IRC logs for #openttd on OFTC at 2019-02-23
⏴ go to previous day
00:00:45 <peter1138> GS could be improved if there was someone interested in writing a GS _and_ improving the API.
00:00:57 <_dp_> like newgrfs have nothing to do with GS not being able to do reliable callbacks
00:01:54 <_dp_> peter1138, it's not just api, GS engine has to change as well
00:02:09 <peter1138> Okay. What needs to change?
00:02:28 <_dp_> peter1138, 1 command / tick
00:03:14 <_dp_> peter1138, and if we talk about callbacks any execution delay in general + operation limit I guess
00:03:39 <peter1138> Ok, that's because there's a delay in how commands are sent :/
00:04:13 <_dp_> peter1138, yeah, I know why :)
00:04:44 <peter1138> network_command.cpp:157
00:04:50 <_dp_> peter1138, doesn't change that it needs to go :p
00:05:38 <_dp_> I was actually thinking to patch citymania servers to get rid of that delay
00:05:46 <_dp_> but then I got rid of GS altogether
00:05:50 <peter1138> I'm thinking that delay should not apply to GS.
00:06:22 <peter1138> I don't know if it actually solves anything though.
00:06:38 <peter1138> It's not going to make the GS be executed again.
00:07:31 <peter1138> The command system kinda works like syscalls on a co-operative multitasking system (e.g. old RISC OS)
00:07:54 <peter1138> Except there is the max opcodes thing as well :)
00:08:17 <andythenorth> bring back RISC OS :P
00:08:21 <andythenorth> and a three-button mouse
00:08:51 <peter1138> Can we speed up the CI? :/
00:09:34 <LordAro> not without giving MS large sums of money
00:10:09 <peter1138> Not sure if it's possible to execute commands on the server and then queue it to send to the clients.
00:11:24 <_dp_> peter1138, should be fine as long as order stays the same
00:11:47 <peter1138> Yes, I was thinking in terms of already-queued commands, dunno though.
00:12:05 <peter1138> Not just the same order, they have to be on the same tick.
00:12:22 <peter1138> Why the heck am I listening to Enigma?
00:13:46 <andythenorth> watching a documentary about Dre though
00:14:13 <_dp_> peter1138, yeah, I basically consider tick change to be another command
00:37:09 *** supermop_work_ has joined #openttd
00:38:00 *** supermop_work__ has joined #openttd
00:45:11 *** supermop_work_ has quit IRC
00:48:44 <peter1138> Ooh. Vivecraft for 1.13.2
01:02:17 *** tokai|noir has joined #openttd
01:02:17 *** ChanServ sets mode: +v tokai|noir
01:31:54 *** Gustavo6056 has joined #openttd
01:35:16 <peter1138> Hmm, CPU usage up :/
01:35:18 *** Gustavo6056 is now known as Gustavo6046
01:48:13 *** drac_boy has joined #openttd
01:48:47 <drac_boy> hi there .. historic question: was 3-axle freight wagons generally only for uk milks or did it abound anywhere else as well? (baggage cars kinda doesn't count in this sorry)
02:00:54 <drac_boy> also just had to wonder after friend asked me. is this just a selective usercase or is there any reason why osx doesn't react to ctrl+apple+c but rather ctrl+option+c instead for the cheat dialog
02:32:48 *** supermop_Home has joined #openttd
02:41:55 <drac_boy> hi superpeter ;) heheh
03:05:56 *** snail_UES_ has joined #openttd
03:36:56 <drac_boy> anyway going sleep soon here
03:43:24 <Samu> subtract_input, add_output, i finally start to see how this works
03:58:54 <Samu> enum IndustryCallbackMask {
04:14:48 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7265: Fix #7004: Mark linkgraph dirty to be rebuilt on next draw call. https://git.io/fhFpA
04:31:21 <Samu> there may be a way to do what I want
04:32:00 <Samu> will have to think about it for tomorrow
05:06:55 *** HerzogDeXtEr1 has joined #openttd
07:42:38 *** Alberth has joined #openttd
07:42:38 *** ChanServ sets mode: +o Alberth
07:57:53 *** sla_ro|master has joined #openttd
08:38:45 *** andythenorth has joined #openttd
08:40:54 <andythenorth> because of vehicle exclusive previews
08:44:23 <andythenorth> vehicle is offered as preview -> vehicle is missing dependencies -> vehicle is uselss
08:44:58 <andythenorth> one argument might that the newgrf is broken in that case
08:45:55 <peter1138> Hmm "if intro date is the same include dependencies in preview"
08:45:56 <planetmaker> exclusive preview is only offered for engines iirc
08:46:11 <planetmaker> if another one uses wagons w/o an engine... is it a fail?
08:49:05 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7265: Fix #7004: Mark linkgraph dirty to be rebuilt on next draw call. https://git.io/fhFj1
08:49:49 <andythenorth> designing vehicles that depend on each other is a fail
08:50:19 <andythenorth> but that means abandoning wagon speed limits
08:50:27 <planetmaker> andythenorth> designing vehicles that depend on each other is a fail <-- I agree. Might be different for wagons. But vehicles... yes
08:51:09 <planetmaker> s/vehicles/self-powered vehicles/
08:51:19 <nielsm> are unpowered wagons ever offered for exclusive trial?
08:51:27 <peter1138> "because of vehicle exclusive previews" is a non-issue then?
08:51:45 <andythenorth> well it depends how completionist we want to be :P
08:52:05 <nielsm> so would have to add another property to vehicles "always introduce together with" somehow?
08:52:23 <andythenorth> I chose the words "it won't work" carefully :P
08:52:33 <nielsm> or "never introduce without X being available"
08:52:44 <andythenorth> I'm still in favour of synchronising the dates using the seed
08:53:11 <andythenorth> I just thought of the edge case it doesn't meet :P
08:54:34 <andythenorth> nielsm: an older idea was exactly a property, so that if one vehicle becomes available, other vehicles are also triggered for introduction
08:54:48 <andythenorth> the Eddi|zuHause solution is pretty neat because no spec extension needed
08:55:19 <andythenorth> zero-effort for authors, nml maintainer etc
08:55:31 <peter1138> The property idea is too rigid as well.
08:56:44 <andythenorth> we _could_ infer wagon introduction
08:57:07 <andythenorth> where intro date matches exactly set(engines)
08:57:17 <andythenorth> if the engine is available, introduce the wagon
08:57:34 <andythenorth> the engine / any engine in the set /s
08:57:46 <planetmaker> we can only set introduction years, can we?
08:57:57 <andythenorth> it's a date afaik
08:58:04 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7225: Add: Various AI/GS functions that may be useful. https://git.io/fh7Gz
08:58:28 <andythenorth> some of us are in the habit of only setting year :P
08:58:30 <planetmaker> oh, well. Then it's a total non-issue
08:58:42 <planetmaker> same date = all together, irrespective of the newgrf they come from
08:58:49 <planetmaker> different date = differently randomized
08:58:51 <peter1138> planetmaker, there's two.
08:59:05 <andythenorth> planetmaker that's what eddi's patch is intended to do yes
08:59:29 <peter1138> Hmm, no, there isn't two.
08:59:41 <peter1138> Well, there are two, but one is relative to 1920, the other is relative to 0
08:59:47 <andythenorth> but both in days?
08:59:50 <planetmaker> peter1138, two properties. But they map to the same internal one
09:00:05 <planetmaker> and granularity is the same, just offset
09:00:06 <peter1138> planetmaker, yeah, one of them might've been years rather than days, I meant.
09:00:21 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhFj5
09:01:50 <andythenorth> so how about a flag to exclude vehicles from exclusive preview offers? o_O
09:01:59 <andythenorth> fixing symptom not cause? o_O
09:02:42 <planetmaker> ok... how does exclusive preview now pose difficulty?
09:02:57 <andythenorth> engines introduced with no wagons available
09:03:02 <planetmaker> if we say "same date = always together" that might apply to exclusive preview, too
09:03:08 <DorpsGek_II> [OpenTTD/OpenTTD] GabdaZM commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhFjN
09:03:21 <andythenorth> to be clear it's an edge case I'm describing, no kittens die
09:04:00 <andythenorth> it improves the common case, and occasionally the game will slap the player around the face a bit with preview
09:04:01 <planetmaker> but I think I'm missing the details how exclusive preview is handled internally :)
09:04:30 <planetmaker> actually it would be cool to have per-company introduction dates :D
09:04:41 <planetmaker> especially for scenario purposes
09:05:05 <planetmaker> that would - at the same time - cover exclusive previews
09:06:31 <planetmaker> (but I'm side-tracking)
09:07:51 <peter1138> Well that sounds something that should be GS functionality.
09:09:13 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhFjj
09:09:41 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7250: K-d tree data structure for spatial lookups https://git.io/fhbee
09:09:46 <planetmaker> peter1138, yes, that's how I see it, too. Exposed to GS only. But it's not internally supported yet
09:10:08 <nielsm> peter1138: heh pulling in opposite directions :)
09:10:26 <peter1138> I've editted mine now.
09:11:19 <peter1138> planetmaker, hmm, I guess these little details are awkward because the GS only runs on the server.
09:11:46 <peter1138> So anything you want to change has to be a network command... and GS can only do one command per tick. Hmm.
09:12:19 <planetmaker> na, it's perpendicular things, nielsm :)
09:12:31 <nielsm> or you'd have to pack more data in the string parameter as an abuse
09:12:46 <nielsm> (if that's even possible)
09:14:26 <andythenorth> so maybe we can finish 16-cargo nmlc support today :)
09:15:50 <nielsm> problem is that the width to search really depends on font size as well
09:17:05 <peter1138> font size and zoom level.
09:17:48 <peter1138> Well, zoom level affects the bounds you need to search, at least.
09:18:44 <peter1138> andythenorth, so maybe we can approve nrt today :)
09:19:16 <nielsm> maybe beta3 or rc1 should be the focus, really
09:20:10 <peter1138> There's a few PRs I'd like to get in, not including my own obviously.
09:20:29 <peter1138> I suppose I better get cycling-dressed.
09:21:31 <peter1138> Ok, I'll mark them.
09:24:15 <nielsm> I'll do some squashing of ai-framerate
09:26:51 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN approved pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fhbe3
09:27:12 <nielsm> hmm I can't click on the last item in the list
09:27:49 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN closed issue #6918: Add option to adjust font size separately from GUI size. https://git.io/fhbes
09:27:50 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fhLFX
09:32:34 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #6931: Change: Prevent town growth from blocking ships https://git.io/fhbeZ
09:55:41 *** Progman has joined #openttd
10:15:34 <LordAro> moin was one of the dwarves in the hobbit, right?
10:35:33 <andythenorth> Alberth: I sent forum DM btw ;)
10:37:15 <Alberth> I cloned the repo, but that's all I did so far
10:39:11 <andythenorth> not sure if hg revs will ever line up with git numeric revs
10:43:01 <andythenorth> my global .gitignore ignores 'bin'
10:43:13 <andythenorth> because python projects
10:43:17 <andythenorth> where bin is generated
10:51:41 <_dp_> yeah, per company introduction dates, patch I never did
10:51:59 <_dp_> it's actually quite easy as flags are already there only API is missing
10:52:15 <_dp_> and 1command/tick is not a big issue here as not many commands are needed
10:52:36 <_dp_> it's not like updating company goals where every line is a command so it takes forever
10:54:04 <Eddi|zuHause> <andythenorth> not sure if hg revs will ever line up with git numeric revs <-- unlikely, because hg counts commits that are not direct ancestors. but you could apply an arbitrary offset for the conversion...
11:17:47 <LordAro> with or without merging nrt?
11:18:22 <TrueBrain> NRT for after 1.9, right? :)
11:18:28 <TrueBrain> bit late to add new shit like that now :D
11:19:13 <LordAro> peter1138 wanted to add it sooner rather than later - thought another year for 1.10 was too long
11:19:37 <LordAro> it's been reasonably well tested by this point, right?
11:19:55 <TrueBrain> but nobody is saying a new release should take a year
11:20:10 <Eddi|zuHause> "well tested" has been a terrible argument, historically :p
11:20:11 <TrueBrain> nothing wrong with pushing a new release in 6 months .. or 3 months ..
11:20:19 <_dp_> imo it's better to release 1.10 early than include something that big at beta stage
11:20:35 <TrueBrain> once you release your first beta, you should be in a feature freeze
11:20:42 <TrueBrain> personally, I thought that was the whole point of the beta :)
11:20:47 <TrueBrain> also the reason we had a milestone etc
11:21:49 <LordAro> i thought that was RC stage, i.e. when it branches
11:22:01 <LordAro> until then, the betas are just renamed trunk
11:22:09 <TrueBrain> what I thought we were doing: push out 1.9 ASAP, so we can merge NRT :)
11:22:31 <LordAro> but if you're happy to break "the release cycle", then ok :)
11:22:34 <TrueBrain> personally, I would pick up the last few things in the 1.9 milestone, and get it out there :)
11:22:39 <LordAro> it can go in the devblog
11:22:54 <TrueBrain> the "release cycle" only exists because people maintainted it
11:23:01 <LordAro> stable releases not in april, whatever next
11:23:11 <TrueBrain> the first 1st of April was a joke on a joke ..
11:23:18 <TrueBrain> the second time was funny because .. it was funny
11:23:23 <TrueBrain> the third time, it enraged a few people
11:23:28 <TrueBrain> after that ... it was "a rule"
11:23:31 <TrueBrain> funny how that works :P
11:23:59 <TrueBrain> if we do release more often, we should consider an auto-updater btw :P
11:24:12 <_dp_> it kinda fit a development tempo last few years as there weren't much features to include even after a year
11:24:24 <TrueBrain> but please, lets not merge NRT, because "otherwise it takes so long" .. rushing shit is never a good sign :)
11:24:53 <TrueBrain> I think it would be fair to make a new release when ever something big happened
11:25:01 <TrueBrain> (lets call them .. feature releases!)
11:25:41 <LordAro> autoupdaters are hard, and only really worth doing on windows anyway
11:25:56 <LordAro> i'd settle for a message on the mainscreen though
11:26:00 <TrueBrain> that reminds me, I want Steam! :P
11:26:55 <Eddi|zuHause> mainscreen turn on
11:27:40 <TrueBrain> hmm, new Azure Pipelines version still isn't rolled out .. 20 days after changelog was published
11:27:46 <TrueBrain> this random time between those is annoying
11:27:48 * LordAro adds NRT to a 1.10 milestone
11:27:53 <TrueBrain> LordAro: or .. 2.0? :D
11:28:14 <TrueBrain> 4 entries in the 1.9 milestone
11:29:13 <TrueBrain> ICU is the biggest in there, I guess
11:29:54 <TrueBrain> btw, LordAro, did you ever manage to get the website PR finished somewhat? (the one where you remove most of the icons :D)
11:30:04 <TrueBrain> (honest question out of interest; not to push you or anything)
11:30:12 <LordAro> ICU will probably slip to 1.10, tbh
11:31:13 <LordAro> i imagine blathjis might complain :p
11:31:43 <TrueBrain> I think everyone agrees we should replace it; but .. if it is not happening now, it is not happening now :)
11:32:04 <LordAro> can't remember how far i got with the website, but it was definitely still a work in progress
11:32:05 <andythenorth> no NRT for 1.9.0
11:32:17 <andythenorth> push it ASAP to nightlies
11:32:22 <andythenorth> do a 1.10 in 3 months
11:32:34 <LordAro> andythenorth: i put 1.10 as October
11:33:00 <andythenorth> anyway we're making a game not a religion, April 1 is not sacred :)
11:33:01 <TrueBrain> if we automate releases a bit more, it also becomes easier to release
11:33:40 <LordAro> so... RC1 this weekend?
11:33:49 <LordAro> does a month of RC(s) sound reasonable?
11:34:21 <LordAro> gives us enough time to work out how branch commits are going to work as well :)
11:34:31 <LordAro> (i imagine cherry-picking, but)
11:34:43 <TrueBrain> cherry-picking is best, otherwise merging, I guess
11:34:59 <LordAro> nielsm: looks like someone needs to answer the question
11:35:14 <LordAro> i think the answer is "yes" ?
11:37:10 <TrueBrain> so .. what else is there to do .. hmm ..
11:37:19 <TrueBrain> guess I should make a REST API for "download versions"
11:38:14 <TrueBrain> _dp_: I spammed your PMs, but never got a reply; you figured out how to build via Azure Pipelines yourself? :D
11:39:47 <TrueBrain> hmm .. what is the easiest way to create a REST API these days ..
11:40:12 <LordAro> something in python, to be consistent, presumably?
11:40:47 <TrueBrain> Flask I guess ... how to document the API ..
11:41:12 <LordAro> swagger's pretty nice
11:41:32 <LordAro> as long as you go swagger -> implementation, rather than the other way around
11:41:41 <TrueBrain> yeah .. that is always the annoying part
11:41:46 <TrueBrain> you rather have documentation from your code
11:41:49 <TrueBrain> as it is .. more up-to-date :D
11:42:35 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7187: Change: Add scrollbar to cargo legend in cargo payment rates window. https://git.io/fhbvY
11:42:52 <TrueBrain> yeah .. Flask-RESTPlus, flask-apispec, ...
11:43:05 <TrueBrain> always difficult to find one that is both maintained and workable :P
11:43:48 <TrueBrain> flask-restplus for example hasnt seen a commit in 6 months
11:44:05 <nielsm> that may also mean it's stable :)
11:44:30 <TrueBrain> given it is version 0.12
11:44:42 <TrueBrain> that makes it hard to estimate
11:45:29 <TrueBrain> I am not the only one wondering :D
11:45:46 <TrueBrain> seems bus-factor of 1 was reached
11:46:03 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh merged pull request #7187: Change: Add scrollbar to cargo legend in cargo payment rates window. https://git.io/fh99m
11:50:09 <nielsm> ai/gs framerate is just expansion of an already new feature, to be more complete, I think it's obvious it needs to go in!
11:51:17 <TrueBrain> we use to keep all the sources of every nightly we ever produced (only the sources)
11:51:21 <TrueBrain> is that really useful, I wonder
11:54:34 <LordAro> a list of commit hashes should be plenty
11:54:56 <TrueBrain> even there, is it useful to keep of those older than .. 90 days?
11:55:00 <TrueBrain> I have them back to r1 :D
11:56:15 <LordAro> perhaps not, but i don't see the harm in it
11:56:23 <LordAro> text file can't be that big :p
11:56:54 <TrueBrain> keeping things for keeping, and nobody ever using it, only decreases maintainability :D
11:57:10 <TrueBrain> so I was more wondering why/when anyone would like to know
11:57:29 <TrueBrain> any crashdump etc will always contain the githash or svn revision
11:57:57 <andythenorth> was it an outdated approach to GPL compliance?
11:58:12 <TrueBrain> no; it was done because we could, and in case subversion went bye-bye
11:58:33 <TrueBrain> (so we could recover from nightlies, as good and bad as that would go)
11:58:36 <andythenorth> job done, no need
11:58:53 <TrueBrain> yeah .. so I am somewhat tempted to only keep record of the last 90 nightlies
11:58:59 <TrueBrain> and just remove anything else
12:00:57 * andythenorth starts FIRS 4 properly now
12:01:06 <andythenorth> feature branches and everything
12:01:13 <TrueBrain> that is what they all say :P
12:05:52 <andythenorth> oof nmlc ERROR: "generated/firs.nml", line 7342: Syntax error, unexpected token "waiting_cargo_1"
12:05:53 <andythenorth> make: *** [generated/firs.grf] Error 1
12:07:30 <andythenorth> so the prod. cb format needs to change?
12:07:45 <andythenorth> I knew this would be like pulling teeth :P
12:07:55 <andythenorth> trying to test the 16 cargo industry stuff against FIRS is dumb
12:08:29 <andythenorth> but nobody else has provided a reference.grf for it :(
12:12:28 <planetmaker> what exactly is the unexpected token there, andythenorth ?
12:12:39 <andythenorth> "waiting_cargo_1"
12:13:04 <nielsm> the old produce syntax is supposed to still work
12:13:23 <andythenorth> trying to figure out how prod. cb detects version
12:13:26 <nielsm> but I'll admit to not testing that
12:16:12 <planetmaker> TrueBrain, I'm sure it will be of interest in some way. And if it's for history reasons :) Yes, we have the repo...
12:17:00 <andythenorth> not sure where version is specified in nml
12:17:29 <andythenorth> I think it's inferred to be version 1, historically
12:18:16 <andythenorth> might be set here?
12:18:26 <planetmaker> andythenorth, there's nml/__version__.py
12:18:34 <planetmaker> which in turn is written by setup.py
12:18:48 <planetmaker> which in turn runs nml/detect_version.py or so
12:18:48 <nielsm> planetmaker not nml version, production callback version
12:23:02 <andythenorth> ok so a few things clicked :P
12:23:22 <andythenorth> 1. this shouldn't break old grfs
12:23:33 <andythenorth> so I should test it with FIRS based on 3 in & 2 out
12:24:02 <andythenorth> 2. I'm really unclear where the canonical PR for this is :P
12:24:27 <andythenorth> I'm using peter's rebase, but we can't push any more fixes to that :P
12:26:24 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh requested changes for pull request #7080: Change: Cache stations and links for whole map for linkgraph GUI to eliminate delay when scrolling or zooming https://git.io/fhbvE
12:28:50 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh closed pull request #7005: Fix #7004: Redraw linkgraph overlay correctly after zoom https://git.io/fhtUO
12:28:51 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7005: Fix #7004: Redraw linkgraph overlay correctly after zoom https://git.io/fhbvg
12:31:04 <andythenorth> the bot doesn't see nml repo?
12:31:10 <nielsm> yay PRs down to two pages
12:31:16 <andythenorth> come on DorpsGek_II, keep up :P
12:31:23 <TrueBrain> andythenorth: because nobody told him to
12:31:29 <TrueBrain> and why are issues disabled on the NML repo?
12:31:37 <andythenorth> DorpsGek_II watch the nml repo :P
12:31:53 <LordAro> TrueBrain: because no one wanted to migrate them from OTTDC
12:31:57 <andythenorth> TrueBrain: because Frosch was going to import from devzone
12:32:02 <andythenorth> he had a test for it
12:32:31 <andythenorth> nielsm: I would have liked to know what it does, like a screenshot or so
12:32:48 <LordAro> TrueBrain: bit late to maintain the issue numbers now, so probably just migrate the open ones
12:32:53 <andythenorth> it's a change with no context, no explanation
12:32:56 <nielsm> I wonder if I can somehow force the issue it appears to solve
12:33:19 <andythenorth> it's poor habits
12:33:36 <TrueBrain> LordAro: I hope frosch is doing that? (as it is not on any of my todos)
12:33:48 <andythenorth> we shouldn't encourage this kind of dump-and-run PR, it puts all the quality checking onto the reviewer
12:34:04 <LordAro> i do recall an image somewhere..
12:34:09 <LordAro> but yeah, should be in the issue
12:34:15 <TrueBrain> hmm .. NML is also not forcing review
12:34:31 <TrueBrain> someone has been slacking ;)
12:34:42 <andythenorth> no maintainer :(
12:34:52 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7194: Fix: Remove desert around lakes upon generation. https://git.io/fhbvr
12:34:52 <andythenorth> does anything actually have a maintainer?
12:34:56 <andythenorth> or is that a broken concept?
12:36:14 <DorpsGek_II> [OpenTTD/nml] TrueBrain commented on pull request #15: Industries: support 16 cargos in / 16 cargos out https://git.io/fhbv6
12:36:31 <TrueBrain> but that repo does need some love in terms of configuration
12:36:51 <TrueBrain> nobody talked to me about any of it, so I don't know if anyone is on top of that, or that it was assumed it "magically" would work :P
12:38:58 <TrueBrain> there are also weird branches in the nml repo :P
12:39:11 <andythenorth> really nml is a bit fucked currently
12:39:21 <andythenorth> it's a pretty critical tool for content authors but....
12:39:33 <TrueBrain> I am more talking about the repo; not the content :)
12:39:35 <andythenorth> - bad repo / CI environment
12:39:37 *** synchris has joined #openttd
12:39:41 <LordAro> shall we get RC1 out of the door first, then worry about nml?
12:39:52 <andythenorth> - no reference grfs
12:39:55 <andythenorth> - spec is a PITA
12:40:03 <andythenorth> - and nmlc is slow as all hell to use
12:40:08 <andythenorth> LordAro: maybe :P
12:40:16 <andythenorth> that 16 cargo stuff is pretty untested tbh
12:40:20 <TrueBrain> LordAro: if you expect people that can help with RC1 are in the same group as NML :D
12:40:22 <andythenorth> and it's shipped
12:40:38 <LordAro> TrueBrain: well, it's you, basically :p
12:40:56 <LordAro> i forsee some sort of issue with creating a release from a branch
12:41:00 <TrueBrain> what am I suppose to do for either one? :)
12:41:27 <TrueBrain> if the branch is correct, I think it just works tbh :)
12:41:31 <TrueBrain> but we will find out :D
12:41:51 <LordAro> and CI setup is basically your area
12:42:19 <TrueBrain> but I am not touching NML currently, to be clear :)
12:42:45 <TrueBrain> well, I am, but the README.md annoys me :P
12:43:55 <LordAro> don't have to touch NML to setup CI for it :p
12:44:08 <planetmaker> issues... are disabled on nml for no good reason, I guess. But it might need migration of devzone issues first
12:44:29 <TrueBrain> LordAro: I assume someone else is on top of that :)
12:44:30 <DorpsGek_II> [OpenTTD/OpenTTD] Nik-mmzd opened issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbvM
12:45:06 <LordAro> TrueBrain: i think you overestimate how much other people understand about the CI
12:45:18 <andythenorth> planetmaker: it's precisely that, frosch wants to move all issues from devzone (or wants someone to :P )
12:45:24 <planetmaker> TrueBrain, so, no, no oversight on your part and no complaint. Migration of nml to github is (obviously) not quite finished. Issues are missing
12:45:28 <TrueBrain> LordAro: I at least assume people ask if that is the case :) Others did for their patchpacks :)
12:45:44 <TrueBrain> planetmaker: cool! As always, let me know if I can help
12:45:54 <planetmaker> and yes, I'd ask, if I wanted you to do something. I didn't consider having nml stuff announced here
12:46:08 <TrueBrain> planetmaker: as long as we are in sync, I am fine :)
12:46:11 <planetmaker> I'm fine with it being announced though (anyone having opinions about that?)
12:46:14 <TrueBrain> as you just noticed, LordAro did make some assumptions :P
12:46:34 <TrueBrain> planetmaker: the announcing is already done :D So no take-backs :P
12:46:44 <planetmaker> LordAro, CI for nml basically is setup. It just needs triggering
12:47:01 <planetmaker> the old CI for devzone still works. Currently just doesn't get a trigger from github
12:47:03 <andythenorth> we should teach it build tests
12:47:20 <TrueBrain> and this is why I ask before I meddle with it ;) Happy to see someone is on top of it :)
12:47:27 <planetmaker> it is modified to build from github though
12:48:18 <TrueBrain> planetmaker: one thing that needs fixing ASAP: "see 'license.txt'"
12:48:22 <TrueBrain> there is no license.txt in the repository
12:49:01 <TrueBrain> planetmaker: is that something you can fix?
12:49:12 <TrueBrain> (README.md points to the file, but it is not there)
12:49:35 <planetmaker> docs/license.txt ;)
12:49:47 <TrueBrain> even GitHub did not pick up on that ;)
12:49:55 <TrueBrain> license files are best in the root directory
12:50:10 <planetmaker> not at all. be my guest
12:50:42 <TrueBrain> it is seriously GPL2 or later? Is that allowed? :D
12:50:52 <planetmaker> why wouldn't it be allowed?
12:51:05 <TrueBrain> thought you need to explicitly dual license it :)
12:51:08 <TrueBrain> never looked into that
12:51:12 <TrueBrain> assumption on my part :)
12:51:49 <planetmaker> which basically is what openttd has, too ;)
12:52:00 <planetmaker> (or has it v2 and no v2+ ?)
12:52:02 <TrueBrain> OpenTTD is licensed under the GNU General Public License version 2.0. For
12:52:02 <TrueBrain> the complete license text, see the file 'COPYING'. This license applies
12:52:02 <TrueBrain> to all files in this distribution, except as noted below.
12:52:13 <TrueBrain> OpenTTD is not dual-licensed
12:54:40 <DorpsGek_II> [OpenTTD/nml] TrueBrain opened pull request #17: Move license file and update readme to be more markdowny https://git.io/fhbvx
12:54:49 <TrueBrain> my OCD hit too often looking at the README.md :P
12:55:23 <TrueBrain> '<filename>' was shown as '' :D
12:56:08 <TrueBrain> pretty sure I missed things, but meh :)
12:56:08 <planetmaker> it needs testing whether the bundle still ships the license :P
12:57:15 <DorpsGek_II> [OpenTTD/nml] TrueBrain updated pull request #17: Move license file and update readme to be more markdowny https://git.io/fhbvx
12:58:02 <planetmaker> anyway... I'm lazy. I approve it. I build it. And will fix it, if it fails to ship it :P
12:58:13 <DorpsGek_II> [OpenTTD/nml] planetmaker approved pull request #17: Move license file and update readme to be more markdowny https://git.io/fhbfv
12:58:44 <planetmaker> But... commit message doesn't fit style :)
12:58:45 <DorpsGek_II> [OpenTTD/nml] TrueBrain merged pull request #17: Move license file and update readme to be more markdowny https://git.io/fhbvx
12:59:35 <planetmaker> everything. Fix: ... or Change: ...
12:59:48 <planetmaker> it should follow the same as OpenTTD
12:59:54 <planetmaker> it's not enforced
13:00:33 <TrueBrain> what are you on about? :D
13:00:54 *** Gustavo6056 has joined #openttd
13:00:58 <planetmaker> I'm... again confused about PR title and commit messages
13:01:12 <TrueBrain> but I did make a boo-boo :)
13:01:37 <TrueBrain> now GitHub picked up on the license too, sweet
13:03:55 <planetmaker> hm... did s/o trigger NML build on devzone?
13:04:28 <TrueBrain> hmm, lunchtime I guess
13:04:40 <planetmaker> because... it just builds w/o me triggering it :D
13:04:58 <planetmaker> so some trigger seems to work
13:05:06 <andythenorth> I didn't trigger
13:05:16 *** Gustavo6056 is now known as Gustavo6046
13:05:24 <andythenorth> frosch might have already made Jenkins trigger?
13:05:47 <planetmaker> I just have to teach it to not show 'failed'. Because it didn't fail
13:06:06 <Wolf01> Hello, from the new internet line \o/
13:06:21 <planetmaker> hi Wolf01 with new internet :D
13:07:55 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7194: Fix: Remove desert around lakes upon generation. https://git.io/fhbft
13:08:26 <nielsm> there, that should show the intention of the patch :)
13:10:02 <LordAro> nielsm: i seem to recall it being possible to reproduce in the scenario editor
13:10:45 <nielsm> LordAro, I think there was a previous bug already fixed where placing river tiles did not convert desert to tropical
13:11:14 <nielsm> but otherwise I don't know of any regular way to trigger the river generator outside automatic terrain generation
13:12:55 <andythenorth> I can't meaningfully review it
13:14:31 <andythenorth> (lack of skill on my part)
13:15:34 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7194: Fix: Remove desert around lakes upon generation. https://git.io/fhbfn
13:16:15 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh merged pull request #7194: Fix: Remove desert around lakes upon generation. https://git.io/fhHqA
13:18:35 <LordAro> i cannot reproduce 7266
13:20:08 <LordAro> i did manage to make the "small news window" shifted up above the toolbar though
13:21:36 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbf4
13:21:52 <LordAro> not for me, with the default font
13:22:30 <LordAro> also the toolbar is unusable at higher font size levels
13:36:29 <DorpsGek_II> [OpenTTD/OpenTTD] Nik-mmzd commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbfD
13:36:36 <DorpsGek_II> [OpenTTD/nml] planetmaker opened pull request #20: Fix/Change: Publish readme and license also separately on the bundles… https://git.io/fhbfy
13:36:42 <DorpsGek_II> [OpenTTD/nml] planetmaker merged pull request #20: Fix/Change: Publish readme and license also separately on the bundles… https://git.io/fhbfy
13:40:04 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbfQ
13:43:11 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbfd
13:48:53 *** sla_ro|master has joined #openttd
13:50:09 <TrueBrain> shouldnt in the issue template also be which OS the user is using?
13:55:20 <andythenorth> I've been thinking for a while about making reference.grf
13:55:38 <andythenorth> 'making' = 'not me alone' :P
13:56:04 <TrueBrain> what is it referencing?
13:59:05 <andythenorth> 1. it contains cases for most of the common uses for each feature type
13:59:23 <andythenorth> 2. it acts as a regression test for nmlc changes
13:59:45 <andythenorth> 3. it provides reference cases for testing OpenTTD changes
13:59:46 <TrueBrain> nml repo has tons of files in regression folder already?
14:00:15 <planetmaker> yes, a few. But the regression coverage still is quite poor
14:00:18 <andythenorth> they're not maintained
14:00:31 <andythenorth> that's fixable :P
14:00:50 <andythenorth> but none of them produce a grf that can be used to test an OpenTTD PR
14:00:55 <planetmaker> well... 'not maintained'... yes. Basically they're for "as long as the regression tests don't fail, they're valid and ok"
14:01:03 <planetmaker> but... new stuff needs new tests :)
14:01:27 <andythenorth> the biggest PITA seems to be extending newgrf spec, or changing things which affect existing newgrf spec
14:01:36 <andythenorth> $someone has to make a grf
14:01:39 <planetmaker> some were made small on purpose, not producing a complete grf, not meant for OpenTTD testing
14:02:05 <andythenorth> currently for OpenTTD PRs, $someone has to make a test grf
14:02:08 <planetmaker> (just explaining their history, not contradicting you :) )
14:02:17 <andythenorth> which is usually a hack of something like FIRS or USSR set
14:02:20 <planetmaker> And I know you know :P
14:02:41 <andythenorth> to make that, $someone usually has to have a patch of nml
14:02:49 <andythenorth> against a spec which is a moving target
14:03:05 <andythenorth> or use grfcodec, which 'just works' but causes the nml part to be neglected
14:03:22 <andythenorth> it's totally unappealing, and makes newgrf spec stuff like pulling teeth
14:03:55 <planetmaker> next step here would probably also to build NML PRs
14:04:39 <andythenorth> for 6867, I have a fork of nml, which is using a remote of peter1138's rebase
14:04:50 <TrueBrain> Tetris 99 is WEIRD :P
14:05:05 <andythenorth> I then build a branch of FIRS, which is slightly broken because the makefile isn't ported for git yet
14:05:23 <andythenorth> I'm using docs from the OpenTTD PR, which may or may not be accurate for the final spec
14:06:30 <andythenorth> we didn't discover until today that those patches break all existing grfs using prod cb
14:06:39 <andythenorth> because there's no CI and no reference grf
14:06:59 <andythenorth> this is all just standard crap, I know
14:07:04 <andythenorth> I'm just trying to figure out the fix
14:07:28 <planetmaker> *ouch (probably is the English translation :D)
14:07:46 <andythenorth> then at the end of all this, me or $someone has to go and edit docs in a wiki
14:08:02 <planetmaker> so... nml wiki inside git repo?
14:08:03 <andythenorth> we can't prep the docs as the spec, alongside the PR
14:08:09 <planetmaker> and newgrf wiki also inside git repo?
14:08:14 <andythenorth> something like thaat
14:08:31 <planetmaker> but... newgrf wiki is edited by way more people than know how to use git
14:08:41 <andythenorth> there are _lots_ of successful nml authors in forums, I am considering trying to recruit them to make reference cases
14:08:52 <andythenorth> but I don't know if my plan has any support or is coherent even
14:09:17 <andythenorth> newgrf is the most important content API imho, and developing for it is a bag of spanners
14:09:35 <planetmaker> the API has serious history issues
14:09:47 <planetmaker> and is... baroque at best
14:10:01 <planetmaker> which is the main cause for all the bag of spanners
14:10:21 <TrueBrain> so integrate Squirrel with NewGRF you say? :P
14:10:21 <planetmaker> factorio is less history - obcessed ;)
14:11:22 <planetmaker> sounds insane. Might be less than it sounds. But that's a 2.0 project, if any at all
14:11:38 <planetmaker> in essence newgrfs could be part of a game script
14:11:52 <planetmaker> it's the same vector: global changes
14:12:25 <andythenorth> it would be a substantial decision to switch API :)
14:12:39 <andythenorth> and it would still have infrastructure issues, like docs, and test cases
14:12:45 <andythenorth> even in a new shiny formaat
14:13:04 <TrueBrain> if done properly, old format should "just work"
14:13:19 <TrueBrain> but I never understood NewGRF, so easy for me to say :D
14:13:44 <andythenorth> newgrf is simples
14:14:03 <andythenorth> just put digits 0-F in order until you get a lot of trains
14:14:28 <TrueBrain> I am guessing that the main difference between GS API and NewGRF is that the latter is called, and the first does the calling
14:15:13 * andythenorth bisects to find where nml is broken
14:15:58 <peter1138> Typical, I go out on a bike ride and I come back to a lot of chat.
14:17:03 <andythenorth> nielsm: my naive assumption is that it's simply using the wrong cb version, so parsing wrong
14:17:11 <andythenorth> but I don't know how to prove that, or try to fix it
14:17:29 <nielsm> parsing comes before picking cb version
14:17:45 <nielsm> it's probably the syntax declaration that's wrong
14:18:02 <planetmaker> traditionally NML does not support multiple ways to do one thing. It just supportes the newest way to do stuff
14:18:39 <nielsm> planetmaker: mainly the thing here is to keep supporting the old syntax I suppose
14:18:58 <nielsm> so you don't have to rewrite large chunks to use the new version of nmlc
14:20:04 <peter1138> nielsm, #7080 introduces a shed load of new code for something very minor.
14:20:05 <planetmaker> I'd keep the old branch with the old version. And new versions require adoption of syntax... bad luck.
14:20:40 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7080: Change: Cache stations and links for whole map for linkgraph GUI to eliminate delay when scrolling or zooming https://git.io/fhbJG
14:20:46 <planetmaker> providing bug-fix versions to support old syntax can be done for a time. We did so with nml 0.2 vs. nml 0.3+
14:21:37 <peter1138> non-rect-catchment for 1.9?
14:21:59 * peter1138 ponders making a patch pack out of un-merged PRs :p
14:22:05 *** drac_boy has joined #openttd
14:22:11 <drac_boy> weekend hi to anyone here ;)
14:22:12 <nielsm> peter1138 I'd be fine with it
14:22:18 <planetmaker> openttd-pr-pack ;)
14:24:19 <TrueBrain> the more you keep adding to 1.9, the more risk you take; balance that :)
14:24:53 <TrueBrain> (the issue of the ever-moving release moment, as new stuff being added that really needs some more testing :D)
14:25:40 <drac_boy> truebrain maybe separate it between 1.9 and 1.9-bleeding? just suggesting randomly ;)
14:26:28 <TrueBrain> isn't bleeding called 'master'? :)
14:30:07 <drac_boy> truebrain so beside thinking as usual what else're you doing atm?
14:30:33 <TrueBrain> "thinking as usual" .. not sure that is an insult or not :D
14:30:51 <TrueBrain> and currently I am mostly enjoying the sun; what is the real question here? :)
14:31:17 <Eddi|zuHause> the correct reply should have been "same thing we're doing every night, pinky"
14:31:38 <TrueBrain> Eddi|zuHause: meh; made that joke so many times :P
14:31:41 <TrueBrain> but you are right :)
14:31:50 <andythenorth> stop adding to 1.9, start adding to 1.10 :P
14:31:59 <TrueBrain> branch 1.9 already, yes :P
14:32:25 <Eddi|zuHause> branch 1.9, then rename to 2.10 :p
14:32:36 <TrueBrain> okay, I made the mistake of opening the NewGRF spec .. I totally forgot it was build to work with TTDPatch :D
14:36:40 <drac_boy> truebrain well I mean a brain if its not thinking then what else is it doing? :)
14:36:57 <TrueBrain> being true, of course
14:38:00 * drac_boy is just trying to find photos of 3-axle wagons as to answer my question but otherwise just simply has a list of loco to sift editing through
14:38:07 <drac_boy> thats about it for my morning atm
14:39:04 <TrueBrain> I love how the documentation in our newgrf code is out of sync with the real code following :D
14:40:02 <TrueBrain> (the pitfall of documentation in code and somewhere else .. something will get out of sync :P)
14:41:22 <TrueBrain> is TTDp dead btw? (honest question)
14:41:56 <drac_boy> to me its 'old stable' truebrain
14:42:16 <TrueBrain> andythenorth: how come?
14:42:30 <drac_boy> theres still one major feature I would had liked but meh programming isn't really my thing so I'm just going leave it as a distant wish for now
14:43:06 <TrueBrain> meh; the moment you remember you are getting old ..
14:43:34 <drac_boy> 2007? which one are you looking at as its dated well past 2007 here
14:44:52 <drac_boy> truebrain its 2013 .. I guess someone broke the links perhaps
14:45:09 <andythenorth> TrueBrain: define dead?
14:45:12 <TrueBrain> you are not giving any sources either, so .. I am still not the wiser :)
14:45:26 <TrueBrain> andythenorth: fair question; active development of any sorts in the last 5 years
14:45:32 <drac_boy> truebrain well I dunno who broke the links .. you'll have to ask them
14:45:46 <TrueBrain> yeah .. this is not helping :)
14:46:09 <LordAro> wasn't ottd hosting the svn repo last i checked?
14:46:20 <LordAro> that had had *some* updates in the last... 5 years?
14:46:39 <andythenorth> afaik there are dedicated players
14:46:44 <andythenorth> I have no recent evidence though
14:46:54 <andythenorth> generally it's thought to be superior to ottd
14:47:14 <TrueBrain> LordAro: we once build the binaries for TTDp, but that was long long ago
14:47:26 <TrueBrain> cannot remember we ever hosted code :P
14:47:35 <LordAro> perhaps that's what i'm thinking of
14:47:35 <drac_boy> andy but if you were to include the jgr patch into the stable release it would be nearly 100% equal tho (minus dos networking support)
14:47:51 <drac_boy> otherwise ottd as it is now does indeed have several noticable features still missing
14:48:02 <andythenorth> like I said, ttdp is considered superior
14:48:08 <TrueBrain> tnx andythenorth, that is a useful link :)
14:48:11 <andythenorth> and jgr is considered superior to ottd
14:48:29 <andythenorth> it's interesting how single-person projects can deliver so much more than a group collaborative effort
14:48:40 <TrueBrain> how do you define "superior" in this scenario?
14:48:44 <LordAro> considered superior by whom?
14:48:48 <andythenorth> do you read forums?
14:48:55 <andythenorth> or even what drac_boy said above ^^ ?
14:49:08 <andythenorth> JGRPP won some time ago
14:49:22 <TrueBrain> player-base, feature-set, ... ?
14:49:38 <TrueBrain> winning things without context is like not wining at all :D
14:49:58 <drac_boy> btw truebrain I rechecked my patch folder and its dated from october 07, 2013 .. author still is by ttdpatch.net in the source comment so .. again .. who broke the wiki links anyway?
14:49:58 <TrueBrain> (I am honestly curious; not judging)
14:50:02 <andythenorth> it's the preferred choice of important soical social influencers on forums
14:50:20 <peter1138> Yes, everyone uses JGR.
14:50:31 <andythenorth> it's unusable, but eh
14:50:33 <peter1138> Development is faster, he doesn't have to care about breaking master.
14:50:37 <TrueBrain> "the latest version is not working" :P
14:50:48 <andythenorth> the point is though that it won
14:50:53 <peter1138> So there's no PRs, there's no real releases, just builds.
14:50:56 <andythenorth> I remember when ottd won
14:50:59 <andythenorth> and ttdp was dead
14:51:05 <drac_boy> nielsm nice .. btw a little more context just so you know tho, yesterday I had asked if the only 3-axle freight wagons were uk milk cars or were there any other non-passenger examples anywhere too?
14:51:09 <peter1138> The thing is, people either play our official release, or they play JGR
14:51:15 <andythenorth> drac_boy: see rule 1
14:51:16 <LordAro> imo we should work to "fix" JGRPP being "superior"
14:51:27 <peter1138> They don't play our nightlies because if they have to go to the bother to downloading something else, they just get JGR's instead.
14:51:39 <nielsm> drac_boy, I haven't been able to find any 3 axle freight cars from a quick browse
14:51:39 <peter1138> LordAro, yeah but we can't do that until 1.9 is out.
14:51:58 <TrueBrain> LordAro: it is not always important to be "superior". You want to be healthy
14:52:01 <andythenorth> LordAro: I dunno
14:52:07 <LordAro> TrueBrain: hence the liberal use of ""
14:52:09 <TrueBrain> it is okay if patchpacks superseed you in features, functionality, etc
14:52:12 <andythenorth> on the one hand, no
14:52:22 <TrueBrain> you just want to cherrypick the good things :D
14:52:22 <andythenorth> on the other, eventually the new competitor always wins
14:52:24 <drac_boy> nielsm yeah thats what I was starting to suspect this morning .. still looking through more photos before I decide for real :)
14:52:38 <andythenorth> like when git beat hg
14:52:44 <drac_boy> nielsm nice site btw .. going use it for misc photo references for other things tho
14:52:55 <andythenorth> or like when Drupal killed Python
14:53:06 <andythenorth> or Windows killed Mac
14:53:08 <TrueBrain> and Wildstar was the WoW killer
14:53:14 <andythenorth> or Windows NT killed Unix
14:53:24 <andythenorth> yeah, the analogy fails to scale
14:53:33 * andythenorth was just trolling
14:53:44 <andythenorth> JGRPP is fine, it has bus factor 1, forum players love it, and it fits our goals
14:53:47 <TrueBrain> personally, I like it that JGRPP is popular; it is good to know people still want more from OpenTTD :)
14:54:33 <TrueBrain> but ... I play most games with mods, given the chance :D
14:54:41 <TrueBrain> Minecraft .. Factorio .. I never play them vanilla :P
14:54:42 <drac_boy> truebrain yeah .. I still haven't tried it but I guess the reason I like the jgr patch (and keep checking that particular forum thread daily) is that it would let me open my gamesaves without a lot of things breaking
14:54:43 <andythenorth> to include providing a stable platform for PP
14:54:52 <drac_boy> anyway to our own nitpicks tho :)
14:55:14 <andythenorth> drac_boy: are you sure you have that the right way round?
14:55:17 <TrueBrain> andythenorth: that is a good idea
14:55:27 <drac_boy> andy .. I pretty much do
14:55:29 <andythenorth> a fast-moving PP with regular failures?
14:55:31 <TrueBrain> it is how I rewrote this whole CI/CD stuff too .. to make patchpacks a lot easier
14:55:43 <andythenorth> your savegames break more often in regular OpenTTD?
14:55:47 <andythenorth> what breaks in them?
14:56:09 <drac_boy> andy .. why does ottd refuse to understand the metalcargo-only signal at my busy rail station? etca
14:56:10 <andythenorth> given that one of the boat anchors on OpenTTD is the savegame promise
14:56:30 <andythenorth> yes, but what breaks in your savegames?
14:56:46 <andythenorth> OpenTTD spends disproportionate effort on maintaining functioning savegames
14:56:58 <drac_boy> signal routing .. train spec .. pretty much a lot of things that ottd just doesn't even have for some reason
14:57:05 <TrueBrain> you cannot load a JGR savegame into it andythenorth :P
14:57:12 <andythenorth> I don't understand
14:57:17 <andythenorth> maybe I should have coffee
14:57:25 <TrueBrain> get me some too plz
14:57:37 <LordAro> drac_boy: that's not "breaking savegames" that's "game doesn't load savegames from a different game"
14:57:49 <andythenorth> oh that's what he said
14:58:03 <andythenorth> OpenTTD can't open saves from patchpacks
14:58:05 <drac_boy> lordaro its the same game silly .. jgr could had loaded it without breaking things
14:58:06 <LordAro> or perhaps "i would like to continue using my existing JGRPP savegames"
14:58:27 <andythenorth> yes he wants to load unsupported features into OpenTTD afaict
14:58:31 <andythenorth> and have them work
14:58:41 <andythenorth> make a ticket for that?
14:58:48 <TrueBrain> right, CMake .. what did I try to do there .. ah ... right .. I created a parser for the sources.list ... I can be such a silly
14:58:53 <LordAro> from the perspective of savegames, they might as well be completely separate games
14:58:54 <drac_boy> andy the only ticket is to roll jgr into a stable version :)
14:59:30 <LordAro> even if every feature in JGRPP was merged into the mainline, the games would still not be compatible
14:59:38 <andythenorth> nor would it be stable
14:59:56 <drac_boy> nielsm on a side note I'm still trying to figure out how to nicely slot the 3-axle coaches inbetween the 2-axle and 2-truck ones .. I think it probably has a hybrid capacity/speed inbetween both most likely
14:59:58 *** Thedarkb-T60 has joined #openttd
15:00:30 <drac_boy> lordaro so why is it that if I remove the station junctions and other non-jgr items .. the map loads without issue?
15:01:29 <nielsm> drac_boy, I think 3 axle is cheaper to build than bogey type cars, while having larger load capacity than 2 axle cars, but also limiting comfort and ability to take sharp curves
15:01:51 <peter1138> JGRPP has a modified saveload system that can handle way more than what we can do.
15:02:09 <Eddi|zuHause> <andythenorth> it's interesting how single-person projects can deliver so much more than a group collaborative effort <-- that's a slippery slope and culminates in people wanting a "strong leader" instead of this stupid "democracy" that is obviously not working.
15:02:33 <peter1138> Unfortunately that same saveload system makes it hard to cherry pick bits, unless we also took that system.
15:02:40 <andythenorth> Eddi|zuHause: welcome to my day job :P
15:02:51 <drac_boy> nielsm yeah seem we have same thoughts
15:03:24 <Eddi|zuHause> andythenorth: do i at least get paid on your day job? :p
15:03:37 <TrueBrain> the moment you see your own cmake code, and realise that if you didnt do cmake for 3 months, you forgot most of it
15:04:28 <Eddi|zuHause> TrueBrain: i occasionally come about bits of code i wrote and think "wait, i knew how to do this once?"
15:04:40 <TrueBrain> yeah ... but in 3 months time?
15:04:56 <Eddi|zuHause> 3 months is a bit short for that, maybe :p
15:04:58 <drac_boy> nielsm on a similar topic I've been thinking to add extra-long 2-axle good vans .. with certain cargo limit .. they're good at being able to go up to a quite fast speed even almost compatable to trucked vans too
15:05:01 <TrueBrain> (okay, in all fairness, the last 3 months I have been doing a completely different job that filled by brain, but still)
15:05:09 <drac_boy> even if not many railways did actually use them (danish and german for two)
15:05:15 <peter1138> TrueBrain, that's like YAPF
15:05:31 <TrueBrain> LordAro: cool, let me use that :D
15:06:10 <TrueBrain> LordAro: my idea was to remove endian testing, I believe, as cmake has a variable for that :D
15:06:36 <LordAro> TrueBrain: i believe i already did that
15:06:41 <TrueBrain> owh, that is TestBigEndian
15:06:46 <TrueBrain> sorry :D Thought it was a ready variable
15:07:04 <andythenorth> drac_boy: how much grf have you coded, and how much is a paper plan? :)
15:07:27 <LordAro> Eddi|zuHause: is there a good way to transport salvage that doesn't have attach points?
15:07:27 <drac_boy> optional question - does anyone here ever care for using non-revenue wagons in their trains? water gin, tools van, non-baggage brake van, etca
15:07:40 <drac_boy> andy heh well lets see...
15:07:47 <Eddi|zuHause> LordAro: been wondering about that as well :p
15:07:49 <andythenorth> always have cabeese
15:08:12 <andythenorth> drac_boy: most grfs don't provide them, so asking people if they like them...is a non-question ;)
15:08:16 <Eddi|zuHause> LordAro: but have been brushing that concern aside as i have no means to process it anyway yet
15:08:21 <andythenorth> rule 1 applies again
15:08:51 <TrueBrain> LordAro: yet-another-findversion-sh? :D
15:09:06 <drac_boy> coded - editted a few grfs for myself .. coded one small grf just all for myself before (learned how to do grfid/etc kind of thing for myself) but long deleted that .. and as for the paper plan umm well I guess I'll have to get back to you in early march when I've cleanred up a lot of details tho ;)
15:09:28 <TrueBrain> findversion.sh is used by external tooling too
15:09:33 <TrueBrain> not sure that removing it is the best thing atm
15:09:51 <peter1138> drac_boy, when I used to play with newgrfs, yeah...
15:09:52 <LordAro> my changes predate the git related changes anyway
15:09:57 <peter1138> Maybe I should do that again.
15:10:11 <peter1138> But I miss the glory days of UKRS 2, when running sounds were a thing.
15:10:15 <peter1138> I really love running sounds.
15:10:16 <TrueBrain> yeah .. going to remove that commit for now; that is giving a conflict :D
15:10:53 <Eddi|zuHause> TF has running sounds :)
15:11:22 <TrueBrain> I seem to have tested this cmake stuff with MSVC .. no clue how ..
15:11:27 <drac_boy> ukrs2 mm .. I've played with that a little bit (ukrs1 more) .. not really my sort of trains but I'm not going to complain about seeing it on other players' maps tho :)
15:11:37 <andythenorth> peter1138: Horse sounds? o_O
15:11:44 <drac_boy> strangely the EM1 was more than often my favorite loco
15:11:53 <andythenorth> so err....what sounds?
15:12:12 <Eddi|zuHause> # what does the horse say?
15:12:13 <andythenorth> oof no, it will make compile times even worse
15:12:26 <andythenorth> Horse is barely possible to develop on already
15:12:35 <andythenorth> I might have to remove features soon
15:12:48 <Eddi|zuHause> you need partial compiles
15:12:54 <andythenorth> there are a lot of varaction2s doing quite silly features
15:13:01 <andythenorth> I should delete them
15:13:05 <andythenorth> also the cargo sprites on wagons
15:13:17 <andythenorth> nmlc is basically unworkable
15:13:22 <Eddi|zuHause> replacing one silly feature by another doesn't sound like the right thing to do
15:13:39 <andythenorth> so you prefer cargo sprites to running sounds? o_O
15:13:43 <drac_boy> have fun getting Iron Horse into the next progress andy ;)
15:14:13 <TrueBrain> LordAro: Unknown CMake command "add_compile_definitions".
15:14:38 <Eddi|zuHause> andythenorth: if that is honestly your question, i think yes, i'd prioritize proper cargo sprites over running sounds :p
15:15:19 <TrueBrain> Debian Stretch is 3.7
15:17:10 <drac_boy> huh that jernbanen site is quite big .. I'm going to spend a long time browsing through it ... fun
15:19:23 <LordAro> TrueBrain: that's a shame
15:19:33 <LordAro> is what i get for developing on arch though :p
15:19:48 <LordAro> i think set(CFLAGS ...) is approximately equivalent?
15:19:53 <TrueBrain> okay, this is funny .. I have a space in my path
15:19:58 <andythenorth> let's try deleting cargo support
15:20:01 <TrueBrain> now it changes some files to a\\ b, instead of a\ b
15:20:11 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbJb
15:20:11 <TrueBrain> LordAro: add_definition(-D..)
15:21:59 <andythenorth> yeah cargo support needs to die in Horse
15:24:22 <TrueBrain> NO_NETWORK is broken again :(
15:24:32 <TrueBrain> toolbar_gui.cpp, line 165
15:24:36 <TrueBrain> ‘NetworkCompanyIsPassworded’
15:24:58 <andythenorth> with cargos 1m3.515s, without cargos 0m29.445s
15:25:02 <andythenorth> it's pretty conclusive
15:25:11 <DorpsGek_II> [OpenTTD/OpenTTD] gregcarlin commented on pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fhbJN
15:26:06 <andythenorth> or I switch back to nfo
15:26:13 <andythenorth> grfcodec takes 0m2.758s
15:26:22 <andythenorth> that's pretty conclusively damning of nmlc
15:26:53 <andythenorth> nmlc would be about twice as slow as this, except frosch put a cache in, and I do some stupid MP stuff in the compile
15:27:08 <drac_boy> I know the game engine doesn't even factor this but any of you know if trailer-style talgo wagons were ever made to "only run fast in one direction" or thats just a silly question? (I don't mean the one where the axle sits inbetween each cars but rather the one where its one axle per car "hitched non-axleend into next car"
15:28:14 <andythenorth> I could switch to m4nfo
15:28:20 *** supermop_Home_ has joined #openttd
15:29:58 <andythenorth> I don't really want to depend on support from an author who does all but call me out by name in his docs
15:30:14 <andythenorth> "In addition, m4nfo natively includes a powerful macro processor suitable for private customizing or templating applications (indeed, m4nfo is written in M4, a very efficient macro processor itself), so there's no need for any external tools, like CPP for macro usage or artificially crafted extra Python layers, resulting in bloated installations."
15:30:36 <andythenorth> not sure I want to join that club
15:31:30 <peter1138> andythenorth, why does build time matter?
15:33:16 <andythenorth> because I build constantly
15:33:33 <andythenorth> I will sometimes build 30 times in an hour
15:33:46 <andythenorth> I can't build more than that because there are only 60mins in 1 hour :P
15:33:46 <peter1138> And that's why you never get anything done...
15:33:59 <peter1138> Why do you rebuild so often?
15:35:04 <andythenorth> it's the only way to see changes
15:35:15 <andythenorth> there's no equivalent of browser developer tools
15:35:20 <TrueBrain> LordAro: I get why you ported it to CMake, the findversion.sh
15:37:22 <LordAro> TrueBrain: i'm glad someone does :p
15:37:37 <TrueBrain> it makes it Linux + Windows compatible :P
15:37:44 <TrueBrain> calling findversion.sh is Linux only
15:37:51 <TrueBrain> so ... Windows is a lot harder this way
15:38:06 * drac_boy deletes 12 wagon lines and just stares at table
15:39:56 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbUU
15:40:53 <drac_boy> hmm was going try code a test grf tomorrow but even then still - any of you know if you only can refit pass/mail or could you for example refit to pass/grain or any non-express cargos to latter slot in a dual-cargo vehicle?
15:41:37 <supermop_Home_> it's perplexingly annoying to make a plywood material in vray
15:41:55 <peter1138> drac_boy, pass/mail only. Dual-cargo is _only_ aircraft.
15:42:18 <supermop_Home_> but so close to having an easy way possible
15:42:26 <peter1138> (It's the plane's shadow that holds the mail...)
15:43:46 <peter1138> Damn, this font is weird-ass
15:44:12 <peter1138> lower-case a is taller than other characters :/
15:47:02 <peter1138> So when it 1.9 going to be branched?
15:47:25 <TrueBrain> LordAro: guess 'rev.cpp' should be created in the build folder with cmake, not?
15:47:33 <TrueBrain> (instead of in src/rev.cpp)
15:47:43 <LordAro> TrueBrain: probably, yeah
15:48:32 <drac_boy> hmm ok guess I'll try one of the other thought I had and make an 'empty' articulated train .. and hopefully that'll get me the refit menu I was looking for
15:48:46 * drac_boy goes to look at something else for now
15:49:03 <LordAro> TrueBrain: that was the point i got to, iirc
15:49:07 <peter1138> drac_boy, articulated rail vehicles are two parts, either part can take any type of cargo you like.
15:49:11 <LordAro> various optional dependencies missing, of course
15:49:17 <TrueBrain> LordAro: yeah, I only had to 'fix' findversion.sh
15:49:22 <TrueBrain> and some specific issues with my setup
15:49:26 <peter1138> drac_boy, but dual-cargo, with the special weird refit case, is only aircraft.
15:50:18 <drac_boy> yeah thats what I had somewhat suspected but couldn't be 100% certain .. till now
15:50:22 <TrueBrain> LordAro: ah, includes fail if it is not in src/rev.cpp :)
15:51:46 <drac_boy> anyway have fun with ottd fonts etc .. I'm off for a while here
15:51:52 <LordAro> TrueBrain: hmm. ideally a cmake build should not modify the source folders
15:52:54 <TrueBrain> LordAro: why did you set the generated property on rev.cpp? (curious what that does)
15:54:29 <LordAro> because it's generated :p
15:54:38 <LordAro> i'm not entirely clear what it does either
15:54:49 <TrueBrain> meh; was hoping you knew :)
15:58:51 <peter1138> Hmm, looking at #7051. Seeing a depot sign is a bit weird :-)
16:01:58 *** supermop_Home_ has quit IRC
16:02:46 *** supermop_Home_ has joined #openttd
16:03:17 <TrueBrain> I have Visual Studio Community 2017 ... and it is telling me my trial is over
16:03:51 <TrueBrain> you have to sign in .. lol
16:03:57 <peter1138> Sign in and it gets renewed.
16:04:24 <TrueBrain> There is a problem with your Microsoft account. To fix this, sign in to account.live.com from a browser.
16:04:27 <TrueBrain> this is going just great :D
16:05:06 <TrueBrain> Your account has been temporarily suspended
16:06:00 <TrueBrain> the account I use for Azure Pipelines
16:06:05 <TrueBrain> I tried to login with that ..
16:06:17 <TrueBrain> now it wants my phone number
16:06:40 <peter1138> Urgh, I can't stop eating :/
16:06:54 <TrueBrain> what a stupid shit is this ...
16:06:56 <TrueBrain> it is just harvesting data
16:07:09 <peter1138> Welcome to the modern age :(
16:07:58 <TrueBrain> yeah ... it just needed ANY phone number
16:08:03 <TrueBrain> wasn't even linked to the account or anything
16:08:06 <TrueBrain> just any would work
16:08:54 <TrueBrain> had to double check a few times I was really visiting Microsoft, as that feels like a scam ..
16:09:00 <TrueBrain> especially as you are bounced over several domains
16:10:50 <peter1138> Hmm, do AI scripts check how many ops they have remaining? Or would that just waste ops?
16:16:20 *** supermop_Home_ has quit IRC
16:20:00 <TrueBrain> LordAro: fun fact, your NOT OPTION_WIN32 didn't work :P
16:20:24 <TrueBrain> can be another CMake 3.12+ thingy :D
16:25:39 <peter1138> Ok, well, seeing as we can't add new features until 1.9 is branched, I think I will go play Vivecraft for a bit.
16:25:53 <DorpsGek_II> [OpenTTD/OpenTTD] Nik-mmzd commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbTe
16:27:21 *** supermop_Home has joined #openttd
16:27:44 <TrueBrain> MSVC is a bit weird
16:27:48 <TrueBrain> with the cmake support
16:27:51 <TrueBrain> not sure what to think of it tbh :)
16:28:02 <TrueBrain> but okay .. slowly it is getting somewhere
16:28:03 <TrueBrain> error C2429: attribute 'fallthrough' requires compiler flag '/std:c++17'
16:28:05 <LordAro> llvm make it work, so...
16:28:33 <TrueBrain> what I was hoping, that CMake and MSVC were good friends, so we can drop the 'projects' folder
16:28:40 <TrueBrain> and have one single way of doing it for all platforms
16:28:45 <TrueBrain> but so far I am not the biggest fan
16:28:51 <LordAro> it should be doable, for sure
16:30:10 <milek7> eh, steam-runtime enviornment uses gcc 4.8
16:31:02 <TrueBrain> okay, MSVC is running ... :D
16:31:17 <LordAro> steam runtime environment is basically just a copy of Ubuntu 12.04 system libraries, last time i checked
16:31:42 <TrueBrain> hmmm .. -D ofc doesn't work
16:31:45 <TrueBrain> and is not translated
16:32:14 <TrueBrain> cmake and MSVC are not bestest of friends
16:32:41 <TrueBrain> error C3861: 'SAFEGUARD_DO_NOT_USE_THIS_METHOD': identifier not found
16:34:42 <TrueBrain> so ... why is it using that?
16:34:54 <LordAro> missing define for an include guard, perhaps?
16:39:10 <TrueBrain> okay, first I found out that the endian stuff was not working :D
16:39:16 <TrueBrain> lets fiddle with that first ..
16:41:54 <TrueBrain> okay, funny, things like -D do work via cmake .. not what I expected, but enfin :)
16:42:14 *** Wormnest has joined #openttd
16:42:36 <TrueBrain> LordAro: I really do not understand the SAFEGUARD error. It happens on _tcsncpy()
16:43:23 <LordAro> i don't know about MSVC though, perhaps nielsm knows more?
16:44:02 <peter1138> Hmm, why is CmdCustomNewsItem() in news_*gui*.cpp :/
16:46:31 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbTq
16:47:32 <TrueBrain> if I disable those lines, I also get:
16:47:32 <TrueBrain> unresolved external symbol main referenced in function "int __cdecl __scrt_common_main_seh(void)" (?__scrt_common_main_seh@@YAHXZ)
16:47:35 <nielsm> what are we talking about? :)
16:47:45 <TrueBrain> CMake MSVC compile of OpenTTD :)
16:47:57 <TrueBrain> it is giving me errors I cannot process
16:48:05 <TrueBrain> one of them the above SAFEGUARD_DO_NOT_USE bla
16:48:07 *** gelignite has joined #openttd
16:48:25 <nielsm> that's supposed to be defined by safeguard.h
16:48:33 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbTY
16:49:17 <TrueBrain> yes .. from what I can see, it is defined there, and that is the reason I get this error?
16:51:15 <nielsm> what does __tcsncat expand to, strncat or wcsncat?
16:51:39 <nielsm> (or, is UNICODE defined?)
16:52:32 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbUU
16:53:04 <TrueBrain> no clue if UNICODE is defined .. possibly that is completely missing :)
16:53:31 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbTc
16:53:56 <TrueBrain> okay, this CMake integration needs work .. let me update first .. (cannot even clean the solution :D)
16:55:24 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbT8
16:58:19 * andythenorth was cutting ivy with bare hands
16:58:25 <andythenorth> now has itchy hands
16:59:59 <Samu> if industries can produce 15 cargos
17:00:09 <Samu> the industry directory will look really bad
17:00:26 <TrueBrain> okay, unrelated to above, but strgen is weird
17:00:37 <TrueBrain> it doesn't look in the folder given with -s for the file you give it
17:01:07 <nielsm> Samu yes, let's solve that issue when an industry set that actually uses this exists! :D
17:01:29 <nielsm> and some GUI mockups of ideas how to better present it would be a good idea too
17:04:56 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbTo
17:06:28 <peter1138> Smallmap looks bad with lots of industry/cargo types.
17:06:59 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbTK
17:07:19 <TrueBrain> okay, it is strgen that is stupid, not so much CMake acting up
17:07:28 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7267: Fix #7003: Freetype fonts were scaled by both GUI and Font zoom levels. https://git.io/fhbUU
17:08:17 <TrueBrain> only for Mingw it replaced / with \\
17:10:53 <peter1138> andythenorth, sub-variants?
17:11:14 <andythenorth> imagine the compile time :P
17:11:36 <peter1138> I'm thinking I can steal code from the groups window to do the, er, grouping.
17:11:40 <TrueBrain> hmm .. "/subsystem:console"
17:11:46 <peter1138> And that supports sub-groups, heh.
17:12:30 <peter1138> I played Vivecraft for about 10 minutes. Minecraft isn't very good in VR.
17:12:32 <TrueBrain> nielsm: the define UNICODE is not set
17:12:47 <peter1138> It's neat being able to see everything in its real size, but, meh...
17:13:05 <TrueBrain> nielsm: MSVC refuses to tell me what it would do with the _tcsncat
17:13:15 <TrueBrain> (tells me it cannot resolve it)
17:14:26 <andythenorth> peter1138: do it for 1.9? o_O
17:14:35 <nielsm> and that definition is in tchar.h
17:14:38 <peter1138> andythenorth, I think TrueBrain has banned new features ;)
17:14:58 <peter1138> (I know that's not what he said)
17:15:35 <nielsm> otoh, that code block _tcsncat appears in is only relevant for ancient versions of windows, which we don't even know if can be made work with the C++11 changes
17:15:40 <TrueBrain> hmm .. I think this has to do with how CMake integrates ..
17:16:05 <TrueBrain> I have another at line 495
17:16:10 <TrueBrain> that are the only two in the whole repo
17:16:12 <peter1138> Ahhh, how obvious, "tig" uses insert/delete keys to scroll text up & down by a line.
17:17:39 <TrueBrain> C:\PROGRA~2\MICROS~1\2017\COMMUN~1\VC\Tools\MSVC\1416~1.270\bin\HostX86\x86\cl.exe /nologo /TP -DTTD_ENDIAN=TTD_LITTLE_ENDIAN -I"C:\Users\Mine\OpenTTD Projects\repositories\OpenTTD\src" -I"C:\Users\Mine\OpenTTD Projects\repositories\OpenTTD\src\3rdparty\squirrel\include" -Igenerated /DWIN32 /D_WINDOWS /W3 /GR /EHsc /MDd /Zi /Ob0 /Od /RTC1 -std:c
17:17:39 <TrueBrain> ++17 /showIncludes /FoCMakeFiles\openttd.dir\src\os\windows\win32.cpp.obj /FdCMakeFiles\openttd.dir\ /FS -c "C:\Users\Mine\OpenTTD Projects\repositories\OpenTTD\src\os\windows\win32.cpp"
17:17:51 <TrueBrain> guess it is missing a flag or something
17:22:50 <TrueBrain> okay, fixed the other issues; it now runs on Windows via MSVC :D
17:33:15 <andythenorth> I have failed entirely at 16 cargo nml :P
17:33:36 *** sim-al2 has joined #openttd
17:37:21 <nielsm> peter1138: fyi that patch did _not_ fix #7266 :)
17:39:03 <TrueBrain> is there any real reason we #ifdef complete files while they are also excluded from building if not used?
17:39:22 <TrueBrain> (for example, allegro)
17:39:45 <TrueBrain> but even there we exclude files, not?
17:39:53 <nielsm> they aren't even included in the project file there
17:40:04 <TrueBrain> WIN32_ENABLE_DIRECTMUSIC_SUPPORT
17:40:12 <TrueBrain> I think I can just safely remove it, not?
17:40:36 <nielsm> it only uses the platform sdk now which you're going to have regardless
17:40:59 <nielsm> and if you don't want dmusic then exclude dmusic.cpp from the build
17:41:19 <TrueBrain> what I was thinking too
17:41:29 <TrueBrain> some are a bit more annoying, like UNISCRIBE
17:41:34 <TrueBrain> other pieces needs to know about that
17:41:50 <TrueBrain> not ideal, but what-ever
17:42:08 <nielsm> actually, some of the sound/music drivers that are not allegro test for allegro and fail if allegro is compiled in
17:42:11 *** snail_UES_ has joined #openttd
17:42:42 <TrueBrain> WITH_ALLEGRO is only used in crashlog.cpp, grep tells me?
17:42:50 <TrueBrain> or is it a runtime check?
17:44:39 <TrueBrain> cool, that is no issue for me :)
17:45:04 <TrueBrain> how to detect 64bit in cmake .. hmm
17:49:26 <Samu> firs is a really different beast to tame...
17:50:29 <Samu> anyways, I tried pikka's town and industry newgrf thing
17:51:04 <Samu> max stockpile cargo doesn't work too well now
17:51:25 <Samu> it's behaving more like a max stockpile cargo per company
17:52:42 <Samu> normal and enhanched production also doesn't work that well
17:53:08 <Samu> it's a "per company" that's ruining it
17:55:03 <Samu> it only wants to know if cargo x was delivered, it doesn't trigger any immediate cargo transformation
17:55:34 <Samu> it enhances the base production
17:55:51 <Samu> base production is always free for all :|
17:57:37 <Samu> im more likely to give up
17:58:30 *** Wolf01 is now known as Guest1543
18:00:37 <Samu> oh well, it was a nice test
18:01:08 <Samu> doesn't work (too well or at all) for newgrfs
18:23:12 <Samu> there may be another way
18:24:13 <Samu> I didn't learn of trackcrossestrack stuff for the 90 degrees before
18:24:42 <Samu> it would not need the table, 95% sure
18:29:51 <Samu> i could compute trackdirs
18:29:54 <TrueBrain> hmmm ... SSE doesn't compile via WSL .. annoying
18:30:35 <Samu> extract the next tiles via trackdir and reversed trackdir
18:30:50 <Samu> then do the 90 degrees crossing right there
18:33:03 <peter1138> 16:37 < nielsm> peter1138: fyi that patch did _not_ fix #7266 :)
18:33:09 <peter1138> nielsm, I never said it did?
18:34:19 <peter1138> I discovered it when looking, but no, it's not.
18:34:36 <peter1138> And I was annoyed with myself because I was going to request those changes and just went with it anyway. :p
18:35:19 <peter1138> So anyway, I don't know why the font doesn't go back to normal size. That doesn't happen for me.
18:39:36 <nielsm> are you using bitmap or truetype fonts?
18:40:52 <peter1138> I've been testing all. Even using Hack, Bold.
18:42:49 <TrueBrain> dmusic.cpp(1125): error C2664: 'char *convert_from_fs(const TCHAR *,char *,size_t)': cannot convert argument 1 from 'WCHAR [128]' to 'const TCHAR *'
18:43:34 * andythenorth played tanks a lot
18:43:43 *** Thedarkb-T60 has joined #openttd
18:45:37 <Samu> found a 90 degrees bug with ships
18:46:14 <Wolf01> I read it as "found a 90 degree ship which bugs me"
18:46:25 <Samu> gonna try without using yapf
18:46:29 <TrueBrain> so minus some unicode issues via MSVC, I can build via CMake with most settings enabled :D w00p :)
18:46:55 <TrueBrain> minor shit that fails, like auto-downloading of OpenGFX
18:46:58 <Wolf01> So, this new internet is only 5 times faster than the old one, but it's really noticeable
18:47:04 <TrueBrain> it just tells me I should do that manually .. so some path is wrong
18:48:07 <TrueBrain> and I need to fix some autodetect rules .. but okay .. most stuff is there now :)
18:48:25 <Samu> peter1138, on the last tile
18:48:36 <Samu> the destination tile, the ship can make 90 degree turns
18:49:00 <andythenorth> 90 deg turn banning is the worst thing ever :P
18:49:03 <Wolf01> Also my router self rebooted because reasons and now it's not using anymore 100% cpu,
18:51:13 <nielsm> Samu: how easy is it to reproduce and will players trigger it by chance very often, and be troubled by it?`
18:51:26 <Samu> ew, im testing this not on master, but it shouldn't be any different
18:51:50 <peter1138> Test it on master ;)
18:52:51 *** Gustavo6046 has joined #openttd
18:53:59 *** snail_UES_ has joined #openttd
18:54:36 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on issue #7266: News font size is bugged if font size was changed for a multiple times https://git.io/fhbIf
18:55:12 <peter1138> Hmm, maybe I need to be in a different language. Urgh.
18:56:59 <nielsm> all I do to reproduce it (on windows) is start a new game, build a train route so I have something to generate "first vehicle arrives" news, then go switch font size to quad and back to normal
18:57:05 <nielsm> and run the train to get some news
18:57:16 <peter1138> And the font stays the wrong size?
18:57:28 <nielsm> the news font stays huge yes
18:57:54 <nielsm> i.e. I change the setting before the news is generated
19:02:51 <Samu> funny, can't trigger it on master, why
19:03:59 <peter1138> nielsm, hmm, I've managed to get to show it unzoomed, despite it being to quad size, but only once.
19:04:15 <Samu> Don't call ship pathfinders if there is no available track…
19:04:37 <Samu> it inadvertedly fixed the problem
19:14:43 <DorpsGek_II> [OpenTTD/OpenTTD] Moth-Tolias commented on pull request #7243: Change: Decrease the min #opcodes https://git.io/fhbIR
19:15:01 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #6931: Change: Prevent town growth from blocking ships https://git.io/fhsRY
19:16:52 <peter1138> Samu, just a rebase?
19:17:40 <Samu> was working with 250 something commits behind
19:18:35 <Samu> i'm thinking of another way to do the same thing
19:19:42 <peter1138> It just looks complex
19:20:15 *** joost[m] has joined #openttd
19:22:35 <DorpsGek_II> [OpenTTD/OpenTTD] Moth-Tolias commented on pull request #7243: Change: Decrease the min #opcodes https://git.io/fhbIa
19:22:43 <peter1138> Seems odd that forbidding 90 deg turns comes into it.
19:23:09 *** Thedarkb-T60 has joined #openttd
19:23:33 <peter1138> I'm not sure you even need to check for tracks.
19:24:24 <andythenorth> oof so how does nml work then?
19:24:29 <andythenorth> it must have a big parse step
19:24:39 <andythenorth> then build some kind actions tree
19:24:48 <andythenorth> then some kind of output step
19:26:14 <andythenorth> so I guess I look there :P
19:26:34 <peter1138> andythenorth, I know nothing about NML.
19:26:53 <Faizi> how to start openttd server?
19:27:48 <andythenorth> peter1138: nobody does :)
19:27:59 <peter1138> andythenorth, I don't really know python either.
19:28:23 <andythenorth> seriously wondering about m4nfo
19:28:28 <peter1138> nielsm, what's it like with the old?
19:28:31 <Faizi> peter1138 i have tried many times but i cannot..
19:28:59 <andythenorth> what was wrong with just nfo?
19:29:19 <peter1138> andythenorth, 87 49 db 37 77 64 78 56
19:29:23 <nielsm> peter1138: waiting passengers soar and ratings tank :P
19:29:32 <andythenorth> peter1138: that's not a valid action, obvs
19:29:48 <nielsm> mainly wanted to show that yes with cdist and large cities you can still get plenty of action ;)
19:30:12 <nielsm> 40 buses are not enough to serve that city
19:30:14 *** Thedarkb1-T60 has joined #openttd
19:30:40 <andythenorth> wondering if I can template nfo?
19:30:46 <andythenorth> it's just replacing constants?
19:30:55 <nielsm> andythenorth that's more or less what m4nfo does
19:31:08 <andythenorth> yeah, if m4nfo had a different author, I'd switch
19:31:13 <andythenorth> the syntax looks better
19:32:45 <andythenorth> so to write my own
19:32:55 <peter1138> nielsm, so it doesn't "fix" it? Hmm.
19:32:55 <andythenorth> I'd need a table of all the actions, and their byte sizes
19:33:10 <andythenorth> and I'd need some classes for the actions
19:33:21 * peter1138 attempts to implement Eddi|zuHause's algorithm in #6931
19:33:21 <andythenorth> and I'd need some stuff for bit mangling
19:33:28 <andythenorth> it's the bit mangling I'd fail at :(
19:33:28 <nielsm> peter1138, the new algo makes it much more manageable but yeah it's still possible to overload a bus network with cdist
19:35:23 <andythenorth> could we just move newgrf editing into OpenTTD?
19:36:15 <Eddi|zuHause> if that goes in textbooks and history as "Eddi|zuHause's algorithm"... :p
19:37:59 <andythenorth> Algorithm|zuHause
19:41:41 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #6965: Add: Option for population-linear town cargo generation https://git.io/fhbIP
19:42:57 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh updated pull request #6965: Add: Option for population-linear town cargo generation https://git.io/fpkqa
19:53:17 <planetmaker> <andythenorth> wondering if I can template nfo? <-- you can. You did. Or ye ancient 2cctrains did
19:53:39 <planetmaker> but you have about 1000% more boiler plate and hassle
19:54:06 <planetmaker> so you basically have the choice: invest time in nml (and have that shit done once and for all). Or invest that time into every NewGRF
19:54:36 <planetmaker> for a one - on - one comparison it's about even. For doing more than one NewGRF: do the math ;)
19:54:45 <andythenorth> I was thinking of writing a compiler
19:54:50 <andythenorth> 'thinking' is a strong word
19:54:59 <andythenorth> I was considering the downsides of writing a compiler :P
19:55:10 <nielsm> you mean designing a new nml? :D
19:55:21 <andythenorth> 'designing' is a strong word
19:55:32 <planetmaker> nmlc started as a tool which output nfo
19:56:19 <planetmaker> nml surely has a lot of shortcomings. The biggest plus is: it's there. And not vapourware
19:56:27 <andythenorth> it's so slow though
19:56:35 <andythenorth> and no-one wants to maintain it :)
19:56:49 <andythenorth> the upsides are large
19:56:54 <andythenorth> but it's kind of dead and dying
19:57:09 <andythenorth> I seriously considered buying a desktop machine just for compiling newgrfs
19:57:22 <planetmaker> seriously... how bad can it be?
19:57:23 <TrueBrain> or .. use azure pipelines :D
19:57:44 <andythenorth> but it won't go much faster, it's single threaded, and single-threaded CPU perf is underwhelming
19:58:16 <planetmaker> and honestly, your firs surely eat twice the time because your wrote a programming language to write nml ;)
19:58:26 <andythenorth> planetmaker: it's about 1m10s versus 2s
19:58:35 <andythenorth> and no, the python steps are marginal
19:58:38 <andythenorth> nmlc is a slow slow pig
19:58:41 <peter1138> Eddi|zuHause, hmm, seems to work.
19:58:45 <andythenorth> and it's not keeping up with OpenTTD
19:58:55 <andythenorth> so I can't make newgrfs to test nightly features
19:59:05 <andythenorth> as a project, it's currently nead ded
19:59:20 <andythenorth> if I want to test nightlies I need to use nfo
19:59:31 <andythenorth> so I might as well go back to that and get the speed benefit
20:00:05 <andythenorth> I did try writing a linker for nml fragments, but it failed on string IDs
20:00:15 <andythenorth> it was much faster, except it didn't work
20:00:21 <peter1138> Hmm, do towns ever raise water to build on?
20:00:56 <Eddi|zuHause> andythenorth: reimplement nmlc in C++? :)
20:01:01 <Samu> they can remove houses that were built on water
20:01:27 <planetmaker> he... it is really the nmlc which takes ages on firs
20:01:30 <andythenorth> wonder how it runs with pypy
20:01:50 <peter1138> This algorithm is a bit simpler.
20:01:57 <Eddi|zuHause> when CETS blew up, it was mostly the lex/yacc part of nmlc that was sloooooow
20:02:14 <andythenorth> I think frosch cached the parsing step somehow
20:02:14 <Eddi|zuHause> it has a problem with large files
20:02:21 <planetmaker> sprite encoding is done in a C-module
20:02:34 <Eddi|zuHause> there's a sprite cache
20:02:34 <andythenorth> anyone got pypy installed?
20:02:34 <planetmaker> frosch wrote it time ago
20:03:37 <Eddi|zuHause> yeah, i vaguely remember a sprite cache
20:03:51 <planetmaker> that. But also the encoding itself
20:04:58 <Eddi|zuHause> "ImportError: No module named chameleon"
20:04:58 <LordAro> if there's a C-module, pypy won't work at all
20:05:11 <DorpsGek_II> [OpenTTD/OpenTTD] michicc approved pull request #7265: Fix #7004: Mark linkgraph dirty to be rebuilt on next draw call. https://git.io/fhbIj
20:05:46 <planetmaker> LordAro, it's not required
20:05:55 <planetmaker> it just is quite a speed-up
20:06:10 <Eddi|zuHause> "ImportError: No module named markdown"
20:06:13 <Eddi|zuHause> what is this magic?
20:06:20 <planetmaker> at least I think. I didn't test for ages
20:06:28 <nielsm> I'm surprised the lex/parse of nmlc is the expensive part, usually I'd expect that to not be too bad
20:06:34 <Eddi|zuHause> also, this is still python2? or
20:06:37 <nielsm> maybe the library used for it is too much meta?
20:07:04 <nielsm> I think it uses some funky docstring magic
20:07:15 <LordAro> what's doing the parsing?
20:07:17 <nielsm> to declare the syntax rules
20:07:18 <andythenorth> Eddi|zuHause: you trying FIRS?
20:07:54 <Eddi|zuHause> andythenorth: it didn't work when i did pip3, only after pip2 it worked
20:08:12 <Eddi|zuHause> or am i having ancient firs?
20:09:00 <Eddi|zuHause> <nielsm> I'm surprised the lex/parse of nmlc is the expensive part <-- iirc it had to do with string handling
20:09:26 <Eddi|zuHause> python doesn't do dynamic strings very well
20:10:30 *** sla_ro|master has joined #openttd
20:11:36 <andythenorth> Eddi|zuHause: that's appalling times :P
20:11:40 <andythenorth> how old is your CPU?
20:11:43 <DorpsGek_II> [OpenTTD/OpenTTD] michicc commented on pull request #7147: Change: Synchronize randomness in vehicle introduction… https://git.io/fhbLU
20:11:47 <Eddi|zuHause> hm... 6 years or so?
20:12:09 <Eddi|zuHause> i wanted to buy a new one, never really got around
20:12:33 <planetmaker> for good factorio fun I'll need a new computer... somewhen. Or at least graphics card :P
20:12:55 <LordAro> planetmaker: there was a recent FFF about how they made the whole thing much more efficient
20:13:14 <planetmaker> LordAro, yes, I read it. Yet... my GPU is... really lacking :)
20:13:18 <Faizi> Can some one tell me the link of original Open TTD download
20:13:28 <Eddi|zuHause> there is no such link
20:13:52 <planetmaker> thank you Eddi|zuHause :P
20:13:52 <LordAro> what do you mean by "original" ?
20:14:07 <LordAro> does that still work?
20:14:37 <Faizi> some links are fake or down version ORIGINAL
20:15:28 <LordAro> Faizi: which links are you finding that aren't working?
20:15:46 <planetmaker> or rather: where?
20:16:15 <peter1138> So I submitted a PR to Samu :p
20:16:42 <peter1138> Assuming Eddi|zuHause's algorithm is correct, that should be it.
20:16:49 <peter1138> And it does certainly seem to work for mel.
20:17:30 <nielsm> so you made a pr to modify a pr
20:17:50 <LordAro> guess we'll never know
20:18:11 <peter1138> This is what distributed version control is about ;)
20:19:45 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7265: Fix #7004: Mark linkgraph dirty to be rebuilt on next draw call. https://git.io/fhFpA
20:19:51 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN closed issue #7004: Cargo flow legend does not get redrawn in additional area when zoomed out https://git.io/fhLjv
20:20:31 <Samu> ah, i was working on it too
20:20:45 <planetmaker> andythenorth, with pypy it doesn't find my ply... not sure how to tell it about it
20:20:50 <Samu> was trying a way which wouldnt need to use tables
20:21:04 <andythenorth> planetmaker: have you got pypy virtualenv?
20:21:17 <nielsm> when they make things simples to express
20:21:30 <planetmaker> andythenorth, no, not yet
20:21:47 <planetmaker> I wonder ... whould work in a conda env
20:26:39 <peter1138> They're not the end of the world.
20:26:55 <peter1138> In this case we need 3 results, and the table means we don't need to calculate it all.
20:27:09 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #7254: Codechange: introduce a few unit tests https://git.io/fhbLs
20:27:36 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7147: Change: Synchronize randomness in vehicle introduction… https://git.io/fhbLG
20:28:46 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7262: Change: Smooth AI CPU usage by spreading out AI ticks in relation to competitor speed. https://git.io/fhbLn
20:29:55 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7147: Change: Synchronize randomness in vehicle introduction… https://git.io/fhbLC
20:32:41 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #7238: Codechange: Remove assert when trying to intersect two tile areas and… https://git.io/fhbLW
20:35:03 <TrueBrain> wait, what, unittests? :P
20:35:13 <TrueBrain> that is to modern, stop doing that :(
20:35:34 <LordAro> TrueBrain: current approach is certainly all wrong :p
20:35:56 <TrueBrain> right, why doesn't the game want to download gfx on its own ..
20:36:02 <TrueBrain> why does it sometimes say: he, I can download shit for you
20:36:06 <TrueBrain> other times it just gives this error message
20:36:08 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7147: Change: Synchronize randomness in vehicle introduction… https://git.io/fhbL8
20:37:38 <LordAro> i've never been certain of the distinction either
20:38:06 <TrueBrain> owh, it complains it cannot find sound sets
20:38:08 <TrueBrain> like that is important
20:39:54 *** Beerbelott has joined #openttd
20:40:21 <Beerbelott> Do you know if there is a way for a password-protected password to avoid sending the list of companies on query?
20:40:52 <peter1138> Second password should be server?
20:42:13 <nielsm> I don't think there is an option to keep the companies list secret for clients that don't have the server password, no
20:42:37 <LordAro> maybe someone should file an issue about it
20:44:59 <nielsm> there's another suggestion to always join a server as spectator isn't there
20:45:07 <nielsm> which would also sort of solve the problem
20:45:13 <nielsm> maybe merge the two idea?
20:45:32 <nielsm> "force join as spectator" which does not present the company list before joining the server
20:48:10 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbLE
20:49:09 *** Thedarkb1-T60 has joined #openttd
20:53:16 <TrueBrain> that moment you follow a manual, and it doesn't work
20:56:34 <TrueBrain> ah .. had to clean the cache
20:56:40 <TrueBrain> vcpkg and cmake are also good friends
20:57:50 <TrueBrain> ah, working directory support from CMake is only in VS2019
20:57:54 <TrueBrain> that is annoying :D
20:58:25 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbLo
21:00:30 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbL6
21:02:28 <andythenorth> so I need to put some prints into nml eh
21:02:33 <andythenorth> I never did learn pdb :P
21:03:00 <nielsm> there is a python debugger?
21:03:31 <nielsm> I always made do with logging, and splitting thing into parts I could try in the repl
21:03:38 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbLX
21:03:57 <andythenorth> I suspect the trouble is in p_produce2
21:04:20 *** snail_UES_ has joined #openttd
21:04:33 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbL1
21:05:21 <peter1138> This is awkard. A PR review is just one notification. But the replies to each point in it are individual notices!
21:05:27 <peter1138> Sorry for spamming :p
21:06:50 <TrueBrain> one of the more annoying things about GitHub :(
21:06:57 <TrueBrain> also depends on how you review, tbh
21:07:40 <andythenorth> hmm, all I can get it to do is print the invalid token :(
21:07:51 <andythenorth> can't see anywhere p_error is called
21:07:53 <TrueBrain> why do I need a soundset to play the game? :(
21:07:55 <peter1138> Should I cycle tomorrow, or just piss about on the computer all morning?
21:07:58 <peter1138> TrueBrain, you don't.
21:07:59 <TrueBrain> why does a dedicated server need a soundset ..
21:08:04 <glx> review can spam notification if you do individual comments ;)
21:08:06 <TrueBrain> "Error: Failed to find a sounds set. Please acquire a sounds set for OpenTTD. See section 4.1 of README.md."
21:08:27 <peter1138> I may need the baseset metadata for no sound, however.
21:08:54 <glx> no sound is present by default IIRC
21:09:41 <nielsm> the build process is supposed to put no_sound.obs in bin/baseset/
21:09:51 <TrueBrain> one thing that always annoyed me about OpenTTD, but I never looked into it .. you need things in places which are .. hard to figure out
21:09:54 <TrueBrain> once it works, it works
21:11:42 <TrueBrain> nielsm: tnx, that was the piece I needed
21:11:53 <TrueBrain> would be nice if the error suggested that :P
21:12:04 <TrueBrain> Base musicset detail loading: description field missing.
21:12:07 <TrueBrain> : /orig_dos.obm readable for the user running OpenTTD?
21:12:30 <andythenorth> wonder if this commit only works at all with the following commit
21:12:43 <andythenorth> should the be squashed together?
21:12:58 <andythenorth> 34ad7c6f51b5840e4f5e79ec783a456cec22ecab
21:13:00 <nielsm> those two lines are those defining the syntax
21:13:04 <andythenorth> nd ec4a9f4e6e1785028c829c05b84323220eab6041
21:13:04 <Samu> TrackToTrackdir is ambiguous
21:13:32 <nielsm> but the p_produce definition above is unchanged and should still allow the original syntax
21:13:37 <glx> it converts a Track into a Trackdir
21:13:44 <andythenorth> that's what confuses me nielsm
21:13:54 <TrueBrain> hmm, 'description' really is missing in the obm files
21:13:59 <nielsm> maybe the syntax needs to be changed in a dumb way
21:14:05 <andythenorth> oh does nml just select on the first 'produce:' it finds?
21:14:14 <andythenorth> I have now idea how the parsing rules work
21:14:17 <nielsm> like make it: PRODUCE LPAREN produce_params RPAREN
21:14:29 <Samu> converts into one of the trackdirs
21:14:34 <nielsm> and then have produce_params be a switch between the three possibilities
21:14:41 <andythenorth> does it look for rules matching the token in the block?
21:14:44 <Beerbelott> nielsm: Thx for the info (sry for delayed answer) I'll file an issue if there is not one already
21:15:12 <nielsm> andythenorth the parser is supposed to try all rules for a definition, as far as I know
21:15:31 <nielsm> so the rule in p_produce and those in p_produce2 are supposed to merge
21:16:33 <nielsm> oh wait, the old produce syntax ends with SEMICOLON and the new one does not
21:16:38 <nielsm> maybe that makes a difference?
21:16:50 <Eddi|zuHause> andythenorth: what is the exact problem you're facing?
21:17:07 <andythenorth> how exact do you want it>
21:17:19 <nielsm> Eddi|zuHause I think it stopped accepting the old syntax
21:17:28 <andythenorth> the goal is to merge the 16 cargos in, 16 cargos out fork
21:17:42 <planetmaker> nielsm, andythenorth just remove the old syntax and implement a new one. No big harm
21:17:42 <andythenorth> currently it's blocked because it won't compile existing grfs
21:17:53 <planetmaker> we maintain compatibility by a (legacy) branch
21:18:09 <glx> TrueBrain: it's ok for me and MSVC
21:18:15 <Eddi|zuHause> andythenorth: example syntax?
21:18:25 <TrueBrain> glx: I am testing this on Linux, so that is possible I guess
21:18:26 <DorpsGek_II> [OpenTTD/OpenTTD] Berbe opened issue #7268: Suggestion: Option not to disclose server information when password-protected https://git.io/fhbLp
21:18:31 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbLh
21:18:31 <TrueBrain> obs is fine, obm is not
21:18:38 <TrueBrain> even on latest master
21:19:08 <planetmaker> treat it like python treats compatibilty :)
21:19:25 <andythenorth> nmlc ERROR: "generated/firs.nml", line 7342: Syntax error, unexpected token "waiting_cargo_1"
21:19:36 <TrueBrain> can any of you test it on linux; do the obm files have a description field?
21:19:38 <LordAro> planetmaker: how much effort would it be for you to make an ogfx 0.5.3(?) with the 4 group sprites its currently missing?
21:19:45 <andythenorth> AIUI, it's now expecting a cargo list
21:20:06 <andythenorth> planetmaker: that's an appealing route, but a big jump :P
21:20:10 <peter1138> They have description, and translations.
21:20:17 <TrueBrain> so ... why don't my obms :(
21:20:22 <TrueBrain> especially as the obs do
21:20:26 <TrueBrain> same codepath, give or take
21:20:47 <TrueBrain> also no error or warning
21:20:50 <Eddi|zuHause> andythenorth: you must merge the produce and produce2 functions, and move the distinction into the semantic analysis
21:21:07 <Eddi|zuHause> andythenorth: the parser cannot distinguish it with the given rules
21:21:53 <LordAro> should we branch then update the changelog & friends (on the branch), or the other way around?
21:22:21 *** Beerbelott has left #openttd
21:22:29 <TrueBrain> executing the awk manually also is without description
21:22:34 <Samu> whatever, gonna test peter1138 stuff instead
21:22:35 <andythenorth> ok so how do we analyse the tokens to figure out which version of the cb we have?
21:22:36 * andythenorth has literally no idea :P
21:23:04 <peter1138> Uh, what... fitbit says I can eat 7000 kcal today. I'm pretty sure that is wrong!
21:23:05 <DorpsGek_II> [OpenTTD/OpenTTD] planetmaker commented on issue #7268: Suggestion: Option not to disclose server information when password-protected https://git.io/fhbtU
21:23:11 <DorpsGek_II> [OpenTTD/OpenTTD] planetmaker closed issue #7268: Suggestion: Option not to disclose server information when password-protected https://git.io/fhbLp
21:23:16 <peter1138> Samu, might as well, I mean I wrote it and tested it and it seems to work.
21:23:22 <peter1138> I didn't even design it ;p
21:23:40 <TrueBrain> so ... what to check why description is empty here?
21:23:49 <TrueBrain> not much debugging info other than: "it does not work" :P
21:23:51 <peter1138> Are your lang files built?
21:24:02 <peter1138> Hmm, no, it doesn't need them built anyway.
21:24:02 <TrueBrain> langfiles.tmp is populated
21:24:07 <glx> english.txt provides the base lang
21:24:15 <TrueBrain> string can be found when I grep for it
21:24:34 <TrueBrain> it pauses for a bit when it should do descriptions, so it is doing something
21:24:42 <Samu> create a merge commit, squash and merge or rebase and merge? what to do?
21:24:53 <Eddi|zuHause> andythenorth: basically, in "produce.Produce()" you must check whether the 2nd and 3rd list entries are cargo lists, then use the new rules, else fall back to the old rules
21:25:07 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on issue #7268: Suggestion: Option not to disclose server information when password-protected https://git.io/fhbtq
21:25:07 <planetmaker> LordAro, not so much. It just needs figuring out which are missing, adding those, and then releasing it
21:25:32 <planetmaker> LordAro, I shall do that before 1.9.0 without any -beta/rc becomes a thing
21:25:42 <planetmaker> but please keep reminding me :)
21:25:43 <TrueBrain> and why does obs work but obm not .. this is weird
21:26:27 <planetmaker> nielsm, sorry, if I was too fast to close the issue :P I really think the main part here is: removing the info is undesirable
21:26:46 <glx> works for mingw too TrueBrain
21:26:50 <Eddi|zuHause> andythenorth: alternatively, move the LBRACKET/RBRACKET out of the produce_cargo_list, into produce2, then that should work like it is now
21:26:53 <Samu> the website is too complex
21:26:57 <TrueBrain> what to do to debug this ..
21:26:58 <peter1138> Oh right, apparently I forgot to stop my ride last night, so even though my Garmin was off, fitbit thinks I was exercising all night.
21:27:05 <peter1138> 4000kcal instead of 300kcal lol
21:29:41 <Eddi|zuHause> andythenorth: so if that reads something like "produce : PRODUCE LPAREN ID COMMA LBRACKET produce_cargo_list RBRACKET COMMA LBRACKET produce_cargo_list RBRACKET COMMA expression RPAREN"
21:31:38 <TrueBrain> there are \rs in my obm files
21:31:40 <TrueBrain> but not in my obs files
21:32:26 <andythenorth> Eddi|zuHause: same error persists :)
21:32:35 <andythenorth> that's interesting
21:32:42 <TrueBrain> that is ... very odd
21:32:51 <TrueBrain> especially as it is on a linux mount
21:33:23 <TrueBrain> it even shows up as a diff
21:33:28 *** Beerbelott has joined #openttd
21:33:31 <glx> we have some broken files in the repo
21:33:49 <TrueBrain> that should give more people issues, not?
21:34:06 <andythenorth> WARNING: There is 1 unused rule
21:34:07 <andythenorth> WARNING: Symbol 'produce_cargo_list' is unreachable
21:34:10 <glx> depends on the file I guess
21:34:22 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #7225: Add: Various AI/GS functions that may be useful. https://git.io/fhbtG
21:34:24 <glx> there's at least 1 cpp with mixed line endings
21:34:25 <andythenorth> and produce2 is not called before the error happens
21:34:27 <TrueBrain> ugh, I cannot commit this, as I have git configured to ignore line endings, and always commit with \n
21:34:39 <Eddi|zuHause> andythenorth: what can i check out?
21:35:30 <TrueBrain> I hate line endings
21:35:38 <Eddi|zuHause> andythenorth: so that approach is not going to work, ats it wants to fit stuff into the p_produce() rule now
21:35:57 <Eddi|zuHause> andythenorth: so you really need the semantic analysis step
21:36:32 <TrueBrain> no clue where the error is, as the line ending keeps on hiding ... it works now, what-ever
21:38:12 <peter1138> Samu, if you want to just test it, "git pr <PR> <REMOTE>" so possibly "git pr 3 origin"
21:38:19 <peter1138> Although... hmm that might be an extension.
21:38:28 <TrueBrain> time to install VS2019 .. see how that CMake integration is doing :)
21:38:38 <Eddi|zuHause> "git: 'pr' ist kein Git-Befehl"
21:38:48 <planetmaker> sounds somewhat non-standard @ peter1138 . Yet like an interesting git extension
21:39:09 <peter1138> Ah yeah, it's in git-extras.
21:39:26 <peter1138> planetmaker, gotta say, it's a godsend.
21:39:36 <faizee> how to start a server in openttd/
21:39:36 <peter1138> I always forget the longhand :P
21:40:06 <glx> oh and of course translations.awk changes EOL on mingw
21:40:18 <TrueBrain> OpenTTD\bin> ..\..\..\..\CMakeBuilds\9a9e0753-b318-cc38-bb77-257a59230490\build\x86-Release\openttd.exe
21:40:31 <nielsm> faizee, what part of setting up the server are you having trouble with?
21:40:47 <faizee> please tell me few steps i cant find any thing on these links
21:40:54 <TrueBrain> glx: is there anything we can do about the awk/vbs .. is there something common between linux/windows .. especially when looking at CMake?
21:40:59 <TrueBrain> we could script it in CMake I guess
21:41:01 <nielsm> faizee, do you have the game installed and can play singleplayer?
21:41:10 <planetmaker> faizee, just start OpenTTD. And start a multiplayer game
21:41:50 <faizee> have you ever started an server?
21:41:55 <planetmaker> you might need to tell your router and your OS to allow connections from extern to your computer. Details... on the URL you had quoted several times
21:42:12 *** planetmaker was kicked by DorpsGek (Wrong channel. Retry in #openttdcoop.)
21:42:17 <Eddi|zuHause> peter1138: as a git-imbecile, how do i make "git pr" work?
21:42:22 *** planetmaker has joined #openttd
21:42:22 *** ChanServ sets mode: +o planetmaker
21:42:23 <TrueBrain> hahahaha :D That is one old command :P
21:42:31 <TrueBrain> the good times people kept thinking this was openttdcoop :D
21:42:35 <peter1138> Install "git-extras" on Debian. I dunno elsehwere.
21:42:45 <glx> I still command the bot yes
21:42:48 *** supermop_Home_ has joined #openttd
21:42:52 <peter1138> You're looking for a package that provides "git-pr"
21:42:58 <peter1138> /usr/bin/git-pr, that is.
21:43:00 <faizee> stoppppppppppppppppppppppppppppppppppppppppppppppppppppppppp
21:43:08 <planetmaker> :) will do. thanks peter
21:43:09 <faizee> staertttttttttttttttttttttttttttttttttttttttttttttttttttttttt
21:43:20 <Eddi|zuHause> peter1138: can't find anything
21:43:28 <peter1138> Eddi|zuHause, what system?
21:44:02 <planetmaker> fetch = +refs/pull/*/head:refs/remotes/origin/pr/*
21:44:02 <planetmaker> <-- I have that in remote "origin". So I can locally checkout each PR
21:44:10 <glx> if no git-extras you still can use alias
21:44:13 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro requested changes for pull request #7193: Fix #6468: Load correct version of AI as specified during the time of its save. https://git.io/fhbtw
21:44:41 <planetmaker> faizee, I guess many here have started a server, both on remote machines as locally... you just need to follow the steps
21:44:46 <planetmaker> and follow them properly
21:46:02 <TrueBrain> that moment you find out VS2019 can't handle spaces in folders when using CMake
21:46:10 <planetmaker> yay... sudo dnf install git-extras :)
21:46:25 <planetmaker> stupid windoze :)
21:46:29 <glx> TrueBrain: for CMake I don't know exactly how it works, I suppose it creates a solution a give it to MSBuild
21:46:32 <Eddi|zuHause> <planetmaker> <-- I have that in remote "origin" <-- more words on that, please?
21:46:43 <TrueBrain> glx: no, not really
21:46:46 <TrueBrain> but it works fine on VS2017
21:46:49 <TrueBrain> they broke it on VS2019
21:46:51 <TrueBrain> it is funny like that
21:47:13 <TrueBrain> (and I have a space in the path exactly for this reason, as I wanted to spot mistakes I made with escaping :D)
21:47:25 <planetmaker> I copied it from fr0sch, iirc
21:48:50 <DorpsGek> LordAro: OpenTTD uses TCP and UDP port 3979 for server <-> client communication, UDP port 3978 for masterserver (advertise) communication (outbound), and TCP port 3978 for content service, a.k.a. BaNaNaS (outbound)
21:48:56 <LordAro> planetmaker: ya numpty
21:49:28 <LordAro> faizee: read the wiki link provided above, then if it doesn't work, tell us what didn't work
21:49:29 <TrueBrain> yippie, VS2019 does do the Working Directory correct
21:49:44 <TrueBrain> okay, I can now build with VS2019 + vcpkg + CMake, and it is even somewhat workable :)
21:49:49 <faizee> my port is 192 168 1 255
21:49:54 <Eddi|zuHause> planetmaker: thx
21:50:06 <planetmaker> faizee, that's your *IP* in *your* local network
21:50:42 <LordAro> not a .255 address it isn't
21:50:52 <LordAro> that's your broadcast address
21:50:58 <glx> TrueBrain: but project files are nice to work with MSVC :)
21:51:02 <LordAro> well, it could be, but it's unlikely
21:51:19 <TrueBrain> glx: possibly we should use folders more :P
21:51:41 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbti
21:51:51 <TrueBrain> as using Visual Studio Code is horrible :P Such a long list of filenames ..
21:52:19 <planetmaker> Eddi|zuHause, backdraw is that "git fetch" pulls *all* pull-requests. But... that's ok for me
21:52:55 <Eddi|zuHause> planetmaker: i usually specify a branch with fetch
21:53:01 <faizee> 119.160.67.150 this is my public
21:53:01 <peter1138> planetmaker, drawback.
21:53:34 <milek7> vs2017 does workdir settings in some json, very un-windows way ;d
21:53:37 <milek7> they changed it in vs2019?
21:53:48 <TrueBrain> LordAro: I pushed a tons of shit to my cmake branch (well, 1 commit, I was lazy)
21:53:52 <planetmaker> faizee, we don't know how to configure your router. Only you know it
21:53:56 <glx> I have pr = "!f() { git fetch -fu ${2:-$(git remote | grep ^upstream || echo origin)} refs/pull/$1/head:pr/$1 && git checkout pr/$1; }; f" in [alias] section of git config
21:54:03 <peter1138> My bottle of wine ran out :(
21:54:14 <TrueBrain> LordAro: feel free to commit in my branch btw
21:54:37 <faizee> where you are doing this/
21:56:27 <Eddi|zuHause> glx: thx, i put that into my global gitconfig
21:56:57 <peter1138> glx, yup, that's what I had before.
21:57:47 <Eddi|zuHause> now... i completely forgot what i tried to do :p
21:59:38 <andythenorth> checkout the nml fork? o-o
21:59:54 <andythenorth> fix the semantic analysis?
22:00:47 <milek7> why would anybody drink alcohol?
22:00:48 <milek7> it has only negative effects
22:01:08 *** Terwa347 has joined #openttd
22:01:21 <planetmaker> I... should have looked at my default .gitconfig
22:01:25 <planetmaker> 20 fetchpr = !sh -c 'git fetch $1 pull/$2/head:pr$2' -
22:01:31 <Eddi|zuHause> milek7: at least historically, it was safer to drink alcoholic drinks than water from some random well
22:01:48 <Eddi|zuHause> ... which may carry all sorts of diseases
22:01:54 <planetmaker> or not default? comment indicates I added it but forgot :P
22:02:18 <planetmaker> 26 # Example usage of "fetchpr":
22:02:18 <planetmaker> 27 # git fetchpr <remote> <pr-number>
22:02:18 <planetmaker> 28 # git fetchpr openttd 6939
22:02:18 <planetmaker> 29 # git checkout pr6939
22:02:34 <Eddi|zuHause> planetmaker: doesn't sound very default-y
22:03:13 <glx> no need to repeat it 3 times ;)
22:05:23 <planetmaker> Faizan, and when you have told your router to do the proper port-forwarding - you might still tell your OS (windows?) to allow that, too in your firewall settings
22:06:34 <TrueBrain> I think it is okay, but I am not sure
22:06:51 <Eddi|zuHause> andythenorth: so, basic strategy: 1) remove p_produce2, 2) p_produce calls a disambiguation function, which checks the types of t[something] whether it's an expresion or a list, and then returns Produce() or Produce2()
22:07:26 <planetmaker> I still don't want two produce... :P
22:07:31 <milek7> Faizan: do you have mobile connection?
22:08:22 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain opened pull request #7269: Fix c3dbe836b4: also compile without ENABLE_NETWORK defined again https://git.io/fhbqt
22:08:27 <TrueBrain> peter1138 is to blame ^^ :P :P
22:08:42 <peter1138> Can we remove that bullshit? :(
22:08:50 <Eddi|zuHause> planetmaker: that would be the next step :)
22:09:13 <peter1138> I mean, the ability to compile without networking.
22:09:26 <TrueBrain> we used to target DOS, which doesn't have networking available
22:09:45 <TrueBrain> possibly a better solution, if we want to keep this, is to stub all the network functions
22:09:50 <TrueBrain> and remove the #ifdefs from all over the place
22:10:20 <TrueBrain> but, removing no-network-support .. doesn't sound like a horrible idea ;)
22:10:26 <peter1138> Yes please. I'll only keep messing it up :-)
22:10:44 <glx> TrueBrain: the change is supposed to activate / to \\ replacement for MSVC ?
22:10:45 <TrueBrain> something for 2.0 :D
22:10:52 <TrueBrain> glx: yes, well, also
22:11:06 <TrueBrain> I was doing non-mingw Windows builds :P
22:11:09 <TrueBrain> (via MSBuild, so yeah)
22:11:10 <glx> because _WIN32 is enabled for all windows compilers
22:11:31 <TrueBrain> yes; and all Windows executables us \ not?
22:11:43 <TrueBrain> so I couldn't really figure out why strgen did do it when compiled with mingw32
22:11:46 <TrueBrain> and not when compiled with msvc
22:11:54 <TrueBrain> meant that if you use the msvc binary in a mingw build, it fails
22:12:15 <TrueBrain> but ... this is exactly why I am asking, as I am not really sure why this came to be :)
22:12:33 <glx> I guess the project files where using \\ only
22:12:51 <TrueBrain> with cmake, things change a bit :)
22:12:56 <TrueBrain> so I take it the change is fine :)
22:13:08 <glx> so enabling it for MSVC should be ok
22:13:38 <TrueBrain> why is there gen-kdev4.sh in the projects folder .. we support that? :)
22:16:11 <glx> TrueBrain: applied change locally, it still works :)
22:16:32 <planetmaker> but... gen-kdev4.sh seems... somewhat buggy:
22:16:37 <planetmaker> ./projects/gen-kdev4.sh: Zeile 19: printf: Fehlende hexadezimale Ziffer nach \x.
22:16:43 <TrueBrain> well, I am going to remove it all, and replace it with cmake :)
22:17:27 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fh5s1
22:18:13 <glx> TrueBrain: and just check on my debian VM, no missing description for obs nor obm
22:18:51 <milek7> Faizan: then it may don't work at all
22:19:12 <milek7> mobile providers typically block incoming connections
22:19:24 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbqn
22:20:13 <glx> even if they don't block all ports you usually don't have access the the configuration
22:20:20 <Eddi|zuHause> andythenorth: oh, other alternative, try changing "p_produce" rule from "expression_list" to "expression COMMA expression COMMA ..."
22:20:59 <Eddi|zuHause> andythenorth: then the parser can distinguish it by the number of commas
22:21:11 <peter1138> LordAro, as a bonus, my non-rect visualisation still works, so nothing broke with those changes. I hope. ;)
22:22:43 *** supermop_Home_ has quit IRC
22:22:58 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain opened pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
22:23:46 <peter1138> Thie removes: configure / Makefile ... Things still to do: remove configure / Makefile
22:23:57 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fhbqC
22:24:19 <TrueBrain> peter1138: cool aint it? :D
22:25:15 <TrueBrain> added details how to test it .. might be nice
22:26:20 <TrueBrain> so far I am pretty happy with CMake .. it really allows to do all our project stuff in a single way, instead of all the different ones we have :)
22:26:44 <planetmaker> TrueBrain, but how do I make it with your PR?
22:26:50 <planetmaker> cmake . does error for me
22:27:17 <TrueBrain> like I just said, I updated it with build details
22:27:22 <TrueBrain> but also, "error for me" is not descriptive
22:27:28 <TrueBrain> if you provide more details, I can help out :)
22:27:36 <glx> oh I already have cmake, need to check if it's the vcpk one though
22:28:03 <planetmaker> it wants me to run it from the bin dir...
22:28:13 <TrueBrain> glx: I haven't tested mingw btw; for MSVC you need to add some lines in the CMakeSettings.json; but vcpkg explains that
22:28:23 <TrueBrain> planetmaker: either follow what I wrote in the PR, or follow what is on your screen :)
22:28:24 <glx> hmm 2.6-patch 4 should be an old version
22:28:31 <planetmaker> and there it tells me the same
22:28:34 <TrueBrain> not sure I agree with LordAro's suggestion to run it from the bin directory :P
22:28:47 <planetmaker> running from bin dir: -1 from me
22:29:03 <TrueBrain> I hate to repeat myself, but again: I updated the PR to tell you how to build it :)
22:29:06 <TrueBrain> let me know if that fails :)
22:29:18 <planetmaker> I ran git fetch 5 seconds ago...
22:29:26 <TrueBrain> the PR; not the code :)
22:29:31 <LordAro> cmake is not supposed to run an "in source" build
22:29:38 <glx> yes I have read the PR :)
22:29:51 <TrueBrain> glx: I was talking to planetmaker, sorry :D
22:29:54 <LordAro> you're supposed to do an "out of source" build, so running cmake <something other than .>
22:30:01 <TrueBrain> LordAro: yeah; but the 'bin' folder is an odd suggestion :)
22:30:02 <LordAro> that can be bin, or it could be any other directory
22:30:10 <planetmaker> TrueBrain, no change...
22:30:11 <TrueBrain> it leaves too many files around
22:30:17 <planetmaker> CMake Error at CMakeLists.txt:4 (message):
22:30:17 <planetmaker> In-source builds not allowed. Please run "cmake .." from the bin directory
22:30:38 <planetmaker> as you write there
22:30:39 <Eddi|zuHause> didn't we have an "objs" dir?
22:30:47 <TrueBrain> planetmaker: now that is odd :)
22:31:09 <andythenorth> Eddi|zuHause: no longer fails :)
22:31:51 <TrueBrain> planetmaker: can you add two lines after the first line in CMakeLists.txt:
22:32:03 <TrueBrain> message(STATUS "${CMAKE_SOURCE_DIR}")
22:32:03 <Eddi|zuHause> andythenorth: side effect is that error messages on wrong number of parameters will likely be less descriptive
22:32:07 <TrueBrain> message(STATUS "${CMAKE_BINARY_DIR}")
22:32:46 <TrueBrain> after the first line
22:33:07 <TrueBrain> (not sure which part you didn't understand, don't ment to be rude by repeating myself :P)
22:33:35 <peter1138> Hmm, towns behave very oddly when they're not allowed to build roads.
22:33:44 <planetmaker> -- /home/planetmaker/ottd/OpenTTD-master
22:33:44 <planetmaker> -- /home/planetmaker/ottd/OpenTTD-master
22:33:54 <planetmaker> yet... pwd is -- /home/planetmaker/ottd/OpenTTD-master
22:33:54 <planetmaker> -- /home/planetmaker/ottd/OpenTTD-master/build
22:34:13 <TrueBrain> I have never seen CMake fucking up his BINARY dir ..
22:34:23 <TrueBrain> basically, CMake disagrees that you are in build :)
22:34:27 <peter1138> build is not outside the tree, is it?
22:34:34 <TrueBrain> remove any stray CMake file?
22:34:50 <TrueBrain> I can guess that your first run created a file it got stuck on
22:35:27 <planetmaker> ah. That... helps
22:35:33 <TrueBrain> pfew :D you were scaring me :D
22:35:41 <TrueBrain> good to know CMake gives that priority ..
22:36:38 <TrueBrain> and flip it to YES :P
22:36:42 <TrueBrain> I did not do all autodetect yet :)
22:36:49 <peter1138> Command not found. Hmm.
22:37:12 <andythenorth> so where do we apply fixes, when I have a remote checkout of peter1138's rebase of nielsm's branch? And we need to fix a broken previous commit?
22:37:16 <Samu> peter1138, you're doing incomplete tests :(
22:37:32 <Eddi|zuHause> i have ccmake, but no cmake-gui
22:37:38 <Eddi|zuHause> (not trying either)
22:37:39 <peter1138> peter1138, yes, that's my diff.
22:37:43 <TrueBrain> Eddi|zuHause: ccmake is ncurses, cmake-gui is QT :)
22:37:43 <Samu> you're only checking if there is any water track
22:37:47 <peter1138> Samu, yes, that's my diff. What's missing.
22:37:55 <peter1138> Isn't that sufficient?
22:37:55 <Eddi|zuHause> TrueBrain: well, i was guessing as much :p
22:38:01 <TrueBrain> planetmaker: ah, yes; remove src/rev.cpp :)
22:38:09 <TrueBrain> I have to fix that :P
22:38:36 <planetmaker> no rule to create ../src/rev.cpp
22:38:45 <Eddi|zuHause> apparently cmake-gui is a separate package
22:38:58 <planetmaker> ah... re-run cmake
22:39:11 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain updated pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
22:39:32 <TrueBrain> planetmaker: fixed; it will no longer pick up rev.cpp that might have been there from older runs :) (in ROOT/src/rev.cpp)
22:39:51 <TrueBrain> still have to fix the MSVC issue with UNICODE something ..
22:39:58 <peter1138> TrueBrain, confusing, some thigns are ON/OFF and some are YES/NO. "ON" and "NO" look very similar.
22:40:20 <TrueBrain> peter1138: they should all be YES/NO
22:40:26 <TrueBrain> what shows you ON/OFF?
22:40:31 <TrueBrain> (and this is a known issue with CMake indeed)
22:40:34 <planetmaker> ew... failed to load titlegame
22:40:38 <planetmaker> with the cmake build
22:40:43 <TrueBrain> planetmaker: cd ../bin && ../build/openttd
22:40:54 <peter1138> Those are the ones I've configured with ccmake.
22:41:02 <TrueBrain> peter1138: ah, so ccmake uses ON/OFF
22:41:05 <TrueBrain> that is .. annoying
22:41:06 <peter1138> (Command not found, I had to install it separately)
22:41:27 <Eddi|zuHause> so can you do ON/NO? :p
22:41:27 <TrueBrain> depends on your OS, but yes, ccmake tends to be installed :)
22:41:38 <peter1138> Hmm, SSE fails to compile.
22:41:45 <TrueBrain> peter1138: I had the same, so I disabled it :P
22:41:47 *** Gustavo6046 has joined #openttd
22:41:48 <TrueBrain> didnt look into it yet
22:42:21 <peter1138> Samu, what case are you checking for?
22:42:27 <TrueBrain> planetmaker: but so it linked for you; good :D
22:42:38 <peter1138> Ah, and it doesn't link fluidsytnh :0
22:42:59 <TrueBrain> -DFLUIDSYNTH is not set yet
22:42:59 <andythenorth> nmlc ERROR: "generated/firs.nml", line 30275: Cargo list must be an array with no more than 2 values
22:43:01 <TrueBrain> so that is possible
22:43:03 <andythenorth> thought that was fixed :P
22:43:16 <peter1138> multiple definition of `IsReleasedVersion()`
22:43:28 <TrueBrain> peter1138: either remove src/rev.cpp, or update to latest version of PR :)
22:43:42 <peter1138> Oh, you updated it while it was building.
22:43:45 <TrueBrain> owh, no, -DFLUIDSYNTH isnt needed
22:43:53 <TrueBrain> yes, because planetmaker also found that bug :)
22:44:06 <peter1138> It's not including the library during linking, is all.
22:44:12 <TrueBrain> guess fluidsynth just needs a find_package call
22:44:14 <Samu> there's water tracks in opposite, there's water track in tile 1 and there's water track in tile 2
22:44:18 <TrueBrain> yeah, that is an easy fix
22:44:22 <Samu> but building on the track
22:44:31 <Samu> doesn't mean the others are connecte
22:44:43 <andythenorth> nml won't accept prod_cargo_types: [cargotype("MNO2"),cargotype("PETR"),cargotype("RUBR")];
22:44:49 <peter1138> Samu, they have to be connectedf.
22:46:07 <andythenorth> seems we didn't test it?
22:46:08 <peter1138> Well, okay, tile 1 could be blocked with a depot or something, but I don't think such a small case warrants making the code overly complex.
22:46:49 <Samu> why not if it works already?
22:47:30 <Eddi|zuHause> andythenorth: look for remaining calls to cargo_list()?
22:47:35 <peter1138> Just because it's already written doesn't mean we should add overly complex code.
22:47:58 <andythenorth> this needs rethought :P
22:48:07 <peter1138> However, yes you are right, opposite is blocking. Hmm.
22:48:14 <glx> oh nice first thing to do after getting the PR is discard 3047 modified files
22:48:19 <planetmaker> I cannot say I like cmake. But the build process definitely looks nicer
22:48:20 <peter1138> Don't need to check tile2/tile1 tracks though.
22:48:29 <andythenorth> using FIRS to test nmlc is clown shoes
22:48:30 <planetmaker> Yet... I don't like that I simply can start it from OpenTTD main dir
22:48:41 <Samu> i guess it really needs to be complex
22:48:48 <TrueBrain> planetmaker: those are easily resolvable issues, which have nothing to do with cmake
22:48:48 <peter1138> I suspect it really doesn't need to be complex.
22:48:49 <Samu> to be 100% certain it works
22:48:57 <TrueBrain> we have to fix that up in the current build process too
22:49:01 <TrueBrain> as we have some weird constraints
22:49:32 <TrueBrain> (Basically, you get a 'make' that runs cmake for you in a folder like 'build' :P)
22:49:53 <nielsm> andythenorth: because the accept/prod cargo types properties were restored to only produce the old GRF data
22:49:58 <TrueBrain> so if that is the only thing you don't like, it is easily solved :D
22:50:09 <nielsm> and there is a new nml property that combines them instead
22:50:16 <Eddi|zuHause> what always annoyed me when projects use cmake is that you can't type "./configure" like everywhere else
22:50:25 <peter1138> Samu, why would you make that sample on land? :p
22:50:37 <TrueBrain> Eddi|zuHause: the more you use cmake, the less you want to try that :)
22:50:45 <Eddi|zuHause> you must first figure out "ah it's using cmake", then "md build", and then it gets complicated...
22:50:48 <TrueBrain> (and given more and more projects are using cmake ... :)
22:51:04 <Samu> it's easier for me to visualize the tracks
22:51:07 <TrueBrain> but yeah, I see many projects having a 'configure' script
22:51:19 <Eddi|zuHause> that's preferable, imho
22:51:21 <planetmaker> I'm +-0 to it in general. Eddi just summarizes it :D
22:51:31 <TrueBrain> Eddi|zuHause: it is just a habbit
22:51:36 <Faizan> how to create a website. which is good to use
22:51:38 <nielsm> it uses the new properties and prod cb
22:51:49 <TrueBrain> so there is a lot of debate if you should move with the habit, or just learn people not to have the habit :)
22:51:58 <Eddi|zuHause> have a "dummy" ./configure that creates the build dir and runs cmake in that
22:52:31 <Eddi|zuHause> TrueBrain: i don't see any harm in the habit. keep it simple
22:52:34 <glx> TrueBrain: looks like the "how to handle \r" in gnutools debate
22:52:45 <TrueBrain> Eddi|zuHause: the harm is technical debt
22:52:49 <milek7> "ah it's using cmake" -> mkdir build;cd build;cmake ..;make
22:52:58 <TrueBrain> by keeping those "glue", you keep adding complexity
22:53:09 <TrueBrain> so there are 2 camps .. just learn the new tool .. or keep the glue
22:53:13 <andythenorth> we should add the tests to nml/examples
22:53:15 <TrueBrain> Python is having the same issue currently with setup.py :)
22:53:17 <Eddi|zuHause> TrueBrain: if a script with 4 lines is "complexity"?
22:53:19 <andythenorth> as we were doing
22:53:29 <TrueBrain> Eddi|zuHause: basically, any line to support N is complexity
22:53:37 <TrueBrain> for the same reason .gitignore should not ignore all the editors
22:53:39 <TrueBrain> sure, they are only lines
22:53:43 <TrueBrain> but ... where do you stop :)
22:53:54 <Eddi|zuHause> TrueBrain: you offload these 4 lines to everyone checking out the source is better?
22:53:57 <TrueBrain> but again, this is exactly what people are debating
22:54:07 <TrueBrain> "everyone" is an overstatement :)
22:54:18 <Faizan> can u send me the server gs download link
22:54:27 <TrueBrain> but I am only pointing out that people are having this debate over and over on the internet
22:54:32 <TrueBrain> just showing you there are 2 camps to the story
22:55:06 <planetmaker> TrueBrain, so the argument for cmake is "meh, some people use it and change is good for its own sake"?
22:55:21 <TrueBrain> why ... do you say that?
22:55:22 <planetmaker> sorry, but that's just how you currently argue pro cmake :)
22:55:27 <glx> the main argument is one script for all platforms
22:55:28 <Eddi|zuHause> TrueBrain: tbh, i see that the ./configure is a bad solution, the better solution would be for a simple "cmake" call should do all this standard setup
22:55:31 <TrueBrain> no .. we were talking about configure bootstrap planetmaker :)
22:55:33 <TrueBrain> nothing more, nothing less :)
22:55:48 <TrueBrain> Eddi|zuHause: I like the 'make' bootstrap personally
22:55:53 <TrueBrain> but lets get there when we get there :)
22:56:01 <milek7> autotools is so awful that I'm surprised it survived that long
22:56:13 <Eddi|zuHause> TrueBrain: so how is a "makefile" better than a "configure"?
22:56:18 <TrueBrain> planetmaker: I seriously wonder how you made that jump :) Eddi|zuHause and me were just talking about having './configure' that calls cmake or not :)
22:56:23 <TrueBrain> we were not debating cmake itself :)
22:56:32 <TrueBrain> Eddi|zuHause: nope; it is just what I personally prefer
22:56:33 <peter1138> Samu, ok, we only need to check 1 track exists.
22:56:35 <glx> ok removed the 3047 useless files from the repo
22:56:38 <TrueBrain> mostly as I add stuff like 'make run' in there
22:56:43 <peter1138> On the opposite tile.
22:56:52 <Faizan> send me the link of server gs game script please
22:57:06 <TrueBrain> Eddi|zuHause: guess that preference comes from: ./configure && make
22:57:15 <TrueBrain> only configure is not useful
22:57:18 <TrueBrain> so I can do without it
22:57:24 <TrueBrain> people are weird :P
22:57:33 <planetmaker> TrueBrain, well, the jump from "there are two parties" to "an argument for and against" is very small
22:57:49 <glx> well first restart powershell as I uninstalled the old cmake then installed the new cmake
22:57:52 <TrueBrain> planetmaker: we really only talked about ./configure bootstrap or not .. there are 2 camps there :)
22:57:54 <planetmaker> that's how I read it. And I didn't mean it offencive. Sorry if I came accross as such
22:58:03 <Faizan> stopppppppppppppppppppppppppppppppppppppppppppp Keeeep QUUUUUUUUUUUUUUit
22:58:03 <TrueBrain> no worries :) I was trying to understand you :)
22:58:06 <nielsm> Faizan: you're better off asking on tt-forums, setting those things up is not simple, and I don't think any of us in here actually know very much about running an OTTD server in practice
22:58:11 <TrueBrain> IRC is difficult :P
22:58:41 <Samu> but ship depots will be the other problem
22:58:46 <glx> D:\developpement\GitHub\glx22\OpenTTD\build [pr/7270]> cmake ..
22:58:46 <glx> -- Building for: Visual Studio 15 2017
22:58:46 <glx> -- Selecting Windows SDK version 10.0.17763.0 to target Windows 10.0.17134.
22:58:46 <glx> -- The C compiler identification is MSVC 19.16.27026.1
22:58:46 <glx> -- The CXX compiler identification is MSVC 19.16.27026.1
22:59:03 <TrueBrain> glx: I did not try that route yet :P
22:59:06 <TrueBrain> let me know what it brings
22:59:07 <planetmaker> <TrueBrain> so there are 2 camps .. just learn the new tool .. or keep the glue <-- that's where I came from
22:59:20 <TrueBrain> yeah, glue was a bit unclear there :P
22:59:20 <nielsm> Faizan: we're working hard on improving the core game, but the practicals of actually running a game server are somewhat different from that :P
22:59:25 <Faizan> who is playing multiplayer at that time
22:59:35 <TrueBrain> but to answer the question anyway: cmake removes all the custom config.lib code we have
22:59:47 <TrueBrain> the thousands of incompatible lines with <insert OS here>
22:59:52 <andythenorth> no-one here really plays MP
22:59:53 <TrueBrain> cmake solved that very much
23:00:06 <planetmaker> yep, starting to see that, TB :)
23:00:21 <TrueBrain> (because every library has its own cmake that comes with it for that OS when installing .. as everyone is using cmake these days)
23:00:25 <planetmaker> yet I also see, that compilation for me does get twice as complicated ;)
23:00:48 <planetmaker> ./configure && make > mkdir build && cd build && cmake .. && make
23:01:02 <TrueBrain> planetmaker: goldplating; we can easily make that just 'make'
23:01:05 <Faizan> tell me your server adress
23:01:05 <TrueBrain> that is not the issue :)
23:01:13 <TrueBrain> and as mentioned, we can even keep the ./configure
23:01:19 <TrueBrain> but that is all just gold plating
23:01:25 <planetmaker> I want my bikeshed goldplated. NOW! :P
23:01:26 <TrueBrain> currently ./configure calls our own config.lib
23:01:36 <TrueBrain> cmake just replaces config.lib
23:01:38 <TrueBrain> and all MSVC projects
23:01:42 <TrueBrain> so no more generating of project files
23:01:50 <TrueBrain> so ... everything is the same, basically
23:01:53 <planetmaker> and I know that "can be done" things often end up "meh, never" :P
23:02:05 <TrueBrain> only do the things people ask for
23:02:09 <planetmaker> (judging from myself) :P
23:02:19 <TrueBrain> so if this PR is done, and nobody notice anymore that they miss configure
23:02:25 <Eddi|zuHause> TrueBrain: i think you now have enough cases of "people asked for it" :p
23:02:26 <glx> hmm adding build to .gitignore is probably required
23:02:29 <TrueBrain> that is why I call it gold plating .. doing some gold now, is silly :)
23:02:49 <Samu> + forbid 90 degrees problem
23:02:54 <TrueBrain> Eddi|zuHause: you now ask for it as it is like: OMG I AM NOT USED TO THIS PANIC :P
23:03:02 <TrueBrain> (which is totally how every human ever works)
23:03:12 <glx> because it now shows 84 modified/new files
23:03:19 <TrueBrain> but the first prio is getting it in a workable state
23:03:26 <TrueBrain> with all the features attached that are a MUST :)
23:03:28 <Eddi|zuHause> TrueBrain: no, i ask for this because i've been through the same steps a few times already
23:03:35 <Eddi|zuHause> TrueBrain: and it's annoying me
23:03:47 <TrueBrain> Eddi|zuHause: fair enough; still, I write it as a COULD, but this is OSS, so you can always contribute :)
23:03:48 <nielsm> Samu: you just proved that "forbid 90 deg turns" is a bad idea for ships, becuase there just is no solution to that layout
23:04:11 <andythenorth> nml cargo syntax has changed?
23:04:12 <TrueBrain> glx: 'build' ... or call it 'sweetypie' .. the name is free :)
23:04:27 <andythenorth> so we need to use the cargotype entity?
23:04:38 <TrueBrain> Eddi|zuHause: just to be clear: I am not saying we should NOT do it; I am just saying, it is goldplating, and not important NOW (to me)
23:04:45 <nielsm> I wonder if that #openttd.dev suggestion may be worth taking for once?
23:04:53 <glx> and of course make fails as there's no Makefile, only vcxproj :)
23:05:04 <nielsm> hmm wait no it's no longer there
23:05:10 <TrueBrain> glx: you created a MSVC project? Nice!
23:05:12 <nielsm> (I did not look at the topic)
23:05:34 <Samu> i thought i proved 90 degrees was also important in the decision
23:05:35 <planetmaker> nielsm, the channel exists
23:05:38 <andythenorth> ok so for 16 cargo FIRS I need to rewrite my nml templates
23:05:40 <TrueBrain> but if the only problem is: I cannot do the autotools way, I am not worried :P
23:05:44 <peter1138> Samu, what's the issue there?
23:06:09 <TrueBrain> I am worried, how-ever, why VS2017 and VS2019 don't want to compile for reasons I cannot understand
23:06:11 <Samu> if 90 degrees are allowed, houses don't build on 'tile'
23:06:13 <glx> ok now I try the other option
23:06:32 <peter1138> Samu, 90 degrees option should not affect how towns expand.
23:06:39 <nielsm> peter1138: ship depots can be traversed by any ship, also those owned by other companies than the depot, but they only allow traversal in one direction (they don't have trackdirs out the sides)
23:06:40 <Samu> if 90 degrees are forbidden, houses build on 'tile'
23:06:40 <TrueBrain> dmusic.cpp(1125): error C2664: 'char *convert_from_fs(const TCHAR *,char *,size_t)': cannot convert argument 1 from 'WCHAR [128]' to 'const TCHAR *'
23:06:40 <TrueBrain> like .... where did I go wrong? :(
23:06:50 <nielsm> so you can make "tunnels" of them where ships can't leave
23:06:59 <nielsm> meaning it can restrict movement in tight spaces
23:07:31 <TrueBrain> glx: are MSVC projects Unicode or Multibyte Character Set?
23:07:36 <TrueBrain> (you happen to know?)
23:07:41 <Samu> because the ship couldn't walk with 90 degrees forbidden
23:07:44 <nielsm> this is another case of, you can't prevent griefing by players intent on ruining the day for others
23:07:59 <Eddi|zuHause> TrueBrain: if i check out source code of an OSS project, and then fight half an hour trying to compile it with incomplete/diverging documentation of "how to compile", then i've lost all interest in contributing to said project
23:08:00 <nielsm> and you can't detect the difference between intentional griefing and unintentional consequences
23:08:12 <glx> would need to open them for that (they are too big for github desktop to show the diff)
23:08:15 <TrueBrain> Eddi|zuHause: yes, I understand what you are trying to say
23:08:30 <TrueBrain> Eddi|zuHause: I also hope you understand what I try to say: first get it to work, than get all features in, than work on goldplating
23:08:37 <peter1138> Samu, that's only because you test the 90 degree setting, and needlessly test tracks on adjacent tiles.
23:08:44 <peter1138> Samu, I've updated the PR.
23:09:03 <andythenorth> what have houses got to do with 90 deg?
23:09:09 <peter1138> andythenorth, exaclty.
23:09:11 <andythenorth> did I drop some pills and not notice?
23:09:23 <peter1138> 22:06 <@peter1138> Samu, 90 degrees option should not affect how towns expand.
23:09:27 <andythenorth> I don't do drugs, have I accidentally taken some?
23:09:45 <andythenorth> houses have fuck all to do with 90 degrees
23:09:55 <andythenorth> and 90 degrees was a realism foamer mistake anywy
23:10:05 <nielsm> you can't retroactively have towns destroy some houses because someone toggled the "allow 90 deg turns" switch mid game
23:10:26 <andythenorth> 90 degrees *needs removed*
23:10:32 <andythenorth> how many times does it need saying?
23:11:02 <andythenorth> it's a stupid feature that is wasting ridiculous amounts of time in discussing irrelevaant PRs
23:11:21 <andythenorth> about 25% of the channel is just spam about 90 deg
23:11:26 <andythenorth> and the rest is me complaining about that
23:11:42 <peter1138> 50% is you complaining about nml.
23:11:52 <nielsm> "disallow 90 deg turns" was never intended to be for anything but ships
23:12:08 <nielsm> who decided it should apply to ships, and when?
23:12:23 <Eddi|zuHause> never made any sense to me
23:12:40 <peter1138> Samu, so is this an "issue" that applies to the PR I gave you?
23:12:49 <planetmaker> so a setting to be removed. cleanup: remove 90° for all but trains
23:12:52 <TrueBrain> glx: found it, it is unicode indeed
23:13:03 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain updated pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
23:13:08 <andythenorth> planetmaker: + lots
23:13:09 <peter1138> I can't see how it can be.
23:13:11 <TrueBrain> meh; wish draft PRs would not spam every time
23:13:21 <peter1138> Samu, you'll need to update, because I pushed to it.
23:13:37 <TrueBrain> MSVC runs c++17, yes
23:13:42 <TrueBrain> the rest runs c++11
23:13:50 <TrueBrain> silly? yes. Microsoft? yes
23:14:08 <nielsm> peter1138, this situation: https://0x0.st/z-Xj.png - if a town builds a house at the tile of the end of the depot, the test in your proposal will say there is a valid path past when there isn't
23:15:14 <nielsm> to which my response is: stop trying to prevent intentional or accidental griefing, you can't
23:16:15 *** drac_boy has joined #openttd
23:16:19 <nielsm> Samu: no check is perfect and the landscape can be modified
23:16:39 <TrueBrain> topic of the day :)
23:16:40 * drac_boy wonders what I stepped into now 0_o
23:16:43 <andythenorth> we have very little history of accepting anti-griefing PRs
23:17:53 <Samu> my checks would detect that the opposite tiles weren't connected and then deny building a house there
23:17:54 <peter1138> Either way, the fact that Samu's original patch tests the 90 degree setting is of course wrong.
23:18:14 <glx> hmm opened CMakeList.txt in VS, clicked build all, and it fails hard
23:18:57 <nielsm> I still think it's silly to try to prevent towns from expanding into half-water tiles on the chance someone might potentially at some point in the past or future want to sail ships nearby
23:18:58 <TrueBrain> glx: check vcpkg installation guide :)
23:19:02 <TrueBrain> hmm, that should not fail
23:19:07 <TrueBrain> it should just not compile PNG and Zlib
23:19:09 <glx> grr why did it tried to use the old mingw headers ???
23:19:31 <peter1138> nielsm, I actually think it's nice just to keep rivers open.
23:19:35 <TrueBrain> either way, I couldn't find a way to have vcpkg work globally for cmake glx :(
23:19:40 <peter1138> I don't particularly care about the ship aspect :-)
23:19:42 <andythenorth> nielsm: it was incredibly irritating when the game did it to me....like once :P
23:19:57 <TrueBrain> hmm .. fluidsynth packages don't have a cmake file .. wwuuttthhh?
23:20:06 <TrueBrain> fluidsynth is a cmake project .. and has a cmake file in their github ..
23:20:21 <andythenorth> I should really learn python
23:20:26 <andythenorth> or go to bed earlier yesterday
23:20:30 <andythenorth> can I still do that?
23:20:49 <peter1138> You are committed to staying up all night.
23:21:00 <drac_boy> do depot-view sprites have to be of same length as on-map sprites or did I think wrong there?
23:21:10 <andythenorth> I watched the Dr. Dre thing
23:21:11 <drac_boy> peter heh you calling him an owl now? :P
23:21:14 <peter1138> nielsm, hmm, and in that case my test is valid :)
23:21:16 <andythenorth> Dre stays up all night
23:21:47 <peter1138> The river is still intact, even if ships can't traverse it due to the stupid depot location.
23:22:00 <nielsm> peter1138 yeah I agree
23:22:14 <nielsm> if a player object is in the way, that sure is a shame
23:22:22 <nielsm> the player could have not built it there
23:22:38 <nielsm> maybe the game administrator should kick the player for griefing
23:22:59 <nielsm> or the singleplayer human should berate the AI author for writing bad AIs
23:23:06 <andythenorth> river grief, mountain high
23:23:22 <andythenorth> did Eddi|zuHause fix rivers yet? :P
23:23:39 <peter1138> nielsm, hah, one of samu's ship ai savegames does indeed have plenty of examples of AIs blocking THEIR OWN DOCKS with depots...
23:23:53 <Eddi|zuHause> and certainly not ready for 1.9
23:24:05 <nielsm> peter1138 yeah, he tried making a PR that prevents you from building depots and docks in stupid places
23:25:11 <drac_boy> nielsm I know its something the author can't do anything about but I always find it funny when a game recreates an enclosed inlet (complete 1-tile barrier separation from actual open waters) and then the industry grf decides for some reason to put the dockside industry right inside that little hole :)
23:25:33 <glx> hmm ok renamed the old mingw dir so it's not in the path, and I still get errors 'D:/MinGW/lib/libpng.lib', needed by 'openttd.exe', missing and no known rule to make it C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\Common7\IDE\CMakeLists.txt C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\Common7\IDE\ninja 1
23:26:04 <peter1138> nielsm, maybe I should PR this as "attempt to prevent town from closing off rivers, but don't give a shit about what stupid players do"
23:26:22 <TrueBrain> glx: did you add vcpkg in the CMakeSetting.json?
23:26:26 <TrueBrain> that solved a lot of things for me
23:27:23 <andythenorth> I have a FIRS industry in game
23:27:29 <andythenorth> producing more than 2 cargos :D
23:27:44 <andythenorth> just needed to read the docs :P
23:27:52 <nielsm> I hope you like the new syntax :)
23:27:58 <andythenorth> it's fine, makes sense
23:28:05 <andythenorth> I have to figure out how to adapt FIRS to it
23:28:38 <andythenorth> I don't know what value to use in 'produce_cargo("COAL", 0.5)' etc when prod. cb handles production
23:28:46 <Faizan> i want to play multiplayer with you
23:29:05 <peter1138> nielsm, yes.... AIs need to check themselves if it's going to block o_O
23:29:51 <Faizan> tell me name of your server please
23:30:32 <Samu> i wanted to keep the industry check though
23:31:54 <Samu> (and revert all the other player made checks)
23:34:43 <glx> ah it does things now I added CMAKE_TOOLCHAIN_FILE
23:35:03 <TrueBrain> glx: I hope they fix that some day .. as that needs a README to do right :)
23:35:27 <TrueBrain> or we have to include that cmake file .. that is not the worst idea .. just annoying
23:35:39 <TrueBrain> glx: update to my latest version, if it fails on dmusic and/or win32 :)
23:36:03 <glx> fails on these one indeed :)
23:36:23 <peter1138> Ok, so ... my bitmap_type.hpp ... should I leave it in src/core/ or move it to src/
23:36:30 <peter1138> Not seeming like a core thing :p
23:36:43 <peter1138> tilearea isn't in core...
23:37:01 <TrueBrain> what are the odds I do 'rm -rf *' in the wrong folder soon?
23:37:31 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain updated pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
23:37:32 <TrueBrain> peter1138: now with fluidsynth support ^^
23:37:57 <drac_boy> hmm nevermind I'll just try find the sprites length thing tomorrow .. lets see about more wagons instead
23:38:55 <LordAro> TrueBrain: cd ..; rm -r build; mkdir build; cd build
23:39:02 <TrueBrain> LordAro: that would be wiser
23:39:41 <glx> oh and "cmake --build" should have work instead of "make"
23:40:32 <DorpsGek_II> [OpenTTD/OpenTTD] planetmaker opened pull request #7271: Cleanup: Ships can always make 90° turns (it's even realistic) https://git.io/fhbmz
23:40:46 <TrueBrain> DENIED; OpenTTD is not realistic
23:41:03 <planetmaker> TrueBelugas, you!
23:41:18 <TrueBrain> why is compiling SSE failing, I wonder ..
23:41:33 <drac_boy> :) truebrain .. realistic != game fun
23:42:05 <planetmaker> heading to bed though. Will be back tomorrow night. gn!
23:42:41 <peter1138> planetmaker, ANOTHER pr for that? lol
23:43:14 <LordAro> clearly an in demand feature
23:44:28 <andythenorth> drac_boy are you listing trains from any specific country?
23:47:01 <glx> now package detection fails
23:47:07 <drac_boy> its more generic so for example a diesel with long high hood and on the other/shorter side of cab a small stub end .. could basically be a lot of things, NSW GP9 PKP etc
23:47:15 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7235: Change: Non-rectangular sparse station catchment area https://git.io/fh5s1
23:47:27 <TrueBrain> so we check SSE support by including some headers and see if that compiles ..
23:47:31 <TrueBrain> I expected more code there
23:48:35 <glx> Gravité Code Description Projet Fichier Ligne État de la suppression
23:48:35 <glx> Erreur CMake Error at C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/Common7/IDE/CommonExtensions/Microsoft/CMake/CMake/share/cmake-3.12/Modules/FindPkgConfig.cmake:659 (message):
23:48:35 <glx> None of the required 'fluidsynth' found C:/Program Files (x86)/Microsoft Visual Studio/2017/Community/Common7/IDE/CommonExtensions/Microsoft/CMake/CMake/share/cmake-3.12/Modules/FindPkgConfig.cmake 659
23:49:45 <TrueBrain> you can remove that safely, line 56
23:50:15 <peter1138> What's the view on FOR_ALL_SOMETHING(x) dosomethingwith(x); ?
23:50:38 <glx> ok no zlib, no png, and no version
23:50:41 <peter1138> Should I add { }, or even put the innards on their own line?
23:50:53 <TrueBrain> glx: first two you can fix with vcpkg :)
23:51:14 *** supermop_Home has joined #openttd
23:51:19 <TrueBrain> check if you have the right OS etc
23:53:33 <drac_boy> andy hmm that lms photo is interesting thanks
23:53:45 <andythenorth> you want other 3 axle trains?
23:56:23 <andythenorth> nielsm: in this 'accept_cargo("IORE", produce_cargo("STEL", 4))', 4 units of STEL from 1 unit IORE?
23:56:57 <andythenorth> whereas here 'produce_cargo("LVST", 8)', 8 units produced each time cargo production loop runs?
23:58:09 <drac_boy> heh I guess I could get away with adding a 3-axle of most freight types and noone'll be any wiser .. thanks for the thoughts :)
continue to next day ⏵
