IRC logs for #openttd on OFTC at 2015-02-28
⏴ go to previous day
00:13:50 <Samu> where in the afterload do I put this code?
00:14:40 <Samu> i don't want to put it in a random place
00:16:41 <Samu> which line do I insert this code?
01:23:28 <Eddi|zuHause> usually you would add this stuff at the end of the function
01:23:39 <Eddi|zuHause> so the conversions go in chronological order
01:24:22 <Eddi|zuHause> you don't know which ancient version might have used that bit for other stuff
01:24:40 <Eddi|zuHause> so you better not insert this code before that conversion was handled
01:42:40 <Eddi|zuHause> that's not the end of the function
01:42:49 <Eddi|zuHause> also, you're missing the check for the savegame version
01:44:02 <Samu> i dont' know which savegame version it wouldn't work
01:45:18 <Samu> i know it works for version 194
01:45:50 <Samu> and i know i don't have to create version 195
01:46:46 <Eddi|zuHause> Samu: this code will currently destroy the information in the savegames you made with the current (and future) versions
01:46:59 <ST2> *** and the oscar goes to.... Eddi|zuHause - crowd applauds enthusiastically \o/ ***
01:47:04 <Samu> they can co-exist without issue
01:47:55 <Eddi|zuHause> then you have a different understanding of "issue" or "co-exist"
01:48:36 <ST2> the "oscar" of helping Samu ;)
01:49:39 <Samu> if bit 4 is 0 and is canal and industry, ignore it, stays 0
01:49:49 <Samu> if bit 4 is 1 and is canal and industry, ignore it, stays 1
01:50:13 <Samu> both savegames are compatible this way
01:51:00 <ST2> spent 2 nights trying to fix some GS issue and solution was in a ".tointeger()" - at least now I won't forget it xD
01:51:02 <Samu> for the other non-canals, it is re-applying 1, which is also intended
01:51:32 <Samu> if it was 0, set to 1, save, then load, if it was 1, set to 1
01:51:45 <Samu> what matters is that it's set to 1, as intended
01:52:14 <Eddi|zuHause> Samu: so what if a future version introduces a different water class that can have owners?
01:52:25 <Eddi|zuHause> you then break that future version
01:54:23 <Eddi|zuHause> like a deep canal for ocean-going ships
01:54:51 <Samu> that re-applying 1 is being handled wrong
01:55:16 <Samu> or do I reall y need to bump savegame version?
01:55:37 <Eddi|zuHause> no. i'm saying you MUST under all circumstances guard any segment of AfterLoadGame with a savegame version check.
01:56:09 <Samu> i don't understand what's that check
01:56:17 <glx> a bump may not be needed but a guard in case a future bump happens is needed
01:57:40 <glx> conversion must be done only for older game
01:59:18 <glx> all afterload code is done that way
02:00:29 <Samu> if (IsSavegameVersionBefore(188)) {
02:00:58 <Samu> but i don't know which version it wouldn't work :(
02:01:39 <glx> I guess it should be for all version before the current one
02:02:34 <Samu> if (IsSavegameVersionBefore(194)) { ?
02:03:05 <glx> 195 if you want to convert current games too
02:03:32 <Eddi|zuHause> that's what the future is
02:03:36 <Eddi|zuHause> it's not there yet.
02:03:45 <Eddi|zuHause> but it will (probably) come some day
02:03:55 <Eddi|zuHause> and then you need to be prepared for it
02:04:45 <Samu> what would happen at then?
02:05:16 <Samu> it will stop converting?
02:05:22 <Eddi|zuHause> nothing, ideally.
02:05:33 <glx> it will still convert old saves
02:06:07 <glx> but new saves won't need the conversion
02:06:31 <Samu> ah, i guess I understand
02:07:10 <Samu> it's a chronologic event
02:07:59 <Samu> because it was already converted
02:17:45 <Samu> i put this before this ?
02:18:44 <Samu> before line 2988 or after line 2988?
02:20:00 <Eddi|zuHause> given that anything after return is ignored, you probably should put it before :p
02:21:05 <Eddi|zuHause> imagine you're sending someone out of your office, close the door, and then tell them how to do something
02:21:11 <Eddi|zuHause> they probably not hear you :p
02:25:47 <ST2> unless we place next lines in CAPS, like screaming trough the door xD
02:35:30 <Samu> i am tremendously slow at these simple edits
02:37:14 <Samu> crap, the 195 version thingy, grrr grrr! sorry
05:21:30 *** chillcore has joined #openttd
05:27:20 <chillcore> I am torn between tackling some of the stuffs I still have to learn the details of or picking some low hanging fruits ...
05:27:42 <chillcore> does not feel like my patch advanced much this week
05:31:31 <ST2> pick the fruits, make juice and enjoy the weekend xD
05:33:10 <chillcore> that seems like a good idea yeah ;)
05:37:22 <chillcore> I stopped drinking completely early 2013 ...
05:37:23 <ST2> had an ass of a task to improve a GS, not made by me, and that was merged in our master GS, that communicates with server software
05:38:18 <chillcore> those server communications can be a pain in the ass yeah
05:39:22 <ST2> gets easier if using xShunter as server controller software
05:39:48 <ST2> many of the functions already there
05:40:01 <ST2> and most of the work is on plugins
05:40:22 <ST2> and I'm not confortable with it :S
05:42:12 <chillcore> That I do not know, I do know that if "a" server just asumes datablobs to be valid (legit) because it is coming from "a" client things may go very wrong without the server owner noticing.
05:42:41 <chillcore> untill the server owner has his nose rubbed into it that is
05:42:55 <chillcore> unrelated to openttd but yeah
05:43:42 <ST2> hehe, that's a funny way to put things ont the table ^^
05:45:25 <ST2> not yet opensource, and servers still need a key to operate with that software
05:45:56 <ST2> but makes server things way much simplier ^^
05:46:20 <chillcore> entering that link on my ipad ... computer sais no ... :P
05:47:25 <chillcore> I am sure, it is just if you look at how they are stealing all these steam accounts ... can't trust noone no more
05:48:19 <ST2> well, no one will ever steal my steam account
05:48:37 <chillcore> hmm don't be sure of that
05:48:55 <ST2> reason: I dnt have a Steam account
05:49:42 <chillcore> yeah second time I got "account restricted untill verification of blabla" from paypall
05:50:41 <ST2> specially when you dnt have nothing in the stuff
05:51:03 <ST2> well, makes me sad on what internet becomes
05:51:22 <chillcore> yeah I looked into it but they needed my bank account nr, my real name ... yada yada so i did not finish registering
05:51:44 <chillcore> but that was like two years ago so I doubt they would start sending me messages now
05:52:50 <ST2> and someday someone will knock your door :P
05:53:34 *** Supercheese has joined #openttd
05:53:40 <chillcore> instead I took an anonymous credit card ... anonymous to the outside world that is, postal office has my real name but they will never give it to anyone
05:55:08 <chillcore> Steam allowed me one purchase, then they tried some of their magic verification (putting money in it) but they do not have the extra "code" so that failed
05:55:39 <chillcore> since then they do not accept that payment method no more
05:56:04 <ST2> follow the wise man advice: dnt create an account there xD
05:56:17 *** Eddi|zuHause has joined #openttd
05:56:51 <ST2> I barely have time to openttd (and not playing)
05:57:12 <chillcore> still do not understand why someone would allow any access to a bank account by whatever company for whatever reason
05:57:25 <chillcore> ye I get your drift
05:57:44 <ST2> oh well, playing now KaM, to kill the missings of it ^^
05:58:12 <chillcore> it is just that I bought a bxed version of HL2 ep1 and 2, did not notice the registering requirment and had to let it rot for two years on the shelf
05:58:45 <chillcore> so I got one and bought a few oldies
05:59:12 <ST2> have to ask: if using an Ipad, had you filled your credit card details there?
05:59:32 <chillcore> e-mail adress and that is it
05:59:48 <chillcore> even the store has nothing
05:59:54 <chillcore> they tried though ...
05:59:59 <ST2> that's a thing on Apple stuff
06:00:29 <chillcore> it was not an apple store but the guy tried filling in the warranty forms
06:00:34 <ST2> note: I only have it on VM's to check if changes I make compiles there too
06:00:40 <chillcore> I gave him my phone nr
06:01:01 <chillcore> then he asked for my name too and I was like "for what?"
06:02:26 <chillcore> my reply was "look in have this invoice right with the date?... when there is a prob that is all I need"
06:03:05 <chillcore> then when it came to registering it with apple just an e-mail adress was good enough and it still is
06:03:18 <ST2> well, gonna watch the other cause of free time lacks sometimes.... TV series
06:03:30 <ST2> new episode of 12 monkeys :)
06:04:22 <chillcore> I threw the cable out years ago ... I can watch adds everywhere else too :P
06:04:29 <chillcore> anyhoo, have fun ;)
06:05:10 * ST2 grabs the pirate costum xD
06:05:58 <ST2> well, I bought DVD's of the series I saw and liked
06:06:21 <ST2> Prison Break, Band of Brothers
06:06:39 <chillcore> I sometimes watch stuff that is on youtube ... much more fun to binge then wait a week for the rest over and over again
06:06:55 * chillcore wants all startrek dvds
06:07:07 <chillcore> there are just so many ...
06:07:42 <chillcore> Live long and prosper. MR. Spock.
06:08:43 <chillcore> RIP Leanord Nimoy, you will be missed.
06:13:57 <ST2> [06:12:01] * +ttd-srv5 * !! EMERSON TRANSPORTES (1/Dark Blue) had a crash (vehicle ID 2038: road vehicle was destroyed by a train at 0x73593). <<-- I hate when this spams the channel only because dnt build bridges :S
06:15:05 <chillcore> also xShunter ... looks cool except for GeoIP
06:15:17 <ST2> it's servers channel - very good to spot when it's a train to crash RV's
06:15:25 <chillcore> I consider that spyware ... but that may be just me
06:16:12 <chillcore> still it retrieves my location
06:16:31 <chillcore> I understand most peeps have no prob with this
06:18:06 <ST2> basically it's a table with IP ranges and ISP's that use them
06:19:00 <ST2> only tells you the country xD
06:20:11 <ST2> something like this: [06:16:21] * +ttd-srv27 * # Player (#11/user_IP/UA) has started company 4 ()
06:21:20 <chillcore> yes I know ... I just don't see the point of it, except for greeting someone in his own language
06:21:38 <chillcore> which can be achieved by looking at the language config
06:21:49 <chillcore> all my progs are set to english
06:22:03 <chillcore> not my country not my native language
06:22:14 <ST2> when you join a ttd server
06:22:24 <chillcore> so being greeted inmy own language looks shady
06:22:29 <ST2> no one knows your language settings
06:22:59 <chillcore> yet, might be a nice patch
06:23:34 <ST2> it's on servers controller software
06:23:47 <ST2> servers dnt need to be patched for that
06:24:27 <chillcore> what I mean is that it would be nice pacth for openttd
06:24:38 <chillcore> send lang config on connect
06:24:41 <ST2> join this IRC server, just to check it: hub.irc.x-base.org:6668
06:25:11 <ST2> and on it, channel #btpro-openttd
06:25:32 <ST2> just to check the work of it ^^
06:26:31 <chillcore> hehe I think I just created my own private btpro channel ... hihi
06:27:39 <ST2> after connect to server, type /join #btpro-openttd
06:28:35 <chillcore> thing is I do not find "hub.irc.x" to begin with
06:28:47 <chillcore> I click networks and it is not in the list
06:30:30 <chillcore> when I enter just hub ... it shows me York univeristy student club
06:30:38 <chillcore> I'll figure this out I am sure
06:31:18 <ST2> well, irc server address is: hub.irc.x-base.org
06:31:30 <chillcore> I am checking prefs now
06:33:35 <ST2> but it's a mibbit link :S
06:35:56 <chillcore> I used mibbit before untill it was no longer alowed here
06:36:53 <ST2> well, above link will work, if you dnt want to make it easier xD
06:37:25 <chillcore> checking on my ipad ... maybe I will get a message about dead frogs maybe not
06:38:21 <chillcore> hmm does not find btpro
06:38:30 <chillcore> I'll enter the link directly
06:39:59 *** Flygon_ has joined #openttd
06:41:18 <chillcore> hmm I saw it then clicked go and it took me back to mibbit's front page
06:42:28 <chillcore> ah ok it opened a tab I did not see before
06:42:47 <chillcore> dr_bibble has entered the building
06:43:17 <ST2> already seeing the servers actions?
07:05:17 *** tokai|noir has joined #openttd
07:05:17 *** ChanServ sets mode: +v tokai|noir
07:20:09 *** Alberth has joined #openttd
07:20:09 *** ChanServ sets mode: +o Alberth
07:25:49 *** Celestar has joined #openttd
08:28:10 *** Progman has joined #openttd
08:50:35 *** Pensacola has joined #openttd
09:20:14 *** chillcore has joined #openttd
09:48:15 <Alberth> argh, how to delete the local tracking branch of a remote repo in git?
09:50:38 <Alberth> as "git branch -D" would be the logical command, that obviously fails to work "branch not found"
09:52:01 <Alberth> "git remote remove" fails with "cannot remove config section"
09:52:30 <chillcore> I once almost commited to trunk untill I changed/removed the push location, this was in the gui however and not quite the same prob you are having?
09:53:27 <chillcore> the act of puhing was stopped by me not being allowed to push .... luckily
09:53:59 <Alberth> yeah, always nice not to have too many powers :)
09:54:42 *** oskari89 has joined #openttd
09:55:20 <Alberth> I added another 'upstream-like' repository in the past where I pulled updates from, but that's done now, so I want to kill all tracking branches from it now
09:56:53 <Alberth> preferably without killing that repo as a possible remote source, in case I need it again
09:57:07 <Alberth> but git refuses to delete the branches
10:02:27 <chillcore> fetch -p ?? removes any remote-tracking branches which no longer exist on the remote
10:02:34 <chillcore> don't take my word for it though
10:03:06 <Alberth> I know, but the remote hasn't dropped the branches :)
10:03:49 <chillcore> hmm small caps "d" ?
10:05:08 <Alberth> ? git fetch has no -d option
10:05:28 <Alberth> and push will fail due to not having push access, as it's not my remote
10:06:22 <Alberth> ah well, it's just git being useless :)
10:07:11 <Alberth> they must have taken a master course in making error messages as non-informative as possible :)
10:07:35 <chillcore> hmm then why does this page sais so? ... "git branch -d branchname"
10:08:18 <Alberth> that works for your own local branches
10:08:43 <chillcore> k. any further help will be just guessing on my part so ... :(
10:08:45 <Alberth> and -D if the branch didn't get merged into something
10:09:26 <Alberth> yeah, thanks for trying though :)
10:09:39 <Alberth> one day we'll find the answer :)
10:15:03 <chillcore> if that it is not it I give up ;)
10:26:49 <Alberth> looks like that is only used when you make a new branch
10:27:25 <Alberth> don't give up, book it under 'needs further study somewhen' ;)
10:27:26 <michi_cc> I don't think you're supposed to delete the remote branch if it still exists on the remote, after all it is a remote tracking branch, and not a local one.
10:28:18 <Alberth> so I have to kill the remote entirely to get rid of the tracking branches?
10:29:16 <michi_cc> You don't actually, reading man pages does seem to help. (i.e look at -r option to git branch)
10:32:16 <michi_cc> git branch -d -r sadger/0.01
10:32:34 <michi_cc> I.e. like git remote -r lists them.
10:32:41 *** Quatroking has joined #openttd
10:37:11 <Alberth> so the "-r" option sort-of adds the 'remotes' prefix
10:37:32 *** Ketsuban has joined #openttd
10:38:14 <Eddi|zuHause> i'm sure that's all perfectly logical in the twisted sense of a git developer
10:39:07 <Eddi|zuHause> but if you don't have that exact same mindset, git is terrible...
10:40:26 <Alberth> yeah, mostly playing around with it in another project to get exposed to it :p
10:42:16 <chillcore> I had some fun with it while doing some never published patchpacks, as long as you do not intend to take out a patch it is kinda cool.
10:43:01 <Eddi|zuHause> the problem is that "the community" is so immersed in that mindset, that they cannot even adapt to people not having this mindset, and understand their problems.
10:43:15 <Eddi|zuHause> it's like talking to a person that only watched FOX News their entirel life
10:47:50 <chillcore> reminds me of the news on filmnet <- for those that know filmnet
10:48:20 <Alberth> I can see how that happens, I already have problems mapping git concepts onto hg, let alone if you never seriously used hg
10:48:46 *** smoke_fumus has joined #openttd
10:50:39 <chillcore> I like git for the option to continue to develop each branch seperatly and then merging them into "main"
10:51:51 <chillcore> as for taking something out of main ... you're better of starting a new "main"
10:52:15 <chillcore> ^^^ and lose the merging stuffs ... sniff
10:54:14 <Eddi|zuHause> pulling cargodist was a nightmare, because every time it was basically a new repo
10:54:47 <Eddi|zuHause> that was before i had "fast" internet
10:55:42 <Alberth> chillcore: that's what 'rebse' is supposed to fix, haven't used that much yet
10:55:50 <chillcore> exactly ... but you avoid the troubs you get with hg ... which can be a pain too if a piece of code keeps getting modified
10:57:22 <Alberth> You sort-of want branches stacked onto each other, where each branch can change indpendently, I think
10:57:41 *** roidal_ has joined #openttd
10:58:40 <chillcore> "did you mean rense?" fuuuuuuuuu
11:01:16 <chillcore> "did you mean reese version control?" double fuuuuuu
11:01:22 <chillcore> got a link alberth?
11:02:36 <Alberth> chillcore: oh, typo, "rebase" :)
11:03:22 <Alberth> ^^ notation for man 1 git-rebase
11:04:49 <Alberth> but rebase where the merge will be non-trivial scares me too much to even attempt, currently :)
11:07:17 <chillcore> not sure if used rebase yet, can't remember anyways ... I am an avid svn user unless I want something comitted someday and do not feel like splitting afterwards.
11:09:15 <chillcore> maybe it is the way I do things but I spend way too much time "fixing" patches with hg
11:09:16 <Alberth> rebase is basically about moving branches around
11:09:50 *** Myhorta has joined #openttd
11:10:08 <chillcore> right ... no never done that
11:10:26 <Alberth> ie you created a branch from master, and update master from upstream. Then your branch has to be moved to the new HEAD. That's where you rebase
11:11:19 <Alberth> ie the branching point between master and your branch moves
11:11:39 <chillcore> hmm ... ye done that plenty of times then, no different then pulling in hg or updating in svn
11:12:22 <Alberth> that's not rebase, a pull in hg doesn't move your branch
11:13:13 <chillcore> before I pull I pop all pacthes and aply to tip ... that is not the same?
11:14:39 <chillcore> with svn I leave patch applied while updating, ususally
11:15:00 <Eddi|zuHause> it may have similar results, but it certainly is not "the same"
11:15:39 <Alberth> hg qpop -a does it somewhat the same, but your mq patches are not in the repository
11:15:46 <Alberth> ie they are not a commit
11:16:04 <Alberth> they live just next to the repository
11:17:22 <chillcore> hg qpop -a == click the double arrow
11:19:15 <chillcore> but I see the difference now thanks. ;)
11:22:41 <Alberth> :O you can actually do that?
11:24:38 * chillcore dislikes spaces in patches names
11:24:44 <Alberth> hmm, "patch repository" no idea what what means
11:25:19 <Alberth> hg qser lists the patch?
11:25:21 <Eddi|zuHause> Alberth: if you "hg init --mq", then the patch queue gets under version control
11:26:16 <chillcore> you forgot the 'test' part eddi?
11:26:16 <Eddi|zuHause> or "hg qinit" might work as well
11:26:17 <Alberth> yeah, tried that one time, but diffs between patch file revisions are less than readable :p
11:28:07 <Alberth> hg qgo "adjacent crossings" ; hg diff -r qparent ?
11:28:11 <michi_cc> That's what interdiff was invented for.
11:29:05 <Eddi|zuHause> Alberth: i'm not sure that that would do what i want it to do
11:29:21 <Alberth> ok, what is it that you want? :)
11:29:50 <Eddi|zuHause> Alberth: show the changes made to the adjacent crossings patch, but not the one to the diagonal crossings patch
11:30:03 <Eddi|zuHause> i can "cd .hg/patches" and do the "hg diff" there
11:30:16 <Eddi|zuHause> but why wouldn't hg diff --mq do that exact same thing?
11:30:31 <Eddi|zuHause> is it really the space that's screwing things up?
11:30:59 <Alberth> find a random patch with a non-space name?
11:32:01 <Eddi|zuHause> doesn't work with "series" either
11:32:10 <Eddi|zuHause> so that's not it
11:33:47 *** luaduck is now known as luaduck_zzz
11:35:01 <Alberth> with a real patch it would be better, but I suspect the result will be the same
11:36:08 <Alberth> however, since mq is quite dead, reporting it as bug won't really help, I am afraid
11:46:56 *** onlyonemac has joined #openttd
11:47:21 <onlyonemac> where can i ask development-specific questions?
11:47:56 <Eddi|zuHause> where can i ask metaquestions?
11:48:07 <onlyonemac> what are metaquestions?
11:48:09 *** gelignite has joined #openttd
11:48:25 <onlyonemac> could someone tell me what the os-specific files are in the openttd source?
11:48:27 <Eddi|zuHause> a metaquestion is a question about asking a question
11:48:29 <Alberth> onlyonemac: asking about asking is a meta question :)
11:49:20 <Eddi|zuHause> the os-specific files are in src/os/
11:49:33 <onlyonemac> what about video drivers?
11:49:40 <onlyonemac> aren't those in src/video/?
11:49:43 <onlyonemac> what else is there?
11:50:49 <Eddi|zuHause> well, in src/video are platform-dependent files, which may or may not apply to multiple OSes
11:51:12 <onlyonemac> ok i think i mean to ask where the platform-dependent files are, not the os-specific files
11:51:29 <Eddi|zuHause> there are also sound-related files like this
11:51:40 <onlyonemac> basically i'm wanting to port to a new os on a new platform
11:51:44 <onlyonemac> what files do i need to look at?
11:52:19 <Alberth> grep for the usual suspects, LINUX WIN32, etc ?
11:52:38 <onlyonemac> i was wondering if there was a centralised list of such files though, so that i can be sure to have everything
11:52:54 <Eddi|zuHause> onlyonemac: if the new platform has SDL, then you don't need to touch the video/sound/music stuff
11:53:07 <onlyonemac> also i don't know what functions in the other parts of the source are used for platform abstraction and which are the platform-specific files
11:53:12 <onlyonemac> no the new platform does not have SDL
11:53:58 <Eddi|zuHause> onlyonemac: additionally to the ones mentioned, src/thread/ is also platform specific
11:54:07 *** JacobD88 has joined #openttd
11:54:10 <Alberth> don't think there is a simple list or directory you can look at
11:54:45 <onlyonemac> are there any other considerations that i should take into account?
11:55:09 <onlyonemac> e.g. the new platform has no stdlib - where are the memory allocation functions that i must rewrite?
11:55:15 <Eddi|zuHause> onlyonemac: you should probably look through configure (config.lib)
11:56:12 <Eddi|zuHause> i don't know anything about stdlib
11:57:10 <Eddi|zuHause> sounds like a horrible job to rewrite that
11:57:39 <onlyonemac> i shouldn't have to rewrite stdlib; i'll just need to change calls to stdlib to match the functions implemented by the new platform/os combination
11:57:50 <Alberth> I would you you have a c++ compiler for the platform
11:58:41 <Alberth> that should handle new/malloc things, wouldn't it?
11:58:53 <onlyonemac> no because those are in stdlib
11:59:10 <onlyonemac> the os has other library functions which do the same things as stdlib but they are called differently
11:59:27 <onlyonemac> so the openttd code would need to be modified to use the new os library
11:59:29 <Eddi|zuHause> onlyonemac: src/stdafx.h has some compiler specific stuff
11:59:40 <chillcore> <Alberth> don't think there is a simple list or directory you can look at <- grep #endif ?
11:59:58 <Alberth> chillcore: that's all .h files :)
12:00:19 <chillcore> not .cpp too ... hmm ok
12:00:24 <onlyonemac> you could grep WIN32 or something like that but that would probably for example miss calls to stdlib
12:00:32 <onlyonemac> maybe i should also grep stdlib or something like that
12:00:46 <onlyonemac> (or grep strings.h/malloc.h/etc...)
12:00:47 <Alberth> run the compiler, it will break on anything you don't have
12:00:54 <onlyonemac> that's the other option lol!
12:01:13 <Alberth> it does a lot better search than you can :p
12:02:00 <Alberth> I use that trick quite often, rename or comment something out, and recompile everything
12:02:14 <onlyonemac> that's probably what i'll do then
12:02:22 <Alberth> the compiler will nicely list where it is used :p
12:02:35 <onlyonemac> i was just wondering if there was a list of such files/functions, but it seems that there isn't
12:11:35 <Eddi|zuHause> btw. the mercurial thing is some working directory bullshit.
12:12:18 <Eddi|zuHause> "hg diff --mq blah" passes ./blah to the command, but ./blah is not inside .hg/patches, so it bails out
12:12:36 <Eddi|zuHause> so i must write .hg/patches/blah
12:20:14 <Alberth> I had some trouble with hg -R in that area too (point to repo other than current directory) iirc, but never bothered figuring it out exactly
12:21:33 <Eddi|zuHause> yes, they said --mq is basically an alias for -R .hg/patches
12:22:46 <Eddi|zuHause> i'll file this under "stupid behaviour"
12:23:33 <Eddi|zuHause> things that should be circumnavigated, but not important enough to worry about
12:32:24 *** Supercheese is now known as Guest114
12:32:28 *** Supercheese has joined #openttd
13:13:58 *** luaduck_zzz is now known as luaduck
13:24:57 *** frosch123 has joined #openttd
13:37:27 *** Myhorta has joined #openttd
13:40:13 *** Biolunar has joined #openttd
13:42:39 <Samu> i submited the patch again, last night
13:44:00 <Alberth> it still doesn't make sense to me
13:44:37 <Alberth> in my view, the value of an unused bit is not relevant to preserve, as it is not used.
13:44:53 <Alberth> therefore the entire patch serves no purpose
13:48:27 <frosch123> can other windows users reproduce that?
14:10:25 *** Klanticus_ has joined #openttd
14:15:16 *** Myhorta has joined #openttd
14:16:03 <Samu> i experienced a similar issue once
14:16:38 <Samu> about right click panning
14:16:50 <V453000> but I cant seem to find it
14:18:51 <Samu> when releasing right click button, the cursor was outside OpenTTD borders
14:19:09 <frosch123> was that before beta2 ?
14:19:29 <frosch123> because the whole thing was changed on the day before beta2
14:19:45 <Samu> it was in one of the testing versions
14:19:57 <Samu> haven't happened yet in beta2 as far as i can tell
14:20:10 <Samu> seemed to happen often when it was lagging
14:20:32 <V453000> you cant pan out of the map?
14:20:35 <frosch123> well, that's adifferent issue then
14:20:55 <V453000> oh yeah now I got it too
14:20:59 <V453000> it keeps me in some area
14:23:18 <Samu> i could pan towards a direction, but then the cursor would show up outside the window openttd border
14:23:59 <Samu> releasing button would make the openttd cursor to disappear
14:25:14 <Samu> it's like there was two programs tracking cursor position
14:27:40 <Samu> it wouldn't retain it's position at the moment the button was pressed down, it would show up somewhere else
14:29:10 <Samu> let me explain again, it's becoming confusing
14:29:33 <Samu> 1 - hold down right click inside OpenTTD window
14:30:23 <Samu> 2 - while hold down, move mouse towards outside borders. the cursor is still in the same hold position in OpenTTD, but another cursor would appear outside oepnttd window
14:31:02 <Samu> I'm just not sure if it was still scrolling when it was outside the borders though
14:33:05 <Samu> 3 - upon releasing right click button, the cursor outside openttd was the true one with the real position, the other inside OpenTTD would disappear.
14:33:54 <Samu> not sure which version of OpenTTD this happened, but it was one or two days before beta2 release
14:37:09 <Samu> testing version was 32-bit
14:57:28 <frosch123> would explain why it only happens sometimes for V
15:02:07 <frosch123> hmm, actually i may have an idea to make it less reliant on exact behaviour
15:02:28 <frosch123> so it can wobble around win overwriting messages sometimes
15:39:07 *** sla_ro|master has joined #openttd
16:06:55 <Samu> sea 00, canal 01, river 10
16:07:40 <Samu> sea 001, canal 01x, river 101
16:09:31 <Samu> canal on river 011, canal on sea 010 hmm
16:16:51 *** zwamkat has joined #openttd
16:19:28 <Samu> sea tile, owner m1=x0010001, canal owner m6=(notacanal)
16:22:20 *** Myhorta has joined #openttd
16:23:57 <Samu> canal tile on sea or on land of company 15, owner m1=x0101110, canal owner m6=11xxxx10 (iscanal)
16:25:01 <Eddi|zuHause> that is complete gibberish.
16:29:12 <Samu> canal tile on sea or on land of owner none, owner m1=x0110000, canal owner m6=11xxxx11 (iscanal, isonsea, isnoowner)
16:31:43 <Samu> hmm, sorry, thinking out loud
16:36:19 <Samu> canal tile on sea or land 010
16:37:00 <Eddi|zuHause> i don't think this is the right place for these ramblings
16:45:31 <Samu> if canals could be owned by towns only, this would be much easier
16:49:17 *** Celestar has joined #openttd
16:50:37 <Samu> to identify a canal built on either river or sea/land, it would be preferrible if the canal owner was town
16:51:01 <Samu> this would allow me to store 3 owners
16:51:38 <Samu> only 2 at a given time, but with the possibility to reconstruct the 3rd owner
16:56:40 <Alberth> lots of people do teddy bear programming, eddi :)
16:57:41 <frosch123> also eddi has a beard, so it suitable to be a teddy bear
17:04:57 <Eddi|zuHause> 5 nights at Eddi's?
17:08:43 <peter1138> Someone wants my 21" CRT, right?
17:08:56 <peter1138> Any gamers still prefering CRTs? :p
17:13:09 <Samu> i don't have means to distinguish between company 1 and owner_none
17:13:11 <DorpsGek> Commit by frosch :: r27172 trunk/src/gfx.cpp (2015-02-28 17:13:07 UTC)
17:13:12 <DorpsGek> -Fix [FS#6238] (r27167): Apparently Windows randomly drops SetCursorPos calls. According to internet due to incorrect/missing synchronisation in threads accessing the event queue.
17:25:40 <Eddi|zuHause> i had a 21" CRT some years ago, and even carried it to a LAN party once
17:26:55 <Samu> ok it's possible after all, with some crude logic to determine owner
17:31:25 *** kais58 is now known as kais58|AFK
17:40:38 *** luaduck is now known as luaduck_zzz
17:42:59 *** JacobD88 has joined #openttd
18:10:44 <Samu> damn, there's 4 possible layers
18:26:28 <Samu> i thought a bare land tile had waterclass invalid
18:27:09 <Samu> they dont have waterclass apparently
18:29:06 <Alberth> it's land, why would it need a water class?
18:29:29 <frosch123> stuff that exists on both land and water has a water class
18:29:32 <frosch123> e.g. industry tiles
18:29:48 <frosch123> if they are removed they revert back to water or land depending on the stored water class
18:29:57 <frosch123> in that case the invalid-water-class counts as land
18:31:00 <frosch123> but plain land has no water class, just like they have no road bits :p
18:32:49 <Samu> it's just going to be quite messy
18:44:53 <heffer> planetmaker: fixed nml 0.4.0 source tarball in sight? :D
18:45:14 <planetmaker> ehm, would be 0.4.1 :)
18:49:33 <planetmaker> it doesn't fix the un-nice version in the filename though. Just that it should build at all ;)
18:52:42 <Eddi|zuHause> call it 0.4.0.1 in memory of openttd's version scheme :p
19:01:28 <heffer> planetmaker: looks good. seems to build without an error :D
19:02:15 *** Celestar1 has joined #openttd
19:04:51 <heffer> hmm nmlc gives me "ImportError: No module named 'nml'"
19:06:46 <heffer> i do python3 setup.py build first
19:07:11 <heffer> then python3 setup.py install -O1 --skip-build --root $RPM_BUILD_ROOT
19:07:33 <heffer> and it doesn't seem to install all files to the correct directory
19:14:13 <alluke> ffffffuuuuuuuucking fuck
19:14:17 <alluke> angerfist in helsinki today
19:24:12 <planetmaker> hm, so setup.py install fails, heffer?
19:25:12 <planetmaker> do you want to write me a patch? :D
19:32:26 *** FLHerne has joined #openttd
19:38:17 *** oskari89 has joined #openttd
19:39:37 <heffer> planetmaker: i'm afraid i won't be able to. I'm currently moving :D
19:42:51 <Eddi|zuHause> alluke: i don't know what an angerfist is. but maybe you should have planned ahead? :p
19:43:07 <Eddi|zuHause> alluke: but it's nice that of all the people on the whole internet, you come to us :)
19:44:59 <Samu> i'm thinking so slow today
19:46:48 <Eddi|zuHause> i have no desire to watch that video :p
19:49:03 *** luaduck_zzz is now known as luaduck
19:59:34 *** lobster has joined #openttd
20:05:35 *** JacobD88 has joined #openttd
20:09:52 <DorpsGek> Commit by frosch :: r27173 trunk/src/newgrf_debug_gui.cpp (2015-02-28 20:09:49 UTC)
20:09:53 <DorpsGek> -Fix [FS#6237] (r27134): Use the regular clipping functions in the sprite aligner instead of some magic.
20:12:51 <__ln___> is it now less magical?
20:29:43 *** oskari892 has joined #openttd
20:38:13 <Eddi|zuHause> now implement sensible alignment across zoom levels, and tracking of changes made?
20:38:58 <frosch123> albert is reviewing
20:38:59 *** Cybertinus has joined #openttd
20:39:24 <Supercheese> Eddi|zuHause: Not monitoring #openttd.dev?
20:39:41 <Supercheese> there was some discussion there
20:40:04 <Eddi|zuHause> there was something annoying about that channel
20:40:19 <Eddi|zuHause> i think it was that i need to register to get +v or so
20:40:39 <frosch123> assuming you would get +v :p
20:40:51 <Eddi|zuHause> i think i was offered +v :p
20:41:10 <Supercheese> Yeah there is a bit of hoopstuff to jump through
20:41:12 <frosch123> maybe just to laugh afterwards :p
20:42:08 <frosch123> anyway, this channel is quite spammy lately, so .dev becomes more active :p
20:42:24 <Supercheese> spam spam wonderful spam
20:42:48 <frosch123> it even shares 3 of 4 letters
21:03:03 *** Myhorta has joined #openttd
21:23:41 <Samu> how do i tell the game that owner with bits 01111 is not town, but either owner_none or owner_water?
21:24:35 <Samu> i wanna store owner_none and owner_water depending on that tile waterclass
21:25:13 <Samu> can i trick the game believing "this" owner is actually "that owner" and behave accordingly?
21:28:38 <Samu> i have to expose this problem clearer, it's becoming too complex without guidance
21:29:49 *** oskari89 has joined #openttd
21:30:46 <planetmaker> Samu: it sounds like total folly to try some kind of trickery in that area. Define the bits to mean *that* in the context of whatever. That's it
21:32:13 <Samu> basically, have a special rule in which all water tiles owner are stored as x1111, that x will be a variable I wanna use for other purpose
21:32:37 <Samu> instead of using 5 bits to store owner, use 4
21:34:26 <Samu> then retrieving owner of that tile wouldn't return owner_town, but it would be constructed to output the exact owner
21:36:40 <planetmaker> samu you really should get into actually doing some stuff and *trying*. Without sharing every thought on every line as it crosses your mind
21:37:05 <Samu> i'm building the bit matrix first
21:38:39 <Samu> yes, i wanna make sure if it is possible to have no bits conflcting with each other
21:38:59 <Samu> it's complex when not looking at it
21:39:59 <Samu> there's 3 possible layers on a given water tile
21:41:12 *** Celestar has joined #openttd
21:41:27 <Samu> each layer can store 2 owners
21:41:46 <Samu> i have 8 bits for owners, 4 at m1, 4 at m6
21:41:59 <Samu> then i have 1 more bit which must be shared by both
21:42:10 *** Sonny_Jim has left #openttd
21:42:18 <Samu> this is the one that's causing most trouble
21:43:43 <Eddi|zuHause> that will never work
21:43:45 <Samu> the third layer uses this sharing bit to identify the firstmost owner of the tile, and the owners are always either none or water, but i have to store them in the form of 4 bits
21:43:53 <Eddi|zuHause> you can't store 10 bits worth of data in 9 bits
21:44:18 <Samu> i can because, the owner of sea is always water
21:44:28 <Samu> the owner of river is always water
21:44:38 <planetmaker> then you do not need to store it
21:44:40 <Samu> the owner of canal, however... complicates
21:46:21 <Samu> i speak of 3 layers because i am also trying to distiguish canals built on river or on sea/dryland
21:47:10 <Eddi|zuHause> that needs 1 bit. why do you want 8?
21:47:13 <Samu> destroying canal reverts to either river, or sea/land
21:49:54 <Samu> can you open excel files?
21:51:33 <Samu> it is not final, nor correct yet
21:51:49 <Samu> i am still fiddling with bits
21:54:09 <Samu> destroying ship depot, reverts to previous layer
21:54:25 <Samu> destroying that layer, reverts to the one before
21:56:12 <Samu> the second is with an error
21:56:24 <Samu> or actually, it is as if owner_none was using 5 bits
21:59:05 <Samu> that line with 4 layers is misleading, i really have to work on this yet
22:00:14 <Samu> being have to revert canals to rivers instead of just getting rid of them, and at the same time know which owner owns what at a given tile
22:03:36 <Samu> that bit 4 at m1 will be of extreme importance, it will allow the game to restore it's first form. I can see this as being possible with 4 bits for identifying 1 owner, instead of 5.
22:04:06 <Samu> with 5, well... I don't know if it's doable, it probably is, but may become too complex
22:05:03 *** kais58|AFK is now known as kais58
22:08:55 *** HerzogDeXtEr has joined #openttd
22:29:56 *** oskari892 has joined #openttd
22:46:00 <Samu> dumb question, can there be rivers on sea?
22:53:35 <FLHerne> Samu: Sea doesn't flood rivers/canals
22:53:42 <FLHerne> If that's the question
22:56:55 <Samu> sorry, i was confusing myself
22:57:18 <Samu> thinking that if removing a river tile that was planted on sea would revert back to sea
23:00:06 <glx> if you remove it it will be flooded
23:01:34 <Samu> that means i can treat rivers like they're the base
23:01:43 <Samu> simplifies this even further
23:01:56 <Eddi|zuHause> river or canal on sea level prevents the sea from flooding
23:02:08 <Eddi|zuHause> so you can have land on sea level that way
23:03:54 <Samu> I see, but I was thinking of something related to ownership
23:04:39 <Samu> don't have to store river owners
23:04:50 <Eddi|zuHause> i don't know what you try to achieve here. there's no difference if the river was built on land or at sea level. river is river
23:05:25 <Samu> there's no 4th layer to take care off
23:05:29 <Eddi|zuHause> excuse my scepticism about this.
23:07:36 <Samu> thus in turn, owner_none becomes the only exception rule, I don't have to care about owner_water, this simplifies it
23:33:06 *** ChanServ sets mode: +v tokai
23:51:36 <Samu> now i got to repeat, but for the case station tile, then later, for the case industry tile
continue to next day ⏵
