IRC logs for #openttd on OFTC at 2018-03-19
⏴ go to previous day
00:08:28 *** Eddi|zuHause2 has joined #openttd
00:12:51 *** Eddi|zuHause2 is now known as Eddi|zuHause
00:32:49 <TrueBrain> import of issues done
00:33:06 <TrueBrain> of which 158 marked as 'bug'
00:33:11 <TrueBrain> euh, first bugs -> issues
00:34:02 <TrueBrain> @calc 78 + 36 + 31 + 46 + 139
00:34:09 <TrueBrain> @calc 330 * 60 / 6686
00:34:09 <DorpsGek> TrueBrain: 2.96141190547
00:34:12 <TrueBrain> 3 seconds per issue
00:34:15 <TrueBrain> total of 330 minutes
00:34:21 <TrueBrain> so indeed, bit over 5 hours
00:46:44 *** Wacko1976 has joined #openttd
00:50:53 *** Thedarkb has joined #openttd
01:28:12 *** supermop has joined #openttd
02:36:23 *** Smedles has joined #openttd
02:39:23 *** Eddi|zuHause has joined #openttd
02:44:04 *** Smedles_ has joined #openttd
02:45:16 *** Eddi|zuHause has joined #openttd
03:29:45 *** supermop has joined #openttd
03:54:07 *** ToffeeYogurtPots has quit IRC
03:54:25 *** ToffeeYogurtPots has joined #openttd
04:29:40 *** LANJesus has joined #openttd
04:35:01 *** supermop has joined #openttd
04:51:58 <LANJesus> COAN codebase scares me. looks like someone butchered some C into C#
05:26:16 *** SeraphAngel has joined #openttd
07:30:10 *** supermop has joined #openttd
07:54:59 *** ToffeeYogurtPots has quit IRC
07:55:29 *** ToffeeYogurtPots has joined #openttd
08:56:47 *** Biolunar has joined #openttd
09:10:18 *** ToffeeYogurtPots_ has joined #openttd
09:10:38 *** ToffeeYogurtPots has quit IRC
10:41:33 *** tycoondemon has joined #openttd
10:43:18 *** tycoondemon has joined #openttd
10:59:57 *** triolus[m] has joined #openttd
11:13:09 *** WWacko1976-work has joined #openttd
11:42:08 <Eddi|zuHause> how is it that when a game comes with an OST, the one song i'm interested in is not part of that OST?
11:45:11 <Eddi|zuHause> ... spent half the night trying to find a .hpk unpacker that runs on linux, ended up installing freepascal and commenting out lines from a windows command line tool written in delphi, until it compiled
11:51:01 <Sacro> is there no fuckit for delphi?
11:51:20 <Eddi|zuHause> what's a fuckit?
11:52:27 <Eddi|zuHause> delphi isn't an interpreted language
11:52:28 <Sacro> Bascally just hotpatches the code to ignore any errors
11:53:43 <Eddi|zuHause> but last time i programmed delphi was like 15 years ago
11:56:03 <Eddi|zuHause> well, the errors were mostly of the kind of missing referenced libraries that i figured i don't actually need for this particular case
11:56:34 <Eddi|zuHause> or some file called "unit1.pas" that wasn't included in the source
11:57:24 <Eddi|zuHause> and then all lines that failed on undeclared identifiers from those...
11:59:05 <Eddi|zuHause> oh, and find the -Mdelphi switch for freepascal
12:06:22 <Eddi|zuHause> Sacro: everything on that page screams "OMG is this horrible!" to me
12:13:03 <_dp_> lol, I just went from reading mypy docs to fuckit ones
12:13:59 <SpComb> which one is the better choice
12:14:32 <_dp_> I'll stick with mypy for now
12:15:03 <_dp_> fuckit doesn't seem to be a wise choice for financial code :p
12:15:17 <Sacro> It's for your frontend too
12:17:53 <_dp_> js is shitty enough even without it :p
12:18:38 <Sacro> I'm a JS dev and I resemble that remakr
12:19:16 <Eddi|zuHause> _dp_: fuckit doesn't seem to be a wise choice for any code that you intend to run more than once
12:19:56 <Sacro> But for that once when you *really* want to plough through regardless
12:20:10 <Sacro> Like the "Pass signal at red" button in TTD
12:20:41 <_dp_> Eddi|zuHause, ... even once
12:21:14 <Eddi|zuHause> yeah, even pressing that button once can result in a fireball
12:21:54 <Eddi|zuHause> Sacro: i still much rather plough through the code and consider what the code was meant to do before commenting it out
12:22:24 *** Wacko1976 has joined #openttd
12:34:44 <Sacro> Eddi|zuHause: and I'd rather automate that :)
12:39:55 <peter1138> Why is it not lunch time yet.
12:40:04 <peter1138> Also, damn, I just missed me o'clock.
12:46:02 <Eddi|zuHause> i actually did notice you o'clock about an hour ago
12:48:27 *** FLHerne has joined #openttd
13:11:15 <peter1138> UTC is the only one that matters.
13:13:37 <__ln__> then you should be called peter1138utc
13:19:18 *** Thedarkb has joined #openttd
13:22:10 <peter1138> Anyway, nearly food time.
13:27:56 <__ln__> anyway, what language will we choose as the preferred channel language after brexit?
13:40:18 *** snail_UES_ has joined #openttd
14:15:51 *** tokai|noir has joined #openttd
14:15:51 *** ChanServ sets mode: +v tokai|noir
14:35:45 *** TheMask96 has joined #openttd
15:29:53 *** supermop has joined #openttd
15:50:28 *** sla_ro|master has joined #openttd
15:55:31 *** SeraphAngel has joined #openttd
16:01:13 *** Wacko1976 has joined #openttd
16:50:52 *** Wormnest has joined #openttd
16:56:46 *** Alberth has joined #openttd
16:56:46 *** ChanServ sets mode: +o Alberth
17:20:31 *** WWacko1976-work has quit IRC
17:48:44 *** andythenorth has joined #openttd
17:50:33 <andythenorth> haven't tested yet
17:50:49 <Alberth> me neither, but nice that it appears like this
17:51:40 <andythenorth> a lot of FIRS 3 was unseen rewriting of code to pure python, no C pre-processor
17:51:50 <andythenorth> which makes it easier for people to hack FIRS
17:52:02 <andythenorth> I'm encouraging forks
17:52:13 <andythenorth> as long as they're hands-free, i.e. no support :P
17:53:09 <Alberth> that seems to work here too :)
17:53:27 <andythenorth> so anyone here done any fuzz testing / chaos monkey ever?
18:20:22 *** Progman has joined #openttd
18:20:27 *** Thedarkb has joined #openttd
18:33:03 <Sacro> andythenorth: I click a lot of things
18:36:55 *** HerzogDeXtEr has joined #openttd
18:41:50 <andythenorth> are you repeatable? :P
18:41:59 <andythenorth> do you log all your actions? o_O
18:46:43 *** Wacko1976 has joined #openttd
18:48:28 <Wolf01> Back from the lego weekend
18:49:07 <Wolf01> andythenorth: I've some pictures of model ships which are really cool
18:52:37 <Wolf01> 6000m2 of lego exposition, 2 days to take pictures of everything S_S
18:59:29 *** frosch123 has joined #openttd
19:00:32 <Alberth> @calc 3000 / (8*60*60)
19:00:32 <DorpsGek> Alberth: 0.104166666667
19:01:35 <Eddi|zuHause> not sure if that is a helpful figure :p
19:01:50 <Alberth> probably not, but it's fun :p
19:14:28 *** Biolunar has joined #openttd
20:11:39 <LordAro> Wolf01: more like when 2 towns merge together
20:39:53 *** Thedarkb has joined #openttd
20:43:17 <nielsm> hmm I get instant crash when selecting hebrew as language
20:43:25 <nielsm> in the ParagraphLayouter code
20:44:05 <nielsm> oddly didn't have problems when I tried Arabic (Egypt)
20:44:27 <frosch123> icu is completely broken on windows lately
20:44:51 <frosch123> i have no idea whether due to new windows, new fonts, or from a compile farm update 3 years ago
20:44:57 <nielsm> uhh... japanese sure selects a huge font size
20:45:11 <nielsm> the game options window can't fit on my 1080p display
21:00:39 *** Stimrol has joined #openttd
21:10:04 *** ATMunn is now known as Guest1010
21:20:22 <nielsm> frosch123 can't be compile farm, I don't think, since this is my own compile ;)
21:21:13 <glx> icu is a monster anyway ;)
21:21:37 <nielsm> although yes I do use the ICU from the dependencies package download offered
21:22:09 <LordAro> frosch123: needs new compile farm
21:26:06 <nielsm> so, when's that new github repos coming online? I want to make some pull requests :D
21:27:12 <LordAro> not for at least a week, probably
21:28:36 <nielsm> okay the hebrew crash happens inside ICU, not in OTTD code
21:28:56 <LordAro> doesn't mean it's not OTTD's bug :p
21:29:01 <nielsm> the call from OTTD causing the crash seems to be correct
21:30:33 <frosch123> nielsm: there are about a dozen fs tasks about those crashes
21:30:45 <frosch123> they started like 2 years ago
21:30:51 <nielsm> even better: seems to be random!
21:30:57 <nielsm> no changes and it didn't crash this time
21:31:38 <nielsm> but then did once I clicked around some more
21:32:16 <frosch123> that is common for invalid reads and writes
21:32:43 <frosch123> only pure null pointer usages are reproducible
21:34:08 <LordAro> probably worth looking through ICU release notes for the last few years, see if anything looks relevant
21:34:30 <LordAro> alternatively, implement a pango layer for the text layout engine, and upgrade it :p
21:35:05 <glx> and all its dependancies ?
21:37:10 <orudge> TrueBrain: ooh, github. I've claimed orudge@openttd.org. ;) If you're adding folk to the project you can add me there.
21:43:37 *** andythenorth has joined #openttd
21:44:04 *** Thedarkb has joined #openttd
21:50:49 <glx> orudge: I tried something but I don't know if it worked
21:52:51 <orudge> glx: it seems to have worked, thanks
21:55:28 <TrueBrain> frosch123: I am going to leave it to you who you want to add as Owner of the OpenTTD organization, and who in the Developer Team
21:55:40 <TrueBrain> both give enough permissions to do what-ever on the OpenTTD repository
21:55:45 <TrueBrain> Owner gives more rights ;)
21:55:57 <TrueBrain> I am unsure it is wise to make every Developer an Owner, as hitting the wrong button is an easy mistake
21:56:25 <glx> wrong buttons are written in red at least ;)
21:57:21 <andythenorth> is the 'add coin-mining malware' button easy to press?
21:57:23 <frosch123> well, i will rely on people to hint at their account :)
21:57:49 <LANJesus> wait, openttd is switching to git/github?
21:58:00 <TrueBrain> frosch123: I was talking about owner vs developer, not adding in general :)
21:58:17 <frosch123> TrueBrain: anyway, switch before or after 1.8.0 tag?
21:58:31 <frosch123> mostly depends on switching the farm, so maybe after?
21:58:34 <TrueBrain> is 1.8 on schedule?
21:58:40 <LordAro> after is probably easier
21:58:58 <frosch123> the branch is done, i wonder which titlegame to add
21:59:12 <TrueBrain> if on schedule, after; that gives us another year to iron out any bugs
21:59:20 <TrueBrain> would be a shame to delay 1.8 because of this
21:59:25 <TrueBrain> also a good start of 1.9
22:00:49 <frosch123> but we can more fs earlier
22:00:59 <frosch123> so we can already redirect people there
22:01:00 <peter1138> Hmm, I already have peter1138@openttd.org claimed, somehow.
22:01:42 <TrueBrain> frosch123: indeed; well, next weekend is the earliest option, but we need people to validate the repo first. I am running a new Issue import as we speak .. tomorrow the git content will follow ..
22:03:38 <peter1138> What's the plan anyway? I've not seen anything written :p
22:04:16 <glx> the main goal is to take advantage of pull requests
22:04:19 <peter1138> There's probably some long forgotten IRC channel that I, er, forgot about.
22:04:27 <peter1138> Anything changing with SVN?
22:04:36 <frosch123> nothing in that irc channel
22:05:20 <frosch123> github will become main repo, svn will be readonly. git will be fast-forward merge only
22:05:53 <frosch123> newgrf and bananas version will drop the revision number
22:06:22 <LordAro> worked out how nightlies will be versioned?
22:06:25 * andythenorth wonders who will do first pull request :P
22:06:41 <andythenorth> probably needs to be something already on github, right? o_O
22:06:41 <LANJesus> version using tags and ISO date code
22:06:57 <frosch123> LordAro: ingame versions will display as "YYYYMMDD-gXYZ"
22:06:59 <LordAro> andythenorth: i could probably come up with something pretty quickly
22:07:10 <LANJesus> nightly-20180319 or so
22:07:22 <LordAro> frosch123: only XYZ? or a longer part of the hash?
22:07:32 <LANJesus> don't need the hash if using tags
22:07:45 <frosch123> LordAro: whatever does not look silly :)
22:07:59 <LordAro> glx: well, 8 is default, until you have enough commits that you start getting conflicts :p
22:08:06 <LANJesus> what is the potential for having more than one nightly in 24hr period?
22:08:12 <LordAro> LANJesus: tag per nightly seems unnecessary
22:08:27 <frosch123> unless you want to break the github gui :p
22:08:38 <peter1138> LANJesus, when the clocks change ;)
22:08:54 <nielsm> would be more reasonable to name the nightly builds after the revision?
22:09:16 <nielsm> both date and revision
22:09:33 <LordAro> 21:06:57 < frosch123> LordAro: ingame versions will display as "YYYYMMDD-gXYZ"
22:09:34 <peter1138> 21:06 < frosch123> LordAro: ingame versions will display as "YYYYMMDD-gXYZ"
22:09:36 <LANJesus> the revision should be master or develop (depending on branching strategy)
22:09:42 <nielsm> just to make it clear, in case someone was pushing changes just around build time
22:10:06 *** sim-al2 has joined #openttd
22:10:11 <LANJesus> nielsm: what do you mean by revision?
22:10:21 <glx> we usually don't commit at nightly time :)
22:10:41 <LANJesus> does openttd have CI currently?
22:11:16 <andythenorth> bears, woods etc
22:11:30 <LANJesus> cool. excuse my ignorance. i've not poked at the openttd official dev stuff in a decade or so
22:11:47 <LANJesus> i've made some branches on github from OpenTTD/OpenTTD
22:11:49 <frosch123> it already had CI 10 years ago :p
22:12:00 <LANJesus> you're expecting me to remember 10 years ago : P
22:12:04 <TrueBrain> it has a CI since 2004 ...
22:12:09 <peter1138> git.openttd.org staying?
22:12:17 <LordAro> LANJesus: you'll need to rebase, that's the old repo with old hashes
22:12:52 <LANJesus> the old repo is still the active repo, yes?
22:13:03 <peter1138> hg.openttd.org? :-)
22:13:10 <LANJesus> then i'll burn that bridge when i get to it ; )
22:13:18 <peter1138> Are those poor mercurial users left out?
22:13:21 <TrueBrain> all VCSes on openttd.org will go read-only peter1138 :)
22:13:21 <peter1138> (Was there any left?)
22:13:30 <andythenorth> mercurial lost :)
22:13:35 <LordAro> peter1138: Alberth & planetmaker, i think :p
22:13:44 <andythenorth> I use it for newgrfs
22:13:46 <LANJesus> git makes it easier to shoot yourself in the foot
22:13:54 <andythenorth> I keep spelling 'hg' as 'git' though :P
22:13:57 <LANJesus> and unshoot yourself after a few googlings, since everyone uses it
22:14:14 <peter1138> I need to build some CI system for work :S
22:14:21 <andythenorth> you just delete the repo if *anything* didn't work as expected
22:14:38 <peter1138> There's a gitea system.
22:14:52 <peter1138> As raw git is slightly painful when your users don't know what ssh is.
22:14:52 <andythenorth> it was years before I dared do a merge in hg :P
22:15:22 <LANJesus> i frequently use sourcetree for a git frontend
22:15:40 <LANJesus> and visual studio's git frontend is getting better. i still find it clunky though
22:17:55 <LANJesus> in git, any commit can turn into a merge
22:21:03 <TrueBrain> frosch123: if this issue import round succeeds, the import scripts are all done and automated
22:21:18 <TrueBrain> found a nice bug in the issue import script .. URLs to attachment with a space didnt work :P
22:21:21 <TrueBrain> but that is fixed now :)
22:21:42 <TrueBrain> PR policy on GitHub can be applied, and works fine
22:21:44 <glx> I saw you closed an issue manually too
22:21:59 <TrueBrain> the thing I am not sure yet what the best approach is, either setting up a new compile farm and linking it to docker, or reusing the current
22:22:15 <TrueBrain> glx: in the last import, yes; in the current I did not touch a thing :)
22:22:20 <TrueBrain> (still has 3k tickets to go)
22:22:40 <frosch123> the docker farm is more uptodate for linux and osx
22:22:44 <TrueBrain> I am very tempted to start a new compile-farm; new Bamboo integrates a lot better with git
22:22:50 <glx> oh I didn't noticed the "spam" ;)
22:22:51 <TrueBrain> yeah .. Windows is the biggest PITA there
22:22:54 <frosch123> so if we can already switch those, that would be great
22:23:22 <TrueBrain> if we move after release, there is more time fixing Windows after migration
22:23:34 <TrueBrain> either way, next weekend I hope to experiment a bit with it
22:23:52 <TrueBrain> also gives us some time to create fake PRs, and see how everything interacts
22:25:29 <TrueBrain> but for now, good night! we will see tomorrow if the import survived :D
22:33:16 <peter1138> "Your bike is currently in a build queue, awaiting assembly. It joined the queue today and this process can take up to 5 days, after which it will be sent to you."
22:34:24 <Wolf01> I purchased a new mattress instead :P
22:35:19 <Wolf01> Nah, 3 types of foam :D
22:43:43 <LordAro> peter1138: what'd you get?
22:45:46 <peter1138> Wasn't particularly looking for another Cube, just this was on sale which is rare to find in XXL sizes.
22:46:41 <peter1138> It's mostly matt actually, heh./
22:46:48 <peter1138> I might stick some reflective crap all over it.
22:47:02 <peter1138> I remember back in the day when you could actually just choose the colour of your bike.
22:47:27 <peter1138> 2x11 is a bit oldschool these days for MTBs but it'll do.
22:58:50 *** supermop has joined #openttd
23:17:47 *** andythenorth has left #openttd
23:25:28 *** CompuDesktop has joined #openttd
23:28:46 <supermop> love these helpful error messages: "[Knmlc ERROR: nmlc: An internal error has occurred: nmlc-version: unknown Error: (AssertionError) . Command: ['nmlc', 'moprv64.nml'] Location: File "nml\output_base.py", line 166, in prepare_dword"
23:30:08 <LordAro> supermop: i recommend not doing that
23:30:59 <supermop> LordAro: recommend not getting weird messages?
23:31:58 <supermop> just added a new recoloring for new layer of a sprite stack, and i am sure i did it wrong, but that error message doesn't really indicate that
23:36:26 <glx> if -0x80000000 <= value < 0: value += 0x100000000
23:36:26 <glx> assert value >= 0 and value <= 0xFFFFFFFF
23:36:54 <glx> the error is in the assert
23:38:36 <glx> so a value in your file is too big to be an unsigned double
23:39:53 <supermop> that's the only thing changed since it last worked
23:40:17 <supermop> trying to figure out how to rewrite my stack to use different recoloring per layer
23:40:41 <glx> I can't help more, I know nothing about nml
23:42:04 <supermop> no problem, but nothing in the switch seems like it would have much to do with dwords
23:42:17 <supermop> as compared to its prior wording
23:42:33 <supermop> so i'm just going to blindly change it till it works
23:42:52 <glx> getbits probably use dword
23:51:29 <supermop> nml specs state that you can set a separate recoloring for each iteration of the sprite stack
23:51:56 <supermop> but it's not clear to me how so i am just guessing where to put that information
23:52:14 <supermop> "When enabled, sprites are resolved multiple times while incrementing an iteration number, that can be read via getbits(extra_callback_info1, 8, 8) Currently this is limited to at most 4 sprites per articulated part. In addition you need to set register 100 as additional result: STORE_TEMP(CB_FLAG_MORE_SPRITES | recolouring, 0x100) if there are more sprites to draw. STORE_TEMP(recolouring, 0x100) if there are no more sprite
23:54:44 <glx> hmm maybe it doesn't like getbits() < 0
23:56:08 <glx> or if this test pass it's always false
23:56:27 <glx> the the result is always 1
23:56:38 <supermop> that compiles without error
23:56:45 <supermop> lets see if works in game
continue to next day ⏵
