions?
09:32:05 <planetmaker> andythenorth: intention or error?
09:32:19 <andythenorth> I don't know how it happened, they are in my ignore file
09:32:39 <andythenorth> I've used hg forget
09:32:39 <planetmaker> I'll temporarily disable the commit hook
09:34:37 <planetmaker> try again, andythenorth
09:35:53 <planetmaker> I'll re-enable it. Or is it not useful (due to needing manual intervention)?
09:36:11 <andythenorth> I think it's useful
09:36:31 <andythenorth> I'm not sure, but I think putting compiled python bytecode on the server might be a possible attack vector
09:38:05 <andythenorth> FISH auto-refit creeps a bit closer
09:38:21 <Ammler> hmm, why didn't you remove the pyc files instead force-push it?
09:38:57 <Ammler> also, you can have a custom ini file in your repo to allow it
09:42:52 <andythenorth> how do I remove the .pyc from an older rev? Except by doing dangerous things with strip?
09:43:10 <Ammler> why should strip be dangerous?
09:44:02 <Ammler> it can't be that old revs as it is in the newer revs you push
09:44:10 <andythenorth> removing revs further down the tree seems dangerous to me
09:44:34 <andythenorth> I'd sooner diff all changes, delete my local repo and clone new
09:45:03 <Ammler> now, it is too late, you have already a "broken" repo on the server :-)
09:47:01 <TomyLobo> Ammler i have no idea what andythenorth's situation is, but you do know "git log -g", right?
09:47:22 <andythenorth> I'd have to install git for that :P
09:48:03 <TomyLobo> if you're using mercurial, then you're probably fucked, yeah
09:48:33 <andythenorth> I like mercurial, but I should probably learn git :P
09:48:45 <andythenorth> I don't like svn much, but I've yet to break a repo with it
09:48:51 <andythenorth> whereas I regularly break hg repos
09:50:00 <TomyLobo> yeah breaking svn *repos*, no
09:50:07 <TomyLobo> working copies? uh.... once a week?
10:19:55 *** eQualizer has joined #openttd
10:28:12 <Alberth> you should explain your tools not to touch stuff in .svn directories :p
10:30:11 <Alberth> wouldn't it be an idea if nml projects used a unique prefix for their language files? Having several "english.lng" files in one download directory causes chaos :(
10:40:20 *** andythenorth has joined #openttd
10:54:45 *** andythenorth is now known as Guest1825
10:54:45 *** andythenorth has joined #openttd
10:55:16 <Alberth> two andys, now you can work twice as fast!
10:56:07 <TomyLobo> Alberth you mean tortoisesvn? :P
10:58:09 <Alberth> TomyLobo: you use a graphical frontend for svn, yet know about log -g which is not even in the short help of git log???
10:58:29 <TomyLobo> the svn command line client sucks
10:58:43 <Alberth> try the git command-line :p
10:58:45 <TomyLobo> all graphical frontends for git suck
10:59:03 <Alberth> they do for every VCS
10:59:05 <TomyLobo> that should explain my preferences :)
10:59:30 <TomyLobo> makes SVN's problems somewhat bearable
11:01:36 <Alberth> I have no problems with svn, just need to be a bit careful with directory shuffling, and branch-updating
11:03:37 <TomyLobo> SVN is slow, it has a confusing way to make branches and tags, you cannot commit locally with most clients
11:03:57 <TomyLobo> and many times, the SVN working base's size even exceeds a git repo's size
11:09:20 <TrueBrain> did he just say SVN is slower than git? Lolz.. that is just a lie :D (sorry, didn't read anything else, that just made me giggle)
11:09:29 <Alberth> how is cp semantics confusing?
11:09:44 <Alberth> and w.r.t. clients, blame the clients, not svn
11:09:48 <TomyLobo> TrueBrain some guy made benchmarks, lemme find the link
11:10:03 <TrueBrain> do a random git clone, and do a random svn checkout ... you will know which one is the slow one :P
11:10:06 <TrueBrain> hint: it is not subversion :D
11:10:42 <TrueBrain> "some guy", now the trustworthy links come :D
11:10:42 <TomyLobo> TrueBrain you do realize that git clone fetches the entire history (by default)?
11:10:55 <TrueBrain> so ....... how does that matter from a user perspective?
11:11:17 <TomyLobo> you can diff and blame locally
11:11:32 <TrueBrain> that indeed is a good reason why svn is slower
11:11:54 <TomyLobo> switch branches locally, retrieve logs locally
11:12:15 <Alberth> make two svn working copies
11:12:15 <TrueBrain> if that is your measurement of 'slow', than you are very good in comparing apples with oranges
11:12:18 <TrueBrain> concratz to that :)
11:12:36 <TrueBrain> like saying Wordpad is slower than Netbeans :)
11:12:48 <TomyLobo> TrueBrain how about oranges and oranges then... git is also faster than hg on most operations
11:13:00 <TrueBrain> "on most", except the checkout and commit operations
11:13:06 <TrueBrain> which are, in my opinion, the most important onces
11:13:21 <TrueBrain> but that is acomplete different story .. SVn was the topic
11:13:24 <TomyLobo> no, it's really only slower on a huge commit
11:13:41 <TomyLobo> think hundreds of files
11:13:50 *** Zeknurn has joined #openttd
11:14:03 <TomyLobo> meh the page with git vs hg comparison vanished
11:14:10 <TomyLobo> now it's only comparison with svn :/
11:14:14 <TrueBrain> its easy to now start comparing Netbeans to Eclipse, but you were comparing it with Wordpad, so ..
11:14:58 <Alberth> haha, yeah, that looks like an objective document :D
11:15:00 <TrueBrain> so your source that SVN is slow, is on a git-specific website?
11:15:04 <TrueBrain> impressive choice ...
11:15:35 <TomyLobo> well, you pick one that proves your opinion
11:15:55 <TrueBrain> open any CLI, do a svn checkout of OpenTTD, do a git checkout of OpenTTD
11:15:58 <TomyLobo> the one that proved mine was apparently popular with the git people and is now hosted by them
11:15:59 <TrueBrain> take a stopwatch if you like
11:16:15 <TomyLobo> TrueBrain you mean it has both svn and git?
11:16:23 <TrueBrain> just read what you are saying .. a page saying Git is better is popular with git people ... owh realy? :P
11:16:23 <TomyLobo> or are you talking about git-svn?
11:16:35 <TrueBrain> OpenTTD has SVN, git and hg
11:16:51 <TomyLobo> and not some kind of gateway?
11:17:19 <TrueBrain> I doubt gateways as such exists, as that would be highly impossible/unlikely
11:17:41 <TomyLobo> github has an svn gateway
11:17:57 <TrueBrain> at best you can import svn commits into a git repos
11:18:07 <TrueBrain> which is not really a gateway in any sense
11:18:09 <TomyLobo> take a look at git-svn
11:18:18 <TrueBrain> git-svn is an IMPORT script
11:18:21 <TrueBrain> (and export, but meh)
11:18:32 <TomyLobo> no, that'd be git-svnimport :D
11:18:43 <TrueBrain> I suggest you read up how git-svn works
11:18:50 <Beul> * andythenorth ponders doing trucks as a railtype
11:18:50 <Beul> * andythenorth is now considering it seriously
11:18:55 *** oskari89 has joined #openttd
11:19:17 <TomyLobo> TrueBrain i could suggest you the same
11:19:25 <TrueBrain> you could;but I already know
11:19:33 <Beul> as I discovered yesterday, you can make some pretty awsome features with new rail types!
11:20:23 <TomyLobo> git-svn marks the imported commits so it can reassociate them with the svn repo. it can push new commits easily that way
11:20:37 <michi_cc> TrueBrain: git-svn is a perfectly good two-way interface. How do you think my commits arrive at the SVN server?
11:21:38 <TrueBrain> michi_cc: well aware; but a two-way interface !== a gateway :)
11:21:58 <TrueBrain> git-svn is a import/export tool, automated, hooked into git (pretty neat btw)
11:22:26 *** Chris_Booth[ph] has joined #openttd
11:22:36 <TrueBrain> a gateway would suggest I can also do a svn checkout of a git repos
11:23:03 <TomyLobo> you can check out svn repos with git-svn
11:23:33 <TrueBrain> and now read what I said, and you see there is a miss-match
11:23:35 <TomyLobo> and you can check out git repos on github with svn
11:23:40 <TrueBrain> hence, not a gateway :)
11:23:53 <TomyLobo> svn.github.com/user/project i think
11:24:30 <michi_cc> And as for the speed, your mileage will always vary, but checkout/clone is the most unimportant command to measure. I'm not doing a clone every five minutes, but I am using status/log/diff/etc, and for everything that does not operate on the very latest revision, svn is principally slower.
11:24:56 <TrueBrain> michi_cc: heavily depends on usage, and many many many other constraints
11:25:02 <TrueBrain> much more important: it is silly to compare apples with oranges
11:25:23 *** Chris_Booth[ph] has quit IRC
11:25:26 *** Chris_Booth[ph]_ has joined #openttd
11:25:36 <TrueBrain> michi_cc: as it is not SVN that is slow. It is the design that demands other conditions
11:26:01 <TrueBrain> it is like saying Wordpad is slower than Netbeans, because in Netbeans you can open files easier
11:26:04 <TomyLobo> the design is from another time
11:26:04 <TrueBrain> it makes little sense
11:27:08 <michi_cc> Of course, it is mostly the network that is slow, but from the end-user perspective only the result counts.
11:27:41 <TrueBrain> michi_cc: get a 100/100 :P
11:27:42 <Alberth> Beul: amazing cablecars newgrf!
11:28:09 *** plastics has joined #openttd
11:28:20 <TomyLobo> the only 2 conditions i'd see that svn fulfills better than git are 1. faster checkout 2. tight access control to the repo
11:28:37 <TrueBrain> euh ... you forget one very important one:
11:28:45 <Alberth> eh, svn is fatser on checkout now?
11:28:52 <TrueBrain> it is the reason why you compare oranges with apples ..
11:28:58 <TrueBrain> it has nothing to do with access
11:28:59 <TomyLobo> Alberth if the repo is large enough, yes
11:29:09 <TomyLobo> on small repos, git is faster again :P
11:29:10 <TrueBrain> decentralized vs centralized are two very different methods of solving this "VCS" issue
11:29:29 <andythenorth> so how does this FISH thing work? :P
11:29:33 <Alberth> TomyLobo: small repos are non-interesting, as it is fast enough either way
11:29:35 *** FLHerne_ has joined #openttd
11:29:44 <andythenorth> seems that some ships can refit PAX and mail only
11:29:48 <TomyLobo> Alberth yeah, hence i only mentioned the large ones
11:29:58 <andythenorth> others refit all cargos, or freight only
11:30:43 * andythenorth wonders how many CMS properties are needed to handle this
11:31:01 <TomyLobo> ok, tsvn took 21s to check out a shallow copy
11:31:31 * andythenorth wonders what the FISH rewrite should do about smoke
11:31:35 <TomyLobo> fatal: dumb http transport does not support --depth
11:31:44 <TomyLobo> do you have ssh-based git too?
11:32:09 *** Chris_Booth[ph]_ has quit IRC
11:33:16 <planetmaker> you're completely re-writing fish and heqs, andythenorth?
11:33:25 <andythenorth> planetmaker: converting FISH to nml
11:33:33 <andythenorth> ideally with no loss of features :P
11:33:42 <andythenorth> due to mistakes :)
11:33:46 <TomyLobo> currently making a full clone
11:33:53 <TomyLobo> and yeah, that takes some time
11:34:00 <planetmaker> nice, andythenorth :-)
11:34:06 <andythenorth> I'll do HEQS later, but it will be a v2, with new features
11:34:14 <TomyLobo> not surprising with a repo containing tens of thousands of revisions
11:34:25 <TomyLobo> and of course the binary files
11:34:43 <TomyLobo> git doesn't delta-compress binary files (yet)
11:34:53 <Beul> Alberth, actualy, the new tunnel grf that I coded could (unintentionally) also add cablecars:p
11:35:28 <TomyLobo> TrueBrain that took like 5 seconds :D
11:35:30 <TomyLobo> git clone git://git.openttd.org/openttd/trunk.git/ openttd-git-shallow --depth 1
11:35:30 <andythenorth> so I should do mining trucks as trains?
11:35:52 <Alberth> Beul: those unintentional side effects are usually giving the best result :)
11:35:58 <TomyLobo> it's as useless as an SVN checkout though
11:36:39 <Beul> Well, they gave me the idea for that grf indeed, but ideally I would like to be able to disable them in the tunnel grf
11:36:48 <TomyLobo> mind if i push the full repo to github and see if i can clone it from there faster?
11:36:57 <Beul> Would that be posible currently?
11:37:16 <Alberth> TomyLobo: I'd be surprised if it is not already there somewhere
11:37:19 <TomyLobo> pushing? sure... but syncing it probably not
11:37:28 <TomyLobo> Alberth likely an old version though
11:37:51 <TrueBrain> so easy to widely guess stuff .. but opening the page would give you an aswer immediatly
11:37:57 <TrueBrain> and you will know your assumption was wrong :P
11:38:03 <Terkhen> andythenorth: I don't think so, I like to be able to run them on normal roads for short distances
11:38:41 <Alberth> Beul: no idea, I don't speak newgrf at all :)
11:38:45 <andythenorth> I could make them enormous though :)
11:38:45 *** Chris_Booth has joined #openttd
11:39:12 <TrueBrain> s/wrong/wrong or not/, hmm, parts fell off
11:39:26 <TrueBrain> I think faster than I type! Where is the fucking bionic interface :(
11:40:00 <Alberth> we wouldn't be able to process that amount of information ;)
11:40:21 <TrueBrain> don't brains only work at like 100 bits per second or something silly?
11:40:37 <andythenorth> I think they go faster, but consciousness lags
11:40:43 <TomyLobo> but highly parallel :)
11:41:04 <TrueBrain> new excuse: "sorry, but my concious was lagging"
11:41:26 <Beul> TrueBrain, I think faster than I can do anything, I want a bionic interface to build real stuff, do engeneering calculations, type and make real nice gaphics
11:41:51 <TomyLobo> pushing the entire repo to github tooki hear reality has nice graphics
11:42:07 <TomyLobo> i hear reality has nice graphics, but the user interface sucks
11:42:29 <Beul> getting a bit out of hand here
11:43:04 <andythenorth> what happens if I set action 0 capacity to 0? No refits?
11:43:30 <TomyLobo> pushing the entire repo to github took about 3.5 minutes
11:47:16 <TomyLobo> git clone git@github.com:TomyLobo/openttd.git openttd-git-github - 2:24 until i get my prompt back
11:48:50 <Beul> on a more ttD related note: is there anybody that knows if is is posible to disable certain new railtype objects from construction?
11:49:21 <Beul> for example constructing track
11:49:56 * andythenorth has nml feature suggestion
11:49:56 * Beul railises it sounds quite stupid to prevent track construction in a new tail type
11:50:17 <TomyLobo> apparently the git protocol is faster for shallow copies
11:50:21 <andythenorth> 'is_refittable' is a stupid property for ships; nml should set it automatically if refittable classes exist
11:51:12 <Beul> andythenorth, good point imo
11:53:24 <TomyLobo> 1:10 until i get my prompt back after "git clone git://github.com/TomyLobo/openttd.git openttd-git-github-gitproto"
11:54:19 * andythenorth wtfs at setting refitted capacity in nml
11:54:40 <Xaroth> TomyLobo: get a faster pc?
11:54:53 <TomyLobo> Xaroth it's reasonably fast
11:55:10 <Xaroth> define 'reasonably' ...
11:55:27 <Xaroth> my grandmother is reasonably fast
11:55:44 <andythenorth> in cb36, nfo var 47 maps to 'cargo_classes_in_consist'
11:55:47 <andythenorth> for the record ^
11:55:56 <TomyLobo> pheonom ii x4 3.2 ghz
11:56:05 <andythenorth> in case anyone else was playing 'guess the var name' with the nml spec
11:56:08 <TomyLobo> pheonom ii x4 955 3.2 ghz
11:56:22 <Xaroth> you're missing the point
11:56:40 <andythenorth> I'm rarely waiting on my vcs
11:56:46 <andythenorth> and if I do, it's thinking time
11:56:47 <TomyLobo> Xaroth what else do you want to know?
11:56:59 <Xaroth> your definition of reasonably.
11:57:12 <TomyLobo> it plays all the games i want to play
11:57:28 * andythenorth runs into the 'wtf is the correct weight of mail, goods etc'
11:57:31 <TomyLobo> it doesnt usually lag on everyday work
11:57:32 <andythenorth> anybody care to make bids
11:58:15 <Alberth> andythenorth: mail is about the same as paper
11:58:33 <TomyLobo> except if someone's mailing rocks
11:59:02 <Alberth> goods are lighter (mostly just air)
11:59:34 <andythenorth> so which of the game's inconsistent uses of mail / goods weight should I be consistent with?
11:59:37 <TomyLobo> dear mother, here is the uranium block i promised to mail you. love, your son
11:59:54 <andythenorth> currently this ship carries 52t freight, or 52 bags mail
11:59:54 <Alberth> andythenorth: the first one that comes to mind
12:00:00 <Beul> Alberth, i beg to differ, as paper wil usualy be stacked much more compact, thus making it heavier
12:00:07 <andythenorth> 1 mailbag = 1t in that case
12:00:33 <Alberth> Beul: some people are sending uranium over the mail :p
12:00:52 <Alberth> but in general, I agree :)
12:02:41 <TomyLobo> 1:22 to make a full clone through git://
12:03:00 <TomyLobo> from the openttd.org repo
12:03:29 <TomyLobo> so that's 4 times the time of an SVN checkout or 1 minute and 1 second longer
12:03:48 <TomyLobo> i'd say that's not too long a wait :)
12:05:11 <Terkhen> with mercurial I keep a clean clone of the remote repo, and then clone all of my repositories from the local "central" repository
12:05:16 <Terkhen> I suppose that you can do the same in git
12:08:12 <TomyLobo> Terkhen yep, shallow copy takes 1/4 the time of an svn checkout :D
12:08:47 <TomyLobo> well i didnt measure it, hold on
12:09:02 <Terkhen> that's on windows or linux?
12:09:15 <TomyLobo> on linux it'd be even faster
12:11:10 <TomyLobo> svn checkout and git working copy+repo are about the same size :D
12:11:58 <TomyLobo> svn: 74 (86 on disk), git: 85 (88 on disk)
12:13:55 <Alberth> Beul: there is a much easier way, just make bridges invisible :p
12:13:56 *** telanus1 has joined #openttd
12:15:39 <Beul> Alberth: that would not be unintended
12:16:34 <Alberth> TomyLobo: that's great and all, but I don't pick a VCS based on how fast or slow it is. "Reasonable" is fast enough in practice.
12:16:48 <TomyLobo> Alberth i don't, either
12:17:13 <TomyLobo> it's just that TrueBrain is saying SVN is faster than git at what it does
12:18:17 <Alberth> TomyLobo: don't you know you can prove anything you like with numbers/statistics?
12:18:19 <TrueBrain> euh .. don't start twisting my words there
12:18:28 <TrueBrain> you said SVN is slow. Which is unfounded, and untrue
12:19:24 <planetmaker> I really love these VCS fundamentalists :-)
12:19:27 <TomyLobo> if i wanted to say something unfounded and (probably) untrue, i'd say you're working for tigris
12:19:45 <andythenorth> how about a patch
12:20:03 <andythenorth> I need to have two grfs with the same ID both active
12:20:04 <Alberth> TomyLobo: subversion has moved
12:20:30 <planetmaker> Alberth: Beul: invisible bridges are... somewhat problematic. You'd remove them for all infra types. And there's only 11 different ones
12:21:11 <Alberth> planetmaker: who need visible bridges, they are just in the way to see what happens under it :)
12:21:14 <planetmaker> Beul: and I'd also like to congratulate you on your nice grf work
12:22:11 <Beul> ty planetmaker, I think Alberth meant makeing bridges transparent not making them invisible like i did with the tracks:p
12:22:35 <planetmaker> invisible = ingame function, transparent = sprite level?
12:22:52 <planetmaker> hm... both is an ingame GUI option... :-P
12:22:58 <Alberth> Beul: they are non-visible either way :)
12:23:03 <planetmaker> but I think you talked about sprite levels, right?
12:23:12 <oskari89> Does NML support waypoints?
12:24:03 <planetmaker> oskari89: no, it doesn't support stations (which waypoints are a special kind of)
12:24:19 <planetmaker> Beul: sorry, nvm me then :-)
12:24:34 <Alberth> nml development is not moving very fast, is it?
12:24:46 <oskari89> May be a big project to include stations to NML?
12:25:02 * Beul also wonders if planetmaker knows a way to disable building tracks for the mew tunnel grf
12:25:06 <oskari89> It would significantly accelerate station projects, though :)
12:25:07 <planetmaker> oskari89: sure. And tbh, anyone would be our guest
12:25:19 <planetmaker> oskari89: it's on my wishlist definitely
12:25:20 <Beul> (don't think it is possible without a patch thoug?
12:25:23 <andythenorth> nfo station spec is a right arse
12:25:35 <planetmaker> Beul: there's no such way
12:25:40 <andythenorth> the station spec lacks sense, it's madness
12:26:40 <oskari89> And there's very little people who have coded stations :P
12:26:47 * Beul braces for bug reports with URTP grf then
12:26:52 <planetmaker> Beul: and tbh, a patch I see for that is rather on the line of a flag for the tunnel portal which says "special entrance" which allows to build new objects (or possibly anything) on top
12:27:00 <Alberth> oskari89: three guesses why :)
12:27:14 <planetmaker> Beul: which would be 50% of the so-called custom bridge and tunnel heads
12:27:33 *** SomeoneMcMe has joined #openttd
12:28:02 <SomeoneMcMe> Anyone what to play on my sever?
12:28:12 <Xaroth> this isn't the place to ask :)
12:28:31 <Alberth> very few players here :)
12:28:31 <Beul> planetmaker, I would shurely agree, but writing patches is beyond my capabilities atm
12:28:59 * andythenorth seeks help with converting FISH to nml
12:29:13 <andythenorth> I have a buy menu string that describes refitted capacities
12:29:24 <andythenorth> can't make my brain figure out how to put the numbers in
12:29:34 <Alberth> Xaroth: do you know a better channel?
12:29:54 <planetmaker> andythenorth: similar to openttd strings:
12:30:08 <planetmaker> STR_BLAH: This displays {NUM} numbers in this string
12:30:24 <andythenorth> thanks. it's wrapped in some if / else condition stuff I need to figure out too
12:30:30 <Alberth> number{P "" "s"} actually :)
12:30:31 <andythenorth> I'm probably nesting a stack of substrings I think
12:30:42 <planetmaker> Right Alberth :-)
12:30:47 <andythenorth> BANDIT does this , but Eddi|zuHause solved it for me :P
12:31:14 <Beul> does using the maintenance_cost callback make a grf incompatible with versions lower than 1.2.0?
12:31:37 <Beul> or will it simply ignore the callback?
12:32:59 <planetmaker> unless you include the maintence cost callback conditionally, depending on the openttd version
12:33:10 <planetmaker> but tbh, Beul: stop caring about pre-1.2.0 OpenTTD ;-)
12:33:29 <TrueBrain> there is no pre 1.2.0 :P
12:33:32 * TrueBrain removes some binaries
12:33:35 <planetmaker> a huge load of newgrf things were added in that version which make newgrf developer life so much easier
12:33:50 <andythenorth> eurasia was always at war with eastasia
12:34:25 <planetmaker> Beul: and the audience plays 1.2.0 or later anyway
12:35:45 <Beul> the reason I would like to be able to use it in earlier versions than 1.2.0 is to use it with CillCore's patchpack
12:36:04 <planetmaker> Beul: for what it's worth: all NewGRFs I changed since December won't work on OpenTTD prior to 1.2.0
12:36:20 <planetmaker> Beul: also ChillCore will create a new version of his PP. Or so I read
12:36:42 <planetmaker> And the next major OpenGFX won't work on pre-1.2.0 OpenTTD anymore either
12:37:06 <Alberth> Beul: in open source, anything older than a year or so is dead :)
12:39:37 <planetmaker> Beul: you mean the maintenance cost property, not callback, right? Or of what thing do you speak of?
12:39:56 * TomyLobo just built 2 airports at the opposing ends of his city
12:40:06 <Beul> yeah, used the worng term once again
12:40:43 <Beul> property is what I meant indeed:p
12:41:28 <Beul> @kamnet - surely it won't be too long (tm) that the patchpack is updated to 1.2.1 compatibility?
12:43:54 <andythenorth> converting FISH to nml is quite quick to do
12:44:44 <planetmaker> it's a relatively simple NewGRF, is it andythenorth?
12:44:46 *** cmircea has joined #openttd
12:45:06 <planetmaker> technically speaking. Except maybe the wood rafts
12:45:17 <planetmaker> and the many wakes :-)
12:45:50 <andythenorth> the wood rafts will need special case
12:46:09 <Beul> planetmaker: saw that thread as well would be very nice to have an new clean patchpack from him
12:46:15 <andythenorth> planetmaker: I need to try converting to the new makefile
12:48:25 <planetmaker> hm... I just wonder whether it's missing the essential parts :-O
12:49:18 <andythenorth> today would be a good day for me to test it
12:49:27 <planetmaker> lol. It is missing the dir with the makefile parts :-P
12:49:56 <TomyLobo> why could a Dinger 200 explode at the end of the landing strip?
12:50:07 <TomyLobo> of an intercontinental airport
12:50:35 <TomyLobo> because planes don't normally go up in flames
12:50:35 <planetmaker> bad weather, bad tires, malfunctioning thrust reverser
12:50:47 <planetmaker> they usually do when they hit an obstacle
12:52:54 <planetmaker> the first linked one is just an example newgrf
12:53:03 <planetmaker> which uses the makefile repo as a sub repo
12:55:23 <andythenorth> so where do I start? :)
12:55:56 <andythenorth> I should rm current makefile stuff?
12:56:17 *** TrueBrain has joined #openttd
12:56:21 <planetmaker> you could. But keep your Makefile.config
12:56:48 <planetmaker> Though personally I'd just overwrite and delete unused files
12:56:57 <andythenorth> what's /scripts do?
12:57:02 <planetmaker> actually, yes, keep the files. and just overwrite
12:57:18 <planetmaker> the example repo I linked uses build-common
12:57:22 <planetmaker> that is the scripts dir
12:57:26 <planetmaker> no need to rename it
12:57:53 <planetmaker> it's amatter of telling it in the Makefile where to find its parts :-)
12:58:24 <planetmaker> just replace build-common by scripts
12:58:45 <planetmaker> or let's start from the beginning:
12:59:07 <planetmaker> change line 16 to point to the scripts dir
12:59:37 <planetmaker> and also mind lines 9-17
13:00:54 <planetmaker> and try whether it builds ;-)
13:01:01 <andythenorth> I should set values for MAIN_SRC_FILE and GFX_LIST_FILES ?
13:01:30 <planetmaker> unless you don't generate sprites. Then uncomment GFX_LIST_FILES
13:02:21 <andythenorth> MAIN_SRC_FILE - what's that expecting?
13:04:01 <planetmaker> the main pnml file which includes all others
13:04:12 <andythenorth> k I don't have one of those :)
13:04:20 <planetmaker> probably src/fish.pnml
13:04:55 <planetmaker> well... that's tricky then
13:05:20 <andythenorth> managed to hack it onto the old makefile for BANDIT ;)
13:05:29 <planetmaker> especially as I don't know what you want to do and plan to do
13:06:48 <andythenorth> in a nutshell: call the python build script if any deps change
13:07:22 <andythenorth> that will generate a complete nml file
13:09:41 <andythenorth> and it works afaict, including for the graphics generation
13:11:01 <FLHerne_> andythenorth: While you're redoing FISH, would it be possible to make the rafts go round corners neatly?
13:11:13 <andythenorth> I'm open to patches for that
13:11:18 <andythenorth> what's the issue?
13:11:21 <planetmaker> ok, andythenorth, keep that Makefile.in in the scripts dir like for bandit
13:11:21 *** FLHerne_ is now known as FLHerne
13:12:12 <FLHerne> Looks strange in rivers and canals when the rafts are sliding across the bank :P
13:12:43 <andythenorth> FLHerne: how would you solve it?
13:13:17 <FLHerne> I don't know :P . Is there any way for a ship to know it just went round a corner?
13:13:39 <michi_cc> Articulated ships! :)
13:14:04 <andythenorth> ^ claims to apply to ships
13:14:32 <planetmaker> they can thus be commented out in the main Makefile. But the targets thereing (nml and lng) must thus be re-defined in the Makefile.in of yours
13:14:33 <michi_cc> So were do you find "adjacent wagon pairs" for ships?
13:14:45 <andythenorth> I wtf-ed a bit at that
13:14:57 <planetmaker> and I'm coming late for BBQ.... sorry, see you later :-)
13:15:02 <andythenorth> planetmaker: np thanks
13:15:17 <andythenorth> michi_cc: I forgot, it's a train game o/c :P
13:15:32 <planetmaker> I shall have a look maybe tomorrow morning in the train, andythenorth
13:15:45 <andythenorth> the conversion is not ready yet
13:24:37 * andythenorth wonders what default model life is for ships
13:25:12 <andythenorth> the code probably knows
13:43:33 <andythenorth> Alberth: variable ship speed? Faster only when 100% unloaded? Or proportional to current load amount?
13:45:15 <Alberth> propoertional for a few km/h ?
13:46:09 <Alberth> you could do +1 for less than 66% and +2 for less than 33% if you want to do proportional, but 0 if >=50% and +x if < 50% is fine too
13:46:43 <TomyLobo> you could make a physical model
13:47:05 <Alberth> or if you want 3 km/h faster, one step every 25 % which is probably easier
13:47:23 <andythenorth> I have python to hand, so I can just do straight maths on it ;)
13:47:36 <Alberth> TomyLobo: yep, right after you provide a patch for non-zero sized ships
13:47:54 <TomyLobo> taking into account the part of the ship that's in the water (and thus friction)
13:48:03 <andythenorth> sounds like an ottd patch
13:48:09 <Alberth> oh, nml implies no computing by hand any more :)
13:48:34 <TomyLobo> as well as the inertia of the cargo and ship
13:49:09 <TomyLobo> and of course the power of the engine
13:49:44 <Alberth> don't for get people rowing to make it go faster :p
13:50:34 <Alberth> oh, and the amount of fuel on board is also important
13:50:45 <TomyLobo> not to forget bribing the captain to look the other way on some state regulations
13:50:57 * Alberth ponders limited range of ships
13:51:10 <andythenorth> similar to planes/
13:51:21 <andythenorth> ships can bunker at sea
13:51:37 <andythenorth> civilian planes don't do that :P
13:51:56 <Alberth> I don't know how often a ship needs to take in fuel, I guess not that often to make it worthwhile
13:52:29 *** HerzogDeXtEr has joined #openttd
13:55:32 <andythenorth> I have to figure out how this interacts with river / sea speeds :)
14:03:15 * andythenorth wonders if that will happen via automagic
14:06:34 * andythenorth needs a volunteer to add ships stats to the CMS :P
14:16:30 <andythenorth> what causes the ship depot window to change size when reloading grfs?
14:19:26 <andythenorth> Alberth: difference between speed laden / unladen? 3mph?
14:20:29 <TomyLobo> is there some max radius for a city?
14:20:40 <Alberth> I suggested 10%, which gets somewhere around that number
14:20:57 <Alberth> but it's a random number, so feel free to adjust
14:25:16 <TomyLobo> my city is kinda shrinking right now
14:25:42 <Alberth> it's collapsing under its own gravity? :p
14:26:10 <TomyLobo> has been almost 300k earlier
14:26:25 <Alberth> I don't know about cities, but people seem to be obsessed about them
14:26:49 <TomyLobo> well i restricted myself to one city
14:27:08 <Alberth> so I am quite sure someone can answer all details about cities that you ever want to know
14:27:22 <Alberth> quite likely it is already at the forum or the wiki
14:28:02 <andythenorth> Alberth: I'll make it a fraction set per ship, that will do nicely
14:28:08 <andythenorth> 1.1 can be typical
14:29:30 <Alberth> hopefully that did not come as a surprise :)
14:31:15 * andythenorth ponders setting 100 discrete values representing 0-100% loaded
14:31:26 <andythenorth> with pynml, such things are entirely plausible :P
14:32:13 <Beul> seems to me like andythenorth is qiute the ambitious type :)
14:32:24 <Alberth> 11 are many more than enough :)
14:32:57 <Alberth> Beul: ambitious in ideas, but also practical to scale down when needed :)
14:32:59 <TomyLobo> yay, it's stable above 300k... and down to 299xxx again
14:33:15 <TomyLobo> ffs stop shrinking you silly city
14:34:04 <Alberth> TomyLobo: what's so special about x493e0 ?
14:34:27 <Alberth> ie it seems stable at 299k :)
14:34:27 <TomyLobo> that the game is teasing me with it
14:34:59 <TomyLobo> now it went up to 205k
14:35:44 <Alberth> new building temporarily do not have inhabitants, that might explain your fluctuations
14:36:09 <Beul> I also do not understand the obsession with cities, they seem to grow easy enoug on normal settings. I actualy find it quite annoying to set up a transport network in a city, expand the rest of my network, only to come back and realise all stations are overloaded with waiting passengers and the whole area is overgrown :p
14:36:15 <TomyLobo> while being construction sites, right?
14:37:16 <Alberth> Beul: there is a setting to forbid towns from making new roads, that helps in not overgrowing :)
14:37:33 <andythenorth> I need to know % loaded, so cargo_count / cargo_capacity
14:37:42 <Alberth> TomyLobo: I think so, but I don't have the details
14:37:46 <TomyLobo> if they're on a 3x3 grid, you can just buy the places they can grow to :)
14:38:17 <TomyLobo> to prevent the city from growing into the mountains
14:38:22 <TomyLobo> before i flatten those *g*
14:38:31 * andythenorth wonders how nml handles range checks
14:38:35 <Alberth> Beul: ie disallow it, and then build roads yourself where you want the city to grow
14:39:02 <Alberth> TomyLobo: you know you can load a (flat) heightmap instead, right?
14:39:19 <andythenorth> x...y: return stuff
14:39:26 <TomyLobo> Alberth where's the challenge?
14:39:28 * Alberth prefers random roads and growing
14:40:16 <Alberth> TomyLobo: flattening a mountain is a challenge?
14:40:23 <andythenorth> hmm, I can just return values of vars in nml it seems
14:40:27 <TomyLobo> before i had those planes, flattening the terrain actually took almost all my income
14:40:27 <Alberth> it just takes a lot of money
14:41:10 <Beul> Alberth, also prefer the random roads and growing, it's just the speed at whitch it happens that I can't handle
14:41:12 <Alberth> so you bought planes for the challenge?
14:41:29 <Beul> and yes, I do know that there is a patch for that
14:41:40 <TomyLobo> no, i bought them to make more money
14:42:08 <TomyLobo> once my city was of sufficient size to support a plane route inside the city
14:42:11 <Alberth> Beul: I don't service towns except where I want to deliver goods and/or food
14:42:43 * Beul loves playiog with cargodist for passengers
14:43:10 <Alberth> I haven't done that enough yet :)
14:45:37 <Alberth> Beul: there are also grfs that change the 'I bought this tile' sign to a grass-like tile, which you can put around your city
14:48:09 <andythenorth> return 16 to nml speed cb results in speed 5mph
14:48:20 <andythenorth> which makes no sense :P
14:50:37 <andythenorth> perhaps it's nfo style
14:50:57 <Alberth> can't you add a unit or so?
14:51:29 <andythenorth> I have to do int(3.2*16)
14:51:53 <andythenorth> I love it when values are not the values you expect :)
14:52:07 <andythenorth> otherwise the grf would get done really quickly, and where's the fun in that ?
14:52:44 <Alberth> perhaps it should depend on its revision number :p
14:53:31 <Alberth> preferably very slowly changing, so authors don't notice it very quickly :P
14:55:07 <Alberth> hmm, we sohould add realistic acceleration first then, so we can blame it :)
14:56:16 <andythenorth> using int(3.2*speed) gives me values that are too low compared to the nfo version of FISH :P
14:56:46 * andythenorth wonders how python rounds
14:57:12 *** TWerkhoven[l] has joined #openttd
14:57:28 <Beul> haha Alberth, good idea: blame it on realistic acceleration :p
14:57:53 <andythenorth> python truncates towards zero
14:57:57 <andythenorth> so I need to call floor / ceil?
14:59:30 <andythenorth> but round seems to round down also
14:59:35 <DorpsGek> andythenorth: 15.9375
14:59:52 <andythenorth> ah, nml might be doing the rounding
15:00:03 <andythenorth> that should come out as 16, but doesn't :P
16:05:27 *** Progman has joined #openttd
16:06:50 *** andythenorth has joined #openttd
16:31:42 * MNIM tosses a tribble at __ln__
16:34:29 <andythenorth> Alberth: max speed to report in buy menu etc - laden or unladen?
16:35:00 <andythenorth> might be more annoying :P
16:35:14 <Alberth> as manufacturer you don't want to underestimate the performance of your vehicles :p
16:35:18 <FLHerne> Laden - that's how fast it'll go when actually in use :-)
16:35:34 <FLHerne> Probably less confusing for AIs, too?
16:36:15 <andythenorth> some of the existing ships change speed when laden; they report laden speed in buy menu
16:36:17 <Alberth> could be an interesting experiment, waiting for bug reports "he, the ship goes faster than specs" :)
16:36:18 <andythenorth> otherwise annoying
16:36:24 <andythenorth> log raft does it
16:36:33 <andythenorth> hovercraft does it
16:36:56 * Alberth mostly looks at reliability
16:37:30 <Alberth> but I am one of the idiots that plays with breakdowns enabled :p
16:37:40 <andythenorth> moi aussi, sometimes
16:37:49 <andythenorth> otherwise it's boring :P
16:46:08 <Alberth> consistency is probably more interesting
16:49:08 <Beul> I do not see any problems in defining max speed unladen
16:49:37 <Beul> 2cc in combination with nutracks allready has some trains that allmost never reach full speed
16:50:18 <andythenorth> but ships work differently to trains
16:50:36 <andythenorth> if you buy a log tug that reports 21mph, but only get 7mph laden, you'll be....annoyed
16:53:29 *** kkimlabs has joined #openttd
16:55:09 <andythenorth> I have rounding errors that make this speed varying thing unworkable right now :P
16:56:05 <andythenorth> purchase menu speed is shown differently to vehicle speed when built
16:56:39 <andythenorth> ach may be maths by me
16:58:10 <Alberth> ships don't have physics
17:05:48 <Eddi|zuHause> andythenorth: report the (un)laden speed in the additional purchase text (via parameter, so mph-km/h switching works)
17:06:12 <andythenorth> I've already got canal and ocean speed in there too
17:06:17 <andythenorth> quite a lot of speeds :)
17:07:18 * andythenorth is now bored of playing 'guess' with floating points and ints
17:07:25 <andythenorth> this is one case where nfo was just easier
17:08:11 <Alberth> I am surprised it is that difficult with nml
17:08:12 <andythenorth> the ability to do stuff like 'return: 1 + (foo / bar)' in nml is over-rated and flakey :P
17:08:33 <andythenorth> "Floating-point operations are only possible when both operands are compile-time constants"
17:08:56 <Alberth> you do give it floating point values, right?
17:09:15 <andythenorth> no, I'm trying to turn them all into ints
17:09:31 <andythenorth> but fractions as ints is...problematic :P
17:09:43 * andythenorth is tempted to do the 'multiply by 8, divide by x' solution
17:09:46 <Alberth> I could have a looksie whether it is easily fixable, but not now
17:10:02 <andythenorth> I can solve it brutally with some range checks
17:10:12 <andythenorth> it fails the one-line code game
17:10:29 * andythenorth has babies to bath anyway
17:11:58 <TomyLobo> int fac(n) { return n==0 ? 1 : n*fac(n-1); }
17:16:41 *** Biolunar_ has joined #openttd
17:18:50 <Alberth> TomyLobo: and now with a lambda function :)
17:19:16 <TomyLobo> cba to look up the syntax
17:19:39 <Alberth> lambda <parameters> : <return-expression>
17:20:10 <Alberth> f = lambda x, y: x + 1 f(1,2)
17:22:57 <andythenorth> this rounding is tiresome
17:24:35 *** FLHerne has joined #openttd
17:25:38 <andythenorth> Eddi|zuHause: feel like lending a hand?
17:25:57 <Eddi|zuHause> if you can repeat the question in 3 lines?
17:26:53 <andythenorth> writing a good question takes time :P
17:28:02 <FLHerne> Has anyone here tried Cinnamon much?
17:28:12 * FLHerne found it a bit unpolished
17:32:40 <Eddi|zuHause> only in combination with apple sauce
17:36:20 <andythenorth> ship max speed is determined as follows: base speed + (additional speed inversely proportional to % loaded)
17:36:35 <andythenorth> needs to be returned as a cb result, including dicking around with multiplying by 3.2
17:36:59 <andythenorth> I have 'speed' and 'speed_unladen' available as vars
17:37:21 <andythenorth> (python vars in scope during the build)
17:38:45 <andythenorth> relevant nml vars at runtime seem to be 'cargo_count' and 'cargo_capacity'
17:43:53 <andythenorth> it would be acceptable for the granularity of this to be low, e.g four bounds of 25% each
17:45:40 *** Chris_Booth[ph] has joined #openttd
17:49:14 *** frosch123 has joined #openttd
17:55:15 <frosch123> back in the old world?
18:00:50 <andythenorth> Eddi|zuHause: ^^^
18:01:05 <andythenorth> (I also have arbitrary python functions in scope if I want, at compile time)
18:02:08 <andythenorth> I'm thinking of just doing dumb range checks tbh; I was trying to get a return value on a single line of nml but I'm bored of that :P
18:02:58 <Eddi|zuHause> andythenorth: attempt #1: switch(cargo_count*100/cargo_capacity) { 0 : (speed_unladen*100+speed*0)/100, 1..25 : (speed_unladen*75+speed*25)/100, 26..50 : (speed_unladen*50+speed*50)/100, 51..75 :(speed_unladen*25+speed*75)/100, 76..100 : (speed_unladen*0+speed*100)/100,
18:03:21 <Eddi|zuHause> granularity easily varied
18:04:34 <Eddi|zuHause> it's beautiful if it does that :)
18:06:05 <Eddi|zuHause> the parts behind the : should be compile-time constants
18:08:22 *** Chris_Booth[ph] has quit IRC
18:10:38 * andythenorth smells that Eddi|zuHause knows which order computers do mathematical operations :P
18:10:45 * andythenorth always puts brackets in, 'just in case' :P
18:11:17 <Eddi|zuHause> not in all cases. but here it's a common case where you learn a few things over time (hopefully)
18:11:37 <Eddi|zuHause> the most important bit is to do the * before the /
18:11:57 <Eddi|zuHause> and that * comes before + you learn in first grade
18:12:10 <Eddi|zuHause> or third grade if you're unlucky
18:12:35 <andythenorth> ok, so speed_unladen is a float, which isn't allowed in a switch block
18:12:39 <andythenorth> I'll trying rounding it
18:18:43 <andythenorth> I'm not allowed to multiply by 3.2 in a switch block
18:19:22 <Yexo> multiply by 16 and divide by 5
18:20:06 <Eddi|zuHause> multiply by 32 and divide by 10
18:20:30 <Yexo> that only matters if you're going to add 5 before dividing by 10 to get better rounding
18:20:32 <andythenorth> right, those both work :)
18:20:38 <andythenorth> I have rounding errors still though
18:20:39 <Eddi|zuHause> so have (speed_unladen*75+speed*25)*32/1000
18:21:13 <andythenorth> maybe it's working actually :)
18:22:06 <Eddi|zuHause> if you want to add Yexo's suggestion: (speed_unladen*75+speed*25+5)*32/1000
18:22:24 <Eddi|zuHause> ((speed_unladen*75+speed*25)*32+5)/1000
18:23:24 <Yexo> doesn't matter in speed, the 32+5)/1000 part will be encoded as a single varaction2 action
18:23:51 <andythenorth> ugliest nml code in the world right now
18:24:32 <andythenorth> I could round the speed vars in the python first I guess
18:24:53 *** FLHerne has joined #openttd
18:24:59 <andythenorth> I need to round 'speed' as well as 'speed_unladen'
18:28:06 <andythenorth> I should add the +5 :P
18:32:36 * andythenorth now has some 7,000mph ships :P
18:33:04 *** plastics has joined #openttd
18:33:14 <SpComb> isn't that what everyone does the first time they dirty their hands with NewGRFs?
18:39:26 <andythenorth> so the only puzzle I have is why the 'speed' is wrong
18:39:32 <andythenorth> the vehicle speed is 16mph
18:40:04 <andythenorth> but when using a cb and doing the 32 + 5 / 1000 adjustment, I get 15mph
18:40:54 <andythenorth> I could actually live with this, but it puts the config file at odds with the result in game
18:45:03 <DorpsGek> andythenorth: Error: Something in there wasn't a valid number.
18:45:55 <Eddi|zuHause> andythenorth: on the purchase list, you probably cannot check cargo_amount, so the switch will produce a random result. you'll need a special purchase list callback returning "speed"
18:46:03 <andythenorth> got that already
18:46:16 <andythenorth> this is something to do with integer maths that I don't understand yet
18:46:26 * NGC3982 is hungover like a sir.
18:46:27 <Eddi|zuHause> and the last one should be /10, not /1000
18:46:48 <andythenorth> although that result should never be reached, so I swapped it for 0 now
18:47:28 <Eddi|zuHause> and why exactly don't you do the *3.2 in python?
18:47:43 <andythenorth> I could, I'll still have the rounding error though
18:47:48 <andythenorth> I tried that already :)
18:48:02 <Eddi|zuHause> what's "the rounding error"?
18:48:04 <andythenorth> 51.2 will always be used by the game as 15mph, not 16mph
18:48:11 <andythenorth> for some integer maths reason
18:49:00 <Eddi|zuHause> there is no 51.2 in integer maths
18:49:18 <andythenorth> and the game always rounds down?
18:49:28 <andythenorth> this isn't new, I've just forgotten how to solve it :P
18:49:36 <Alberth> integer division does
18:49:38 <Eddi|zuHause> that depends on what you mean
18:49:56 <Eddi|zuHause> honestly you're not making a lot of sense
18:50:10 <andythenorth> forget my thinking - here's the problem:
18:50:21 <andythenorth> 16mph * 3.2 = 51.2
18:50:22 <Alberth> do you want rounding up? (a + b - 1) / b
18:50:40 <andythenorth> 51.2 used as a speed value in newgrf cb will result in 15mph
18:50:50 <andythenorth> so problem: how do I reliably round up (or down?)
18:51:38 <Eddi|zuHause> so "speed" is 16, and you want the ingame speed to be 16 as well?
18:52:11 <Eddi|zuHause> if you want "round up", then change the +5 to a +9
18:52:20 <TomyLobo> hmm i was thinking... if i left a temperate map running for a billion years or something, allowed multiple industries of the same type per city
18:52:35 <TomyLobo> i'd end up with a map full of power plants, right?
18:52:55 <Alberth> depends on the industry newgrf
18:53:07 <Eddi|zuHause> people have been reporting towns full of water towers and banks
18:53:30 * andythenorth is going to assume that rounding up is always safe
18:53:31 <TomyLobo> although... dont those close?
18:54:23 <Eddi|zuHause> tropic banks are different from temperate banks, though
18:54:44 *** kkimlabs has joined #openttd
18:54:45 <TomyLobo> ah, cayman islands, right
18:59:38 <Eddi|zuHause> the "road vehicle queueing" tooltip should possibly explain what "quantum effects" are
19:00:01 <Terkhen> it's easy; quantum tunneling, schrodinger box and so on
19:00:02 <andythenorth> Eddi|zuHause: :)
19:00:14 <andythenorth> also, + 9 doesn't solve the rounding :P
19:00:18 <Eddi|zuHause> Terkhen: well _I_ know that...
19:00:46 <Eddi|zuHause> andythenorth: you sure?
19:00:49 <andythenorth> it's a shame that nml cbs don't support values in the same way as props
19:00:57 <Terkhen> so do I, it seems to be common knowledge and therefore it does not need an explanation :)
19:03:32 <andythenorth> Eddi|zuHause: no not sure, weird results :P
19:04:00 <andythenorth> @calc 57.609/3.2
19:04:00 <DorpsGek> andythenorth: 18.0028125
19:04:08 <DorpsGek> andythenorth: 17.8125
19:04:22 <DorpsGek> andythenorth: 16.875
19:08:44 <Yexo> andythenorth: rounding up is not always safe, but it's better. Less people will complain if they buy a vehicle that said 16mph and it will actually go 17mph then when they buy a vehicle that said 16mph and it actually goes 15mph
19:09:06 <andythenorth> @calc 51.209 / 3.2
19:09:06 <DorpsGek> andythenorth: 16.0028125
19:09:40 <andythenorth> game returns 15 for that calculation ^
19:09:50 <andythenorth> rounding the 51.209 I guess
19:10:43 <Yexo> you can't put 51.209 in your newgrf
19:10:48 <Yexo> so it'll be 51 there I guess
19:11:04 <TomyLobo> maybe round to closest integer?
19:11:19 <andythenorth> that will round down
19:11:30 <andythenorth> which nml doesn't offer afaik
19:11:35 <TomyLobo> ceil is -floor(-x) :D
19:11:48 <Alberth> Yexo: how difficult is it to write a reverse engineering routine to compute the right magic value?
19:12:02 <Yexo> it's already included in nml
19:12:09 <Yexo> but not available for callback results
19:12:32 <andythenorth> I could import nml to my python build script, use the routine?
19:12:44 <Yexo> the problem is that in a callback result nml has to know the type of vehicle
19:13:28 <Alberth> we're talking compile time, right?
19:13:50 <Yexo> it's definitely possible, just not yet done
19:14:08 <Alberth> that's probably a long list :)
19:23:00 <Eddi|zuHause> <Yexo> the problem is that in a callback result nml has to know the type of vehicle <-- the switch has a feature entry, right?
19:24:00 <NGC3982> if trains normally load 10% if they stay two days on a station, shouldnt load increase (as long as productions follows) if the trains stays for a longer time?
19:24:36 <Eddi|zuHause> NGC3982: ENoContext
19:25:12 <andythenorth> is that a new reddit thing or something?
19:25:18 <andythenorth> I suddenly keep seeing it :P
19:25:38 <Eddi|zuHause> andythenorth: that's the naming scheme for Delphi exceptions
19:25:49 <Eddi|zuHause> so actually old :)
19:26:00 <andythenorth> most memes are, they just take a while :P
19:26:21 <Eddi|zuHause> andythenorth: it has been used here for years
19:27:37 <Terkhen> first time I notice it :P
19:27:40 *** LordAro has joined #openttd
19:27:53 <Terkhen> how are your exams? :)
19:28:05 <LordAro> exams? finished them a month ago :P
19:28:30 <LordAro> current life purpose = university applications
19:29:29 <Terkhen> good luck with both the results and the applications
19:29:51 <NGC3982> Eddi|zuHause: i fail to understand. google talks about zaphod beeblebrox.
19:29:58 <Alberth> the results are already decided no matter what you do :)
19:30:39 <DorpsGek> andythenorth: 17.8125
19:30:39 <Alberth> NGC3982: you've not read enough hitch hiker to the galaxy
19:31:31 <Alberth> then you'd know zaphod :p
19:31:32 <Yexo> NGC3982: if you force vehicles to stay in a station by using a timetable, I'm not sure they keep loading all that time
19:31:40 <NGC3982> Alberth: i know zaphod, much and much.
19:32:13 * andythenorth might have to round a bit more :)
19:32:16 <Yexo> NGC3982: ENoContext = Error, Not enough Context provided
19:32:46 <Alberth> LordAro: you were gone too quickly this morning, but I have no clue whatsoever what that code is about
19:33:14 <NGC3982> Yexo: and yes, it seems like they dont load more then usual.
19:33:24 <Yexo> so you do use timetables?
19:34:01 <LordAro> Alberth: :) would i perhaps get more help on the forums?
19:34:47 <NGC3982> Yexo: yes, of course. i didnt know of any other way to make a train wait <amount> of days in a random station? :)
19:35:41 <Yexo> NGC3982: I don't think there is a good solution for you currently
19:36:06 <Yexo> apart from writing a patch so vehicles keep loading more cargo when they are waiting due to their timetable
19:36:20 <Yexo> I do think the current behavior is sub-optimal
19:36:52 <NGC3982> well, ill solve it with more trains.
19:36:57 <NGC3982> usually solves most things
19:37:38 <Yexo> try full-load orders, those usually work for me
19:38:11 <NGC3982> im using UKRS2+ with ECS
19:39:03 <NGC3982> and the industry level jibberish is a bit confusing :)
19:39:21 * NGC3982 conducts an experiment.
19:43:11 <NGC3982> the thing is (or seems to be), that using full load on trains to an ECS oil field lowers the amount of transported oil.
19:43:20 <NGC3982> thus, lowering the level, thus, lowering the production
19:43:39 * NGC3982 scans for a wiki entry
19:47:44 <NGC3982> the wiki entry for the grf doesnt really tell me how transported goods work
19:47:49 * NGC3982 scans for more wiki entries.
19:50:51 <NGC3982> since i just change how long time x trains should wait for cargo, they still pick up the same amount
19:50:57 <NGC3982> thus not affecting the transported rate
19:51:09 <NGC3982> ergo ipsum: moar trainz!
19:53:06 <Zuu> If you do "rm -r *&" and then "cp -r ../some_catalog .", is the * of rm expanded by the shell or rm when that command is issued or will some_catalog be removed too if the cp command finnish before?
19:54:04 <Zuu> I realize that the best is to force the shell to expand * before hitting enter :-)
19:54:04 <Yexo> I'd expect that some_catalog always stays, but I'm not 100% sure
19:55:31 <Zuu> That what I expect/hope too. :-)
19:57:48 <Zuu> ... It appears to have worked :-)
19:57:52 <Beul> can an if block be placed iside an item block?
19:58:08 * NGC3982 fails to increase transported goods.
19:58:23 <Zuu> NGC3982: Do you have a statue?
19:59:55 <Zuu> If the city that the industry (or station?) belongs to have a statue, you get 10% rating bonus for that station.
20:00:15 <Beul> NGC3982, in my experience, using shorter trains that leave the pichup station more regularly increases transported cargo with ecs vectors
20:00:34 <Beul> have not figured out why thoug, dut it seems to work
20:01:04 <NGC3982> im already using 5 tile trains
20:01:42 <Beul> make them even shorter. when production is still low, I use 3 or 4 tile lengt trains
20:01:56 <NGC3982> Beul: with full load?
20:02:09 <andythenorth> default cargo for freight ships should be coal?
20:02:50 <NGC3982> Beul: ill set them to load 50%
20:02:56 <Beul> NGC3982, 333k litres is not much indeed, they should be able to go a lot higher
20:03:23 <Yexo> <Beul> can an if block be placed iside an item block? <- yes
20:03:48 <Yexo> item (..) { if (a) { property { ... } } else { property { ... } } } <- that is valid
20:05:44 *** kkimlabs has joined #openttd
20:08:18 * NGC3982 just tried a zip of jack daniels cola.
20:09:01 *** Chris_Booth has joined #openttd
20:11:40 <Beul> what revision number does 1.1.0 have? (or where could I find that?)
20:17:26 <Eddi|zuHause> Beul: the revision of a release is very uninteresting. you won't get the release if you check out that revision of trunk
20:18:08 <Beul> I am not planning on doing so Eddi|zuHause, I only what to condionally disable the grf
20:18:37 <Eddi|zuHause> Beul: releases have always "revision 0x80000" in the newgrf version
20:19:14 <Yexo> nml probably does not have a nice way to expose that
20:19:43 *** Supercheese has joined #openttd
20:20:04 <Rubidium> but isn't saying ottd_version(1, 2, 0, <revision that added the feature you need) inside an if the thing you should check for?
20:20:28 <Yexo> yes, but that's harder to find
20:20:42 <Yexo> especially if you use multiple features included in 1.x.0
20:20:49 <Eddi|zuHause> but then checking for "ottd_version(1,2,0,0)" will suffice
20:20:59 <Eddi|zuHause> since that is larger than any 1.1.x release
20:21:15 <Yexo> yes, but the 1.1.x releases are suddenly not valid
20:22:03 <Yexo> if features have been introduced in 1.1, so 1.1 can support the grf, checking for 1.0.0r0 will allow all nightlies from before 1.1.0, some of which won't support the features you need
20:23:01 <Eddi|zuHause> i was reading it like he wants a feature that is not in 1.1
20:23:24 <Yexo> ok, checking for 1.2.0r0 will allow the grf for every nightly after the 1.1 branch
20:23:41 <Yexo> which is not the intension, as early nightlies won't have support for the feature
20:25:11 <Eddi|zuHause> yes. but the premise was to not bother to find out which revision
20:25:13 <Rubidium> rejects all non-stable releases of 1.2 (nightly, beta, rc), everything before 1.1 and everything in trunk after 1.2 branching
20:25:48 <Eddi|zuHause> Rubidium: trunk is increased to 1.3 almost immediately after 1.2 branch
20:25:52 *** Rhamphoryncus has joined #openttd
20:26:24 <Yexo> the "rejects beta,rc" part is not so nice though
20:26:54 <Yexo> which brings us back to: find proper revision of feature inclusion, or, since it's easier, find revision of 1.2.0
20:27:16 <Rubidium> hmm, though only for the 1.2 branch and 1.2.0 beta/RCs (I think)
20:27:28 <Rubidium> as 1.2.1 RCs would have version 1,2,1,<rev>
20:29:27 *** FLHerne has joined #openttd
20:29:28 <Beul> Eddi|zuHause, Yexo, I allready have: if (openttd_version > version_openttd(1, 2, 0, 22713)) {
20:29:52 <TomyLobo> "openttd_version > version_openttd" :D
20:29:54 <Beul> but aditionally I want to give a fatal error for versions lower than 1.1.0
20:30:34 <Beul> It looks ridivilous TomyLobo, but it works :p
20:30:48 <TomyLobo> yeah just nice naming
20:33:04 <NGC3982> Beul: i made a ridiculous mess out of it.
20:34:07 <NGC3982> andythenorth: please spare some for me. :)
20:37:02 <andythenorth> using {NBSP} to space vehicles names in the buy menu causes horrible side effects
20:38:29 <andythenorth> also nml doesn't appear to respect default_cargo_type for ships
20:39:44 <andythenorth> hmm, only wrt PASS
20:40:03 <andythenorth> PASS is in my cargo table
20:40:31 <andythenorth> frick, maybe it's not in this game
20:40:55 <andythenorth> FIRS claims to provide it
20:43:03 <Beul> has anybody in here got the latest version of chillcore's patchpack?
20:43:25 * andythenorth solves stupid issue
20:45:04 <Yexo> and isn't defautl_cargo_type unsupported anyway?
20:45:21 <Hirundo> It is supported (though wasn't)
20:45:53 <andythenorth> mistake in my code
20:45:57 <andythenorth> don't trust magic :P
20:46:53 <Supercheese> Night, I guess. Time zones are strange
20:47:02 <Eddi|zuHause> Beul: chills patchpack is based on a really old trunk revision
20:52:41 <NGC3982> Beul: worked like a charm.
20:56:21 *** MinchinWeb has joined #openttd
20:57:12 <MinchinWeb> Is there a way, with NoAI, to find out how much cargo a vehicle is currently carrying?
20:57:43 <Hirundo> andythenorth: why do you need it?
20:57:55 <andythenorth> buy menu offsets for FISH
20:58:54 <Beul> NGC3982, the ecs thingie you mean?
20:59:28 <andythenorth> SETX isn't found in my nml src
20:59:56 <Hirundo> t openttd modified that it takes the sprite width into account?
21:00:07 <andythenorth> apparently not :P
21:00:07 <Hirundo> damn ' next to enter button :S
21:00:26 <Supercheese> That cursed apostrophe key, leading to so many premature Entries
21:01:14 <Hirundo> andythenorth: I can find no trace of it in the NML code or docs, so the answer is no
21:01:39 <andythenorth> maybe I should try and patch it
21:01:54 <andythenorth> I recall some discussion about deprecating it?
21:03:36 <andythenorth> grfstrings.py definitely misses it
21:04:07 <Hirundo> It is indeed sortof deprecated and no longer exists except for NewGRFs
21:04:26 <andythenorth> seems like the game should just handle this case
21:05:07 <andythenorth> why does the depot view window change size sometimes?
21:05:25 <andythenorth> i.e. the boxes in the grid change size depending which ship grfs I have loaded
21:05:52 <planetmaker> it supposedly adjusts to the ship size
21:07:00 <andythenorth> 4am was not pretty
21:07:21 <andythenorth> we can't really deprecate setx I guess
21:07:31 <andythenorth> as newgrf spec may never deprecate things
21:07:39 <andythenorth> so maybe I should patch it for nml tomorrow
21:09:05 *** andythenorth has left #openttd
21:09:51 *** LordAro has joined #openttd
21:13:34 *** Zeknurn has joined #openttd
21:17:10 <CIA-4> OpenTTD: frosch * r24366 /trunk/src/script/api/script_vehicle.hpp: -Doc: Improve description of ScriptVehicle::GetCargoLoad.
21:32:13 *** kkimlabs has joined #openttd
22:13:32 <NGC3982> if i have multiple trains at a single pickup station
22:14:20 <NGC3982> what parameter dictates when more then one of the trains pick up cargo simultaniusly
22:14:36 <NGC3982> with wait for full load, of course.
22:14:50 <NGC3982> i notice that it happends, but only sometimes?
22:14:55 <Supercheese> The loading algorithm, in conjunction with how much cargo is waiting
22:16:14 <Supercheese> Stations > Cargo Handling > Use improved loading algorithm
22:16:26 <NGC3982> ah, i see. i have "no" cargo waiting. since i have trains waiting at the station, the cargo goes almost instantly to a train.
22:16:57 <Supercheese> But yeah, if you have very little cargo, only 1 train loads at a time
22:17:13 <Supercheese> with the improved setting on anyway, I've never tested with it off
23:09:00 <Beul> NGC3982, if irc with improved loading enabled two trains wil start loading simultanously if the cargo is sulpplied faster than the loading speed of the vehicle loading
23:10:25 <Beul> most newer caraiges/busses/trucks/trams have a property that defines how much cargo can be loaded per tick
23:13:12 <CIA-4> OpenTTD: michi_cc * r24367 /trunk/src/ (7 files): -Feature: [NewGRF] Customisable signals for rail types.
23:17:56 <Beul> that was exacly what I need to make 'invisible'signals for my cablecars :D
23:29:15 <Eddi|zuHause> i'd suggest making them not invisible, but instead disguise them as something else
23:30:33 <Beul> well that is why I say "invisible":p
23:31:09 <Beul> I could indeed imagine a couple of problems with realt invisible signals :)
23:34:07 <planetmaker> but they surely will come ;-)
23:34:50 <Beul> well to be honest development from my part might be interupted just as quick as it began...
23:35:49 <Beul> Next week this time I'll be some 10.000ft in the air somewhere above norhtern africa, only to return at the end of august
23:40:03 <Beul> Nice michi_cc, I think some railset developers will also thak you for that ;)
continue to next day ⏵
