IRC logs for #openttd on OFTC at 2019-01-27
⏴ go to previous day
00:00:15 <peter1138> nielsm, is that enough screenshots in 7108 now? ;)
00:00:17 * andythenorth wonders if PRs need stricter criteria
00:01:23 <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6m3
00:01:33 <andythenorth> like a quota or something :P
00:01:52 <andythenorth> proof that you can program
00:02:09 <LordAro> PRs would still end up getting opened
00:02:26 <LordAro> unless you're planning on blocking people, which seems ultimately pointless
00:02:36 <andythenorth> the problem I have is that I closed some, which didn't have viability
00:02:41 <andythenorth> and now they're just back, split up
00:02:47 <andythenorth> so now we have more PRs, not fewer
00:03:08 <andythenorth> and they're still going nowhere
00:03:52 <LordAro> i don't think they're back as such
00:04:15 <LordAro> just that the large one has been split up into multiple, as was asked
00:06:04 * andythenorth should test the mac thing
00:06:06 *** andythenorth has left #openttd
00:06:09 <supermop_Home_> buy menu groups?
00:11:54 <peter1138> Some feature that was pondered a few months ago, but has not seen any coding yet.
00:13:22 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep opened pull request #7114: NPF bails out when source tile is also a destination tile https://git.io/fh6mK
00:30:50 <Samu> resets non locked settings now
00:31:07 <Samu> instead of just plain out disabling the reset button
00:38:59 <peter1138> Why does my game run at 1.02x consistently?
00:39:13 <Samu> hmm, i guess the reset button could be disabled if all settings are locked, what do you think?
00:39:19 <peter1138> I thought it was just an oddity because I was running in a Hyper-V VM and displaying via VNC.
00:39:25 <peter1138> But this laptop also runs at 1.02x
00:39:49 <nielsm> the timing calculation is slightly broken in possibly every video driver
00:40:02 <nielsm> the win32 one tends to run at 0.98x
00:40:13 <peter1138> I remember having a patch to use the std hrtimer.
00:40:14 <nielsm> SDL one tends to run at 1.02x
00:40:24 <Samu> I have a dilema: should reset button also reset invisible settings
00:40:28 <peter1138> Hmm, I wonder where that went.
00:40:35 <nielsm> or otherwise do something to make up the difference
00:40:57 <peter1138> I have a patch that should allow catch-up, but when it runs at 1.02x that doesn't achieve much.
00:42:55 *** snail_UES_ has joined #openttd
00:44:56 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on issue #7091: Dropdown menu can set value to a parameter of a different AI https://git.io/fh6YM
00:45:19 <glx> yes same message but it's to double link :)
00:47:13 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
00:47:52 <Samu> what to expect from a button that says "Reset"
00:48:44 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7092: Fix #7091: Close dropdown menu windows after rebuilding AI/GS settings https://git.io/fh6Yd
00:49:09 <peter1138> 1.00x using the high resolution clock.
00:49:09 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7093: Fix #7090: Close Query String window after rebuilding AI/GS settings https://git.io/fh6YA
00:49:21 <peter1138> Well, 1.00x - 1.01x
00:49:43 <nielsm> peter1138 that would be better yes :)
00:49:52 <nielsm> and then using the same algo for timing on all platforms
00:50:04 <nielsm> (maybe splitting it out to a class of its own)
00:52:03 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fh6Of
00:52:47 <peter1138> Back to 1.02x when I use only milliseconds.
00:52:53 <glx> hey they are not closed Samu
00:52:59 <peter1138> I guess it needs sub-millisecond accuracy.
00:53:37 <Samu> part of my ai gui rework was making it not close windows unless it's really necessary
00:53:41 <glx> but I think it will be better to fix all problems at the same time
00:54:03 <Samu> it currently closes too much
00:54:10 *** Thedarkb-T60 has joined #openttd
00:54:40 <Samu> (it was also not closing in situations it would crash, of course)
00:54:53 <glx> well for now when I click on a different line, the settings are still showing the previous selected line
00:56:18 <glx> most of the time if you change selected line you should close the windows, because there's no way to pass it the new data
00:56:29 *** Thedarkb-T60 has joined #openttd
00:57:17 <Samu> if the script is still the same, it's still gonna change the correct parameter
00:57:33 <glx> well there is a way but it's not done actually
00:57:48 <Samu> in this situation, i wouldn't close
00:58:23 <glx> a possible option would be to pass the new slot in InvalidData
00:58:31 <LordAro> many of the issues arise because the windows refer to a particular slot in the list, rather than an actual newgrf/script config
00:58:54 <LordAro> that's how i was going to go about solving the ai textfile issue, anyway
00:59:04 <glx> yes but at least it's a number so it's possible to give it the new slot
00:59:21 <LordAro> closing all textfile windows because a slot gets changed seems a bit.. nuclear
01:00:05 <glx> yes this text windows uses a slot too so it's doable
01:00:38 <glx> for newgrf I needed a more aggressive way because it's using directly a grfconfig
01:01:00 <LordAro> so why was that breaking?
01:01:04 * nielsm ponders how to structure music code that can select between multiple drivers depending on music format
01:01:13 <nielsm> going to sleep on that
01:01:13 *** Thedarkb-T60 has joined #openttd
01:01:33 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6On
01:10:37 <peter1138> Damn, I've got some coding to do :p
01:12:19 <Samu> wondering if this do what I want it to do
01:13:01 <Samu> not entirely sure I need the free it first
01:14:35 <glx> on removal the grfconfig is deleted
01:22:16 <glx> hmm maybe instead of passing this->active_sel or this->avail_sel, I could pass their address, so if they change the window gets the new value too
01:23:06 <glx> but still require closing in some cases
01:28:58 <Samu> you could check how i handled some of it on my overhaul topic
01:29:09 <Samu> to get some ideas, I suppose
01:29:27 <glx> I'm talking about newgrf window
01:29:42 <Samu> oh, I haven't touched it, sorry
01:45:04 <Samu> I'm a terrible bit-wise mathmathician
01:50:07 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7106: Fix #7089: close NewGRF textfile windows when their data are invalid https://git.io/fh631
01:54:39 *** Thedarkb-T60 has joined #openttd
02:04:15 <glx> my idea should work, just changing 7 lines made the textwindow update the caption
02:04:54 <glx> of course it crashes when I click when I shouldn't :)
02:13:59 <Samu> this shall do, for now, due to me being a noob in bit-wise operations
02:15:58 <glx> ah you mean boolean algebra
02:16:37 <Samu> and some flags bitmask stuff
02:17:41 <Samu> developer settings are hidden from gui by default
02:17:51 <Samu> the reset button shall not reset them
02:18:30 <Samu> but if they're visible, then reset them, if they're not locked
02:18:45 <Samu> slightly confusing explanation
02:21:58 <glx> editable is menu or ingame stuff
02:22:30 <glx> visible is dev or not a dev setting
02:22:51 <glx> gui_dev or not a dev setting
02:23:33 <Samu> a dev setting is hidden on gui when ai dev tools is disabled
02:23:45 <glx> then you just need if (editable && !visible)
02:25:16 <glx> and that way the code explains by itself
02:31:11 <Samu> if editable and not visible is slightly misleading
02:31:42 <Samu> i only want to reset if they're editable and visible, not not visible
02:34:13 <Samu> looks better, but the visible part is quite ugly atm
02:37:27 <glx> make it visible = ai_dev_tool || not a dev setting
02:39:32 <glx> so visible is true if the developper settings are not hidden, or if it's not a developper setting
02:40:33 <glx> and you want to remove the hidden settings I think
02:41:23 <Samu> no no, it's hidden, I dont want to reset parameters that i dont see
02:44:35 <glx> ah right, your previous code wasn't clear enough ;)
02:45:25 <Samu> bool visible = _settings_client.gui.ai_developer_tools || ((*config_item).flags & SCRIPTCONFIG_DEVELOPER) == 0;
02:45:25 <glx> so yes editable && visible
02:47:09 <glx> at the end of the function the list will contain only hidden settings
02:48:26 <glx> and the not editable settings
02:49:11 <Samu> it's a list of settings that are not to be reset
02:49:58 <glx> but you are removing them from the setting list
02:50:25 <glx> this->settings will only contain the not to be modified settings
02:51:06 <glx> I guess the other will be readded with default values later
02:51:14 <Samu> yes, exactly that the intention, some other function will use this list to know where not to touch
02:51:30 <glx> the function name is misleading then
02:55:32 <Samu> not sure how would I name it
02:55:48 <Samu> ResetEditableVisibleSettings
02:56:55 <glx> it's more a RemoveEditableSettings
02:57:23 <glx> hidden are not editable ;)
02:58:44 <Samu> sec, let me find in the code the comment
03:00:06 <Samu> the list that is created by that, is the list I'm working on
03:01:46 <Samu> hmm the hidden ones may pose a problem, indeed
03:02:37 <glx> editable && visible takes care of that
03:02:53 <Samu> if they are in this list, yes
03:03:08 <Samu> if they're not... it means they're at their defaults already
03:03:36 <glx> they are editable but not visible so considered not editable
03:06:34 <Samu> brb testing, to make sure
03:13:10 <Samu> in main menu it's working correctly, now testing in-game
03:18:08 <Samu> also working as I intended
03:18:36 <Samu> one could argue about hidden parameters not being reset though
03:19:15 <Samu> part of me says it's wrong to reset something that is not visible
03:19:55 <Samu> my other part would like them to also be reseted without resorting to ai dev tools setting required
03:20:25 <Samu> so hmm... what is the right thing to do?
03:54:27 *** Progman_ has joined #openttd
03:55:26 *** Eddi|zuHause2 has joined #openttd
03:57:54 *** Progman_ is now known as Progman
04:11:43 <Samu> i named it ResetSettingsGUI
04:14:48 <Samu> RemoveEditableSettings hmm
04:16:01 <Samu> ResetEditableAndVisibleSettings
04:16:13 *** Eddi|zuHause2 is now known as Eddi|zuHause
04:16:53 <Samu> I click the reset button and this function is called
04:18:05 <Samu> not sure about the 'free((*it).first);' code
04:19:15 *** techmagus has joined #openttd
04:33:55 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened pull request #7115: Fix #6452: Reset only editable and visible settings from GUI https://git.io/fh6nJ
05:05:47 <Samu> ah snap, a small problem, will get it fixed asap
05:12:15 <Samu> it's not a small problem, it's quite big, damn it :|
05:14:00 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick closed pull request #7115: Fix #6452: Reset only editable and visible settings from GUI https://git.io/fh6nJ
05:14:01 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7115: Fix #6452: Reset only editable and visible settings from GUI https://git.io/fh6nV
06:11:23 *** techmagus has joined #openttd
07:42:00 *** Alberth has joined #openttd
07:42:00 *** ChanServ sets mode: +o Alberth
07:44:20 *** techmagus has joined #openttd
08:10:24 *** sla_ro|master has joined #openttd
08:40:37 *** andythenorth has joined #openttd
08:42:48 <Alberth> good news, updated site, new nightlies
08:43:04 <Alberth> congrats to you and TB
09:08:52 <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth commented on issue #6916: Re-implement building binaries via compile farm https://git.io/fh6lR
09:25:58 <andythenorth> bendelnos or hard to draw
09:26:56 <andythenorth> I found a bug in AV9.8
09:32:43 <Alberth> mass air transport with zellepins :)
09:34:25 <andythenorth> maybe I should make a plane grf :P
09:35:02 <Pikka> I was thinking maybe I should add the zellepin
09:35:11 <andythenorth> AV9.8 is pretty much perfect
09:35:17 <andythenorth> apart from the major flaw
09:35:30 <andythenorth> probably one is enough
09:35:46 <andythenorth> just make it reasonably hench
09:37:27 <andythenorth> HEQS Revival set?
09:37:42 <Pikka> whether I get to it, who knows
09:38:30 <Pikka> it just needs to be all steam tractors
09:38:33 <andythenorth> ha ha, I didn't try plane liveries yet
09:38:53 <andythenorth> moar company colour needed Pikka
09:44:19 * andythenorth considers automatic rainbow patch
09:44:33 <andythenorth> every time you build a vehicle, it gets the next rainbow colour
09:55:06 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh68R
09:55:46 <andythenorth> that icon though :(
09:55:55 <andythenorth> get pikka to redraw it?
09:56:54 <andythenorth> I did the best I can, it follows house style
09:57:10 <andythenorth> it's the right colours and size, but meh
09:58:24 <andythenorth> maybe I can fix it another time
09:58:38 <andythenorth> so is this where vehicle intro dates get randomised?
09:58:41 <peter1138> Huh. When I tested last night the window title was company colour. Now it's not. What the heck?
09:59:31 <andythenorth> I didn't update from your remote before I made screenshots
09:59:45 <andythenorth> or are we talking about something else?
09:59:56 <peter1138> I haven't updated, I tested just now and it's grey too.
10:00:06 <peter1138> Was I that sleepy I was imagining it?
10:00:16 <andythenorth> petern/colour-scheme-window-title
10:00:28 <andythenorth> and didn't merge it?
10:01:07 <peter1138> I intended to fix it in that branch, but it was already the right colour. So it just contains some other fixes.
10:01:39 * peter1138 recompiles that one.
10:02:55 * peter1138 waits for the Core 2 Duo to warm up.
10:03:37 <peter1138> If I wasn't on mobile data tethering I could probably just develop over VNC to my fast PC at home.
10:05:14 <peter1138> I do like the fact that this laptop shipped in an era when 32bit Windows was still around, so it actually only has 3GB memory.
10:05:38 <andythenorth> so is GB(r, 0, 9) reading 9 bits from random, offset 0?
10:05:42 * andythenorth totally guessing
10:06:25 <andythenorth> so if that's in days, it's about 16 months?
10:06:50 * andythenorth wonders if the 9 can be exposed to newgrf in action 0
10:06:52 <peter1138> I see, it's already company colour in master...
10:06:59 <peter1138> That's broken in group-livery :p
10:07:31 <andythenorth> which window title?
10:07:38 <andythenorth> oh yes, I see :)
10:07:58 <andythenorth> yes, this would be the window of all windows to use CC in
10:08:50 *** Progman has joined #openttd
10:09:34 <andythenorth> so what property number is free for all vehicle types? :P
10:09:58 <Borg> peter1138: develop over VNC? ugh.. with VNC tech you actually mean?
10:10:24 <Borg> every one I used.. just sucks... RDP itself is far better
10:10:46 <andythenorth> I guess we don't really keep common numbers for properties across vehicles
10:11:27 <peter1138> Borg, not in Linux land.
10:11:29 <andythenorth> can I get a negative result out of GB(r, 0, n) ?
10:12:26 <andythenorth> otherwise n/2 - GB(r, 0, n) I guess
10:13:21 <peter1138> Something like how we discussed the other day with that 14 bit callback result.
10:13:25 <andythenorth> train prop 2E: word, max range for random intro date, days, signed
10:14:03 <andythenorth> @calc (65536 / 2) / 30
10:14:03 <DorpsGek> andythenorth: 1092.26666667
10:14:06 <dwfreed> peter1138: x2go > VNC, fwiw
10:14:06 <peter1138> Did you want it negative and positive, or only negative?
10:14:31 <andythenorth> I want to allow randomised vehicles to be introduced later or earlier than ei->base_intro
10:14:54 *** HerzogDeXtEr has joined #openttd
10:14:55 <andythenorth> 500 months random is quite a lot eh :P
10:15:00 <andythenorth> 16 bits might be overkill
10:15:44 <peter1138> Oh, you're talking NewGRF spec, not patching OpenTTD, right?
10:16:03 <andythenorth> rather than a new prop 'force intro date to be intro date'
10:16:10 <andythenorth> it would be better to just specify the random range to be used
10:16:16 <dwfreed> 500 months is 41.667 years
10:16:27 <peter1138> I have a patch that adds a "force intro date to be same as this other vehicle"
10:16:41 *** techmagus has joined #openttd
10:16:45 <andythenorth> I think that's a rare case, and should just be 'set both to range 0'
10:17:34 <andythenorth> I don't code all those EMU sets with special coaches
10:18:11 <andythenorth> I'm more interested in cutting down the random range on wagons
10:18:15 * peter1138 waits for this CPU to finish again. Maybe I should put the kettle on, it's probably run out of steam.
10:18:48 <andythenorth> it would be interesting to have some joker vehicles that have variation of ~+/-10 years
10:19:47 * andythenorth tries to magic up Eddi|zuHause
10:43:02 <peter1138> Yeah, so I broke it. Hmm.
10:45:29 <andythenorth> I should fix my tt key
10:45:39 <andythenorth> without breaking itt
10:55:24 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fhK9F
10:55:32 <peter1138> Ok, window caption correct colour now.
10:56:03 <peter1138> FinishInitNested(company) resets this->owner to INVALID_COMPANY.
10:56:11 <peter1138> I assume there's a reason, but I donm't know it.
10:56:43 <andythenorth> this is so much easier than getting patches of pastebin :P
11:01:31 <DorpsGek_II> [OpenTTD/OpenTTD] andythenorth commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6Bl
11:04:26 <andythenorth> hmm bendelino nearly drawn
11:07:54 *** gelignite has joined #openttd
11:08:24 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7116: Change: Add company name to colour scheme window title https://git.io/fh6BP
11:09:28 <peter1138> So did Pikka make better icons? :D
11:10:14 <peter1138> This laptop battery is constantly at 'Estimating' :/
11:17:40 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh closed issue #6574: Heliport "hangar" can cause stuck helicopters if it replaced a helidepot/helistation. https://git.io/fh6Rt
11:17:55 <andythenorth> they're all split up now, right? o_O
11:18:35 <nielsm> btw if adding new graphics to the extraset included, it's probably a good idea to make a new image file for the new/changed sprites
11:18:49 <nielsm> to avoid modifying the existing spritesheet
11:19:26 <nielsm> i.e. even if modifying an existing sprite, copy that sprite to a new graphics file and leave the original (wrong) sprite in the original spritesheet
11:19:37 <nielsm> that way slowly separating out the sprites that do change
11:20:29 <peter1138> nielsm, that's what group-livery does.
11:22:49 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7100: Fix #6574: Remove go to hangar orders when rebuilding airport https://git.io/fh6Rz
11:29:49 <peter1138> It's why I was asking earlier in the week (or maybe before) if I should use one file per sprite or one file per grouping :)
11:30:09 <nielsm> I may have been out at that time
11:31:18 <peter1138> Turns out there's more to hidden railtypes than just hiding it from the list :p
11:35:03 *** techmagus has joined #openttd
11:38:12 <TrueBrain> planetmaker: you could tweet, not? mind sending a tweet about the latest post?
11:38:16 <TrueBrain> (especially that nightlies are now available)
11:39:55 <planetmaker> tweet... there is some account, yes... I've never used it though, I think
11:40:00 <planetmaker> I didn't setup it
11:40:40 <TrueBrain> I thought you were the last to find out how it worked :P
11:40:50 <TrueBrain> I remember you saying before that you never used it :D
11:40:53 <planetmaker> hm... might actually...
11:40:55 <TrueBrain> hmm .. so who here does know how to tweet :P
11:41:13 <planetmaker> my browser remembers openttd twitter account :D
11:41:27 *** supermop_Home_ has quit IRC
11:42:32 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh60m
11:43:19 <TrueBrain> planetmaker: so it was you that tweeted last :P
11:43:48 <planetmaker> did I say that I don't like twitter? :P
11:43:56 <planetmaker> but yes, I'm on it
11:44:03 <peter1138> I wonder if a hidden railtype needs to be 'available', just not buildable (nor shown in lists)
11:44:03 <planetmaker> your message is far too long :P
11:44:23 <TrueBrain> make something nice out of it :D
11:44:27 <peter1138> I have a feeling if I mask it everywhere it will simply not exist and all the related stuff will not work either :D
11:44:46 <TrueBrain> planetmaker: the topic alone should be enough, with a link, for all I care :P
11:45:56 <Alberth> "nightlies are back!"
11:46:52 <nielsm> hmm I wonder if there would be a way to have msbuild build the baseset definition files while compiling the main source code, since they technically don't depend on each other
11:47:02 <nielsm> they just both need to be present for the game to be runnable
11:47:50 <TrueBrain> that works! Tnx planetmaker :)
11:51:22 *** andythenorth_ has joined #openttd
11:51:31 <TrueBrain> planetmaker: I did try to redirect most URLs,just not all. This seems to be one of them :D
11:51:45 <TrueBrain> hard to fix, that instance
11:51:50 <TrueBrain> your name is wrong andythenorth_ :P
11:51:57 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7116: Change: Add company name to colour scheme window title https://git.io/fh60r
11:52:20 <andythenorth_> my phone will kick me out in 180 seconds
11:52:37 <andythenorth_> ios power saving kills network connections
11:53:40 <andythenorth_> so intermodal containers...
11:54:01 <TrueBrain> hmm, no, I can fix the news ..
11:54:03 <andythenorth_> how many TEU per 1/8 vehicle length?
11:55:59 <DorpsGek_II> [OpenTTD/website] TrueBrain opened pull request #46: Fix: [nginx] redirect news URLs without the language tag to the right… https://git.io/fh60X
11:56:00 <andythenorth_> twenty foot equivalent unit I think
11:56:26 <andythenorth_> 8/8 is about 64 foot in horse, usually
11:56:42 <andythenorth_> except when it’s 50 foot or 80 foot
11:57:08 <planetmaker> so... in essence 8/8 is 64 foot. Except when it isn't :P
11:57:19 <peter1138> Hmm. So I set "RAIL" to be hidden.
11:57:25 <peter1138> I now can't build it, woo :D
11:57:47 <planetmaker> peter1138, what are you trying to solve?
11:59:10 <andythenorth_> so 8/8 is probably a 40 foot + 20 foot, or two 30 foot
11:59:32 <peter1138> It was a feature request a while back, but we didn't do it because we only had 16 types.
11:59:46 <planetmaker> what do hidden railtypes give me as player? (honest question)
12:00:56 <peter1138> Something to do with mixed power engine.
12:01:18 <peter1138> nielsm, that's not diff, that's Samu ;)
12:01:28 <nielsm> "okay so the function was renamed and most of the code in it deleted... and then a new function with the original name and code inserted below
12:01:58 <andythenorth_> it’s more about cross compatible vehicles
12:02:09 <andythenorth_> mixed power is already done
12:02:41 <Alberth> likely samu wrote a new function above it, and diff gets confused
12:02:59 <DorpsGek_II> [OpenTTD/website] TrueBrain merged pull request #46: Fix: [nginx] redirect news URLs without the language tag to the right place https://git.io/fh60X
12:03:01 <andythenorth_> e.g universal vehicles or similar :)
12:03:20 <planetmaker> ah, I see. ty. peter. Interesting
12:03:22 <Alberth> we have railtypes but we dont :p
12:04:02 <peter1138> planetmaker, I don't know how it'll actually work to know if it's a GOOD or BAD feature ;)
12:04:28 <andythenorth_> specifically I want it for Hog, to permit certain go-anywhere vehicles with NRT
12:04:40 <andythenorth_> and George has some use case too
12:06:19 <peter1138> Is "Add #xxxx: " permitted?
12:07:39 <peter1138> nielsm, oops, I didn't generate widgets.
12:08:29 <andythenorth_> did someone fix obs file crap? I didn’t see them yesterday
12:08:52 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN dismissed a review for pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh60m
12:08:53 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fhK9F
12:09:20 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7111: Codechange: Rework automatic service for helicopters (#6493) https://git.io/fh6E4
12:09:38 <peter1138> We ought to call generate widgets as part of the build process.
12:10:45 <nielsm> wait did I not read that PR properly... is its _only_ purpose to do weird things when that dumb serviceathelipad setting is enabled?
12:11:05 <peter1138> nielsm, it's only purpose is to fix things when it's enabled, yes.
12:11:18 <nielsm> oh, well may as well just reject and close it then
12:11:31 <peter1138> If your answer is "don't play with that option" then... maybe we should remove the option?
12:11:44 <nielsm> it'll break some ancient savegames
12:11:47 <peter1138> I wasn't aware that it was a legacy compatibility option.
12:11:57 <peter1138> You can still enable it, can't you?
12:12:13 <andythenorth_> I wish we could stop trying to boil the ocean
12:12:17 <nielsm> but it's only there because ttdpatch implemented it way back
12:12:26 <peter1138> IIRC it only breaks autoreplace.
12:12:34 <TrueBrain> make it so you can only unselect it, but never select it again? :P
12:12:39 <peter1138> (and maybe autorenew)
12:12:57 <peter1138> TrueBrain, voids your warranty!
12:13:15 <TrueBrain> planetmaker: your news link now works again!
12:14:13 <andythenorth_> software is like a building
12:14:23 <andythenorth_> some doors don’t shut properly
12:14:42 <andythenorth_> but does the building work?
12:14:54 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6EK
12:15:03 <andythenorth_> why do we need to fix all these edge and corner cases?
12:15:26 <andythenorth_> they’re not causing crash reports or failure demand
12:15:48 <andythenorth_> there’s more demand from all these frigging PRs about not much
12:15:52 <TrueBrain> andythenorth_: if they are trivial and someone wants to fix it, what-ever
12:16:16 <TrueBrain> but .. the trivial part is important :P
12:16:22 <peter1138> Hmm, if you disable service at heliport, and there's not airports with hangars in its route, does it cope with servicing?
12:16:44 <peter1138> andythenorth_, usability. Bug fixing isn't only about crashes.
12:17:14 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6EX
12:18:23 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh closed issue #6493: Helicopters do not autorenew/autoreplace while in routes consisting of heliports https://git.io/fh6EN
12:18:24 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on issue #6493: Helicopters do not autorenew/autoreplace while in routes consisting of heliports https://git.io/fh6EA
12:18:32 <andythenorth_> do these PRs actually fix bugs? or just add string with high chance of new bugs?
12:19:38 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7111: Codechange: Rework automatic service for helicopters (#6493) https://git.io/fh6uT
12:19:39 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh closed pull request #7111: Codechange: Rework automatic service for helicopters (#6493) https://git.io/fhKhu
12:21:18 <peter1138> andythenorth_, which PRs?
12:21:33 <andythenorth_> the 14 or so samu ones
12:22:02 <andythenorth_> most of them don’t seem easy for reviewers to understand?
12:22:22 <peter1138> I just haven't bothered looking in detail at most of them.
12:22:38 <andythenorth_> I’m +1 to fixing bugs generally, just not ocean boiling
12:22:39 <peter1138> Don't worry yourself over it, we have the bot now ;p
12:23:05 <andythenorth_> natural attrition :p
12:23:44 <andythenorth_> so ‘loading gauge’ for railtypes? o_O
12:24:25 <andythenorth_> some intermodal containers are taller than others :p
12:26:01 <peter1138> Okay, I actually need to rework quite a bit for this hierarchical group livery stuff.
12:26:26 <peter1138> And in doing that, I can probably implement both Default and Inherit as options.
12:26:33 <nielsm> andythenorth_, trains that can't go through tunnels? :D
12:26:44 <nielsm> or require larger/more expensive tunnels
12:27:32 <nielsm> or cannot run under catenary
12:29:40 <TrueBrain> yippie, almost no 404s worth mentioning :D
12:29:51 <TrueBrain> robots.txt is the highest 404 :P
12:29:57 <TrueBrain> and 'ads.txt' ... lol
12:30:15 <andythenorth_> India runs doublestack containers on conventional wagons, under catenary :o
12:30:41 <TrueBrain> only 30 404s, of which 3 are solved. The rest seems to be junk :)
12:30:53 <nielsm> afaik US electrified railways have catenary significantly further up than european ones
12:31:24 <andythenorth_> 64 railtypes anyone? :)
12:31:38 <nielsm> (I once tried putting a german loco on an american electrified route in train simulator, and the pantograph didn't reach the wires :P)
12:33:11 <peter1138> So I need to ... duplicate part of the groups GUI.
12:33:23 <peter1138> Or somehow extract the code into a separate class.
12:33:37 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7102: Fix: Direct the aircraft to the correct location of the hangar when skipping to a go to hangar order https://git.io/fh6ua
12:33:52 <andythenorth_> collapsible groups :p
12:33:56 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh merged pull request #7102: Fix: Direct the aircraft to the correct location of the hangar when skipping to a go to hangar order https://git.io/fhKJN
12:34:31 <peter1138> This code uses smallvectors and stuff.
12:34:52 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh merged pull request #7116: Change: Add company name to colour scheme window title https://git.io/fh6BP
12:34:52 <peter1138> I wonder how to convert it to a generate hierarchical lis.
12:35:26 <peter1138> Because, you know, build vehicle groupings at some point.
12:35:37 <DorpsGek_II> [OpenTTD/OpenTTD] J0anJosep commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6uK
12:35:51 <andythenorth_> how did 3298 do it in the collapsible groups patch!
12:37:12 <peter1138> It's probably still specific to the groups window.
12:37:15 <peter1138> That's what I want to avoid.
12:39:22 <peter1138> Yeah, that's very much specific to vehicle groups only.
12:39:39 <peter1138> The folded state is stored in the vehicle group.
12:39:53 <peter1138> Means it persists, which is nice, but it's ugly and non-generic.
12:42:23 <peter1138> I did hierarchical groups didn't I?
12:42:30 <peter1138> How does my code work?
12:44:15 <peter1138> Ah it uses a temporary list.
12:54:22 <peter1138> AddParents() should be named AddChildren()
13:00:24 *** andythenorth_ has joined #openttd
13:07:15 <peter1138> Ok, so I just duplicated the code :s
13:07:18 <peter1138> Ok, so I just duplicated the code :s
13:29:01 *** frosch123 has joined #openttd
13:30:45 *** Thedarkb1-T60 has joined #openttd
13:33:48 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fh4oP
13:34:44 <LordAro> otherwise known as "random keyboard mashing because github requires a dismiss message"
13:36:19 <Alberth> missing the opportunity to enrich the world with a deep philosophical essay about the meaning of life
13:37:03 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7117: Fix group hierarchy not being visible in RTL languages. https://git.io/fh6gK
13:38:04 <peter1138> I mean, it's a separate change, but also...
13:38:11 <peter1138> Related but not really related.
13:38:53 <peter1138> Also.. reducing indenting in a function designed to set up indenting... :D
13:39:23 <LordAro> #7003 works well for me, anyone got any opposition to it?
13:40:02 <peter1138> Did the 2 - _font_zoom change?
13:41:07 <peter1138> THe "invert" to be consistent doesn't make sense, because gui_zoom isn't inverted.
13:42:42 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fh6gy
13:44:28 <frosch123> TrueBrain: in case you want to tweet. info@ drafts folder has everything
13:44:39 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fh6gH
13:47:11 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7003: Feature #6918: Add option to adjust font size separately from GUI size. https://git.io/fh6gA
13:48:00 <LordAro> how about #7005 vs #7080 ? the latter is a lot more code...
13:48:38 <LordAro> and also #7025 vs #7047
13:50:45 <peter1138> I'm not keen on the "lot more code"
13:51:02 <peter1138> But I've not looked at both in detail.
13:51:37 <peter1138> Oh, the last two I'm pretty sure need more work.
13:51:48 <peter1138> They are tagged as wip of course.
13:52:16 <peter1138> It'd be better not to have to add a load of storage for it.
13:52:22 <peter1138> But if necessary...
14:16:44 <peter1138> I wonder if there is an argument for just storing the local authority in the map array for more tiles?
14:17:05 <peter1138> There are other places where it's needed and it is expensve to find.
14:25:51 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN updated pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fhK9F
14:26:34 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7106: Fix #7089: close NewGRF textfile windows when their data are invalid https://git.io/fh6af
14:30:08 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7108: Feature: Group liveries, and livery window usability enhancements. https://git.io/fh6am
14:35:50 *** andythenorth has joined #openttd
14:39:50 <TrueBrain> nielsm: adding bytes in the middle of a packet might give a bit more work than you would expect; as I assume you have to bump the network protocol version for that?
14:39:55 <TrueBrain> also possibly means updating MSU
14:40:03 <TrueBrain> (not a big deal, just hasn't been done in a long time)
14:40:09 <nielsm> TrueBrain nope, the reading code handles that
14:40:21 <TrueBrain> so it is send as string? :)
14:40:21 <nielsm> it reads a 0-terminated string of variable length
14:40:28 <nielsm> not a fixed width field
14:40:40 <nielsm> I checked that already :)
14:40:48 <TrueBrain> in the database it is a TINYTEXT
14:41:44 <Gabda> can someone help me testing PR #7064?
14:41:44 <nielsm> of course, a git revision hash is 40 hex characters
14:42:03 <nielsm> so it still can't be included in full, unless it gets encoded in some evil format
14:42:54 <glx> a partial git hash seems to be enough
14:43:30 <glx> if it contains a minimum number of chars
14:43:40 <Eddi|zuHause> well, a 40-hex-string equals a 20 byte integer, but then you can't transmit that 0-terminated :p
14:44:04 <nielsm> base 64 the binary then it fits in 30 bytes
14:44:09 <TrueBrain> 10 chars for date + twice a -. Branch name + git tag has 22 chars left
14:44:23 <TrueBrain> in case of master, it means 16 chars from the git hash
14:45:07 <Eddi|zuHause> so if you take, say, 8 chars from the hash, you have 14 chars for branch name?
14:45:24 <Eddi|zuHause> or 13, if you include the "g" as we used to
14:45:33 <TrueBrain> if you take the first 12 chars from the git hash, it is "unlikely" to have a collision
14:45:42 <TrueBrain> but .. that value was 7 a few years ago :P
14:45:47 <glx> don't forget the "M" for modified repo
14:46:13 <TrueBrain> pretty sure the M always drops off
14:46:15 <Eddi|zuHause> TrueBrain: that number grows 1 per year?
14:46:17 <TrueBrain> given it is at the end :P
14:46:29 <glx> yes but I think it shouldn't
14:46:56 <Eddi|zuHause> TrueBrain: the entire point is to not drop the important bits by cutting out stuff in the middle? :p
14:47:23 <nielsm> or we could completely rework the server info format and pack a bunch of the fields :D
14:48:56 <glx> hmm date should not be part of the server revision
14:48:58 <TrueBrain> or reinvent the protocol a bit that not everything has to be in a single UDP packet :)
14:49:07 <Eddi|zuHause> so what happened to the idea to have an internal version for multiplayer comparison and a display version?
14:49:35 <glx> because you can compile a valid version another day
14:49:52 <TrueBrain> but short term vs long term, increasing the length now might be a good solution no matter what :)
14:50:55 <Eddi|zuHause> the CI once again not picking up builds?
14:51:13 <peter1138> glx, date of revision, no?
14:51:30 <andythenorth> duck pate on toast?
14:51:46 <glx> ah maybe, can't remember exactly the generation step
14:51:48 <peter1138> Right, off to visit the MIL again. Back on tonight when I'm home.
14:55:16 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7118: Add #5006: Flag to hide rail type from construction. https://git.io/fh6Vk
15:01:25 <TrueBrain> Eddi|zuHause: why it doesnt pick up 7108 is a mystery .. it ispicking up new ones just fine
15:01:38 <TrueBrain> owh, now it does dequeue
15:03:15 <Eddi|zuHause> we definitely need a solution to make this more reliable
15:04:23 <TrueBrain> glx: I fixed your comment :P
15:05:01 <TrueBrain> (if you go in github to the file you want to link to, click on the line, click on the three dots, click on "Copy permalink", you can use that link to get this :)
15:06:02 <TrueBrain> I just did :D As it was annoying me I didnt know how to do it :P
15:07:05 *** supermop_Home has joined #openttd
15:07:10 <TrueBrain> so ... MSU .. I should move that to the new infrastructure
15:07:13 <glx> hmm we could just reorder stuff in _openttd_revision
15:07:14 <TrueBrain> but the current implementation is a bitch
15:08:09 <TrueBrain> just keep in mind we also use this stuff to display to users which version the server is on etc
15:08:17 <TrueBrain> might not be ideal if that is only a part of a git hash ;)
15:09:14 <glx> but git hash is only 10 chars
15:09:21 <TrueBrain> OpenTTD Coop server should upgrade to latest nightly!
15:09:30 <glx> so there's some chars left after it
15:09:39 <TrueBrain> when did it became 10?!
15:10:06 <TrueBrain> a git hash really is 32 :)
15:10:47 <TrueBrain> anyway, just keep in mind that how we currently do it, it is also a user-friendly thingy
15:11:14 <LordAro> at the point at which 10 chars is insufficient to differentiate revisions, it can be further adjusted :p
15:11:18 <TrueBrain> but okay, I would suggest splitting long-term solutions from short-term solutions :)
15:11:24 <nielsm> how about if the 'g' prefix for git hash is kept, use it to indicate modified-ness?
15:11:35 <TrueBrain> nielsm: G if modified :P
15:11:37 <nielsm> e.g. 'm' instead of 'g' for a modified revision
15:11:51 <nielsm> and 'u' for unknown if modified (??)
15:12:33 <TrueBrain> "There are 186 clients, 283 IPv4 servers and 45 IPv6 servers." <- we always had more servers than players :P
15:12:57 <Samu> aren't servers clients as well?
15:13:13 <frosch123> i miss the 0.6.3 server
15:13:16 <TrueBrain> welcome in 2019, where dedicated servers are a thing for over 10 years :P
15:13:32 <TrueBrain> frosch123: yeah, what happened to it!
15:13:44 <frosch123> maybe someone remote-crashed it
15:14:20 <nielsm> but yeah someone should set up a server running nightly!
15:15:21 <TrueBrain> OpenTTDCoop .... why you so slow :(
15:15:32 <TrueBrain> I blame planetmaker :)
15:15:38 <glx> hmm we don't use short hash, we just take 8 first chars of the hash
15:15:55 <frosch123> andythenorth: there was a more elaborte spec, but it had issues
15:16:19 <andythenorth> I thought of 3 or so options so far :P
15:16:26 <frosch123> i think the idea was to randomise introdates, but make sure to not reorder introduction
15:16:46 <frosch123> which implies, that vehicles with same introdate keep the same introdate, randomised as a group
15:17:17 <glx> using --short gets a 9 chars hash
15:18:45 <Eddi|zuHause> frosch123: i'm not really so fond of that idea
15:19:06 <andythenorth> the crudest solution is to expose the random range to newgrf, signed
15:19:15 <andythenorth> then for simultaneous introduction, set 0
15:19:20 <frosch123> i am sure you can find a closed issue with more thought on it :p
15:19:47 <frosch123> yes, but now dorpsgek
15:19:51 <andythenorth> the crude solution loses random dates for many vehicles, if the grf needs a lot of simultaneous vehicles
15:20:04 <andythenorth> in my grfs, that would work, but YMMV etc
15:20:05 <frosch123> i am not a fan of ad-hoc solutions
15:20:14 <Eddi|zuHause> andythenorth: that is also not a good solution
15:20:38 <andythenorth> so then we seem to need to decompose two issues
15:20:45 <andythenorth> 1. simultaneous introduction
15:20:52 <andythenorth> 2. control over random range
15:21:04 <andythenorth> 1 > 2 in this case :P
15:21:15 <frosch123> 2 is for realism-nerds. i am not a fan of that
15:21:16 <andythenorth> 2 was an interesting side effect, not a goal
15:21:24 <glx> the main problem is the revision string was created in svn time and svn revs are way shorter than git hashes
15:21:25 <Eddi|zuHause> frosch123: i was saying yesterday that, similar to my suggestion on grouping vehicles, there should be a property for a "master" vehicle which gets used for introduction. once the "master" vehicle gets introduced, the others get introduced as well
15:21:28 <frosch123> same people who disable breakdowns :p
15:21:32 <andythenorth> I would use it for joker trains, setting 10 years or so
15:21:47 <andythenorth> every game is more different
15:22:00 <andythenorth> oof I should re-enable breakdowns now that RVs are fixed
15:22:40 <Borg> btw.. its kinda sad that you can move trains to groups.. from station lists
15:22:47 <glx> oh and there's still no solutions for _openttd_newgrf_version
15:22:51 <andythenorth> so 1. just forces simultaneous introduction, seems easy?
15:22:55 <andythenorth> but what about deadlocks?
15:23:07 <andythenorth> do vehicles have to be put into sets?
15:23:45 <andythenorth> if simply A wants B to be available, and B wants A available, that deadlocks?
15:24:18 <frosch123> glx: i think stable version is enough
15:24:45 <Eddi|zuHause> you can implement (1) by generating the random range based on game seed and intro date property, then all introdates get a reproducible random range
15:25:10 <Eddi|zuHause> so all vehicles introduced on the same day will stay on the same day, even if randomized
15:25:20 <andythenorth> so they're pinned
15:25:39 <andythenorth> no newgrf spec extension needed?
15:26:06 <andythenorth> it has a definite appeal
15:26:38 <andythenorth> ha I never set days, just year, but eh
15:26:42 <andythenorth> that's changeable :P
15:28:05 <Eddi|zuHause> also a side effect: you don't get new introdates by setting "reset_engines" or "reload_newgrfs"
15:28:20 <andythenorth> not sure if that's positive or negative
15:28:23 <andythenorth> call it neutral :P
15:28:47 <nielsm> glx, how about newgrf version is savegame version + some constant?
15:29:01 <andythenorth> supermop_Home: yo
15:29:16 <Eddi|zuHause> nielsm: savegame version isn't really related to newgrf version
15:29:37 <Samu> wow, i got approved patch requests, wow nice thx :)
15:29:50 <frosch123> i would like to see more newgrf experimental branches
15:30:43 <frosch123> so, either we go for individual feature-tests, or we stay with stable-release only
15:30:54 <Eddi|zuHause> frosch123: the newgrf people already were semi-opposed to including patchpack stuff in the wiki
15:31:00 <frosch123> linear versioning between stable is pointless
15:31:01 <andythenorth> frosch123 long-running ones, towards a new grf version? Or just throway experiments?
15:31:54 <andythenorth> there is a high impedance mismatch to extending grf spec
15:32:07 <frosch123> andythenorth: throw-away experiments.
15:32:45 <Eddi|zuHause> i understand it, there is lots of resitance to support experimental stuff from a newgrf developer perspective
15:33:29 <frosch123> my subjective impression is: newgrf guy wants to try something but cannot code themself, and requests it for master implementation. then it is either ignored or implemented. when it is implemented it turns out as not-as-expected, so noone continues to use it
15:34:47 <Eddi|zuHause> that's why there's so many George requests
15:35:38 <frosch123> the pattern is even somewhat reinforcing itself. when half-baked ideas are rejected, people try to work around them by requesting minimal changes, that make stuff even worse
15:36:02 <frosch123> like people requesting that vehicle-reverse becomes an integer :p
15:36:09 <glx> newgrf guy should find someone to create a fork then PR once it's ok
15:36:11 <nielsm> (I hope the 16 cargoes in/out stuff becomes used)
15:36:54 <Samu> was 4096x4096 maps "not-as-expected"?
15:37:15 <planetmaker> [14:32:45] i understand it, there is lots of resitance to support experimental stuff from a newgrf developer perspective <-- it's a steep thing to write a NewGRF and then throw everything away. Not the usual attitude of NewGRF people
15:37:17 <Eddi|zuHause> frosch123: we should permaban the next person who requests that :p
15:37:39 <andythenorth> ^^^ I tried to sum up the problem
15:37:44 <andythenorth> not with solutions either :P
15:38:16 <andythenorth> who wanted vehicle-reverse as an integer? That's nuts :P
15:38:39 <andythenorth> I wanted an arbitrary counter, byte, incrementing on ctrl-click, and vehicle-reverse is mod(2) of the counter :P
15:39:10 <glx> well to be fair 2 forks may be needed, 1 for openttd, 1 for nml
15:39:43 <andythenorth> I still do want an arbitrary counter, because I think it's funny
15:39:52 <planetmaker> nml is as easily handled as OpenTTD wrt forks...
15:40:03 <planetmaker> it's python... so you don't need anyone to compile anything really
15:40:23 <frosch123> planetmaker: does nml support 16-in/16-out yet? :p
15:40:24 <glx> but it's always better to try the idea before asking for implementation in master and discard it after testing
15:40:39 <andythenorth> I am getting (very nice) newgrf author DMs on the forums asking about that
15:40:45 <andythenorth> really polite ones
15:41:09 <frosch123> planetmaker: language barriers. you need to know openttd, nfo, nml, python, wiki, ...
15:41:12 <andythenorth> mostly it's not done because I didn't make a comprehensive industry test grf in the nml docs
15:41:26 <andythenorth> because I don't really understand the new prod cb format
15:41:32 <Eddi|zuHause> planetmaker: concerning george, there was more than one incidence where i went like "here, try out this patch" and he came back with "i can't compile, must be in a nightly to test it"
15:41:34 <andythenorth> and I can't remember what we did about the various props
15:41:38 <planetmaker> indeed... a lot of stuff to consider when adding newgrf things
15:41:38 <frosch123> reason why i personally abandoned nforenum
15:41:56 <andythenorth> also I spent a lot of time making rainbow trains :P
15:42:08 *** Thedarkb1-T60 has joined #openttd
15:42:21 <planetmaker> where does nforenum come into the picture you painted, frosch123 ?
15:42:34 <planetmaker> you mean additional to nfo addition in grfcodec?
15:43:32 <frosch123> a few years back we used to maintain nforenum, and add every new property, callback, action ...
15:43:47 <andythenorth> how unrewarding :)
15:43:59 <frosch123> now nforenum purely exists to compute the "sprite count" in openttd.grf
15:44:00 <planetmaker> well. You and rubi did. And... it's a pita from what I see when I read the patches
15:44:06 <andythenorth> how many people left who know how to use grfcodec then?
15:44:14 <andythenorth> Pikka ^ you on nml yet?
15:44:38 <Eddi|zuHause> we established earlier that Pikka is not a sane person :p
15:44:40 <planetmaker> grfcodec is easy. Just writing plain nfo is not
15:45:00 <andythenorth> it's just so much easier when developing newgrf features
15:45:02 <andythenorth> no nml patch needed
15:45:14 <andythenorth> grfcodec just accepts whatever it's given, if it's valid
15:45:30 <andythenorth> but counting bytes :(
15:45:42 <planetmaker> well. You can do stuff like accessing arbitrary var in NML, too
15:45:44 <frosch123> also, patching nml is actually not easy
15:45:54 <Eddi|zuHause> anyone tested the last #7000 yet with more complex RTT?
15:45:55 <frosch123> various grf people need windows binaries insteady of python
15:46:10 <glx> yeah writting nfo is like compiling asm manually
15:46:10 <frosch123> and only devzone (and me who copied it from devzone) can compile nml binaries
15:46:12 <planetmaker> but adding stuff to NML ... yes... not exactly the easiest thing. NewGRF internals never are easy
15:46:22 <planetmaker> there's also too much backward compatibility in NewGRF land
15:46:37 <planetmaker> a zillion ways to do stuff. And it varies from feature to feature what and how it is done
15:47:14 <Eddi|zuHause> i'm pretty sure compiling eddi-nml was broken last time i tried
15:47:19 <planetmaker> but throwing away all newgrf stuff and inventing newnewgrf... not a good option either (see netscape - anyone recall?)
15:47:32 <andythenorth> it's not a big enough problem to throw it away
15:47:39 <andythenorth> although I do like a bonfire
15:47:49 <andythenorth> it's not broken, it's just high-friction
15:48:05 <planetmaker> the single problems are not big. But NewGRF additions approach unmaintainable
15:48:24 <andythenorth> the primary issue is the lack of active devs imo
15:48:38 <andythenorth> when there were 3 or 4 people advancing the spec, problems get reduced
15:48:45 <planetmaker> that it is... yes
15:49:28 <andythenorth> recently, if I don't write a test grf, and possibly the nml patch and docs too
15:49:31 <andythenorth> newgrf features don't move
15:49:42 <andythenorth> but I am not the only newgrf dev :P
15:50:12 <andythenorth> oops turns out andythenorth is feeling guilty about the 16-cargo-industries thing :P
15:50:49 <andythenorth> ha ha, and if I don't help with the website the CF doesn't get finished as fast :)
15:51:01 <LordAro> andythenorth: also, #7109
15:51:25 <andythenorth> if I don't test mac support stuff, mac support dies
15:51:36 <andythenorth> how do I find all this time to play tank games? :P
15:52:02 <planetmaker> you actually deserve a medal, andythenorth
15:52:14 <frosch123> well, thanks for the new website :)
15:52:16 <planetmaker> your assessment is quite accurate
15:52:19 <glx> andythenorth: not the first time mac support is menaced, happened when we lost bjarni
15:52:33 <Samu> they now take time turning around, nice touch
15:52:51 <andythenorth> they do, nice isn't it :)
15:53:44 <andythenorth> so the choice of Jekyll for the website was driven by the idea that someone like Kamnet could then do the news
15:53:55 <andythenorth> it's just a git PR to make a post
15:54:05 <andythenorth> not sure the same approach works for newgrf dev
15:54:19 <Eddi|zuHause> did we post news about new nightly builds yet?
15:54:33 <glx> anyone can write a post now, it's a good thing
15:54:43 <frosch123> Eddi|zuHause: i read it on twitter first
15:54:56 <Samu> just spotted some misalignment doing the turn to the right when the ship is facing south
15:56:12 <Samu> it's reproduced in the main menu ships
15:58:08 <Samu> needs someone with good eye to spot the problem
15:59:45 <glx> it's still on the old infra
16:00:13 <glx> I know because the icon shows ipv6 ;)
16:01:18 <Samu> how am i reporting this, it's difficult to spot
16:01:47 <Samu> or maybe i'm just being too incisive
16:05:06 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fh6rh
16:08:43 <Eddi|zuHause> i don't think that's how it works
16:09:25 <andythenorth> the ships, or the impedance mismatch? o_O
16:09:49 <andythenorth> does it work at all currently? o_O
16:09:58 <andythenorth> what's the goal?
16:11:35 <Eddi|zuHause> it needs a "holistic" approach? :p
16:11:48 <Eddi|zuHause> talk more, make it easier to test stuff, ...
16:12:13 <andythenorth> the spec development is driven by
16:12:27 <Eddi|zuHause> there's lots of parts to that
16:12:35 <andythenorth> - george and michael asking for things via ticket, and not getting them
16:12:40 <Eddi|zuHause> like making compiled versions from PRs
16:12:43 <andythenorth> - me and V asking for things here, and getting them
16:12:48 *** synchris has joined #openttd
16:13:17 <Eddi|zuHause> or having GRFs containing experimental stuff also work in trunk/master
16:13:26 <andythenorth> so the reason we have so many NRT grfs is because TB made the binary, and frosch made a windows nml binary
16:13:40 <andythenorth> single biggest intervention we could make is repeating that
16:14:04 <andythenorth> ok so your second point
16:14:07 <Eddi|zuHause> but doing that manually is overkill for smaller things like #7000
16:14:27 <andythenorth> ha 7000, another debt I have to pay :)
16:15:08 <andythenorth> so when I load NRT Road Hog in trunk, it's disabled
16:15:19 <andythenorth> rather than having features disabled
16:15:27 <andythenorth> that's possibly the most correct thing to do, right?
16:15:51 <nielsm> missing the actual real network revision fix ;)
16:16:05 <nielsm> but gets rid of the buffer size warnings about the gamelog
16:16:11 <Eddi|zuHause> it might need some deeper spec considerations to make master "forward compatible" with grfs containing experimental features
16:17:10 <andythenorth> it might be over-engineering
16:17:21 <andythenorth> let's explore some more
16:17:28 <andythenorth> so I manage that by having hg branches
16:17:41 <andythenorth> we'll ignore my incompetence with hg, and hg's inadequacy
16:17:46 <andythenorth> it's still a hassle
16:17:54 <andythenorth> most authors likely don't even have vcs
16:18:02 <andythenorth> so branching is a non-thing for them
16:18:32 <andythenorth> and then there's the distribution issue: sharing grfs for testing is a PITA
16:19:03 <andythenorth> I rarely post OpenTTD bugs I see because I'd have to make a clean savegame without unreleased grfs
16:19:24 <andythenorth> impedance mismatch on someone else getting those grfs is too high
16:22:25 <andythenorth> new bananas? o_O
16:22:41 <andythenorth> nml binaries on openttd CF?
16:29:10 <supermop_Home> 75 mph large flat cars show no cargo when full of pig iron andythenorth
16:32:13 <glx> nielsm: but you don't get the full hash, I think the substring used for githash still contains g[first 8 chars of hash]
16:33:59 <glx> nielsm: ignore me, I incorrectly readed the scripts
16:36:24 <planetmaker> nml binaries via the same channels as openttd, grfcodec, basesets would be awesome
16:47:39 <Eddi|zuHause> andythenorth: that doesn't help with people who don't really want to share their unfinished works
16:49:50 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7100: Fix #6574: Remove go to hangar orders when rebuilding airport https://git.io/fh6KV
16:50:21 <glx> an it's understandable, they don't want to get bug reports for unfinished work
16:50:38 <Samu> damn it, i was thinking it would turn it into a comment~
16:51:48 <glx> oh use the other inupt box then ;)
16:51:51 <Samu> it's a link to a reply to a comment :(
16:52:00 <Samu> i wanted to link to the comment instead
16:52:58 <Samu> guess im just copy pasting
16:53:08 <Samu> avoids needless confusion
16:53:34 <TrueBrain> planetmaker / andythenorth: there should really be no issue doing that (except possibly for some dollars per month; lolz)
16:59:07 <planetmaker> TrueBrain, do we pay per build?
16:59:22 <planetmaker> or how do additional builds add to our bill?
17:00:25 <TrueBrain> owh, yeah, I could be a bit more verbose there :D Lol :) No, builds are free (well, Azure gives 10 agents free for Open Source)
17:00:38 <TrueBrain> but publishing of the binaries goes on the CDN, which costs a tiny bit of money per download
17:01:18 <TrueBrain> like .. $0.01 dollar per GB transfered
17:02:26 <TrueBrain> so nothing to worry about in any real sense
17:06:43 <Samu> dang 27th january doesnt open
17:08:20 *** Wormnest has joined #openttd
17:08:39 <Samu> yesterday i closed a pr of mine, need to get a link to it, its lost in the log
17:11:01 <LordAro> why do you think we keep insisting stuff goes there?
17:12:31 <glx> they exist for a reason :)
17:14:00 <glx> on PR page, there's a search box, click on filter, select "your pull requests", the click on "xxx closed" on top of the list
17:14:57 <TrueBrain> planetmaker: do you have any influence which version openttdcoop runs on its server?
17:15:56 <Gabda> Can I ask for a WIP tag for #7120?
17:16:00 <planetmaker> TrueBrain, yes. What in particular?
17:16:13 <LordAro> Gabda: is this your 3rd PR for the same feature? :p
17:16:52 <planetmaker> I pay the bill and thus have HV access :P
17:17:13 <TrueBrain> planetmaker: I meant the OpenTTD server; it is not running the latest nightly :)
17:17:51 *** sla_ro|master2 has joined #openttd
17:17:56 <Gabda> I plan to drop the first one, I don't like that, but the conversation under the PR is good
17:18:22 <TrueBrain> glx: if that is a PR people like, someone would have to make the changes in that repository so it is supported by the Linux machines
17:18:37 <Gabda> and this 3rd one a little different, it can be merged with the 2nd one
17:19:25 <Samu> fixing 7115 is very easy on my ai gui overhaul patch. Just need to ask 'IsConsideredDead(this->slot)' and pass it to the reseter function
17:19:29 <Gabda> and after the merge it can have some new features, like when building a new industry, it indicates in real time on mouseover if that is a good place, or you cannot build it for that town
17:19:39 <Samu> but on master, its gonna be hard~
17:19:56 *** snail_UES_ has joined #openttd
17:20:15 <glx> yup that's what I was reading
17:20:36 <Samu> IsConsideredDead is something I magicated specially for my overhaul needs
17:20:43 <Samu> can't simply copy what it does
17:20:50 <Samu> need to mimic master behaviour
17:21:07 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fh666
17:21:11 <Samu> and master behaviour is buggy as hell in the scenario editor
17:21:41 <TrueBrain> glx: seems he is also on the case :)
17:21:59 <planetmaker> hehe... it needs updating all the scripts... err-no-svn :P
17:22:05 <nielsm> so, what shall we do about the network revision string? keep current format, do some hacks to make sure part of the git revision is always included, make an entirely new format?
17:22:41 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fh66D
17:23:03 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain commented on pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fh669
17:23:12 <planetmaker> in essence a new format might not be bad which actually sends three separate entities: branch, hash and newgrf version
17:23:30 <Gabda> now that I think about it, I think PR #7047 is finished. I thought it would be better with this voronoi, but it is good without it.
17:23:47 <Gabda> Can I have the WIP tag from #7047 removed? :)
17:24:30 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain approved pull request #25: Include SDL 2.0 dev headers in debian https://git.io/fh665
17:24:52 <Samu> do i fix the way i fixed for mine? or do i fix and keep the buggy behaviour in the scenario editor
17:25:13 <DorpsGek_II> [OpenTTD/OpenTTD] Gabda87 commented on pull request #7047: Add #6887: Highlight tiles within local authority of towns https://git.io/fh66N
17:25:15 *** Gustavo6046 has joined #openttd
17:25:32 <Gabda> thanks for the tag/untag :)
17:26:06 <TrueBrain> We are your bot; command us! :D
17:27:00 <TrueBrain> we are happy to help :)
17:27:21 <glx> it's simpler to request it that to add a useless comment to ask on github I think
17:28:10 <glx> of course works only if people in charge are present ;)
17:28:42 <Samu> are you guys aware of scripts running in the scenario editor? some settings can be changed even when the widgets are disabled
17:28:53 <Borg> btw guys... is there a plan to allow making groups from vehicles lists from stations?
17:29:09 <TrueBrain> nielsm: if I were you, I would do a short-term fix to fix 90% of the cases. After that, we can work on a more real fix, with splitting what the user sees from what we send over the network (for example, only git hashes over the network, that stuff). Possibly even revamping the whole UDP protocol? :D
17:29:29 <Gabda> in this case can someone help me test PR #7064?
17:29:52 <planetmaker> TB likely has a point ;)
17:30:03 <Gabda> I have only one computer, and I am not familiar enough with the multiplayer mode to test it on one PC
17:30:25 <TrueBrain> Gabda: start OpenTTD twice, and connect via LAN / localhost :)
17:30:50 <DorpsGek_II> [OpenTTD/CompileFarm] nikolas opened pull request #26: Change: Remove subversion and mercurial from base docker image https://git.io/fh6ik
17:31:07 <Samu> oh i "fixed" #7064 it, on the basis that only the server has access to the screen that settings is and that only the server can start the companies
17:31:13 <Gabda> the same build, or I will need two copies in different folders?
17:31:30 <Samu> aka, "just change it directly"
17:31:52 <Gabda> so you cannot edit from two clients?
17:31:59 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain commented on pull request #26: Change: Remove subversion and mercurial from base docker image https://git.io/fh6iI
17:32:03 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain approved pull request #26: Change: Remove subversion and mercurial from base docker image https://git.io/fh6iL
17:32:39 <Samu> but if someone logs in with server credentials or whatever the heck
17:32:43 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh opened pull request #7121: Fix #7021: Better revision strings for network and gamelog https://git.io/fh6im
17:32:49 <Samu> dont know if it has access to that
17:32:58 <TrueBrain> Gabda: as far as I know, you can just launch OpenTTD twice from the same folder :) Start one as server, the other connect as client
17:33:40 <nielsm> yes you can absolutely play network games with yourself
17:34:51 <glx> you can even start a dedicated local server if you don't want to have 2 game windows
17:35:18 <glx> Samu: no issue when connecting to localhost
17:35:27 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain commented on pull request #25: Include SDL 2.0 dev headers in debian https://git.io/fh6is
17:36:11 <Gabda> I need two windows, as I need the GUI on both
17:36:12 <glx> the more usual issue is to connect to a local advertised server
17:36:20 <TrueBrain> I am happy with people who don't only fix that what is in front of them, but also a bit around it :) That is nice :)
17:36:45 <glx> when the router doesn't want to connect to the outside IP
17:37:50 <Eddi|zuHause> there used to be an issue with temporary files?
17:37:58 <Eddi|zuHause> maybe that was running 2 servers
17:38:04 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain commented on pull request #7121: Fix #7021: Better revision strings for network and gamelog https://git.io/fh6ic
17:39:18 <glx> well the config file conflicts when used by 2 servers at the same time, mainly because they need different ports
17:39:49 <Gabda> can I join to a server as client and have admin rights, so I can see and set he AI/Game Script Configuration window from the client as well?
17:40:20 <glx> only server can manage scripts via GUI
17:40:37 <glx> clients can do stuff via rcon, but it's less friendly
17:41:11 <Eddi|zuHause> glx: i was thinking things like the temp savegame when joining a server
17:41:15 <Gabda> can you link me a description on that?
17:41:56 <Gabda> I want to modify the maximum number of competitors from anywhere but the GUI
17:42:32 <glx> you can change settings via rcon
17:42:57 <Gabda> even for not dedicated servers?
17:43:45 <TrueBrain> especially for dedicated servers :D
17:43:57 <glx> not dedicated server just has GUI
17:45:36 <glx> note that rcon requires to se a password in cfg
17:46:29 <glx> unless you have physical access to the running server to do it in its console
17:47:23 <Gabda> it seems like I cannot change the number of maximum competitors from rconú
17:47:42 <Gabda> or I missed it when I read through the commands
17:48:08 <glx> I'm quite sure it's possible to change it
17:48:35 <glx> don't forget the quotes ;)
17:49:45 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain merged pull request #26: Change: Remove subversion and mercurial from base docker image https://git.io/fh6ik
17:49:56 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain commented on pull request #26: Change: Remove subversion and mercurial from base docker image https://git.io/fh6Pe
17:53:14 <Gabda> no, I didn't find it, it was something else
17:53:54 <Samu> hmm how am I gonna report this. scenario editor lets me change settings of in-game parameters even when the widget is shown as disabled
17:54:22 <Samu> i dont know what's wronger
17:54:40 <Samu> being able to edit any running setting in the editor
17:55:11 <Samu> yeah, i need to report this somehow
18:00:36 <glx> Gabda: difficulty setting.max_no_competitors
18:00:55 <nielsm> dammit TrueBrain just won't let me do my stupid C tricks in peace! :D
18:00:56 <glx> oups setting difficulty.max_no_competitors :)
18:01:22 <TrueBrain> nielsm: tricks are fine; unbounded tricks are not :P
18:01:42 <Gabda> I missed the difficulty. part
18:02:17 <glx> yeah network settings don't need the prefix
18:04:07 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh updated pull request #7121: Fix #7021: Better revision strings for network and gamelog https://git.io/fh6im
18:06:02 <TrueBrain> much better nielsm :)
18:06:10 <Gabda> while I searched for it, I found a bug in my modification
18:06:24 <Gabda> I doesn't work when paused...
18:07:09 <Gabda> multiplayer is a mysterious thing
18:07:23 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
18:08:36 <Eddi|zuHause> i feel like i once tried to patch it so rcon doesn't need quotes, but passes the parameters
18:09:28 <Eddi|zuHause> "13. Aug 2010 rcon3.diff"
18:09:46 <LordAro> well it's not like there's already too many PRs
18:09:55 <glx> pff stupid github not considering a force push for the notifications
18:10:13 <LordAro> glx: ? it normally does
18:10:22 <Samu> need some guidance in how I'm fixing #7115. knowing of the problems of #7122. What should be the result of the fix for #7115?
18:10:36 <glx> for me I get notifications only when a comment is added
18:10:51 <Samu> glx, can you take a look
18:11:00 <Samu> you were on it yesterday night
18:11:16 <LordAro> glx: oh, actual github notifications, rather than github irc notifications
18:12:01 <Eddi|zuHause> i've never used website notification thingies
18:13:39 <Samu> do i fix 7115 and mimic the buggy behaviour displayed on 7122?
18:15:04 <Samu> in the scenario editor, some settings are shown as disabled, but they really aren't... what should 7115 do in this situation?
18:15:55 <glx> I think it's a different issue, if the GUI show them as disabled, you are not supposed to be able to change them
18:16:28 <Samu> it's closely related though
18:16:54 <glx> not tried to follow your indications, but I guess the GUI shows disabled, you click on value and it opens a querybox
18:17:10 <Samu> yes, and it does change them
18:17:33 <glx> it should not open a querybox
18:17:47 <Samu> it's not just queryboxes
18:17:58 <Samu> it's bools, left/right arrows, dropdowns, all
18:18:27 <Eddi|zuHause> ci failure report still horrible
18:18:43 <glx> yes but if the state is disabled the OnClick should enforce it, if it doesn't it should be fixed there
18:18:45 <Eddi|zuHause> also, page not loading
18:19:50 <Samu> do i fix 7115 knowing in advance that 7122 will also be fixed?, or do i first make a fix that mimics the buggy behaviour and then fix all of 7122 afterwards?
18:24:35 <Samu> should be _game_mode != GM_MENU
18:26:16 <glx> see, with some quick analysing it does'nt need a lot of work to be fixed
18:26:54 <Samu> hmm, it's a deeper problem
18:27:11 <Samu> instances could be running in editor or not be running in editor
18:27:29 <glx> the first step is to have the gui correctly follow the disabled state
18:28:27 <glx> then you can handle the next issue, which may need to change how disabled state is determined
18:28:58 <glx> but the first step is still the inconsistency between display and click
18:30:21 <glx> a private function to determine the editable state seems a good idea here
18:30:29 <Samu> i have fixed it "my way" on the gui overhaul PR, it touches many of these issues, believe me, it's a deeper problem that it looks
18:31:16 <Samu> however, "my way" is just that... what I think it's best
18:31:21 <glx> anyway I think I'll fix this one, I see how to do it easily
18:32:50 <Samu> instance could be running, or not running, and if running, it could be dead, or alive
18:33:08 <Samu> the slot could also have the config, or not have the config
18:33:09 <Gabda> i could use rcon without quotes
18:33:26 <glx> but that's something unrelated, now the gui is inconsistent
18:36:16 <Samu> imo it's related, in the whole
18:36:34 <DorpsGek_II> [OpenTTD/OpenTTD] Gabda87 commented on pull request #7064: Fix #6438: raising/lowering 'Maximum no. competitors' setting in multiplayer https://git.io/fh6XK
18:37:12 <Samu> how should the widget behave when the instance is there, but the script is dead?
18:38:19 <Samu> the current checks, IsValidAiID isn't a complete check imo
18:43:14 <Samu> i'd like to show how I fixed it. I invented a function called "IsConsideredDead", which is called in many of the widgets
18:43:41 <Samu> it's a centralized function to achieve consistency
18:44:02 <Samu> it's, as you say, "very intrusive"
18:44:15 <Samu> it touches almost all aspects of ai_gui.cpp
18:45:29 <Samu> I know, I know... it's my code, nobody likes to look at, i touch too many things for no reason, etc... :( but i feel it was needed to
18:47:08 <Gabda> Samu: I haven't read the conversation, but does it have any relation with issue #6438?
18:48:05 <Samu> no, this one is about the List of Settings of AIs
18:55:45 <DorpsGek_II> [OpenTTD/OpenTTD] Gabda87 updated pull request #7064: Fix #6438: raising/lowering 'Maximum no. competitors' setting in multiplayer https://git.io/fhCGC
18:57:44 <DorpsGek_II> [OpenTTD/OpenTTD] Gabda87 commented on pull request #7064: Fix #6438: raising/lowering 'Maximum no. competitors' setting in multiplayer https://git.io/fh61E
18:59:46 <Samu> IsConsideredDead - 17 found occurences
19:00:17 <Samu> the majority to decide widgets state and onclick state
19:15:22 <Samu> I wonder how glx will solve
19:18:43 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 opened pull request #7123: Fix #7122: OnClick() and DrawWidget() editability was inconsitent in … https://git.io/fh6MC
19:25:39 *** KenjiE20 has joined #openttd
19:27:20 <andythenorth> and I can close an issue also
19:29:02 <Eddi|zuHause> andythenorth: last time i tried, i drew a blank on rewording the commit messages
19:29:28 <Eddi|zuHause> andythenorth: can you try?
19:31:09 <Eddi|zuHause> also why does it say conflicting? my rebase just went through with no issues
19:32:50 <Samu> rebase upstream/master ?
19:32:55 <andythenorth> LordAro: ^ I am missing what's wrong with the commit message? ^^ :)
19:32:58 <Samu> or rebase origin/master?
19:33:19 <andythenorth> I was reading the PR title, not the 6 commits
19:33:24 <LordAro> Eddi|zuHause: capitalise the first word, and "sorlists"
19:33:59 <LordAro> also yes, fix the conflicts
19:34:12 <LordAro> as Samu says, likely you're rebasing against your master, not the remote one :)
19:34:39 <Eddi|zuHause> i did git fetch origin
19:34:43 <glx> Samu: origin is your own fork
19:35:11 <Eddi|zuHause> my origin is OpenTTD github, not my github
19:35:40 <Samu> i get into less trouble with upstream than origin
19:36:07 <LordAro> Eddi|zuHause: did you push after you last rebased? :>
19:36:18 <Eddi|zuHause> i didn't push yet
19:36:51 <Eddi|zuHause> and master on my github is ancient
19:37:10 <nielsm> "git pull --rebase upstream master"
19:37:36 <nielsm> to rebase your branch on the upstream master
19:40:13 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z updated pull request #7049: Fix #6599: Can still click on buy button in vehicle selection window even if no vehicle is selected https://git.io/fhnEu
19:41:43 <Eddi|zuHause> i haven't addressed the switch yet
19:43:19 <Samu> thx glx, just tested, works fine in menu, in-game before and after ai starting
19:43:29 <Samu> in se... there's still a problem :|
19:43:42 <Samu> before starting, it is disabled
19:43:54 <Eddi|zuHause> ah, now i get a conflict
19:43:54 <Samu> which at least reflects the button state, but hmm...
19:44:09 <LordAro> Samu: Put. It. On. Github.
19:44:23 <Samu> oh, my bad, i'm too used to chat
19:45:33 <Eddi|zuHause> i'm not sure i understand enough about that code to resolve the conflict
19:45:35 <glx> anyway Samu it fixes #7122
19:47:28 <glx> making sure the tests will always match, even if the test is modified in the future
19:48:07 <Samu> yes, it reflects the button state now
19:48:31 <Eddi|zuHause> the conflict is just an added function, or what?
19:48:56 <Samu> the problem however, is that the issue now is that it should be enabled for whenever there's no instance running in that slot
19:49:42 <Samu> imagine you start the scenario editor, and set some AI slots
19:49:49 <Samu> the configs behave like you're just in-game
19:50:43 *** snail_UES_ has joined #openttd
19:50:56 <Samu> ok i'm gonna post there, as LordAro suggested
19:50:57 <Eddi|zuHause> it looks different in my conflict view, but it doesn't seem to compile
19:51:02 <baneos> Hello guys , new to ttd and trying to figure out the patching with MAC can someone help me with the basics ? it seems complicated thanks.
19:51:23 <andythenorth> where are you starting from? o_O
19:51:33 <andythenorth> no staying power :P
19:53:43 *** Gustavo6046 has joined #openttd
19:54:32 <andythenorth> opening the 'all trains' window destroys FFWD perfomance
19:55:00 <Samu> nevermind, glx, false alarm
19:55:06 <andythenorth> goes from about 400+ fps to around 100fps
19:55:12 <Samu> was looking at this wrong
19:55:56 <andythenorth> window-shading it knocks the performance back up
19:56:05 <andythenorth> anyone know how to profile that? :P
19:56:22 <Samu> actually, i'm confused... sorry glx, need to investigate this better, there's something not quite right yet
19:57:25 <Eddi|zuHause> ah, i think i understand it
19:57:31 <Eddi|zuHause> it's a function removed in the patch
19:57:39 <Eddi|zuHause> but the translation change was in that function
19:57:45 <Eddi|zuHause> i need to replicate that elsewhere
19:58:25 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z updated pull request #7049: Fix #6599: Can still click on buy button in vehicle selection window even if no vehicle is selected https://git.io/fhnEu
19:58:43 <Eddi|zuHause> i can only hope that is right
19:59:40 *** Gustavo6046 has joined #openttd
19:59:49 <Eddi|zuHause> at least it compiles now
20:00:26 <Eddi|zuHause> now, the switch layout
20:01:11 <Samu> aha, I get it now. When there's a human company currently active in a config slot, the non-in-game parameters are disabled.
20:01:49 <Samu> i have them enabled on my version of the gui, sorry about the confusion
20:03:32 <Samu> to be quite honest, I prefer "my way"
20:08:17 <peter1138> I see eints has undone my translation removal... Oh well.
20:09:03 * andythenorth draws lights on trains, for the doors, in stations
20:09:43 <Samu> there really are more issues than just what i'm mentioning in the SE, move up and move down buttons, shifting configs of one ai into a currently active ai instance
20:10:00 <Eddi|zuHause> LordAro: did you only want the "break" on a new line, or also the "return"?
20:10:20 <Samu> too many many issues in se :|
20:11:03 <Samu> but for what it's worth it, glx thx for the fix, plz put it into master so that I know how to proceed from then on
20:11:42 <LordAro> Eddi|zuHause: i think it shouldn't be a oneliner at all, break it up into if (foo) {\nreturn false;\n}\nbreak;
20:12:05 <LordAro> as per usual case/if statements
20:12:08 <Eddi|zuHause> so also the return
20:12:25 <LordAro> see how it looks ¯\_(ツ)_/¯
20:12:53 <Eddi|zuHause> well, "if (blah) return false" is a very common pattern in the code
20:13:12 <Eddi|zuHause> so i thought that should be enough
20:15:42 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z updated pull request #7049: Fix #6599: Can still click on buy button in vehicle selection window even if no vehicle is selected https://git.io/fhnEu
20:15:55 <Eddi|zuHause> i kept the return on one line for now
20:18:06 <Eddi|zuHause> i feel like i lost a change somewhere
20:19:37 <nielsm> oh then it is fixed :)
20:20:35 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7049: Fix #6599: Can still click on buy button in vehicle selection window even if no vehicle is selected https://git.io/fh6SR
20:25:47 <Samu> do i really have to create an issue per problem? i kinda dealt with all of them in 1 go.
20:26:01 <Samu> but having to describe them 1 by 1 is daunting
20:26:10 <LordAro> Samu: depends on what the issue is
20:26:29 <LordAro> if they're all the same root cause, then only one issue would suffice
20:26:51 <LordAro> even something like "ai gui does weird things with settings" would do as a single issue, as long as it lists all the issues within it
20:27:09 <LordAro> Samu: what was your verdict on #7123 ?
20:27:53 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z updated pull request #7049: Fix #6599: Can still click on buy button in vehicle selection window even if no vehicle is selected https://git.io/fhnEu
20:28:09 <LordAro> Samu: that's glx's PR
20:28:27 <LordAro> given i'm looking at it, and you were talking about it earlier, i disagree :p
20:29:09 <Samu> my veredict is... approved, but there's more underlying issues
20:29:55 <LordAro> can you elaborate on those issues?
20:31:01 <Samu> its due to slots being currently occupied by a human company, the settings show disabled, and are disabled with that fix
20:31:32 <peter1138> What are you trying to do, separate companies and AI configuration slots?
20:31:49 <Samu> imo, they should be in this case, not showns as disabled, and be editable
20:31:56 <Samu> but... it's really an opinion
20:32:25 <Samu> it's how i handled it on my gui patch
20:32:30 <peter1138> Funny, people asking for subsidiaries, when that was a patch about 12 or so years ago...
20:33:17 <LordAro> Samu: not sure i quite understand, can you add screenshots to your original 7122 issue?
20:34:56 <Samu> for the 7122 issue, the slot was really occupied by an AI company
20:35:40 <Samu> there are various situations
20:36:12 <Samu> config of the slot is occupied by ai company the company instance is running and alive
20:37:29 <Samu> config of the slot is occupied by hu company, no ai instance in this case, which imo, and only my opinion, the settings should be editable
20:38:32 <Samu> config of the slot is occupied by no company at all, no ai instance, all settings are editable (current master behaviour is correct)
20:40:33 <Samu> config of the slot is occupied by ai company, the company instance is running, but dead
20:42:19 <Samu> don't remember what i did on my version for this last situation, let me check quickly, brb
20:44:15 <Samu> just checked, settings are editable
20:45:00 <Samu> let me check what master do, brb
20:48:51 <Samu> then, you have the main menu game mode, scenario editor game mode and normal game mode to add to the confusion
20:50:51 <Samu> situation #4 in master, settings are disabled
20:58:52 <Samu> what I'm arguing about is that I have already taken care of all this with my gui overhaul patch...
20:59:02 <Samu> it's just that it's too big
20:59:31 <Samu> having to splice it appart is so daunting, expose the issues one by one... only to be fixed without considering the whole
21:02:59 <peter1138> I think the issue is you don't explain clearly what you are trying to achieve.
21:03:39 <peter1138> It has all this list of changes. Okay, but why?
21:04:38 <Samu> goal is to have a much better gui
21:04:48 <peter1138> And one comment from one person isn't law.
21:06:06 <peter1138> Oooh that beef steak fillet was good.
21:06:54 <Samu> well, in sum, it solves many of the issues on SE
21:07:01 <andythenorth> but must wash children
21:07:14 <LordAro> Samu: we don't want a summary, we want details
21:07:20 <LordAro> what have you done, why have you done it
21:07:50 <LordAro> andythenorth: just get someone else to resolve the conflict, like you normally do :p
21:08:01 <andythenorth> well my wife is away
21:08:13 <andythenorth> usually this is a 2-person operation
21:08:28 <andythenorth> single-manning has gone ok so far, but I failed to plan food correclty
21:08:39 <andythenorth> permanent single parenting must be quite military
21:09:15 <LordAro> sounds like they should be taught to wash themselves :p
21:09:27 <Samu> random ai is another peculiar situation that could have improvements
21:09:52 <Samu> I thought that it would nice to know which of the slots started as random
21:09:57 <andythenorth> peter1138: obviously all OpenTTTD ships have azimuth thrusters or whatever
21:10:04 <andythenorth> which is why they can push sideways
21:10:48 <Samu> it's... so boring to explain
21:12:18 <Samu> i guess i have better answers when i'm asked why did you do this thing
21:12:35 <andythenorth> the issue is the occasional displacement of the sprite sideways for a frame or two?
21:12:44 <Samu> but explaining everything is just too complicated
21:13:15 <LordAro> Samu: if you can't adequately explain what you've done, it's just going to end up getting rejected
21:13:22 <glx> LordAro: #7122 is simple, some settings are drawn as disabled but are not for OnClick()
21:13:26 <Samu> major goal was indeed to make it less crashable, less buggy in se, more easier to edit configs
21:13:45 <Samu> more easier to notice which ais run as random
21:13:52 <peter1138> That's still a summary.
21:14:22 <peter1138> I did "this thing" because "thing other thing" happened.
21:14:29 <LordAro> glx: yeah, yours is fine, i'm just trying to work out what other issues Samu found
21:15:34 <peter1138> Why? Do you not know yourself?
21:16:50 <Samu> the very first thing I did was
21:16:58 <Samu> change max no competitors cap from 14 to 15
21:17:26 <Samu> so that I can actually start 15 ais in multiplayer servers
21:17:59 <peter1138> Why was that so hard to get out of you?
21:18:01 *** andythenorth is now known as Guest1682
21:18:02 *** andythenorth has joined #openttd
21:18:07 <Samu> that title however remained for the rest of my work
21:18:23 <andythenorth> ok so rice, with carrots and chicken, all cooked in one pot with butter and lemon
21:18:26 <peter1138> Yes, that was a mistake.
21:18:34 <peter1138> andythenorth, ooh that sounds nice.
21:18:35 <Samu> next thing i wanted was to allow all slots to be editable
21:18:50 <Samu> so i made them all 15 visible
21:18:57 <peter1138> I'm outta kale, aubergine and courgette, so that might be on the agenda tomorrow.
21:19:15 <peter1138> Are all slots not editable?
21:19:16 <LordAro> andythenorth: your "failed to plan food correctly" is better than most things i plan
21:19:22 <Samu> instead of what the current behaviour do, which is lock out editability when there's a hu or a ai on it
21:19:30 <peter1138> LordAro, that sounds like a damn good plan, not failed at all.
21:20:12 <Samu> it's very annoying to have a running game and trying to edit slots only to have it start and be locked out, window closing
21:20:20 <Samu> so, i wanted to improve on that
21:20:23 <peter1138> Okay so it's locked out when an AI is running. Why is that so?
21:20:39 <peter1138> Do changes made there affect the running AI/
21:21:31 <peter1138> nielsm, that's another example of a PR that doesn't specify what it does :D
21:21:44 <peter1138> I've heard of voronoi, but I've no idea what it means in that context.
21:21:55 <Samu> about the buttons on the config window and ai settings window, i had to make major modifications there, to accomodate the ability to have everything editable
21:22:08 <peter1138> I didn't ask about that.
21:22:51 <peter1138> Everything you do so far sounds like "I made some changes because I made some changes" :p
21:22:59 <Samu> changes made there affect the running ai? yes
21:23:09 <andythenorth> the thing I have learnt
21:23:15 <andythenorth> is cooking anything in butter is good
21:23:20 <Samu> but i made sure the changes that can still be made, aren't going to cause a crash
21:23:24 <andythenorth> and butter is redeemed now, so it's fine
21:23:25 <peter1138> Samu, do they? Does the AI expect that?
21:23:42 <Samu> yes, ais have some settings that can't be editable while running in a game
21:23:51 <Samu> i took that into the design
21:24:05 <Samu> some can be, some can not
21:24:26 <peter1138> So the AI system understands that, but the AI configuration window doesn't, at the moment?
21:25:03 <Samu> by AI system, you mean the instance
21:25:33 <Samu> when i say ai, i am also including gs's of course
21:25:36 <peter1138> I mean AI as a whole.
21:26:06 <LordAro> frosch123: so today's eints commit. any idea why it did that?
21:26:31 <peter1138> So I spotted an existing bug. If you have the AI parameters window open and then the AI starts, the settings are still editable.
21:26:40 <peter1138> I didn't test actually changing them.
21:27:00 <Samu> some are changeable while it's running
21:27:12 <glx> but I don't know if it's a good idea
21:27:30 <peter1138> glx, that's the point I'm trying to get at, yes.
21:27:40 <glx> most windows need to be looked at I think
21:27:43 <frosch123> LordAro: because it is implemented like that?
21:27:50 <Samu> the window should refresh the noneditable parameters into disabled
21:27:53 <frosch123> eints does not delete translations
21:27:59 <frosch123> it marks them as invalid or outdated
21:28:05 <Samu> unless the ai author actually allows them to be editable
21:28:23 * andythenorth tries the town voronoi diagram
21:28:34 <peter1138> As it stands, the window should close, because AI settings are not changing while an AI is running.
21:28:36 <andythenorth> picture really is compelling
21:28:42 <frosch123> feel free to patch it to not include invalid translations in the download :)
21:29:02 <Samu> the window isn't closed, it goes through an invalidatedata check
21:29:06 <peter1138> And then you need a compelling argument that being able to change AI settings in a running AI is a good idea.
21:29:09 <glx> well closed, or the window can be aware of the state change
21:29:32 <peter1138> glx, currently you can't open the window if the AI is running
21:29:35 <Samu> because it can already be done via ai debug window
21:29:38 <peter1138> glx, so that'll be inconsistent.
21:29:47 <Samu> you can via debug window
21:30:00 <andythenorth> just pause the game with the window open, and enforce modality :P
21:30:01 <peter1138> glx, an obvious first step is to allow the settings window to be opened on running AIs, but disallow editing.
21:30:13 <andythenorth> or we need pub-sub state updates for the GUI? :P
21:30:17 <peter1138> Samu, debug window is not really expect to be used by players.
21:30:19 <glx> yes of course, step by step
21:30:43 <Samu> why would you say that? :|
21:30:53 <andythenorth> because it's true?
21:30:59 <peter1138> andythenorth, hmm, it highlights stuff. But I don't know what!
21:31:04 <frosch123> i think the string knows about its status
21:31:05 <Samu> regardless, it's possible there, so i made it possible on the config window too
21:31:28 <peter1138> Debug window is for AI developers to, er, debug.
21:31:32 <andythenorth> wow, the demolish tool is insane with voronoi :)
21:31:46 <peter1138> If you don't have developer tools enabled, I suppose there is no debug tools?
21:31:46 <Samu> forget that the name is called "debug"
21:31:50 <andythenorth> some towns have huge radii
21:32:05 <Samu> there is, it's still there, enabled or not
21:32:15 <andythenorth> are these really the town radii?
21:32:19 <andythenorth> I assumed they were smaller :P
21:33:01 <Samu> besides, it's for changing the ai settings
21:33:39 <Samu> like editing newgrf parameters
21:33:45 <Samu> made it also editable for ais
21:34:03 <andythenorth> I don't really understand the goal
21:34:12 <peter1138> andythenorth, of voronoi? I don't know what it's doing.
21:34:21 <andythenorth> it's showing town you're in
21:34:27 <andythenorth> it's clever, but doesn't work yet
21:34:38 <nielsm> which town will "claim" the station I guess
21:34:39 <peter1138> Is this an alternative to the other patches he's done?
21:34:41 <andythenorth> you still don't know which town it is, and the landscape is too obscurred
21:34:58 <andythenorth> voronoi goal is obvious to me :P
21:35:04 <andythenorth> I was referring to Samu though
21:35:06 <peter1138> I know that I saw a really cool landscape generator that used voronoi, but I didn't know how to C++ify it.
21:35:24 <andythenorth> Samu: are you still trying to prevent asserts?
21:35:29 <Samu> my goal was really to make a more compelling ai config window
21:35:39 <Samu> something more easier to manage
21:35:41 <andythenorth> ok that's a fair statement
21:35:45 <Samu> make it less restrictive
21:35:57 <andythenorth> what's currently restrictive?
21:36:08 <andythenorth> FWIW, I just choose some AIs, and close it
21:36:19 <peter1138> I don't even choose. Random AI.
21:36:28 <glx> andythenorth: in game, not in intro
21:36:39 <andythenorth> yes, in game it does nothing
21:36:42 <andythenorth> so problem solved?
21:36:43 <Samu> for someone who likes to test ais like me, the current way is too cumbersome, limited
21:36:45 <andythenorth> nothing to see here?
21:36:55 <Samu> i wanted to viabilize it in-game
21:37:07 <Samu> i added the start/stop button
21:37:22 <Samu> which is one of the "great" features
21:37:23 <andythenorth> oh don't piss around with UI for all that
21:37:27 <andythenorth> just add console commands
21:37:34 <peter1138> Does the window already understand (non)editable settings?
21:37:45 <peter1138> It's difficult to see if it does, because you can't open it on a running AI.
21:38:17 <andythenorth> can AIs even handle config being changed once initialised?
21:38:17 <peter1138> Hmm, I ought to fit these brake pads. Should I eat a doughnut?
21:38:27 <andythenorth> peter1138: discs?
21:38:32 <LordAro> C:\msys64\home\LordAro\OpenTTD\src\lang\spanish.txt:1816: warning: STR_LIVERY_CAPTION: Param idx #0 '<empty>' doesn't match with template command 'COMPANY'
21:38:35 <LordAro> well that's not irritating at all
21:38:53 <LordAro> frosch123: i fear i don't know anything about eints
21:38:57 <Samu> if the config is changed to some of the other ai, no. That will cause cause problems. this problem currently exists in scenario editor on master version
21:38:58 <andythenorth> my advice, when pissing around with disc brakes
21:39:08 <LordAro> where's an Alberth when you want him? :)
21:39:13 <andythenorth> never spin the wheel to see alignment, then grab the disc with bare hands
21:39:19 <andythenorth> I nearly took a finger off
21:39:31 <andythenorth> clean cut, but very deep
21:39:38 <glx> yes settings have flags to tell when they are allowed to be changed
21:39:48 <Alberth> just plain python lord aro :)
21:39:54 <peter1138> With the state of my bike this winter, it wouldn't be a "clean" cut :p
21:40:14 <peter1138> glx, okay, so in that case it's up to the AI author to get that right.
21:40:44 <frosch123> LordAro: well, create an issue
21:40:45 <andythenorth> shall we close some PRs :)
21:40:45 <peter1138> If the AI can control it, then the AI should.
21:40:54 <andythenorth> it's more fun than yak-shaving
21:40:54 <Samu> yes, but there are a few limitations when switching from a config from one ai into another, it must be made sure that this is not happening
21:41:13 <peter1138> Samu, who's talking about switching config from one ai into another?
21:41:23 <andythenorth> why would you do this?
21:41:23 <peter1138> Weren't we talking about changing the settings of an existing AI/
21:41:27 <frosch123> most work when fixing eints is updating the test environment, but after that fixing multiple things is easy
21:41:37 <peter1138> Is that not what happens?
21:42:16 <Samu> yes, but during design, i had to ensure i would not let configs be changed to configs of others
21:42:21 <LordAro> frosch123: i would, but seems issues are disabled on the eints repo :p
21:42:26 <LordAro> i guess they're still over on ottdc?
21:42:52 <frosch123> i did not fancy having two trackers for nml and eints
21:42:52 <Samu> random ai for example, posed a somewhat difficult task
21:43:07 <Samu> but i think i handled it fine
21:43:07 <frosch123> though argueably noone looks at them
21:43:34 <Samu> random ai is a config that is usually just start_date, but once it starts, it switches the config to the actual AI
21:43:47 <Samu> i had to be extra careful with this
21:43:52 <LordAro> frosch123: just get TB to migrate them all :p
21:45:31 <Samu> also, some parts of the current code, does not add the start_date parameter, it was causing me a headache on enabling/disabling the configure widget
21:47:12 <glx> if the window is refreshed properly, and initialised with the right data, it can handle all changes without side effects
21:47:28 <Samu> sec, let me show this part of the code
21:47:54 <Gabda> nielsm: thanks for the comments, and for the gif
21:48:08 <Gabda> i really like the gif :)
21:48:11 <glx> but usually it needs a global thinking, else you end up with a lot of changes everywhere
21:49:17 <Samu> this here is what adds the start_date parameter to random ais once an ai that died and started as random, is cleared up
21:49:48 <Samu> without that, I was not able to change the Configure widget state
21:50:17 <glx> one is adding close everywhere, the other let the window handle it by itself
21:54:12 <Samu> sorry, dinner is waiting, be back later
22:00:52 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7123: Fix #7122: OnClick() and DrawWidget() editability was inconsitent in … https://git.io/fh6Fb
22:01:11 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro merged pull request #7123: Fix #7122: OnClick() and DrawWidget() editability was inconsitent in … https://git.io/fh6MC
22:10:05 <peter1138> Well, fail my life.
22:10:39 <peter1138> The bolt for the rear pads is seized and it's made of cheese metal, it seems.
22:10:50 <LordAro> peter1138: ooh, mine did that!
22:10:58 <andythenorth> just get a new bike?
22:11:20 <peter1138> I gave it some WD40.
22:11:24 <peter1138> Gotta find a heat source.
22:11:37 <peter1138> Front one was fine.
22:12:03 <peter1138> I guess whoever last changed them (bike shop, now dissolved) didn't bother greasing the bolt.
22:12:32 <glx> never trust "professionals" ;)
22:12:37 <andythenorth> I got my pads changed by a shop after slicing my finger the first time
22:12:42 <andythenorth> they didn't put a split pin in properly
22:12:44 <peter1138> Or used a broken screw remover?
22:12:48 <peter1138> I have a dremel, but...
22:12:49 <andythenorth> pads fell out in traffic
22:12:49 <LordAro> the hole is only slightly bigger than it used to be
22:13:04 <peter1138> Then of course I need to find a replacement bolt.
22:13:05 <LordAro> i bought dremel bits for it specially
22:13:16 <LordAro> (it was either that or get a new caliper)
22:13:21 <LordAro> peter1138: did the pads not come with one?
22:13:26 <peter1138> On the other bike, it's a allen key type bolt.
22:13:36 <peter1138> Just pads and spring.
22:13:55 <peter1138> Going to be annoying as the thread is... on the wrong side.
22:14:01 <glx> LordAro: brute force often works :)
22:14:12 <peter1138> Yeah, might need a new caliper.
22:14:16 <peter1138> Which comes with pads ;/
22:14:27 <peter1138> But I need the bike tomorrow to commute!
22:15:43 <LordAro> glx: so my idea of making ScriptTextfileWindow use a config instead of a slot doesn't work, as the config just points to an array of slots
22:16:50 <glx> can't you change the slot on invalidate ?
22:17:31 <LordAro> possibly, i didn't investigate other solutions
22:17:47 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #7106: Fix #7089: close NewGRF textfile windows when their data are invalid https://git.io/fh6bM
22:19:02 <glx> it's like changing the slot value on invalidate
22:19:21 <LordAro> i'm not sure i like the whole window being updated when the slot changes
22:19:38 <glx> but in newgrf case it's not a slot it's the full grfconfig
22:19:55 <LordAro> well, more or less the same
22:20:32 *** andythenorth has left #openttd
22:20:54 <glx> that way the window is in sync with active selection, even if there's in fact no data to display
22:21:22 <LordAro> i'm not sure i like that as a solution though
22:21:42 <glx> but it's still possible to just close the window without changing anything
22:21:50 <LordAro> in my mind, you should be able to open many textfile windows, even if they were opened from the same slot
22:22:13 <glx> store the previous config, compare with the new one, and close by itself if they don't match
22:22:26 <glx> less special cases to handle elsewhere
22:23:52 <glx> many textfiles is not very compatible with window class system
22:24:17 <peter1138> £42.74 from Amazon for a replacement caliper.
22:24:31 <glx> oh but there's window number
22:24:42 *** erratic has joined #openttd
22:24:46 <peter1138> Window number indeed.
22:25:27 *** HeyCitizen_ has left #openttd
22:25:56 <glx> but I think opening 2 windows for the same exact info is incorrect
22:26:26 <glx> having one for license and one for changelog, ok
22:26:37 <glx> but 2 for license of the same slot no
22:26:58 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro commented on pull request #7106: Fix #7089: close NewGRF textfile windows when their data are invalid https://git.io/fh6NO
22:39:56 <peter1138> Ah, git reset --hard origin/
22:40:01 <peter1138> For those force-updates.
22:40:28 <peter1138> origin/branch, I mean.
22:51:31 <Samu> yeah reading the history, my main concern was indeed making the interaction to actually work without issues, i believe i did a good work on there
22:51:48 *** sla_ro|master2 has quit IRC
22:52:36 <Samu> v3 was when i decided it would be best to have all config slots editable
22:52:58 <Samu> and it was really the best decision
22:54:09 <Samu> interaction between max no competitors and editable slots is gone
22:55:10 <peter1138> The thing is, do you want to be able to distinguish between running config and startup config?
22:56:12 <Samu> erm... yes, but keep it minor
22:57:03 <Samu> there are some few things that work different in main menu than in game
22:57:29 <peter1138> No, I mean running AI config and startup AI config.
22:58:08 <peter1138> Running config is what the AI is using now.
22:58:15 <Samu> ah, each parameter can be defined with IN_GAME flag, which lets them to be editable while running in a game
22:58:17 <peter1138> Startup config is what an AI would use if it was to start up in that slot.
22:58:26 <peter1138> That's still not what I'm talking about :p
22:59:52 <Samu> not sure i understand the question
23:01:23 <peter1138> running config is the configuration that is used now, while it is running
23:01:34 <peter1138> startup config is the configuration is would use if it was to be restarted
23:01:52 <Samu> it restarts with the running config
23:02:01 <Samu> in the case of random ai, it doesn't
23:02:14 <Samu> it randomizes another ai to start
23:05:16 <peter1138> But what if you wanted it to restart with some changes to settings that aren't editable in the running config?
23:06:04 <Samu> notice some of the settings became disabled
23:06:09 <Samu> the others can be cahnged
23:06:23 <Samu> i can stop the ai, it won't go back to its initial config
23:06:57 <Samu> let's say I change min size of towns to be considered to... 1000
23:08:43 <Samu> the min size of towns is .... 1000
23:08:53 <Samu> the initial config is gone
23:09:08 <Samu> if that's what you were referring to
23:12:25 *** Thedarkb1-T60 has joined #openttd
23:12:42 <Samu> had you configured this in the main menu, and then you edit in-game, you may still go back to the config previously set in main menu, but have to exit the current game
23:14:53 *** FLHerne has joined #openttd
23:15:46 <Samu> for random ai, the behaviour is different
23:15:57 <Samu> do you want screenshots?
23:17:00 <Samu> oh snap i triggered a crash apparently :|
23:17:45 <Samu> assert(strcmp("start_date", name) == 0);
23:17:54 <Samu> yeah, again the issue of the missing start_date parameter
23:19:42 <Samu> i want to avoid closing the config window when i stop a random ai
23:28:46 <peter1138> Ok, maybe I should take the bike up to the bike shop to replace the caliper.
23:29:05 <peter1138> I don't have the proper tools and fluids to do a bleed.
23:30:39 <Samu> AI::Stop in ai_core.cpp I can't call InvalidateWindowData(WC_AI_SETTINGS, company) just yet, i need to push the "start_date" first
23:39:43 <Samu> awesome, that did it! no more crash
23:47:31 <Samu> this means i can delete that for cycle
23:49:20 <Samu> start_date parameter is pushed right as it is stopped
23:50:58 <Samu> this->SetWidgetDisabledState(WID_AIC_CONFIGURE, this->selected_slot == INVALID_COMPANY || GetConfig(this->selected_slot)->GetConfigList()->size() == 0 || (_game_mode == GM_EDITOR && !IsConsideredDead(this->selected_slot))
23:51:17 * peter1138 grumbles at his code duplication.
23:51:21 <Samu> GetConfigList()->size() is no longer 0
23:51:36 <Samu> the start_date parameter now exists
23:52:04 <Samu> tell me where it is, i can try improve it
23:53:23 <Samu> GS'es can never be stopped, nothing to do on game_core.cpp
23:57:48 <Samu> they can crash, but they're not considered Stopped
23:59:15 <Samu> as far as I know, i can't remove a GS from a running game
23:59:33 <Samu> looked everywhere if it was possible, didn't find a way
23:59:45 <Samu> so i assume they're never stoppable
continue to next day ⏵
