IRC logs for #openttd on OFTC at 2021-04-17
⏴ go to previous day
00:49:17 *** Progman has quit IRC (Remote host closed the connection)
01:16:30 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8982: Feature: text filter for the industry directory https://git.io/JO86H
01:17:00 *** didac has quit IRC (Remote host closed the connection)
02:01:09 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
02:14:25 *** Sophon has quit IRC (Remote host closed the connection)
02:22:43 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
02:24:12 *** Wormnest has quit IRC (Quit: Leaving)
04:29:17 *** snail_UES_ has quit IRC (Quit: snail_UES_)
05:16:05 *** MAvrick has joined #openttd
06:06:08 *** NausiyanNyan[m] has joined #openttd
06:17:58 *** andythenorth has joined #openttd
06:22:05 *** sla_ro|master has joined #openttd
06:51:03 <peter1138> Hello. Another cold start.
07:03:53 <Rubidium> good morning peter et al.
07:22:49 *** Progman has joined #openttd
07:33:44 <peter1138> Damn, removing sprite offsets assumes that the offsets are there to make the sprite centred.
07:47:48 <peter1138> Hmm, do we just ignore fix_graphics.png?
07:51:58 <andythenorth> even is it what?
07:52:48 <peter1138> Fixes to the original graphics.
07:57:52 <peter1138> I wonder if I can define a NewGRF action that takes an original sprite and pads it out to the intended size, without having to distribute a copy of the modified sprite data.
07:58:40 <peter1138> I can do it in code but that requires the code to know which baseset is in use, and still requires a list of sprites and sizes.
08:16:04 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 opened pull request #9047: Fix #6322: Attempt to crash the script instead of the whole game https://git.io/JO42l
08:27:51 <andythenorth> but I can't figure out how to use them in nml :)
08:29:59 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9047: Fix #6322: Attempt to crash the script instead of the whole game https://git.io/JO4VJ
08:31:50 <andythenorth> tile_powers_railtype("ELRL") not [tile_powers_railtype(railtype("ELRL"))]
08:35:16 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 updated pull request #9047: Fix #6322: Attempt to crash the script instead of the whole game https://git.io/JO42l
08:45:32 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #9047: Fix #6322: Attempt to crash the script instead of the whole game https://git.io/JO4rk
09:45:31 <michi_cc> Hmm, 1.11.1 this weekend, or are we still waiting on something important?
10:02:08 <andythenorth> FLHerne I will add those to the nml docs later today
10:02:11 <FLHerne> You shouldn't have closed that issue on inline documentation :p
10:02:32 <FLHerne> Having a separate wiki that people forget to update is silly
10:02:36 <andythenorth> it really really is
10:02:41 <andythenorth> but there are reasons
10:03:19 <andythenorth> I think it's a brake on grf spec evolution but eh
10:03:24 <andythenorth> I do not know how to fix that :)
10:04:03 <andythenorth> I don't think it's silly have wiki for actual grf spec
10:04:12 <andythenorth> but nml vars are such a PITA
10:04:49 <andythenorth> like...I need to add these new vars
10:05:53 <FLHerne> Why do both 7.2 and 7.3 even exist?
10:06:09 <FLHerne> Hang on, they're identical
10:06:18 <andythenorth> I am looking in history
10:06:20 <FLHerne> Someone's managed to duplicate that paragraph :p
10:06:24 <andythenorth> I made a lot of edits there, probably me
10:06:31 <andythenorth> I absolutely can't use wiki format at all
10:06:58 <andythenorth> they're literally identical yes?
10:07:33 <andythenorth> worrying how much of the docs is written by me
10:07:34 <FLHerne> Unless I've missed something...
10:07:38 <andythenorth> given how little I understand
10:08:08 <andythenorth> how do I use wiki history?
10:08:33 <FLHerne> I looked again very carefully, and I still think they're literally identical
10:08:59 <andythenorth> if only somebody had made a git based wiki
10:09:17 <andythenorth> FLHerne fancy deleting one of the duplicates then?
10:10:21 <FLHerne> I hesitate to remove it in case you created it for some important purpose I just don't understand :p
10:10:32 <andythenorth> no it will literally be mistake
10:10:58 <Rubidium> I should not have opened that specs page... it kinda trigged something OCD like with the foot notes; the first two foot note references do not belong there as far as I can see
10:11:07 <andythenorth> the nml docs are great
10:11:10 <andythenorth> the nml docs are shit
10:11:10 <FLHerne> Oh, the "Vehicle variables" paragraph is duplicated too
10:11:32 <andythenorth> the nml docs really are quite good
10:11:41 <andythenorth> but they're just bad enough that nobody wants to touch them ever
10:13:48 <FLHerne> Because AI/GS came later when someone realized it was a bad idea :p
10:13:59 <FLHerne> Hm, actually NML isn't that old
10:14:19 <andythenorth> it all made sense at the time
10:15:19 <andythenorth> what's the sphinx convention? Docstrings?
10:15:24 * andythenorth considers things
10:16:08 <andythenorth> docstrings aren't for properties though?
10:23:47 *** tokai|noir has joined #openttd
10:23:47 *** ChanServ sets mode: +v tokai|noir
10:28:50 *** WormnestAndroid has quit IRC (Remote host closed the connection)
10:29:03 *** WormnestAndroid has joined #openttd
10:29:36 <andythenorth> is it any way possible that using tile_powers_railtype() blitzes the text stack or similar?
10:29:56 <andythenorth> cb 23 (buy menu text) is failing, but nothing has changed relating to it in my code
10:30:42 *** tokai has quit IRC (Ping timeout: 480 seconds)
10:34:16 *** frosch123 has joined #openttd
10:37:49 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
10:38:02 *** WormnestAndroid has joined #openttd
10:42:25 <FLHerne> (or if it does, it's completely beyond my understanding)
10:44:09 <FLHerne> The NML implementation of it is very simple and has no connection to the textstack
10:58:43 <andythenorth> that's what I hoped
10:58:57 <andythenorth> local bug seems more plausible
11:06:58 <andythenorth> hmm, I updated nml
11:07:03 * andythenorth wonders if anything changed there
11:09:21 <andythenorth> oh that's interesting
11:09:44 <andythenorth> my use of cb 23 is broken with nml master
11:10:03 <andythenorth> I compiled Iron Horse 2.25.0 with an older nml and pushed to bananas
11:10:17 <andythenorth> the bananas version, cb 23 (buy menu text) works
11:10:31 <andythenorth> locally compiled 2.25.0, cb 23 fails
11:10:57 <andythenorth> cb 23 returned unknown / invalid result 0x5000
11:30:50 <andythenorth> so is that an nml problem, or an Iron Horse problem? :)
11:35:59 *** HerzogDeXtEr has joined #openttd
11:50:28 *** Progman has quit IRC (Remote host closed the connection)
12:10:15 <andythenorth> how can I revert a commit, without doing a revert commit?
12:10:43 <andythenorth> is there some -ve cherrypick method?
12:11:18 <nielsm> you can do an interactive rebase and delete the commit from the list
12:11:29 <nielsm> if it's just the latest commit you can do a hard reset
12:12:16 <andythenorth> it's a fair way down the tree
12:12:21 <michi_cc> Or the literal answer: git rever -n <commit>
12:12:53 <andythenorth> -n looks plausible
12:12:59 <nielsm> I think what andy is asking for it delete a commit, not make a new commit that undoes the change
12:13:14 <andythenorth> I can't now build Iron Horse
12:13:24 <andythenorth> it depends on recent features in nml
12:13:26 <nielsm> -n means "prepare the changes but let me edit it and manually commit it"
12:13:30 <andythenorth> but other recent commits broke the grf
12:13:44 <andythenorth> so I want to eliminate the broken part
12:13:52 <andythenorth> I guess I could just revert in a branch
12:14:29 <andythenorth> this is very likely to lead to shipping broken to prod :)
12:14:30 <michi_cc> interactive rebase allows you to make a commit go away, but then it is really away and if that was on a public branch, people might get angry at you.
12:14:41 <andythenorth> yeah that would be unwanted
12:14:48 <andythenorth> I'll just do it in a branch locally
12:19:33 <DorpsGek> [OpenTTD/nml] andythenorth opened issue #204: Commit 597685d67869554a5dd6e390ca2be72c78448ecf causes text cb to fail https://git.io/JOBTf
12:51:44 <FLHerne> andythenorth: Of course, the commit you want to remove fixes another bug that, IIRC, you reported :p
12:53:05 <andythenorth> see, never report bugs :)
12:53:34 <FLHerne> So you get to pick your breakage
12:54:45 <andythenorth> so no Horse release then :)
13:05:36 * andythenorth wonders what causes the broken
13:15:33 *** supermop_Home has joined #openttd
13:15:40 <supermop_Home> hovercraft time
13:18:33 <andythenorth> if you stand behind it on takeoff
13:18:38 <andythenorth> it does tend to blow sand in your face
13:18:55 <frosch123> meh, you cannot skip live streams
13:19:03 <andythenorth> like multiplayer
13:19:17 <DorpsGek> [OpenTTD/team] frosch123 updated pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JO8vi
13:19:39 <DorpsGek> [OpenTTD/team] frosch123 commented on pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JOBOR
13:19:41 <andythenorth> hmm but previous Horse releases were made without that commit
13:19:45 <andythenorth> so how broken can it be? :P
13:21:18 <DorpsGek> [OpenTTD/team] TrueBrain approved pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JOBO9
13:21:30 <frosch123> the 14:!5 departure is 6 minutes late?
13:21:45 <supermop_Home> it already left frosch123
13:21:51 <frosch123> oh, the line got shorter again :p
13:22:19 <supermop_Home> notice the wet spot where it was
13:23:10 <DorpsGek> [OpenTTD/team] frosch123 dismissed a review for pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JOBO9
13:23:13 <DorpsGek> [OpenTTD/team] frosch123 updated pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JO8vi
13:23:32 <DorpsGek> [OpenTTD/team] TrueBrain approved pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JOB3L
13:23:33 <frosch123> supermop_Home: yes, i noticed that hovercraft-shaped wet spot, once you mentioned it left :)
13:24:31 <andythenorth> oof now I do the AWS dance again for grf.farm
13:25:54 <DorpsGek> [OpenTTD/team] frosch123 merged pull request #193: Change: mark translator teams with 10 members or more as 'full'. https://git.io/JO8vi
13:26:58 <andythenorth> funny how slow this grf.farm compile is
13:28:05 <supermop_Home> this dawlish line seems to get a fair number of trains
13:28:25 <supermop_Home> considering it has no wires
13:28:34 <andythenorth> the webcam is pleasing
13:28:40 <andythenorth> I sometimes look in the morning when the sun is coming
13:28:50 <andythenorth> hmm maybe python shutil.copy is slow?
13:29:08 <supermop_Home> some 43s and 150(?) on it
13:29:27 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #9019: Automatically decide which translations are finished https://git.io/JOB3j
13:30:19 <DorpsGek> [OpenTTD/OpenTTD] frosch123 updated pull request #9019: Automatically decide which translations are finished https://git.io/JOIs3
13:30:30 <frosch123> ^^ don't read, only a rebase
13:32:22 <peter1138> "We accept bikes on all flights"
13:32:34 <peter1138> So is the hovercraft flying or sailing...
13:35:51 * peter1138 ponders throwing away some old bits.
13:35:57 <peter1138> Do I need this A7000+?
13:36:22 <peter1138> Well, it did last time I powered it on.
13:36:39 <andythenorth> my brother texted me about 5 years ago to ask if he could sell our childhood A3000 on ebay
13:36:49 <andythenorth> I said yeah, yeah it's old junk
13:37:01 <andythenorth> chocks away revival when?
13:37:03 <peter1138> I mean, it is, but...
13:37:12 <peter1138> Reverse engineer it?
13:37:40 <peter1138> OpenCA (the letters do not stand for anything...)
13:38:12 <andythenorth> should I just message the author on linkedin?
13:38:18 <andythenorth> ha what a terrible idea, linkedin
13:38:34 *** skrzyp1 has joined #openttd
13:38:36 *** skrzyp has quit IRC (Remote host closed the connection)
13:39:56 <peter1138> Here comes the hovercraft
13:40:22 <andythenorth> hovercraft landing!
13:41:04 <andythenorth> they really are quite cool, shame it's not the future we got
13:42:59 <peter1138> Surprising amount of people on it.
13:43:56 <DorpsGek> [OpenTTD/OpenTTD] frosch123 updated pull request #9019: Automatically decide which translations are finished https://git.io/JOIs3
13:44:37 <peter1138> 10:45 < michi_cc> Hmm, 1.11.1 this weekend, or are we still waiting on something important?
13:44:59 <peter1138> ^ seems a good idea, don't know the answer to the question though.
13:45:35 <andythenorth> peter1138 server game?
13:46:30 <peter1138> That sounds like a terrible game
13:50:28 <supermop_Home> does 2 hourly mean every 30 min or every 120 min?
14:15:33 <peter1138> Hovercraft leavingh
14:15:56 <peter1138> We need fix the game :p
14:19:52 <andythenorth> more hovercraft?
14:20:11 *** Synck has quit IRC (Quit: Page closed)
14:38:06 <supermop_Home> where's this craft
14:40:08 <FLHerne> Hovercraft time again?
14:44:03 <supermop_Home> noone has any luggage
14:45:29 <frosch123> note how there are both busses and hovercraft, but no trains
14:45:54 <FLHerne> There would be trains, but the line is closed to rebuild it for 'new' trains
14:46:10 <FLHerne> It's going to have 1970s London Underground trains instead of 1930s ones
14:46:31 <supermop_Home> what will they do with the old old ones?
14:47:18 <FLHerne> I assume the steam railway on the island will get one?
14:47:22 <frosch123> FLHerne: how do 50 year old vehicles compare to 90 year old ones?
14:47:51 <supermop_Home> FLHerne do any of the other lines on the island have 4th rail?
14:48:11 <FLHerne> The old ones are a bit rattley, I went on one a while ago
14:48:34 <FLHerne> New ones have been refurbed with air conditioning and things, so are probably nicer
14:48:56 <supermop_Home> hovercraft leaving
14:49:00 <FLHerne> (well, 3rd rail, they were modified for that)
14:49:19 <frosch123> weird, how can it turn in that angle?
14:49:23 <supermop_Home> why have 2 if they just turn the same one in 4 minutes
14:49:44 <FLHerne> There's a more frequent service when it isn't Sunday
14:49:53 <glx> andythenorth: if you can write a very simple grf to show the issue for nml#204 it would be a lot simpler to trace I think
14:49:58 <supermop_Home> it isn't sunday today either?
14:50:11 <FLHerne> Isn't lockdown, maybe?
14:50:18 <FLHerne> Why isn't it Sunday today?
14:50:38 <peter1138> It's off-peak season, certainly.
14:50:51 <glx> related to funerals maybe
14:53:07 *** Wormnest has joined #openttd
15:08:16 <peter1138> Finally, MS FS finished its 22GB update...
15:10:42 *** Synck has quit IRC (Quit: Page closed)
15:13:31 <glx> andythenorth: anyway I'll try with iron horse to compare nfo output
15:15:07 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9045: Fix: Corrupted savegame could crash the game by providing invalid gamelog enums. https://git.io/JOBgk
15:17:00 <supermop_Home> nice 153 at crewe
15:18:09 <supermop_Home> forbid 90 degree turns for ships should not apply to hovercraft
15:26:28 *** snail_UES_ has joined #openttd
15:26:45 <glx> oh I should compare output logs too
15:43:02 *** Speeder has joined #openttd
15:45:37 <glx> andythenorth: does it touch all vehicules or only some ?
15:45:52 <andythenorth> all that use buy menu cb
15:46:02 * andythenorth is afk mostly, but BBL
15:48:58 <andythenorth> yeah it's horrible though
15:49:07 * andythenorth wonders if there's a test case in examples
15:49:43 <peter1138> supermop_Home, what version are you plaing that still has forbid 90 degree turns for ships?
15:50:04 <glx> but it contains good search terms to look in generated nml and check optimisation warnings and resulting nfo
16:03:01 <glx> ok I can see the difference (and how it's wrong)
16:05:51 <andythenorth> any devs want to...actually play?
16:06:15 <andythenorth> pikka has a server, nocargoal, 10 years, takes about 2 hours
16:10:13 <peter1138> Also supports voice chat
16:11:16 <DorpsGek> [OpenTTD/nml] glx22 commented on issue #204: Commit 597685d causes buy menu text cb to fail in Iron Horse https://git.io/JOBTf
16:15:30 <peter1138> andythenorth, are you?
16:16:04 <andythenorth> peter1138 goes it MP
16:16:10 <frosch123> ah, magic string ranges :)
16:18:14 <DorpsGek> [OpenTTD/nml] frosch123 commented on issue #204: Commit 597685d causes buy menu text cb to fail in Iron Horse https://git.io/JOBTf
16:19:13 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
16:20:21 *** gelignite has joined #openttd
16:21:26 <frosch123> yay, summoning translators via team-names works :)
16:21:41 <glx> yeah looks I forgot a case to exclude for switch self rewrite
16:22:13 <glx> but iron horse as a test case is slow ;)
16:25:19 <glx> anyway I now where the error is and how to fix it
16:34:19 <glx> ah the diff between both nfo already looks better (no diff for the string actions)
16:43:43 *** EER has quit IRC (Remote host closed the connection)
16:52:49 <DorpsGek> [OpenTTD/nml] glx22 opened pull request #205: Fix #204: Consider string() as numeric constant when optimising switches https://git.io/JOBy7
16:54:30 <glx> oh I should do it for random switches too
16:55:29 <FLHerne> glx: It's not obvious to me why that omission causes it to be broken and not merely unoptimised?
16:55:47 <FLHerne> (I haven't looked very closely yet)
16:56:08 <glx> it's like a constant, so it should just be replaced inline in the caller
16:57:20 <peter1138> The comment on #205 suggests it's just an optimisation.
16:57:34 <frosch123> glx: what about line 401?
16:57:50 <frosch123> ther are many more places checking for ConstantNumeric+SpriteGroupRef
16:58:24 <glx> but instead the callee was rewritten to prevent return action creation with wrong scope
16:58:25 <FLHerne> peter1138: yes, and so is the code as far as I understand it, but it fixes #204
16:58:34 <FLHerne> So clearly I don't understand it far enough :p
16:58:44 <glx> frosch123: you mean 385 ;)
16:58:55 <glx> I'm working on amending my commit
16:58:56 <frosch123> yes, i have some weird version in my worknig copy
16:59:46 <glx> hmm I should add a function to easily check without forgetting a location
17:00:43 <frosch123> or some abstract member in Expression?
17:00:50 <frosch123> "is_constant" or something
17:00:57 <frosch123> instead of checking for types
17:01:22 <FLHerne> Make those things subclasses of ConstantNumeric?
17:01:36 <FLHerne> Might break the world :p
17:01:41 <frosch123> i would not dare that change :p
17:01:48 <frosch123> too many isinstance all over the palce
17:03:04 <DorpsGek> [OpenTTD/nml] glx22 updated pull request #205: Fix #204: Consider string() as numeric constant when optimising switches https://git.io/JOBy7
17:03:46 <glx> anyway scope for a change like that is bigger than this fix
17:04:12 <FLHerne> But seriously, this needs some better comments on how this works and why it's a fix
17:04:29 <FLHerne> I can see how it's *better*
17:07:35 <DorpsGek> [OpenTTD/nml] frosch123 approved pull request #205: Fix #204: Consider string() as numeric constant when optimising switches https://git.io/JOBHc
17:08:14 <peter1138> Magic language constructs
17:10:11 <frosch123> glx: maybe Boolean too? i have no idea why there are so many different "constant" classes
17:12:21 <glx> maybe, I don't know, at least a switch rewrite for boolean is not breaking
17:12:53 <FLHerne> glx: Why is it breaking for String?
17:13:16 <glx> because for String magic happens
17:14:26 <frosch123> FLHerne: String evaluates to a different numeric value when it is in a "return" compared to when it is in a "STORE_TEMP"
17:14:29 <glx> and the rewrite move the string in switch expression with a "return computed value" result
17:15:10 <glx> discarding the special handling
17:15:35 <FLHerne> Can you *please* put that in either a comment or at least the commit message?
17:16:45 <FLHerne> It's this kind of implicit magic that makes NML a nightmare to work on without breaking stuff
17:17:34 <frosch123> i guess one could add a comment to the "String" class, why it exists
17:17:44 <FLHerne> I mean, the commit message is about optimisation, the comment is about optimisation
17:17:47 <frosch123> but the magic about String is unrelated to this fix
17:18:16 <FLHerne> Nowhere does it say "we *must* 'optimise' this or the answer is wrong"
17:18:28 <FLHerne> That's not what optimisation means :p
17:19:56 <FLHerne> frosch123: I mean, the comment needs to say what glx says
17:20:08 <FLHerne> combined with what you said :p
17:21:12 <frosch123> wel,l if you find a wording for it :) i have a hard time explaining in switch.py that there is some fragile logic in action2var.py due to string.py
17:22:01 <glx> I guess I could at least say how self.optimised is used
17:23:00 <frosch123> FLHerne: the "bug" is that "parse_result" assumes the 99.9% case, and is broken for the 0.1% case
17:23:20 <frosch123> unfortunately the incomplete constness check in the optimiser created the 0.1% case
17:24:14 <frosch123> glx' fix changes the optimiser to corretly trigger the 99.9% case again. but noone fixes the original 0.1% case, because that's likely impossible and not helpful
17:25:56 <glx> basically, self.optimised is used as a replacement in the caller, but for complex expressions a return action is then created when parsing caller's results
17:26:37 * FLHerne still reading the bug, trying to actually understand what the problem is
17:26:39 <glx> and this return action uses caller's scope instead callee's one, that was the first bug
17:27:15 <FLHerne> frosch123: but then write *that* in the commit message
17:27:18 <glx> so instead, now the original callee is rewritten like the return action (but with the correct scope)
17:28:04 <glx> and String was catched in the rewrite while it can be used as a direct replacement
17:28:27 <frosch123> commit message is probably fine, after all it's a "fix #204"
17:28:35 <frosch123> but in the code, any comment would be unrelated and confusing
17:29:57 <FLHerne> I think you're right about the code comment
17:30:29 <FLHerne> But the commit message should explain how #204 happens and why this fixes it, like you've said here
17:30:40 <FLHerne> (where 'you' includes glx)
17:31:24 <FLHerne> It's not "So there's no need to rewrite switches for them."
17:33:17 <FLHerne> It's "rewriting switches for them may cause them to be evaluated to a different value because of <reasons>. The value can be directly substituted instead."
17:48:32 <DorpsGek> [OpenTTD/nml] glx22 merged pull request #205: Fix #204: Consider string() as numeric constant when optimising switches https://git.io/JOBy7
17:48:37 <DorpsGek> [OpenTTD/nml] glx22 closed issue #204: Commit 597685d causes buy menu text cb to fail in Iron Horse https://git.io/JOBTf
17:50:49 <DorpsGek> - Update: Translations from eints (by translators)
18:04:12 <peter1138> Andy will have to find another bug.
18:09:03 <andythenorth> oof do I have to? :|
18:10:59 <FLHerne> If you don't, NML will be declared Officially Bug-Free and all future complaints will be user error
18:11:27 <FLHerne> That's how it works for grf spec, right?
18:11:45 <FLHerne> The correct result is whatever it does
18:12:28 <FLHerne> If you expect something different, your code and the docs are both wrong
18:13:07 <frosch123> FLHerne: when ttdp was still alive, we made majority decisions from ottd, ttdp and specs :)
18:13:16 <frosch123> but with only ottd and specs, it's always a tie
18:14:08 <DorpsGek> [OpenTTD/bananas-api] erenes opened pull request #88: Change: Follow redirect convention for login also for developer flow https://git.io/JOBxj
18:16:27 <DorpsGek> [OpenTTD/bananas-api] erenes commented on pull request #88: Change: Follow redirect convention for login also for developer flow https://git.io/JOBpn
18:17:16 *** ChanServ sets mode: +v tokai
18:18:55 <DorpsGek> [OpenTTD/OpenTTD] LordAro closed issue #7513: Squirrel can stack overflow when GCing deep structures, leading to crash https://git.io/fjm1f
18:18:58 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9040: Issue 7513: stack overflow due to recursive GC marking and release https://git.io/JOcQU
18:19:10 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9041: Change: Improve layout and spacing of vehicle group widgets. https://git.io/JOCsI
18:19:22 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9045: Fix: Corrupted savegame could crash the game by providing invalid gamelog enums. https://git.io/JO8IE
18:19:40 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9046: Fix: Corrupted savegame could cause heap corruption by writing outside link graph edge matrix. https://git.io/JO8I9
18:19:51 <LordAro> well no one else was doing it...
18:20:13 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9044: Fix: Inconsistent button sizing on AI/GS setting window depending on scale settings. https://git.io/JOBpQ
18:20:25 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9044: Fix: Inconsistent button sizing on AI/GS setting window depending on scale settings. https://git.io/JOlhg
18:21:41 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9021: Fix: Do not unlock railtypes when enabling wagons with GameScript https://git.io/JOBph
18:24:09 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
18:25:41 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9047: Fix #6322: Attempt to crash the script instead of the whole game https://git.io/JOBho
18:26:20 <LordAro> frosch123: Rb noticed it :p
18:30:46 <andythenorth> thanks for fixing nml :D
18:30:51 <andythenorth> so appreciate :)
18:31:46 <peter1138> Hah, I check on my dinner and everyone is gooooone
18:33:37 <andythenorth> built a lot of hovercraft in a circle :P
18:37:34 <peter1138> Yup, I was there ;)
18:37:47 <peter1138> Right, I have dinner (and a beer) now.
18:38:04 *** tokai|noir has joined #openttd
18:38:04 *** ChanServ sets mode: +v tokai|noir
18:45:04 *** tokai has quit IRC (Ping timeout: 480 seconds)
18:56:03 <frosch123> does anyone know what 7912 did for newgrf? noone seems to have used/needed it in 16 months, so we could also just revert it again
18:57:14 <frosch123> oh, no, that's when the pr was opened... it was only merged in december :p
18:59:16 <frosch123> is the comment on tile_is_railtype english?
19:00:30 <andythenorth> it bothers me that railtype and tracktype are used interchangeably
19:00:31 <frosch123> there should probably be a comment on "current_railtype"
19:01:03 <frosch123> something like "this variable is useless due to equivalent railtypes, use tile_xxx_railtype instead"
19:01:57 <frosch123> andythenorth: i started to use "tracktype" when all 3 of road/rail/tramtype are meant
19:02:53 <frosch123> maybe it's easier to merge theose variables into a single row
19:03:14 <andythenorth> ok that makes sense
19:03:33 <frosch123> current_railtype\ncurrent_roadtype\ncurrent_tramtype | Entry from rail/road/tram translation table or 0xFF |...
19:04:40 <supermop_Home> looks very different with the tide out
19:07:26 <frosch123> supermop_Home: looks like the different timezone is reasonable. it's already dark here
19:08:34 <Wolf01> peter1138 what's your thought on the vr humble bundle? I already have borderlands 2 vr, but the price difference from mid and top tier is just $0.20
19:10:50 <frosch123> looks good. though there are infinite things to improve on that page :)
19:11:52 <FLHerne> andythenorth: Certainly in NML code, 'tracktype' is what frosch123 said
19:11:57 <FLHerne> I think I introduced that
19:12:07 <FLHerne> When I was de-copypasting the roadtypes code
19:12:29 <andythenorth> I have to overcome my hesitancy about wiki every time I touch that page
19:12:39 <andythenorth> obviously we're unlikely to replace it
19:12:40 <glx> oh there are probably some variables to rename on the wiki (4 TTD string, and MAP_TYPE_RECTANGULAR)
19:12:47 <andythenorth> but it's a disincentive to improvement
19:13:38 <andythenorth> anyway more to add :)
19:14:50 <andythenorth> I somewhat prefer TrueWiki to edit in, at least the UI is better :P
19:15:20 *** jottyfan has joined #openttd
19:18:15 <supermop_Home> frosch123 its probably slightly north of you as well?
19:19:05 <supermop_Home> York looks lighter than Ryde
19:19:16 <frosch123> oh yes, it is. i confused isle-of-man with isle-of-wight
19:20:24 <frosch123> eh, no. i confused it twice now
19:20:29 <frosch123> the hovercraft is isle-of-wight
19:20:40 <frosch123> so, no, it's only very little mor enorth
19:22:45 <frosch123> less than a degree more north
19:23:27 <frosch123> but 10.4° more west
19:23:54 <peter1138> LordAro, balls, #9042 doesn't fix all cases (makes the regular case slightly worse :/)
19:24:23 <frosch123> which is 40 minutes local-time
19:25:08 <supermop_Home> the mud flats make it really clear why they use hovercraft instead of regular boats
19:25:19 <supermop_Home> (or why that pier is so long)
19:27:03 <FLHerne> Yeah, there are normal boats that dock at the end of the pier
19:27:13 <andythenorth> tile_has_catenary - no parameter, right?
19:27:23 <supermop_Home> FLHerne when we went in 97, we took a regular ferry
19:28:32 <frosch123> also, yay, that docs thingie was somewhere on my todo list :)
19:30:52 <andythenorth> there are some I don't really understand
19:30:57 <andythenorth> but I'll do the ones I can
19:31:27 *** NausiyanNyan[m] has left #openttd (User left)
19:37:11 <supermop_Home> here comes last one of the day
19:53:29 <andythenorth> frosch123 these vars need the silly coloured [nml 0.x] labels?
19:53:56 <andythenorth> and ottd version I guess
19:54:14 <frosch123> i think ottd labels are enough
19:54:23 <andythenorth> one day I will figure out how we just version the spec
19:54:24 <frosch123> the nml labels are more for syntax-changes within nml itself
19:54:33 <andythenorth> and don't have all this weird label shenanigans
19:54:56 <frosch123> even if you version the spec, those labels are still helpful
19:55:18 <frosch123> though ideally nml would know what vars need which ottd version, and generate a check / bananas info file
20:00:06 <andythenorth> I've done tile_has_catenary
20:01:06 <peter1138> LordAro, oh. I think I've found a suitable fix.
20:01:10 <andythenorth> wait, where has 'vehicles with parameter' gone?
20:01:15 <andythenorth> wiki editing makes literally no sense
20:01:20 <andythenorth> entire section has disappeared
20:02:03 * andythenorth just reverts everything
20:03:07 <andythenorth> wikis are such a technical format
20:03:20 <andythenorth> the target audience is very narrow
20:03:52 *** snail_UES_ has quit IRC (Quit: snail_UES_)
20:08:28 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
20:09:30 <glx> hmm reading parse_result() I think I can also add supported_by_actionD in the rewrite exclusion test
20:09:35 <DorpsGek> [OpenTTD/OpenTTD] PeterN opened pull request #9050: Fix #9042: Make multiplayer server list height auto-fill window. https://git.io/JORmE
20:10:24 *** Wormnest has joined #openttd
20:10:45 <andythenorth> I should ... do something about that ^
20:10:53 <andythenorth> (a) docs (b) use it
20:11:00 <peter1138> ^ This would've fixed the original issue as well, but it's nice to have the correct fix for the details panel as well.
20:12:48 <andythenorth> so if inflation is on, disable Iron Horse?
20:13:08 <peter1138> Why would you do that?
20:13:30 <andythenorth> because bug reports
20:13:33 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
20:13:37 <andythenorth> 'grf is too expensive'
20:14:14 *** WormnestAndroid has joined #openttd
20:15:01 <frosch123> add a "hard mode" to the title
20:15:02 <andythenorth> also I promised a test case for 8427
20:17:45 <peter1138> Wolf01, VR... waste of time and money for me. I've barely touched mine.
20:18:33 <Wolf01> Me too, but for other reasons :(
20:19:30 <andythenorth> does this have any obvious order?
20:19:36 <andythenorth> shall I just append 'inflation' at the end?
20:21:45 *** Progman has joined #openttd
20:22:57 <glx> but filtering with supported_by_actionD would be suboptimal if optimised switch is used in multiple location, like the scope incorrect version was
20:23:36 <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #9050: Fix #9042: Make multiplayer server list height auto-fill window. https://git.io/JOROf
20:23:47 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #9050: Fix #9042: Make multiplayer server list height auto-fill window. https://git.io/JORmE
20:24:55 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #9047: Fix #6322: Attempt to crash the script instead of the whole game https://git.io/JO42l
20:24:58 <DorpsGek> [OpenTTD/OpenTTD] michicc closed issue #6322: AI can allocate more memory than the system has, crashing the game https://git.io/JeRb5
20:26:14 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #9021: Fix: Do not unlock railtypes when enabling wagons with GameScript https://git.io/JOIaF
20:29:27 <peter1138> Hmm, what's the point of all these routers at home if I don't set up OSPF/BGP?
20:30:25 <glx> andythenorth: I think override flags, when enabled will discard any change decided by newgrf callback (and maybe even the callback call itself)
20:31:44 * andythenorth will have to read the full diff I guess
20:32:46 <glx> oh callback is called it seems, but depending on flags it can discard production increase, production decrease and closure
20:34:01 <andythenorth> maybe the intent of the var is to allow the grf to read the flags the GS is setting?
20:34:25 <andythenorth> the flags are just bits yes?
20:35:11 <andythenorth> I have no mental model of where this fits in the control flow
20:35:50 <andythenorth> but I guess it nerfs cb 29 and cb 35 in favour of GS deciding?
20:37:04 <glx> I'm reading the diff (with extended hidden code to see the full function)
20:37:31 <andythenorth> there's no sample GS afaict
20:37:41 *** snail_UES_ has joined #openttd
20:37:47 <andythenorth> oh samu made something
20:39:06 <andythenorth> ok so this lets GS do things like protect an industry from closure
20:39:25 <andythenorth> so var 0x47 is _probably_ so that newgrf can co-operate with that
20:39:46 <andythenorth> this means we now have to keep GS and newgrf in sync :)
20:40:31 * andythenorth hasn't found how it interacts with cb 29 and 35 though
20:41:44 <glx> ok finished reading, I think the new var is there to inform newgrf its decisions may be discarded
20:42:42 <glx> CBID_INDUSTRY_MONTHLYPROD_CHANGE and CBID_INDUSTRY_PRODUCTION_CHANGE still happens
20:43:17 <glx> but resulting production changes can be overriden by GS choice
20:43:37 <andythenorth> oh in terms of just the callback result?
20:44:17 <glx> result is analysed as usual, but modifications are not applied if GS said so
20:44:47 <andythenorth> that's interesting
20:44:53 <andythenorth> not sure why newgrf needs to know that
20:45:20 <glx> to know it doesn't need to do complex calculation for nothing probably
20:46:54 <andythenorth> can't imagine bothering to code that
20:47:00 *** sla_ro|master has quit IRC ()
20:47:20 * andythenorth tries to imagine why newgrf would bother writing and testing code for that
20:47:35 <andythenorth> oh because openttd can't just not run the cb
20:47:45 <andythenorth> it's specific results that are discarded
20:48:13 <andythenorth> well this will be interesting to write docs for
20:48:51 <andythenorth> "after calculating whether your result is increase, decrease or closure, you can check var 0x47 flags to see if you needed to do your calculation"
20:48:59 <peter1138> Imagine how efficient it could all be if it was designed from the ground up...
20:49:25 <andythenorth> I am misunderstanding right?
20:49:38 <andythenorth> this is a post-hoc var to check flags to see if you needed to run the check of flags?
20:50:04 <glx> idea is to check flags before, but well :)
20:50:30 <andythenorth> but you don't know if the result is increase or decrease or closure or static :)
20:50:47 <andythenorth> well it is what it is
20:50:52 <andythenorth> and what it is ... is undocumented
20:51:35 <frosch123> andythenorth: newgrf can implement production changes without using those callbacks. so if they wanted to cooperate with gs, they would need some info from the gs. maybe it's about that
20:51:52 <frosch123> does firs have production changes? last time i played, it didn't
20:52:05 <andythenorth> I want to add them
20:52:22 <andythenorth> but I can never remember your newsletter about which economy is bad (smooth?)
20:52:24 <frosch123> but player keep on complaining why it's so hard? :p
20:52:52 <andythenorth> usually they want it to be more hard, so XIS: exists
20:53:21 <frosch123> an hour ago you said horse is too hard :p
20:53:27 <andythenorth> players say so yes
20:53:55 <frosch123> with sufficient number of players, you can find any opinion
20:54:03 <frosch123> but it would recommend to ignore forums :p
20:54:14 <andythenorth> I balance forums, discord, steam and reddit now
20:54:24 <andythenorth> the only opinion I ignore is...irc, yes? :)
20:54:34 <frosch123> also a good idea :)
20:54:47 <andythenorth> ok I am not the person to document var 0x47, at least not tonight
20:54:51 <frosch123> forums really got a weird demographics
20:55:04 <andythenorth> no comment on individuals
20:55:11 <andythenorth> but I don't recognise them as people
20:55:28 <frosch123> andythenorth: let the PR author document the basics. they should know the intention
20:55:30 <andythenorth> steam is just standard gamerz, it's like every conversation is written by my kids
20:55:36 <glx> seems direct prod_level changes are allowed
20:56:09 <andythenorth> I was hoping to have a sensible discussion about newgf <-> GS interaction soon :P
20:56:44 <andythenorth> mostly regions :P
20:57:06 <peter1138> (how would you do that?)
20:57:10 <andythenorth> I found the spare ranges, using the livestream from frosch
20:57:24 <peter1138> 2 separate layers of 2cc
20:57:40 <andythenorth> is the biggest problem just the UI?
20:57:58 <peter1138> The biggest problem is "why?"
20:58:09 <andythenorth> wagons different colour to engines in groups
20:58:14 <andythenorth> groups are the winningest
20:58:42 <peter1138> Insteasd of removing livery schemes, Let's implement the livery schemes inside groups!
20:58:55 <andythenorth> using company colours
20:59:06 <andythenorth> I sort of have evil ideas about that
21:00:06 <andythenorth> and then forces certain colours
21:00:29 <andythenorth> I did have lots of them, but the colour combos started to remind me of playing Mortal Kombat
21:00:32 <andythenorth> only more boring
21:05:09 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
21:28:33 <peter1138> Weird how the network server list remembers its content, but the "content downloading" window (what a title, I've just noticed) doesn't.
21:31:28 <peter1138> What's so special about 25x game speed?
21:31:48 <peter1138> Why does my fast-forward nearly always settle on that?
21:32:27 <peter1138> That exists. Well that explains it.
21:33:48 <peter1138> Heh, if you change the setting with FFWD active, it has no effect.
21:37:33 <peter1138> Hmm, okay, transparency options window does resize, but the toolbars don't.
21:38:10 <peter1138> Rather, don't resize to below how they started.
21:49:38 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
21:53:45 *** nielsm has quit IRC (Remote host closed the connection)
22:03:03 <peter1138> Okay, fixed that. Incorrect "base" call.
22:05:01 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:11:42 *** snail_UES_ has quit IRC (Quit: snail_UES_)
22:12:57 *** Progman has quit IRC (Remote host closed the connection)
22:15:58 *** jottyfan has quit IRC (Quit: jottyfan)
22:17:25 *** gelignite has quit IRC (Quit: Stay safe!)
22:22:42 *** EER has quit IRC (Remote host closed the connection)
22:24:28 *** Wormnest has joined #openttd
22:27:57 <peter1138> I guess dozing off in front of VS Code suboptimal.
22:35:07 <andythenorth> same, except google
22:35:14 *** andythenorth has quit IRC (Quit: andythenorth)
23:17:28 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
23:53:32 *** iSoSyS has quit IRC (Remote host closed the connection)
continue to next day ⏵