IRC logs for #openttd on OFTC at 2020-04-15
⏴ go to previous day
00:18:41 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain approved pull request #3: Change: embed the main headings of all pages in header-elements https://git.io/Jfver
00:20:24 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain commented on pull request #3: Change: embed the main headings of all pages in header-elements https://git.io/JfveX
00:32:55 *** WormnestAndroid has quit IRC
00:33:33 *** WormnestAndroid has joined #openttd
01:03:30 <FLHerne> _dp_: Reddit S1 just kicks them, they either get the hint or don't :P
01:04:48 <_dp_> we don't have that many players to afford kicking them left and right xD
04:55:30 <Eddi|zuHause> we probably should make that popup mandatory for anyone that attempts to join a server
07:25:31 *** andythenorth has joined #openttd
08:03:45 *** sla_ro|master has joined #openttd
08:41:42 *** arikover has joined #openttd
08:50:28 <peter1138> Frozen vegetables, eh?
09:23:29 *** gnu_jj__ has joined #openttd
09:25:18 <DorpsGek_III> [OpenTTD/bananas-frontend-web] auge8472 commented on pull request #3: Change: embed the main headings of all pages in header-elements https://git.io/JfvnC
09:47:25 *** k-man__ has joined #openttd
10:22:57 *** rotterdxm has joined #openttd
10:34:49 <rotterdxm> top o´the mornin samu
10:43:58 <Samu> they wake me up at this time...
11:44:03 <andythenorth> institutional care
11:47:37 <DorpsGek_III> [OpenTTD/OpenTTD] James103 commented on issue #8049: Suggestion: Allow Game Scripts to set company loan beyond max loan and independent of cash on hand https://git.io/Jv7Jj
11:57:35 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain merged pull request #3: Change: embed the main headings of all pages in header-elements https://git.io/JfejJ
11:58:16 <DorpsGek_III> [OpenTTD/OpenTTD] James103 commented on pull request #8008: Feature: Double-click to rename vehicle https://git.io/Jfv47
12:02:43 *** gelignite has joined #openttd
12:23:28 *** tokai|noir has joined #openttd
12:23:28 *** ChanServ sets mode: +v tokai|noir
12:31:04 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh commented on issue #8049: Suggestion: Allow Game Scripts to set the max loan of a company. https://git.io/Jv7Jj
12:38:27 <Samu> just gained a tiny little bit of time once more
12:38:49 <Samu> removed the "this." AyStar
12:43:28 <Samu> or maybe not, i have so many files open, I'm getting confused
12:54:12 <Samu> 'this.' or no 'this.' it made no difference, after all
12:54:28 <Samu> can't squeeze more time :(
13:00:13 <peter1138> Why would you think removing that would improve performance?
13:12:14 <Samu> because ... I thought so
13:12:57 <Samu> I'm still getting ideas from ShipAI author, he changed AyStar a bit
13:14:45 <Samu> he removed the check direction callback
13:14:59 <Samu> I also did that on my AyStar, but he went even further
13:21:20 <Samu> very fishy results, almost too good to be true
13:21:54 <DorpsGek> Samu: 0.231293229931
13:22:16 <Samu> i must confirm the roads are still being built properly
13:57:29 <rotterdxm> that double click to rename vehicles sounds great, not gonna lie
13:58:30 <_dp_> why not just use hotkey for that?
13:58:50 <_dp_> don't think double click is used anywhere in openttd yet so it will be entirely new concept
14:00:14 <_dp_> pencil icon button would be another solution, that't somewhat commonly used in other apps
14:08:54 <Wolf01> _dp_, next hidden feature is long click :P
14:08:55 <Samu> it greatly increases performance
14:09:11 <Samu> but I don't understand what it's lost if those lines are removed
14:09:58 <Samu> actually, it starts earlier
14:17:26 <Samu> I don't understand, there's no downside to it?
14:18:25 <Samu> the check in line 124 returns always !false
14:19:13 <Samu> but removing the entire while loop is confusing me
14:23:35 <peter1138> Have you set a check direction callback?
14:57:03 <Samu> the callback returns false all the time
15:14:32 <Eddi|zuHause> weird... i downloaded 5 files 2 days ago, but only 4 of them are stored on my disk
15:29:37 *** andythenorth has joined #openttd
15:33:01 *** snail_UES_ has joined #openttd
15:42:26 <peter1138> Samu, if calling the callback is taking time, then you could optimise that, I suppose.
15:42:59 * andythenorth calls peter1138 back
15:43:05 <andythenorth> I had eggs, was it lunch?
15:43:28 <andythenorth> I need more salad
15:43:33 <andythenorth> for eating reasons
15:43:50 <peter1138> And a chewy fruit bar thing. "Healthy" and "natural" they call it. "Processed" and "sugars" I say.
15:43:55 <peter1138> I need less salad, tbh.
15:44:02 <andythenorth> gets a bit rabbity
15:44:02 <peter1138> Like, still the same frequency, just less of it.
15:44:08 <andythenorth> try eating a rabbit?
15:44:11 <andythenorth> it's like refined salad
15:44:31 <peter1138> Bit rare to come by.
15:44:32 <Samu> what's taking time is not the callback per se, it's the looping over the whole path time and time again.
15:45:07 <Samu> since the callback is always returning the same answer, ShipAI decided to remove the entire looping over the path
15:45:19 <peter1138> Samu, exactly, and if you decide you don't need the callback...
15:46:30 <Samu> it's too much of a performance gain, there must be some caveat here...
15:47:54 <nielsm> there can be value in taking less than optimal decision much faster
15:48:39 <peter1138> Samu, always returning false in the callback is not the same as not checking for mismatches (which is what removing the loop would do)
15:49:00 <peter1138> (Whatever a "mismatch" is in this case, I don't know)
15:49:37 <peter1138> It looks like it's checking for a path that crosses itself
15:50:06 <peter1138> But I don't know A* enough any more :p
15:50:28 <andythenorth> Samu peter1138 only knows about newgrf docks now :)
15:56:42 *** arikover has joined #openttd
15:56:49 <Samu> just managed yet another milimetric gain
15:57:56 <LordAro> Samu: that is definitely a meaningless change
15:58:01 <LordAro> run it repeatedly without changing anything
15:58:07 <LordAro> you will see significant variance
16:05:46 <Samu> and i have another coming
16:20:50 <Samu> just realised the Native Heap uses more memory, way more memory than binary heap
16:22:01 <Samu> unsure which kind of heap it is, but I think it's bucket queue
16:27:04 <Samu> while (this is happening) { do stuff } else { do other stuff }
16:27:32 <andythenorth> that would imply a perpetual loop
16:27:42 <andythenorth> I think it would be problematic to fulfill
16:27:51 <nielsm> nah it could mean "if the loop never run then do the else"
16:28:26 *** supermop_Home has joined #openttd
16:29:29 <supermop_Home> hmm this meeting is via hangouts
16:29:49 <supermop_Home> on Monday this client wanted to use bluejeans
16:30:05 <supermop_Home> internally we use slack or zoom
16:30:14 <supermop_Home> I guess no one uses skype?
16:31:58 <supermop_Home> migrations gs kind of weird
16:33:33 <Samu> from path.GetDirection() to path._direction, to dir = path._direction and replacing path._direction everwhere with dir
16:34:21 <Samu> oh and stuff with > 0 or != 0 or == 0 also replaced
16:35:16 <Samu> faster-> while (this._open.Count() && (iterations == -1 || iterations--)) {
16:35:19 <Samu> slower -> while (this._open.Count() > 0 && (iterations == -1 || iterations-- > 0)) {
16:37:16 <andythenorth> supermop_Home migrations builds roads to industries nicely though
16:43:53 <supermop_Home> andythenorth I like that
16:44:25 <supermop_Home> i don't like coming back to a mine 40 days after starting service from it to find a megalopolis has sprung up around it
16:44:59 <supermop_Home> i do like that is builds towns that are have a more interesting distribution of density
16:46:04 <andythenorth> for realism it's great
16:46:12 <andythenorth> for gameplay, industries are the last place I want towns
16:46:34 <supermop_Home> mines should get pretty small towns
16:46:56 <supermop_Home> also i like the hollowing out of non-industrial cities
16:47:12 <supermop_Home> but it too quickly reduces them to zero
16:47:24 *** ChanServ sets mode: +v tokai
16:48:58 <supermop_Home> sometime an existing town is fairly close to an industry, but it will build a new town 10 tiles away and hollow the original one to zero
16:58:21 <TrueBrain> I hope frosch can forgive me :)
16:59:10 <TrueBrain> the "oops I forgot to run flake before push" commit :D
17:21:28 <Samu> same route, built in less days
17:21:48 <Samu> I always complained my AI had a slow pathfinder, I guess no more!
17:23:30 <_dp_> there are two logical ways of implementing else on loops and I never remember what uses which
17:25:04 <nielsm> python has the no break version
17:42:07 <TrueBrain> The never ran doesn't prevent a silly variable, the no break does. Happy they added it :)
17:57:09 *** Laedek_ has joined #openttd
17:57:36 *** frosch123 has joined #openttd
17:58:22 *** Xaroth65 has joined #openttd
17:58:41 *** Xaroth65 is now known as Xaroth
18:56:49 <DorpsGek_III> [OpenTTD/bananas-frontend-web] auge8472 opened pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvP2
18:59:23 <Samu> tiles with a bridge over it, are considered buildable
18:59:38 <Samu> I don't think that's a good idea
19:12:03 *** arikover has joined #openttd
19:17:08 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 commented on pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvXS
19:18:32 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 commented on pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvXQ
19:32:06 *** Progman has joined #openttd
19:35:03 <TrueBrain> frosch123: flask comes with click built-in
19:35:10 <TrueBrain> I have a patch if you like that adds arguments, on top of my PR?
19:35:32 <TrueBrain> frosch123: why is there a 'make run' and 'make debug'? Shouldn't run just always be 'debug'?
19:35:34 <frosch123> only tus-js-client missing then
19:36:24 <frosch123> TrueBrain: i already forgot how the wsgi stuff worked, but unless that goes into "make run", then you are right
19:37:28 <frosch123> i just like makefiles instead of c&p from readme or looking in shell history
19:37:53 <TrueBrain> I am totally fine with that
19:37:58 <TrueBrain> I personally don't do it / use it
19:38:08 <TrueBrain> but .. I have no argument against it either :D
19:39:36 <TrueBrain> description is optional
19:39:36 <TrueBrain> webclient/templates/manager_package_info.html", line 30, in block "content"
19:39:36 <TrueBrain> {% for l in package["description"].splitlines() %}
19:39:42 <TrueBrain> "login" is weird, as it stays like that even if I am logged in
19:39:43 <TrueBrain> rename to "manager"?
19:41:53 <frosch123> i only show the manager/display-name on the /manager pages, so that all /package sites are static, and could be cached
19:42:11 <TrueBrain> hence the suggestion to simply rename "login" :)
19:42:22 <frosch123> i guess that also works
19:42:27 <frosch123> logout won't be displayed either way
19:44:37 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain opened pull request #11: Change: update the wishlist with what was found in bananas-frontend-web https://git.io/Jfv1F
19:47:01 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #1: Add: a very basic CLI frontend for BaNaNaS API https://git.io/JvjCk
19:47:14 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain dismissed a review for pull request #4: Minor bits and pieces https://git.io/JfvX0
19:47:37 <TrueBrain> did not see you already approved it; otherwise I would have put it in another branch
19:48:03 <TrueBrain> if you like, I can still do that, ofc
19:49:26 <TrueBrain> basically, for click, I followed their guide on the 'Custom' method, which seems to be the most clean for our purpose
19:49:44 <TrueBrain> mostly as it removes the need for an env-variable, which makes deployments a lot easier :D
19:52:07 <frosch123> ah, now i understand the relevance of "flask is using click"
19:52:36 <TrueBrain> their documentation is shit for this btw
19:52:54 <TrueBrain> they kinda assume the whole world is using click or something
19:54:34 <frosch123> you import "pages" before "main"
19:54:37 <frosch123> does that even run?
19:54:59 <TrueBrain> sure, why not? somewhere in pages main is included again
19:55:03 <TrueBrain> who ever comes first initializes app
19:55:14 <TrueBrain> that I love about this form of imports .. order is irrelevant :D
19:55:45 <frosch123> oh, i missed thtere is a __main__ now
19:56:04 <TrueBrain> yeah .. maybe we should rename main.py now
19:56:36 <TrueBrain> yeah, I am going to do that if you don't mind
19:58:03 <frosch123> also add a unit (seconds) to the expire options
19:58:35 <TrueBrain> ` --csrf-expire SECONDS Time for the CSRF token to expire.
19:58:46 <DorpsGek_III> [OpenTTD/bananas-frontend-web] auge8472 commented on pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvM8
19:59:07 <frosch123> ah, that's the meaning of metavar
20:00:11 <nielsm> it casts a Money (OverflowSafeInt64) to a plain int64 and then multiplies with another int64
20:00:19 <nielsm> intentionally throwing away the overflow-safe-ness?
20:05:34 <TrueBrain> frosch123: I never settled on a way to module variables
20:05:40 <TrueBrain> but I will adapt to your coding style there :)
20:05:57 <frosch123> nielsm: those are base prices, if they overflow you can as well call abort() :)
20:07:20 *** arikover has joined #openttd
20:07:48 <frosch123> TrueBrain: you missed the session->csrf type
20:08:50 <frosch123> TrueBrain: afaik "_" is pretty standard for private vars
20:09:05 <TrueBrain> not on module level .. not sure anyone has a real opinion about it
20:09:08 <TrueBrain> besides: don't use it
20:09:13 <TrueBrain> but .. meh .. effort
20:10:03 <frosch123> sounds like c++ advice. does python also start to develop static-initialisation dilemmas? :p
20:12:36 <DorpsGek_III> [OpenTTD/bananas-frontend-web] auge8472 updated pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvP2
20:22:00 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 approved pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvDG
20:22:07 <Eddi|zuHause> "Reboot is required to ensure that your system benefits from these updates." i'm scared
20:22:21 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 merged pull request #5: Change: <thead> and <tbody> for tables where suitable, <th> for row names https://git.io/JfvP2
20:22:45 <frosch123> TrueBrain: so, who checks tus-js-client?
20:23:05 <TrueBrain> I know what needs doing, just today don't feel like doing :P
20:25:47 <frosch123> great, i can already not start it anymore :p
20:26:32 <TrueBrain> well feel free to do start it, but I gathered you are not really looking forward to it :P
20:26:51 <TrueBrain> you can look at other things on the wishlists .. or look at the frontend-cli and server PRs? :D
20:27:03 <frosch123> did you add python 3.8 stuff?
20:27:19 <frosch123> ImportError: cannot import name 'pages' from '__main__' (/home/frosch/daten/ottd/bananas3/webclient/__main__.py)
20:28:40 <TrueBrain> that makes very little sense to me .. relative imports should work
20:29:42 <frosch123> well, the same line was in __init__ before
20:30:13 <TrueBrain> I .... am a bit at a loss
20:32:01 <frosch123> somehow it thinks that __main__ is now the package name
20:32:08 <TrueBrain> works for me on Python 3.7
20:32:15 <TrueBrain> how are you launching it?
20:32:26 <TrueBrain> and otherwise your repo is clean?
20:32:38 <TrueBrain> you did install the latest requirements.txt ?
20:32:43 <TrueBrain> there was a new version of flask
20:32:51 <TrueBrain> works even on Python 3.6
20:33:13 <frosch123> i was using the system packages before
20:33:18 <TrueBrain> but .. I really cannot piee together what happens
20:33:28 <TrueBrain> that import should work, even in 3.2
20:33:58 <Xaroth> are you running in a venv?
20:34:00 *** WormnestAndroid has quit IRC
20:34:14 *** WormnestAndroid has joined #openttd
20:34:25 <frosch123> venv changes nothnig
20:34:51 *** Progman has joined #openttd
20:35:35 <TrueBrain> I don't even have a clue where it could be failing for you frosch123 , sorry :( But that import is really just a normal import ..
20:35:45 <TrueBrain> what does "python -m webclient" return?
20:38:47 <frosch123> ah, the problem is that the MAkefile does not use .env/bin/python3
20:39:04 <Xaroth> You're supposed to activate the venv.
20:39:15 <TrueBrain> the README is wrong there indeed
20:39:24 <TrueBrain> as I said, I never use Makefiles :P
20:39:40 <TrueBrain> so you used Python 2.7 :D
20:40:47 <frosch123> it used the global python3
20:41:13 <Xaroth> what version is your global py3?
20:41:15 <TrueBrain> hmm .. really cannot remember they changed the import stuff; owh well :D
20:41:15 <frosch123> i thought it only installed packages, not an entirely different python
20:42:21 <frosch123> anyway, it does affect the PYTHONPATH somehow
20:42:36 <frosch123> "from webclient import pages" did not work either
20:42:49 <frosch123> so, who nows where it was looking up stuff
20:43:46 <Xaroth> it probably gives an import error because it was trying to locate packages not globally installed
20:43:53 <Xaroth> because it wasn't running in your venv
20:55:20 <frosch123> TrueBrain: make can even setup the venv for you
20:55:38 <TrueBrain> I don't like Makefiles in Python projects :)
20:55:41 <TrueBrain> I think they are evil ;)
20:57:40 <frosch123> are you Xaroth' brother?
20:58:23 <TrueBrain> lets keep this game up
20:58:31 <TrueBrain> within 6 bilion guesses you should be there
20:58:35 <xahodo> I'm also not an android.
21:00:14 <xahodo> I'm trying to get fonts to work correctly in openttd. But the openttd will not cooperate. :(
21:00:54 <xahodo> Do I need to include full paths? Or should freetype arrange that?
21:02:04 <frosch123> medium_font = DejaVu Sans, bold <- i use that
21:03:35 <nielsm> actually we aren't linking fontconfig or anything similar, in non-windows non-mac builds there is nothing available to look up fonts by name, so don't you need to specify them by filename then?
21:04:20 <frosch123> nielsm: ofc we are linking fontconfig
21:04:41 <frosch123> any i have been using linux exclusively for 12 years
21:05:29 <xahodo> and, yes, I have arial and hack installed.
21:06:07 <frosch123> do other fonts work?
21:07:46 <xahodo> However it doesn't complain about the monospace font (hack).
21:08:45 <milek7> is 'Regular' actually part of name?
21:09:13 <xahodo> No, checked without that though. Didn't change anything.
21:10:16 *** Eddi|zuHause has joined #openttd
21:10:17 <frosch123> ottd splits the config value on ","
21:10:26 <xahodo> It gives this error: Unable to use 'Arial' for medium font, FreeType reported error 0x1, using sprite font instead
21:10:26 <frosch123> the first part is passed to FcNameParse
21:13:09 <xahodo> Well, if I pass it a full path, it finds the font.
21:13:25 <xahodo> Perhaps I got something misconfigured somewhere?
21:14:07 <glx> without full path I think fontconfig is asked
21:16:15 <xahodo> fc-list works as expected and shows me all installed fonts.
21:17:01 <glx> on windows I use "medium_font = arial unicode ms" and that works, but of course we don't use fontconfig nor freetype there, but with freetype it worked too
21:23:16 <xahodo> I happen to have the noto fonts installed, so when I typed fc-match -a "arial" I got a bucketload of results...
21:24:49 <xahodo> but just "fc-match arial" returns 'Arimo-Regular.ttf: "Arimo" "Regular"'
21:26:08 <Samu> "Since Squirrel 3.0 is possible to declare a table using JSON syntax"
21:26:15 <Samu> hmm is this something good for openttd?
21:26:55 <milek7> openttd uses squirrel 2
21:27:36 *** Wormnest has joined #openttd
21:30:23 <Samu> how do i delete a value from a []
21:30:49 <Samu> i have [tile1, tile2], i wanna remove tile2, make it [tile1]
21:32:47 <Samu> i know how to add values though, it's .push
21:37:12 *** Eddi|zuHause has joined #openttd
21:37:15 <glx> pop() should also work to remove the last entry
21:44:40 *** gelignite has joined #openttd
21:49:30 <Heiki> xahodo: apparently, Arial has localized names for its variations, and apparently fontconfig uses them – the following works with a Finnish locale: medium_font = Arial, Normaali
21:51:31 <Heiki> but when starting openttd saying e.g. “LANG=C openttd” I get “Unable to use 'Arial, Normaali' for medium font, FreeType reported error 0x1, using sprite font instead”
21:55:03 <Heiki> “medium_font = Arial, Regular” works for “LANG=C openttd”
21:56:48 <frosch123> added that to the wiki
21:58:33 <Heiki> I wonder whether those Microsoft fonts are the only ones using localized names
22:03:01 <frosch123> fc-list -v lists adds a language to the name
22:16:57 <andythenorth> everyone else shrinks heatshrink directly with the soldering iron, right?
22:18:23 <andythenorth> the smell is probably healthy
22:18:27 <andythenorth> invigorating fumes
22:18:46 <glx> not worse than solder fumes
22:22:12 <Heiki> FreeMono, FreeSerif, FreeSans have localized style names too
22:36:59 *** arikover has joined #openttd
22:39:52 <nielsm> this is worse than watching grass grow
22:39:58 <nielsm> I'm watching sea floor erode
22:51:35 <_dp_> desync issue solved I guess xD
23:08:52 <nielsm> "takes place over centuries" really does mean "in a hundred years you will barely see any meaningful change"
23:09:38 <nielsm> and "over a few years" means in a decade you will have a full gradient from coast to max depth
23:11:08 <_dp_> nielsm, both sound bad tbh
23:11:18 <_dp_> nielsm, something in between would make more sense imo
23:14:54 <nielsm> _dp_ I'll try something in between then
23:15:26 <andythenorth> rivers can erode land next :D
23:24:21 *** mcbanhas has joined #openttd
continue to next day ⏵
