IRC logs for #openttd on OFTC at 2020-04-22
⏴ go to previous day
00:00:30 <Xaroth> nml/util.py in my branch
00:02:09 <FLHerne> Oh, it's part of the big f-string commit
00:02:26 <FLHerne> (please don't merge it upstream like that!)
00:02:27 <Xaroth> I can split it off at some point, it's just a giant WIP branch
00:02:46 <Xaroth> it doesn't do much atm it seems, speed-wise
00:06:35 <FLHerne> Xaroth: Silly question, are you running it with nfo output?
00:06:51 <Xaroth> just running the regression suite atm
00:07:14 <andythenorth> I ran both nfo output and regression suite
00:07:37 <Xaroth> grab my last ammend, andythenorth, there was a flaw in my util.py. it should be fixed now
00:08:15 <Xaroth> 23.2s is the time to beat now
00:08:32 <Xaroth> 23.9, so slightly faster
00:10:25 <FLHerne> I don't see why grf writing wouldn't be affected also, but performance with `nml --nfo=<filename>` is the original issue
00:12:00 <andythenorth> make test times are unchanged with latest commit
00:12:09 <andythenorth> but Iron Horse nml -> nfo is 5s faster
00:12:32 * Xaroth sneaks in a fix for something he overlooked
00:12:48 <Xaroth> right, that branch should be testable now
00:12:50 <andythenorth> if I was squinting and being super-optimistic, the fastest run now is 50% quicker than the slowest previous run
00:13:43 <andythenorth> being strict it's about 41% faster on average
00:14:02 <andythenorth> it's getting the same 'Writing output' times as FLherne's patch
00:15:17 <Xaroth> andythenorth: those on the py3.6 branch I made or the clean pypy-stringbuilder branch?
00:15:59 <Xaroth> the pypy-stringbuilder branch _only_ has the stringbuilder change
00:17:26 <andythenorth> yeah pypy-stringbuilder is consistently fast
00:17:48 <andythenorth> I'd say the fastest, but we're talking 0.1-0.5s differences from the 3.6 branch or the FLHerne patch
00:18:06 <andythenorth> 40% is definitely faster
00:21:44 <andythenorth> thanks everyone who has looked at nml stuff recently :)
00:21:54 <andythenorth> content keeps the game alive
00:22:03 <DorpsGek_III> [OpenTTD/nml] Xaroth opened pull request #98: Change: Use PyPy's StringBuilder to speed up the use of StringIO. https://git.io/JfkCI
00:22:40 <Xaroth> one for when he wakes up then.
00:31:12 *** glx is now known as Guest23004
03:01:39 *** ChanServ sets mode: +v tokai
03:44:37 *** Smedles has joined #openttd
04:34:52 *** Wormnest_ has joined #openttd
05:15:47 *** Smedles has joined #openttd
06:02:38 *** mindlesstux6 has joined #openttd
08:03:36 *** sla_ro|master has joined #openttd
08:21:20 *** andythenorth has joined #openttd
08:58:49 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on pull request #98: Change: Use PyPy's StringBuilder to speed up the use of StringIO. https://git.io/JfkgX
09:06:51 *** Smedles has joined #openttd
09:24:44 *** Smedles has joined #openttd
10:03:07 *** rotterdxm has joined #openttd
10:04:29 * peter1138 mumbles about the price of unbuffered DDR3 ECC :s
10:08:50 <andythenorth> costs more than your eye teeth?
10:12:21 <andythenorth> I never paid much attention, but AIUI frosch though nml used a lot of memory relative to what it should be doing
10:17:33 *** gelignite has joined #openttd
10:18:09 *** andythenorth_ has joined #openttd
10:20:14 <Xaroth> Would be fun to check out
10:20:28 <Xaroth> Also to see how #98 affects memory usage
10:26:51 <andythenorth_> not sure how to reliably track memory use of a python process
10:34:54 <Xaroth> mprof seems to do something
10:35:13 <Eddi|zuHause> try to compile one-file-CETS and see if it still explodes?
10:44:01 <planetmaker> @LordAro, aye, I shall see that I can get that API key done. If not done by end of week, please remind me again.
10:49:14 <Xaroth> andythenorth_ I can check a bit of it with mprof
10:49:20 <Xaroth> running some comparisons between python versions atm
10:50:40 *** hythlodaeus has joined #openttd
10:56:41 <LordAro> planetmaker: sorry, out of date now - i added openttd user to the nml package and TB added the key
10:57:11 <planetmaker> a new user or additional login?
10:57:33 <planetmaker> or rather: will there now be 2 nml or not?
10:58:04 <LordAro> still just the one nml
10:58:15 <LordAro> just added a new "openttd" user as an owner of the package
10:58:53 <planetmaker> ok... I don't get it. There's two e-mails registered to the pypi account...? But not an openttd one?
11:00:46 <planetmaker> ah, ok... now I do :) nvm. Yes, that's a good solution
11:10:08 <andythenorth_> Xaroth thanks :) there are clearly changes from the context manager, but I'm not sure they tell us much :)
11:14:03 <Xaroth> They all are within the white noise region
11:14:08 <Xaroth> but no strong changes, which is good
12:37:46 *** sla_ro|master2 has joined #openttd
12:40:01 <peter1138> I opened that link but the last comment is from 2019...
12:40:42 <LordAro> is github still having issues?
12:40:51 <LordAro> i got some emails with the text in it...
13:27:56 *** snail_UES_ has joined #openttd
13:28:50 *** WormnestAndroid has quit IRC
13:29:02 *** WormnestAndroid has joined #openttd
13:36:34 <nielsm> those comments are on a random line of code
13:37:01 <nielsm> so they are appended to a review thread github by default collapses
13:45:11 *** WormnestAndroid has quit IRC
13:45:27 *** WormnestAndroid has joined #openttd
14:17:52 <hythlodaeus> i think this is one of the default linux fonts, but i'm not sure
14:28:30 <nielsm> that looks like the bitmap font in opengfx
14:30:26 <hythlodaeus> it's a bitmap? not a truetype?
14:30:58 <nielsm> the baseset graphics for ottd (and the original game) contain a bitmap font
14:32:03 <hythlodaeus> hum, if it's part of the base graphs there might be trouble. I was wondering if you guys would be interested in having that font expanded to include russian/greek/hebrew characters
14:32:40 <hythlodaeus> so it wouldn't have to go and fish those fromother system fonts
14:33:57 <nielsm> I think if you are not using truetype fonts in your configuration, but select a game translation that requires characters not in the bitmap font, it automatically switches to some random truetype font
14:34:12 <hythlodaeus> yeah that's what i'm saying
14:34:32 <hythlodaeus> so my proposal here is to expand the amount of characters on this bitmap font
14:34:36 <nielsm> and that's *all* text that gets switched
14:34:59 <hythlodaeus> because it's something that's very easy to do and I've done myself for other games in the past
14:35:32 <hythlodaeus> so it would make this font also compatible with greek/russian/hebrew
14:35:53 <nielsm> I have personally toyed with the idea of having a good default TTF ship with ottd, which also contains vector versions of the various text-icons (like train, bus, etc. symbols)
14:36:13 <hythlodaeus> well, i might also be able to arrange for that
14:36:33 <hythlodaeus> although not myself doing that personally. I know a couple of font designers
14:37:03 <hythlodaeus> and I also have an idea of a ready-made font we could use as a starting base
14:39:27 <hythlodaeus> we could use it and upgrade it with additional chars.
14:41:01 <hythlodaeus> but how would you feel if I were to upgrade the mono bitmap with additional characters?
14:41:26 <hythlodaeus> would it disrupt compatibility with original TTD graphics?
14:41:42 <nielsm> there are some issues in that every baseset needs to ship at font, and there are several basesets
14:42:08 <nielsm> original baseset will need an extension grf with the extra sprites
14:42:24 <nielsm> those can possibly also be used for fallback for other basesets that haven't been extended with it yet
14:43:05 <nielsm> yeah there's just that newgrf action for it actually
14:43:26 <nielsm> so basesets can add glyphs for any characters, and there doesn't have to be any parity between them
14:44:07 <hythlodaeus> so it would require the creation of a newGRF basically
14:44:26 <hythlodaeus> well that's not so bad
14:44:32 <nielsm> that's how it already works
14:45:05 <hythlodaeus> i might expand the mono bitmap then
14:45:16 <nielsm> remember there's four bitmap fonts
14:45:28 <nielsm> normal, small, large (serif), and monospace
14:46:25 *** supermop_Home_ has joined #openttd
14:46:36 <hythlodaeus> what are large and monospace used for?
14:46:48 <nielsm> monospace is displaying readme files
14:53:19 <hythlodaeus> Yeah I think I should be able to do it
14:53:51 <hythlodaeus> if you'd like a truetype I can arrange someone to work on it too, like i mentioned
14:57:09 <nielsm> some well-hinted truetype glyphs of the various icons would be really great to have IMO
14:57:38 *** andythenorth_ has left #openttd
14:59:33 <hythlodaeus> for the glyphs we should be able to fish compatible dingbats from freefonts
14:59:52 <hythlodaeus> unless you really want custom dingbats but
15:00:24 <hythlodaeus> there should be plenty of freely licensed dingbats picturing trains, boats and plains
15:04:10 <hythlodaeus> curious now, are we loading the glyphs as plain pictures or as UTF-8 dingbats?
15:17:16 <hythlodaeus> hmm they show up as emojis in my browser
15:17:49 <hythlodaeus> but if I type these unicode symbols in my notepad they have chars of their own
15:18:07 <hythlodaeus> meaning there's free fonts we can easily load these from
15:22:39 <LordAro> those sentences don't quite make sense
15:23:06 <LordAro> "they show up as emojis in my browser" "if I type these unicode symbols in my notepad they have chars of their own"
15:23:11 <LordAro> that's what emojis are
15:40:00 <supermop_Home_> hopefully the chars are somewhat similar, and not like train = high heeled shoe
15:42:05 <supermop_Home_> though i'd be down with being able to name my company *minidisc*
15:42:41 <supermop_Home_> I wonder if my minidisc Walkman still works
15:51:41 <Samu> I think I managed a breakthrough on my canal pathfinder! Now takes half the time to pathfind. turns out _GetDirection callback plays an important role
15:52:17 <Samu> there were many directions, causing "collisions" to happen much less frequently (ie more paths to search)
15:52:25 <Samu> I reduced the number of directions
15:53:07 <Samu> now I'm checking if this is fine, or if it breaks something
16:44:15 *** supermop_Home_ has quit IRC
16:46:03 *** supermop_Home has joined #openttd
17:12:15 <hythlodaeus> LordAro: what I mean is whether the current font in the browser displays them as emojis or as color emojis or as dingbats
17:16:07 <hythlodaeus> where can i find the small font file exactly
17:18:40 <hythlodaeus> along with monospace
17:23:06 <nielsm> the (western european) small font is part of the base-baseset, along with the large font, the monospace font is entirely an ottd extension
17:23:39 <nielsm> though mono and the existing (minor) character set extensions are also part of the baseset, just as extensions so to say
17:29:36 <planetmaker> hythlodaeus, get OpenGFX source and look at it. It is reasonably documented
17:30:01 <hythlodaeus> I am looking at it ahah
17:30:09 <planetmaker> some grep on the source files should reveal mono, small, big and normal font sprites and locations
17:30:31 <planetmaker> (maybe it's called large and not big... not sure. There might also be tiny)
17:31:24 <nielsm> they seem to be somewhat mixed and split between files
17:31:48 <planetmaker> yes. There's a base set of glyphs which TTD already had. And all other glyphs are in the extra file
17:32:22 <planetmaker> Especially those which accomodate non-English, but latin-script languages with unusual characters with accents etc
17:32:25 <hythlodaeus> you guys made monospace and large from liberation mono and liberation serif repectively right? at least the readme confirms that abot monospace
17:33:56 <planetmaker> if the readme/docs say so... it probably is right then :)
17:34:19 <planetmaker> dunno who actualy made the glyphs. Zephyris?
17:35:07 <hythlodaeus> i was just wondering which process you guys used so i could replicate it for additional characters
17:35:43 <hythlodaeus> bc liberation mono/serif has additional alphabets too
17:36:28 <planetmaker> My best guess is: look at the source/docs who actually made the glyphs and ask that guy/gal
17:36:37 <planetmaker> If he is still around :|
17:36:42 <planetmaker> it's 10 years ago
17:55:41 <Wolf01> Bad day is bad... headache for the entire day
17:59:59 *** ChanServ sets mode: +v tokai
18:12:38 <FLHerne> > Improve performance of io.StringIO(). It should now be faster than CPython in common use cases.
18:12:53 <FLHerne> I think they might have just made your patch redundant :P
18:13:48 <FLHerne> (testing with 7.3.1 on Linux, I can't measure a difference between the current code and the patched version)
18:13:56 <FLHerne> Retrying again in case I screwed it up
18:16:29 <FLHerne> Nope, still no difference, without your patch is as good as either patch was before
18:16:40 <FLHerne> Odd, though, wasn't andy using 7.3?
18:21:38 <FLHerne> He says 7.3 in the github comment
18:22:32 <FLHerne> Hm, I'm really confused by pypy's changelog
18:22:58 <FLHerne> That's in the "7.3.0+" section, which is separate from both "7.3.0" and "7.3.1+"
18:28:42 <FLHerne> Yes, it's merged between 7.3.0 and 7.3.1
18:33:14 <DorpsGek_III> [OpenTTD/nml] FLHerne commented on pull request #98: Change: Use PyPy's StringBuilder to speed up the use of StringIO. https://git.io/JfkNF
18:55:44 *** Progman has joined #openttd
18:56:36 *** frosch123 has joined #openttd
19:17:42 *** andythenorth_ has joined #openttd
19:27:35 <Xaroth> nothing we can't fix with a version check I guess
19:32:32 <FLHerne> I mean, it re-raises the question of how much we need it
19:32:50 <FLHerne> If anyone who cares can just use up-to-date PyPy
19:34:02 <FLHerne> My other concern was whether someone might try to use the file in some other way
19:34:20 <FLHerne> In which case calling it `StringIO` might be misleading, because it really isn't
19:34:26 <andythenorth_> I read 7.3.1 changelog last night, but didn't upgrade, got too late :)
19:35:05 <Xaroth> We mimic the StringIO interface though
19:35:20 <andythenorth_> is this a 'move on, nothing to see', or is there a worthwhile change?
19:35:31 <Xaroth> 7.3.1 fixes the io.StringIO issue
19:36:26 <Xaroth> so the PR could be closed completely, or adapted to work with < 7.3.1
19:36:32 <Xaroth> your call, andythenorth_
19:36:54 <andythenorth_> we don't need to support older pypy, it's fast moving target
19:37:06 <Xaroth> then you can close the PR :)
19:37:13 <andythenorth_> let's fix other things :)
19:37:44 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on pull request #98: Change: Use PyPy's StringBuilder to speed up the use of StringIO. https://git.io/JfkxD
19:37:44 <DorpsGek_III> [OpenTTD/nml] andythenorth closed pull request #98: Change: Use PyPy's StringBuilder to speed up the use of StringIO. https://git.io/JfkCI
19:37:53 <FLHerne> Xaroth: StringIO inherits from TextIOBase, so it has quite a large interface; I don't think StreamBuffer or your patch implements most of it?
19:38:08 <Xaroth> only the important bits :P
19:38:13 <Xaroth> writing to it, and getting its value
19:38:18 <FLHerne> To be fair, binary output uses `array.array`, which is also not really a file
19:39:43 <FLHerne> Xaroth: The problem I'm envisioning is when in a couple of years someone uses some other part of the API, and doesn't realize that the "StringIO" might not be that
19:40:26 <FLHerne> And of course most people don't test on PyPy, so it would stay broken for a while
19:40:59 *** frosch123 has joined #openttd
19:40:59 <Xaroth> It's a moot point now :)
19:41:05 <FLHerne> i.e. if we did have that patch, `util.StringIO` should be called something else
19:43:49 <andythenorth_> do we know if nml has any alpha/beta/rc release steps?
19:44:04 <andythenorth_> looks to me like I only need to write some docs before we can tag a version
20:09:43 <TrueBrain> and if you have a better name
20:09:51 <TrueBrain> not sure "ape" would make a good PyPi package
20:10:31 <frosch123> if you want to put it on pypi, it should probably include "openttd" in the name
20:10:57 <frosch123> openttd-bananas-client-cli
20:11:39 <frosch123> openttd-bananas-client
20:12:04 <TrueBrain> was hoping on some creative thoughts :P
20:12:30 <frosch123> in our bubble we can name stuff crazy, but on pypi people need to be able to find it
20:12:44 <TrueBrain> openttd-bananas-cli I guess works for that
20:13:10 <TrueBrain> but in that case I guess it is best to rename "ape" to "bananas-cli" too
20:15:28 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JvjCk
20:15:40 <TrueBrain> only the client-id is now "ape"
20:22:33 *** Xetalim has joined #openttd
20:27:28 <andythenorth_> bananas, FIRS, nml, or SEO?
20:27:37 <andythenorth_> SEO is like actual work I have to do that I haven't done
20:29:54 <frosch123> do all 4, switching tasks every 30s ?
20:30:12 <andythenorth_> that works pretty well most of the time
20:30:20 <andythenorth_> never causes mistakes ;P
20:37:00 <andythenorth_> interfaces that load data into tables over xhr
20:37:01 <FLHerne> andythenorth_: I had a vague idea for nml
20:37:09 <andythenorth_> with aggressive caching
20:37:16 <andythenorth_> in a new release that often sets all the values to '0'
20:37:23 * andythenorth_ has told the vendor :P
20:37:54 <andythenorth_> I guess I'm not doing SEO then
20:38:35 <andythenorth_> I'm really not in the mood for writing 311 meta descriptions anyway :)
20:41:46 <FLHerne> No, it was a bad idea
20:43:51 <andythenorth_> rewrite in perl?
20:47:45 <FLHerne> No, you want JS for a rewrite
20:48:14 <FLHerne> Ridiculously optimized JIT compilers, and you can run it in a browser
20:49:59 *** gelignite has joined #openttd
21:02:33 <andythenorth_> I haven't mentioned pypi yet, I don't know the process for that
21:03:11 <andythenorth_> oh the credits are a bit lolcats also :)
21:12:08 <LordAro> andythenorth_: more or less fine
21:13:37 <andythenorth_> someone else can sort credits out, I find creds weird and I delete them from most places I find them :)
21:16:35 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on issue #42: Document how to release nmlc, including building a windows binary https://git.io/Je8vQ
21:19:52 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] frosch123 commented on pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JfIvI
21:24:02 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] frosch123 commented on pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JfIvG
21:25:49 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JvjCk
21:27:06 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] frosch123 commented on pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JfIv4
21:27:10 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JvjCk
21:27:51 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain commented on pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JfIvu
21:28:14 <TrueBrain> tnx a lot frosch123 :)
21:29:23 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] frosch123 approved pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JfIva
21:30:08 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain merged pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JvjCk
21:30:33 <TrueBrain> bit of infrastructure .. lot of testing .. CSS
21:30:38 <TrueBrain> I can control the first two, not the latter ..
21:30:49 <LordAro> andythenorth_: yeah, definitely out of date since the github migration, if nothing else
21:30:51 <TrueBrain> but nice work frosch123 :D This is getting somewhere :)
21:32:37 <TrueBrain> nice idea frosch123 :)
21:37:50 <frosch123> TrueBrain: i wondered about adding a tiling background image with some bananas
21:38:14 <TrueBrain> I would accept that PR :)
21:39:17 <frosch123> really? what about a gif with a rotating banana?
21:39:32 <TrueBrain> you can't do that in a CLI :P
21:40:01 <frosch123> the login success page can show it
21:40:34 * andythenorth_ needs to find the flask 'reload templates' parameter :P
21:42:06 <frosch123> for me it does that by default
21:42:25 <frosch123> or does it need the debug option
21:42:36 <frosch123> try "make debug" or "make run", not sure which version
21:42:46 <frosch123> one of them sets a env var
21:46:02 <andythenorth_> make debug works nicely
21:46:55 *** k-man__ has joined #openttd
21:48:32 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] frosch123 opened issue #4: Display command help text without authentication https://git.io/JfIfG
22:10:41 <andythenorth_> anybody know how the html5 sectioning stuff works?
22:10:50 <andythenorth_> the docs on MDN are self contradictory
22:11:04 <andythenorth_> it seems weird to be setting semantics for something no browsers use
22:11:41 <frosch123> it's probably for blind people
22:11:48 <andythenorth_> no it's for semantics nerds
22:11:57 <andythenorth_> I doubt the AT implements it
22:12:37 <andythenorth_> are we supporting older IE? Like 7 or so?
22:12:39 <frosch123> troll them and use footer at top and header at bottom?
22:12:42 <andythenorth_> it will need polyfills
22:12:57 <andythenorth_> frosch123 that's actually valid, if it's semantically correct
22:13:09 <andythenorth_> the weird world of html
22:13:33 <frosch123> andythenorth_: ottd barely supports win xp, no idea how old ie7 is
22:13:55 <andythenorth_> oh it's mac crash time I think
22:14:01 <andythenorth_> if I disappear, it died
22:14:13 <nielsm> maybe you can run ie7 on vista
22:14:26 <andythenorth_> apple impose a crash every 13-15 days
22:14:34 <andythenorth_> I think it's to keep system logs short
22:14:37 <nielsm> don't bother supporting anything older than IE11
22:29:56 <Eddi|zuHause> ie7 sounds newer than winXP
22:31:25 <nielsm> XP shipped with 6 iirc
22:31:40 <Eddi|zuHause> i was thinking IE 5.5 initially
22:32:32 <andythenorth_> I remember 5 launching on the mac
22:32:40 <andythenorth_> the fastest, most standards compliant browser ever
22:42:16 <nielsm> yea there was a period where IE was actually genuinely the best browser
22:43:27 <nielsm> (it was while netscape was stuck on the v 4.x tech and slowly retooling the browser into mozilla)
22:49:26 <Eddi|zuHause> nielsm: yes, and then IE was stuck on 6 for a decade
22:50:13 <DorpsGek_III> [OpenTTD/OpenTTD] twpol opened issue #8095: Feature suggestion: option to automatically space out vehicles in a shared order https://git.io/JfIUi
22:50:17 <Eddi|zuHause> which is how mozilla made a comeback
22:51:30 <LordAro> what's with the sudden surge in people wanting actually working timetable separation?
22:52:58 <Eddi|zuHause> what? they've been wanting that for like 15 years now
22:52:59 <_dp_> LordAro, social distancing in openttd
22:53:27 <FLHerne> I imagine social distancing probably has something to do with it :P
22:54:19 <Eddi|zuHause> social timetable distancing?
22:54:21 <FLHerne> Lots of people stuck at home with nothing to do but play TTD
22:54:53 <FLHerne> "I have an algorithm worked out and have been poking around the code" is promising
22:55:53 <FLHerne> What's the problem with the version JGR has? The way it's prone to cascade jam-ups?
22:56:49 <andythenorth_> frosch123 that marquee though :D
22:57:32 <Eddi|zuHause> i don't know what JGR has, but separation patches used to have a problem that there were two competing settings that were difficult to set up right
22:58:03 <frosch123> andythenorth_: it was not mentioned today yet...
22:59:58 <andythenorth_> LordAro nml release tomorrow then? o_O
23:00:07 <LordAro> andythenorth_: feasible
23:15:43 <TrueBrain> andythenorth_: why are you doing the CSS now? :P
23:16:14 <frosch123> surely because of the marquee
23:16:43 <andythenorth_> because of the marquee
23:17:28 <andythenorth_> I was surprised marquee worked in my browser
23:19:31 *** sla_ro|master2 has quit IRC
23:42:07 <Samu> i discovered a very very funny bug
23:43:20 <Samu> it's about my canal pathfinder. map size is 256 x 256, mad bridge length is 4096.
23:44:16 <Samu> it tries to find a bridge from tile 16258, going in X axis
23:45:27 <Samu> which is allowed to make a bridge, but 16259 to 16515 is going in axis Y direction1
23:45:42 <Samu> I was only looking for a bridge in X axis
23:47:24 <Samu> i know how to fix this, just got no time atm, will do it tomorrow
23:50:45 <DorpsGek_III> [OpenTTD/nml] FLHerne opened issue #100: Development version format is "invalid" to setuptools. https://git.io/JfIkD
continue to next day ⏵
