IRC logs for #openttd on OFTC at 2019-01-25
⏴ go to previous day
00:01:41 <Samu> TextFileWindow is also used by NewGrf
00:02:39 <LordAro> Samu: specifically, ScriptTextfileWindow stores a "CompanyID slot". Instead, it should store a specific ScriptConfig (i.e. the return value of GetConfig(slot))
00:04:17 *** andythenorth has left #openttd
00:07:29 <Samu> ok, gonna try pass GetConfig(slot) from wherever it's called
00:08:32 <LordAro> Samu: i don't think that's specifically necessary, just do it in the constructor of the ScriptTextfileWindow
00:08:42 <LordAro> it might be better your way though, i'm not sure
00:10:22 <Samu> the idea I had for an alternative solution was "just close it"
00:10:37 <LordAro> not very productive :p
00:11:43 <LordAro> looks like you'll need at least a bool as well as the ScriptConfig, to determine whether it's a GS or AI
00:11:51 <LordAro> Samu: sure, force push
00:13:12 <Samu> i mean, what should the window display?
00:14:27 <LordAro> ooh, i thought you were talking about the PR :p
00:14:48 <LordAro> just closing it is also a perfectly good option
00:15:47 <Samu> there is essentially nothing to display
00:16:14 <Samu> just a large purple square
00:16:19 <LordAro> when the window refers to a "slot", yes
00:16:33 <LordAro> but if the window is referring to a particular AI instead, it makes more sense to keep it open
00:37:22 <Samu> looks like it needs OnInvalidateWindowData, it does not have it
00:41:34 <Samu> virtual void OnInvalidateData? I wonder ... how is it called
00:44:44 <LordAro> i suspect you're going in the wrong direction, but look at other windows to see what they do, there are plenty of examples
01:20:46 *** snail_UES_ has joined #openttd
01:38:19 *** Thedarkb-X40 has joined #openttd
01:41:23 <Samu> no, I'm lost, i'm not going anywhere with this
01:44:11 <Samu> I keep getting into the same problem: there's no way to invalidate the text that's shown on this window
01:45:35 <Samu> TextFileWindow is used in too many places
01:46:18 <Samu> if I edit it, i may break the rest
01:46:37 <Samu> it's even used for network server lists
02:20:58 <Samu> i'm finally getting somewhere
02:58:25 <Samu> looks like I don't have to edit textfile_gui.cpp after all
03:09:41 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick updated pull request #7094: Fix #7088: Retrieve an appropriate name for a non-existant AI/GS when… https://git.io/fhrqY
03:22:29 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7094: Fix #7088: Retrieve an appropriate name for a non-existant AI/GS when… https://git.io/fhrpn
03:37:00 <Samu> it's already hard as is to build water connections in-land :|
03:49:06 <Samu> it doesn't make sense closing that one
03:49:18 <Samu> I don't understand what you want
03:50:58 <Samu> if I want to build a dock in front of an aqueduct, now I have to flatten the land, place a canal, build the dock, demolish canal, make it inclined again and build aqueduct
03:55:22 <Samu> the half-tile water with one corner raised, I don't get it why is it wrong...
03:56:03 <Samu> i can already raise a flat sea into a one corner half-tile in front of the dock
03:56:42 <Samu> you guys don't play ships...
03:58:05 <Samu> the buoy is another one... i can place buoys in front of docks, but docks can't be placed in front of buoys?
03:59:28 <Samu> do you actually build canals in land?
04:01:01 <Samu> all these changes are to help canal construction, if you would actually play ships and build canal routes, you'd get stuck into these types of non-sense situations quickly
04:03:35 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #6926: Change: Allow dock to be constructed in more locations https://git.io/fhrhy
04:16:47 *** Wormnest has joined #openttd
04:31:27 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
04:54:24 <Samu> im getting a crash when starting a game in multiplayer with master
04:55:21 <Samu> doesn't crash in single player
04:56:09 <Samu> wait a minute, it does...
04:56:58 <Samu> Expression: cannot dereference end map/set iterator
05:00:14 <Samu> did you make changes to terraform cmd or tileheight?
05:01:26 <Samu> it's crashing when setting tile heights of some tiles
05:06:37 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
05:17:05 <Samu> doesn't it crash for you?
05:17:16 <Samu> start a new game, and bam, crash
05:17:38 <Samu> release build doesn't crash
05:19:25 <Samu> Expression: cannot dereference end map/set iterator
08:12:01 *** HerzogDeXtEr has joined #openttd
08:18:09 *** sla_ro|master has joined #openttd
08:32:24 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on issue #7098: Crash: cannot dereference end map/set iterator https://git.io/fhoLk
08:40:57 *** andythenorth has joined #openttd
08:49:10 <andythenorth> well I have 9 failing tests, but that's not news
08:51:43 <andythenorth> our pretendolino sketch petered out
08:51:55 <andythenorth> the forums must be waiting keenly for next installment
08:57:03 <andythenorth> what colour is it?
09:00:24 <Pikka> depends on what you mean by "it" and "done".
09:00:35 <andythenorth> you were drawing bigger trains?
09:00:38 <andythenorth> to suit bigger eyes
09:01:23 <Pikka> I did the greyhound today, it's done-ish
09:03:54 <Pikka> and I had a brainwave that - of course - recolour sprites still work with 32bpp, so I don't have to make separate black steam loco sprites. So that's done too.
09:06:04 * andythenorth draws a sort of class 91
09:06:44 <Pikka> can you draw one for me too?
09:08:46 <andythenorth> if you are happy for it to be 1x zoom, yes
09:09:33 <andythenorth> I think your A3 is in there somewhere
09:11:04 <peter1138> How should I solve the GUI for (group) liveries?
09:11:11 <peter1138> Cos it's shitty at the moment.
09:11:38 <peter1138> It's in the company colours window, which is logical for setting colours, but maybe it should be in the group window, because groups...
09:12:23 <peter1138> Hmm, maybe a button in the group window can open the company colours window and select the right thing.
09:12:40 <peter1138> And I can reuse those lovely icons that somebody made.
09:13:20 <andythenorth> should be in the group window yes
09:13:27 <andythenorth> there was some reason you didn't
09:13:32 <andythenorth> like scope or some crap
09:14:09 <peter1138> Nah, window resizing messing up.
09:14:16 <peter1138> Oops, 8:14, I ought to head to work.
09:14:23 * andythenorth already is working
09:25:55 *** sla_ro|master has joined #openttd
09:57:25 <peter1138> Ok, I made it to work.
10:00:40 <peter1138> It's not, It's foggy and a heady 4.5°C
10:02:30 <andythenorth> I set the sparks effect for every coach in a high speed train
10:03:36 <andythenorth> looks quite cool
10:29:01 *** tokai|noir has joined #openttd
10:29:01 *** ChanServ sets mode: +v tokai|noir
10:49:12 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7097: Fix: enable DPI-awareness for MINGW builds https://git.io/fhosr
11:12:08 *** andythenorth has joined #openttd
11:59:00 <Eddi|zuHause> it's of course böing
12:00:55 *** JacobD88 has joined #openttd
12:13:37 *** JacobD88 has joined #openttd
12:53:16 <TrueBrain> I blame andythenorth
12:55:31 *** andythenorth is now known as Trubrain_
12:55:39 <Trubrain_> I blame andythenorth too
12:56:02 <TrueBrain> sad you cant even write TrueBrain
12:56:49 <TrueBrain> I like the dedication :)
13:10:51 *** JacobD88 has joined #openttd
13:15:27 *** WWacko1976-work has quit IRC
13:16:39 *** WWacko1976-work has joined #openttd
13:29:41 <TrueBrain> okay, fixed the feedback for the website .. guess that should be good enough to launch now :D
13:30:22 <TrueBrain> you do need to review it first :P
13:30:57 *** Trubrain_ is now known as andythenorth
13:43:54 <peter1138> div id based on now/date ?
13:45:05 <TrueBrain> hmmm .. NoAI docs return a near-empty index.html
13:45:21 <peter1138> Hmm, I see it the CSS.
13:46:27 <TrueBrain> its a very old gimmick, a few people spotted over the years
13:46:37 <TrueBrain> it makes the website a bit more alive, in some sense
13:47:02 *** Mahjong2 has joined #openttd
13:47:36 <peter1138> I mean, I don't know what the scripting system is there but it looked sane...
13:47:44 <peter1138> {{ }} likes like jinja.
13:47:55 <TrueBrain> Liquid, but different name, same shit
13:48:27 <peter1138> If there was <?php in there I'd be horrified :D
13:48:57 <TrueBrain> lol .. Doxygen fails if there is a space in the folder name
13:49:49 <TrueBrain> okay, locally doxygen for NoAI does work
13:54:38 <TrueBrain> ah .. it needs gawk
13:54:42 <TrueBrain> and it silently continues
13:56:01 *** sla_ro|master has joined #openttd
14:01:04 <peter1138> What built the stats before?
14:01:32 <TrueBrain> the current stats are already fake, now I come to think of it
14:01:36 <TrueBrain> it fetches the data from the WT3
14:01:41 <TrueBrain> which .. hasnt been updated in months, ofc
14:01:52 <TrueBrain> not sure if eints can be polled for this data
14:02:07 <TrueBrain> the frontpage and WT3 are in the same application, so it could just fetch the data like it was nothing :)
14:03:13 <peter1138> Could be built from the lang files some how.
14:03:23 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain opened pull request #23: Add: also create 'noai' and 'nogs' documentation tarballs https://git.io/fhouM
14:05:13 <TrueBrain> 100 MiB for source docs, unextracted ..
14:05:17 <TrueBrain> guess I won't be publishing that
14:05:24 <LordAro> site needs a redesign :p
14:05:41 <LordAro> TrueBrain: that seems high, what on earth is it?
14:06:03 <TrueBrain> I just run 'doxygen', and it gives me that :P
14:06:29 <TrueBrain> images, lots of them
14:06:49 <LordAro> ah yeah, lots of graphs
14:07:32 <TrueBrain> 2115865 Jan 25 14:00 newgrf_8cpp_source.html
14:07:52 <TrueBrain> the source files are eating up all the disk space
14:09:22 <peter1138> I just ran doxygen and got... 190MB :/
14:09:46 <TrueBrain> 0 .. 9 .. not always easy to spot
14:12:35 <TrueBrain> why is it nogo.openttd.org, and not nogs.openttd.org?
14:12:51 <TrueBrain> guess nogo is a bit of an old name ..
14:13:33 <TrueBrain> but it became GameScript
14:14:39 <TrueBrain> is the source tarball
14:14:49 <TrueBrain> -docs-noai, -docs-ai, -docs-aidocs, ..
14:14:50 <peter1138> Could change it to nogs and redirect nogo to that?
14:15:01 <TrueBrain> and: -docs-nogo, -docs-game, -docs-gamedocs, ..
14:15:24 <peter1138> -docs-ai -docs-game
14:16:04 <peter1138> -docs-ai / -docs-gs ?
14:16:15 <peter1138> -docs-gamescript works.
14:16:21 <peter1138> -docs-game could be ambiguous.
14:16:36 <TrueBrain> I like -docs-ai / -docs-gs
14:16:51 <TrueBrain> description: Documentation for AI scripting (xz/lzma archive)
14:16:51 <TrueBrain> description: Documentation for Game scripting (xz/lzma archive)
14:17:19 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain updated pull request #23: Add: also create 'noai' and 'nogs' documentation tarballs https://git.io/fhouM
14:18:18 <DorpsGek_II> [OpenTTD/website] TrueBrain opened pull request #38: Add: there are now three more tarballs for documentation https://git.io/fhozi
14:18:27 <TrueBrain> that should fix that :)
14:21:23 <DorpsGek_II> [OpenTTD/CompileFarm] PeterN approved pull request #23: Add: also create 'noai' and 'nogs' documentation tarballs https://git.io/fhogf
14:22:10 <peter1138> docs.zip is still relevant?
14:22:21 <TrueBrain> yeah; it contains the full history of any extension we ever had :)
14:22:29 <TrueBrain> helps when rendering old folders
14:22:34 <DorpsGek_II> [OpenTTD/website] PeterN approved pull request #38: Add: there are now three more tarballs for documentation https://git.io/fhogL
14:22:46 <DorpsGek_II> [OpenTTD/website] TrueBrain merged pull request #38: Add: there are now three more tarballs for documentation https://git.io/fhozi
14:22:59 <TrueBrain> another one bites the dust \o/
14:24:04 <peter1138> Linux deb release images appeared on the CI, I see.
14:24:22 <TrueBrain> they have for months
14:24:34 <TrueBrain> like I said yesterday, Debian and Ubuntu deb files are created on release
14:24:44 <peter1138> Oh right, it's the CF CI, not OpenTTD CI.
14:24:45 <TrueBrain> but they ONLY work on releases, not for nightlies
14:25:21 <TrueBrain> we could also add stuff like rpms, debs, .. for nightlies
14:25:23 <TrueBrain> that would work too
14:25:27 <TrueBrain> but .. os/debian is a maze to me :D
14:25:53 <peter1138> I know how to bodge them :)
14:26:47 <TrueBrain> it just doesn't work (at all) in its current form
14:29:14 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain merged pull request #23: Add: also create 'noai' and 'nogs' documentation tarballs https://git.io/fhouM
14:29:37 <TrueBrain> right, that should produce those binaries tonight
14:29:43 <TrueBrain> we will see if they do :P
14:36:12 *** snail_UES_ has joined #openttd
14:47:36 *** andythenorth has joined #openttd
15:05:22 <Samu> can you reconsider opening #6926
15:09:02 <peter1138> I think it'd be better to revamp docks.
15:14:00 <Samu> btw, i can't test my other PRs atm with that dereference crash always getting in the way
15:18:13 <peter1138> Just change 316 from != to ==
15:21:47 <peter1138> (Just don't commit that)
15:35:20 <supermop_work_> happy invasion day Pikka
15:35:46 <supermop_work_> melbourners seem to only celebrate australia day when they live in Brooklyn
15:36:23 <supermop_work_> also apparently we have a New York Magpies
15:36:43 <supermop_work_> the only thing that doesn't surprise me about that is that it isn't the Bushwick Magpies
15:37:14 <supermop_work_> i will be eating Tim Tams and drinking Coopers this evening
15:41:49 *** andythenorth has joined #openttd
15:42:30 <andythenorth> industrial trains are BAD FEATURE?
15:42:36 <supermop_work_> i guess hi to Antheus too, no harm being friendly
15:42:55 <supermop_work_> andythenorth: idk. its something i sometimes want
15:43:08 <Samu> andythenorth plz reopen 6926, plzzoh
15:43:45 <supermop_work_> sometimes i just want to build a little metro with various short turn termini and branches
15:43:55 <andythenorth> Samu: no, you'll have to convince someone else
15:44:00 <supermop_work_> and then industries are just noise
15:44:06 <andythenorth> 6926 makes no sense to me
15:44:27 <Samu> how does it not make sense?
15:44:30 <supermop_work_> in those cases i guess industrial trains might make sense as decorative noise
15:44:53 <andythenorth> how many patches have you made to try and prevent ship routes being blocked?
15:45:47 <supermop_work_> i still haven't built a network of 3-track mainlines that has been able to convince me there is any sensible way to use them
15:45:50 <andythenorth> - the '3rd tile' blocks ship routes
15:45:58 <andythenorth> - the aqueduct head blocks ship routes
15:46:17 <supermop_work_> can we just remove ships for real
15:46:53 <andythenorth> replace with NRT
15:47:09 <andythenorth> supermop_work_: you should start a screenshot thread in forums
15:47:10 <Samu> that patch aim is not about blocking routes
15:47:23 <Samu> it's to facilitate canal construction inland
15:47:36 <planetmaker> Samu, I think the argument is: introducing that patch even blocks more ship routes
15:47:41 <planetmaker> makes blocking them easier
15:47:43 <andythenorth> in what way is that goal related to the content of 6296?
15:47:52 <andythenorth> the patch doesn't affect inland canals
15:48:38 <peter1138> And we gave our reasons.
15:48:46 <andythenorth> my point is you don't seem to understand your own goals
15:49:00 <andythenorth> you provide many well-intended attempts to prevent ship routes failing
15:49:12 <andythenorth> then you ask for a PR to be accepted that trivially blocks ship routes
15:49:24 <andythenorth> 6296 doesn't solve any useful problem
15:49:28 <Samu> how is that even related
15:49:53 <andythenorth> it is related, but I will have to re-explain
15:49:59 <Samu> you don't seem to understand my aims
15:50:14 <andythenorth> I am prepared to listen again
15:50:23 <andythenorth> but my theory is that *you* don't understand your aims
15:50:28 <planetmaker> what's your aim? Not that of a single patch or PR. But in general?
15:50:35 <andythenorth> ^ what planetmaker said
15:51:05 <Samu> 6926 aim was to ease water constructions constraints a bit
15:51:49 <andythenorth> 7078 isn't rejecte
15:52:04 <andythenorth> I agree that issue is worth solving somehow
15:52:19 <planetmaker> ^^ (and I agree with the comment concerning foundations)
15:53:21 <Samu> the one about lock pricing is not about because of their costs being high, but because of a flaw in how the costs are added up
15:53:32 <Samu> you misunderstood that one
15:53:35 <peter1138> Hmm, how should I signal to an open window to change its current selection?
15:54:16 *** dihedral has joined #openttd
15:54:38 <andythenorth> Samu: for 6233, I checked with frosch, who had reviewed it. I didn't review it myself
15:54:59 <andythenorth> you have about 30% of the open PRs, and they are consuming a lot of the available reviewer time
15:54:59 <Samu> #6933 is also aimed to ease water construction constraints
15:55:06 <andythenorth> reviewer time is a scarce resource
15:55:43 <Samu> you really need to build canals inland to understand the reasonings behind these changes
15:55:57 <andythenorth> how do you know I don't? o_O
15:55:59 <Samu> if you don't, it's easy to dismiss what I do
15:56:15 <Samu> you get into those issues easily
15:56:57 <andythenorth> Samu: I've made 2 of the 5 most popular ship grfs :)
15:57:00 <andythenorth> I know about canals
15:57:01 <Samu> it's very cumbersome at the moment
15:57:41 <andythenorth> I'm not rejecting your PRs because canals are good
15:57:48 <andythenorth> but these PRs are not well formed
15:57:57 <andythenorth> and you are occupying a lot of reviewer time with them
15:58:03 <andythenorth> and very few make progress
15:58:19 <andythenorth> and it's ultimately not fair on other contributors
15:58:32 <andythenorth> it might seem unfair to you that I close them
15:58:40 <andythenorth> but there is a greater unfairness to other people if I don't
15:58:47 <andythenorth> I don't like doing it, but eh
15:59:04 <peter1138> Make well-formed smaller PRs. In the PR, describe what you are trying to achieve.
15:59:26 <peter1138> Don't stack things together just because they're related, or make sense to you.
15:59:39 <planetmaker> I don't think it can get smaller than e.g. #6233
16:00:03 <peter1138> JGR didn't make one huge PR for improving cargo flow drawing.
16:00:43 <peter1138> 6233 is an issue, not a PR.
16:01:00 <peter1138> And I meant in general.
16:01:55 <andythenorth> given that I trust frosch, why waste more people's time trying to do something with it?
16:02:44 <andythenorth> it's been there since October 2018, so it's hard to make the case that anybody else really cares about that one
16:03:00 <peter1138> "if (!IsWaterTile(tile)) c->infrastructure.water++;"
16:03:06 <peter1138> Yeah... you can't build a lock on water.
16:03:36 <peter1138> (Or can you on a river? Hmm)
16:03:50 <planetmaker> it only changes the infrastructure count. Not the ability to build
16:04:34 <peter1138> But nobody owns a river.
16:05:30 <Samu> if it's a river, don't count
16:05:50 <Samu> river is the only water tile
16:05:52 <peter1138> The lock is aleady counted.
16:06:15 <planetmaker> One can argue for locks both ways, both equally valid IMHO: building a lock in a river to make it shippable, is a HUGE endeavour. As such you definitely add infrastructure in that case
16:06:29 <planetmaker> But you don't add new water....
16:06:42 <peter1138> planetmaker, well, you should get the cost of the river tile being cleared
16:06:56 <peter1138> but ongoing, the cost of the lock is the same, so infrastructure cost doesn't change
16:07:16 <Samu> i mimic what happens to the north and south parts
16:07:37 <Samu> went with that line of thought
16:07:42 <Samu> and applied to the middle
16:08:54 <peter1138> I think there is an issue in that the delta tiles could be river, or could be canals (owned by whom?)
16:10:15 <peter1138> If they are river, they stay river, therefore you don't have infrastructure cost for them
16:10:49 <peter1138> You always pay for the infrastructure cost of the centre lock tile. That cost is more than a regular water tile, though.
16:10:50 <nielsm> imo the outer tiles of a lock should be converted from river to canal when building a lock on a river
16:11:17 <nielsm> that's what the graphics hints
16:11:25 <peter1138> What about sea water?
16:11:27 <planetmaker> the important one is the center... but yes, graphics hint that
16:11:35 <planetmaker> the lock extends into sea @ peter1138
16:11:37 <peter1138> Should that become a canal tile too?
16:11:41 <planetmaker> so... yes, there, too. Why not
16:11:51 <peter1138> Should it stay a canal tile when you remove it?
16:12:21 <planetmaker> though canal at sea level is a convenient hack to fend-off land at sealevel
16:12:34 <nielsm> sea tiles convert back I guess
16:12:37 <peter1138> Oddly when removing a lock placed on a river, it gets converted back to a river.
16:12:57 <planetmaker> there's IMHO two choices: a) convert upper and lower tile to canal - and destroy them upon lock removal
16:12:58 <nielsm> that sounds bad, converting back to river
16:13:11 <planetmaker> or b) keep them the water whatever they are. And don#t destroy them upon removal
16:13:21 <peter1138> b is what we do at the moment
16:13:28 <peter1138> a might make more sense
16:13:32 <planetmaker> the water class of the tiles is not changed currently. So they stay what they are
16:13:52 <Samu> except when it's not water to being with
16:13:55 <peter1138> the lower lock tile on sea can then just be converted to dirt, ready to flood again
16:14:57 <peter1138> Hmm, what happens when you build a lock on a competitor canal? Is that possible?
16:15:13 <planetmaker> maybe a) is more consistent indeed. Not sure I really like it though... it destroys rivers needlessly
16:15:29 <Samu> upper and lower parts belong to the canal owner
16:15:38 <Samu> middle tile belongs to lock owner
16:15:39 <planetmaker> but then it just shows another problem: that rivers don't restore their "natural" path
16:15:58 <planetmaker> which is somewhat out-of-scope for this discussion
16:16:16 <nielsm> if you demolish a lock, regardless of which tile of it you click, it should convert to a click on the center tile of it
16:16:29 <nielsm> and canal ends on top and bottom remain
16:16:48 <peter1138> 14:57 < andythenorth> and you are occupying a lot of reviewer time with them
16:17:00 <Pikka> obviously removing rivers is the only option :D
16:17:01 <nielsm> should rivers even be destroyable at all?
16:17:46 <planetmaker> nielsm, that's another discussion... but you would want to keep the option to build locks into them
16:17:48 <Samu> i had a patch some years ago that would restore rivers upon demolishing canal
16:17:52 <planetmaker> (or not?... maybe)
16:17:56 <Samu> was rejected by yours truly
16:18:34 <peter1138> Yes, that's silly.,
16:18:43 <andythenorth> and does it make any difference to gameplay?
16:19:01 <andythenorth> or is it amusing or fun?
16:19:30 <planetmaker> now you're being mean @ andy
16:19:36 <Samu> it was making rivers more "important" to play around at least
16:19:53 <andythenorth> planetmaker: not intending to be mean :)
16:19:56 <Samu> not just a thing that you remove just so you can build more rail tracks
16:19:57 <nielsm> introduce a "river spring" tile, which is indestructible, it propagates "flow bit" through all adjacent river and canal tiles, a river tile with flow bit only connected to one other water tile will random walk into adjacent tiles expanding the river, and if other water tiles are nearby (e.g. 3 tiles distance) gravitate towards those
16:20:15 <nielsm> except that river flow can't propagate upwards in elevation
16:20:40 <planetmaker> not a new idea. And not necessarily a bad one
16:20:47 <planetmaker> Unfortunately not implemented
16:20:54 <nielsm> possibly even make river flow not a bit but a value so multiple rivers flowing together creates stronger flow
16:20:58 <peter1138> nielsm, I welcome a PR ;)
16:21:00 <planetmaker> does it destroy houses / industries / infrastructure?
16:21:11 <peter1138> nielsm, yes, needs to be a value. We have lots of spaces now.
16:21:27 <peter1138> planetmaker, griefing too, yes.
16:22:01 <planetmaker> so... just add an option "Allow destroying rivers" and it's settled
16:22:14 <planetmaker> if it's set to 'false', such griefing cannot happen
16:22:19 <nielsm> or maybe make it possible to destroy river springs, but it will cost a fortune and all towns in a 128 tile radius will instantly hate you :D
16:22:32 <DorpsGek_II> [OpenTTD/OpenTTD] nikolas updated pull request #7086: Change #6173: Update SDL driver to use SDL 2.0 https://git.io/fhamZ
16:23:20 <planetmaker> Maybe those springs should rather be like antenna masts: un-removable
16:23:31 <nielsm> yeah that's the other option
16:24:33 <peter1138> You'll be getting on to rivers drying up
16:24:38 <peter1138> Sediment shifting...
16:24:52 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7097: Fix: enable DPI-awareness for MINGW builds https://git.io/fhoP1
16:25:01 <andythenorth> just walk them back up from the coast, add some random extra tiles when they turn
16:25:13 <peter1138> When the terrain generator can reproduce cheddar gorge I'll be impressed.
16:25:21 <andythenorth> cheddar's not all that
16:25:24 <andythenorth> I mean it's nice
16:25:33 <nielsm> I also want to have unfillable ocean tiles
16:25:40 <andythenorth> cheddar would fit in 64x64
16:25:46 <andythenorth> if you left out the lake
16:25:49 <peter1138> There used to be deep-ocean patches.
16:25:55 <andythenorth> grand canyon is more like 4096x4096
16:26:10 <peter1138> andythenorth, are you applying a scale to this game?
16:26:15 <nielsm> yeah, deep ocean and that's required for oil fields
16:26:31 <peter1138> Maybe that can be resurrected. It was kinda cool.
16:26:34 <nielsm> kill the trains to oil fields!
16:26:45 <nielsm> (I do it all the time myself)
16:27:00 <peter1138> Urgh, yes... "valid gameplay" :/
16:27:01 <andythenorth> so how do I reverse the ship pathfinder?
16:27:04 <planetmaker> we've FISH for that :)
16:27:09 <andythenorth> so it always starts on coast tile, and goes uphill?
16:27:22 <peter1138> [12:05pm] peter1138: someone archive this convo please, for 5 years time when i decide to revisit it
16:27:42 <andythenorth> I've archived that too :P
16:28:15 <peter1138> My comment about the docks ...
16:28:26 <nielsm> if I suddenly stop participating it's because of birds taking control of my hands
16:28:30 <peter1138> That would fix up the need for 6296
16:31:59 <peter1138> Maybe I should look at that after... er... everything else.
16:32:39 <peter1138> I should try to find my multistop docks patch, that was necessary for it.
16:35:43 <andythenorth> allegedly it's already in the spec
16:36:04 <andythenorth> but only for TTDP
16:36:36 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #6925: Fix #6574 #6636 #5405 #6493: Aircraft hangar issues https://git.io/fhoXQ
16:37:12 <nielsm> have there been experiments with stations having capacity for waiting cargo?
16:37:47 <peter1138> Rather than 'unlimited'?
16:38:18 <nielsm> and making many decorative station elements be meaningful
16:39:15 <andythenorth> I knew I didn't hallucinate it
16:39:20 <andythenorth> frosch told me years ago
16:39:26 <andythenorth> ID 04 in action 3
16:39:41 <andythenorth> dunno if that was ever implemented anywhere, or if it's just wiki wishful thinking
16:40:27 <peter1138> nielsm, probably not but it's a good idea.
16:40:42 <peter1138> It would limit those single bus stops with thousands of waiting passengers...
16:43:55 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on pull request #7078: Feature: Raise a corner of land automatically when building aqueducts https://git.io/fho1a
16:44:20 <supermop_work_> flat docks are nice for those insane polish servers with flat rivers everywhere but insane track costs and unfathomable terraforming costs
16:44:34 <Samu> about blocking routes because of AI behaviour, I accept that AI authors would need to take more care about where they build their stuff. That I understand.
16:44:53 <supermop_work_> where the only way to make money is a horde of hovercraft around Gdansk
16:49:24 *** sla_ro|master has joined #openttd
16:49:31 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on issue #6322: AI can allocate more memory than the system has, crashing the game https://git.io/fhoMO
16:50:01 <Samu> talking about #6927 being closed. I'm okay with this.
16:50:46 <Samu> #6926 is the one that hurts me most :|
16:50:53 *** ZirconiumX has joined #openttd
16:54:57 <peter1138> Arguably, I'd say completely remove the 3rd tile check, and leave it up to the player.
16:55:23 <peter1138> Maybe keep it for AI.
16:55:58 <planetmaker> keeping it for AI would ... be somewhat unfair towards AI. But might make sense nonetheless
16:56:20 <peter1138> planetmaker, just thinking about expecting current behaviour.
16:56:32 <peter1138> Could be toggleable for them.
16:56:39 <planetmaker> The AI wouldn't mind, I guess
16:56:49 <Samu> AI's would have to adapt :o
16:56:58 <planetmaker> Just players who swear when an AI involuntarily blocks passage by these means
16:57:27 <peter1138> That can happen by raising/lowering land anyway.
16:57:46 <planetmaker> yes. So by that argument that's a non-issue
16:57:50 <peter1138> But yeah, single dock tile _next_ to water would be my solution to that :)
16:58:22 <planetmaker> we should have docks which don't extend into the water but are built simply on the coast tiles
16:58:25 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh commented on issue #7098: Crash: cannot dereference end map/set iterator https://git.io/fhoDI
16:58:44 <peter1138> planetmaker, right. That's where my comment in that ticket andythenorth linked to earlier was going.
16:59:03 <planetmaker> ah, I didn't see that
16:59:09 <andythenorth> the building on the aqueduct exit tile blocks the exit
16:59:47 <andythenorth> I don't mind if we allow blocking
17:00:08 <andythenorth> BUT it makes no sense in context of these other boil-the-ocean PRs about preventing blocking
17:00:13 <andythenorth> like, decide what we want, or something
17:00:59 <andythenorth> tell them I said you could
17:01:44 * planetmaker just arrived home :)
17:04:34 <peter1138> nielsm, hmm, I thought we already did some XDG stuff. Is it DATA_DIRS specifically?
17:04:57 <Samu> you can't block a dock ith aqueduct
17:05:10 <Samu> the slope doesn't permit it
17:05:16 <Samu> and they have no foundations
17:05:47 <Samu> it would force the 2nd dock tile to have a non flat slope
17:05:52 <nielsm> peter1138 I didn't actually check
17:05:57 *** Wormnest has joined #openttd
17:06:35 <planetmaker> Samu, what about this suggestion: we do not care whether a waterway path is being blocked by any building (players and AI can take care of that - not OpenTTD's task. It's the same as with road building basically)
17:07:08 <planetmaker> but we can agree on removing most building restrictions for docks and aqueducts.
17:07:35 <planetmaker> So not special casing the build places, but only blocking building when really absolutely needed
17:07:57 <planetmaker> as to really make things a bit simpler
17:19:44 <Samu> that's ok, but most of the AIs will relly on the old behaviour which checks the 3rd tile
17:20:03 <Samu> ai authors would have to update them
17:20:42 <planetmaker> arguably that's correct, yes
17:22:11 <Samu> i have come to accept that ais are just like human players after all
17:23:09 <planetmaker> well... yes. Adopt or amend an AI library which does these checks for them. And AIs just need to use that library
17:24:08 *** HerzogDeXtEr has joined #openttd
17:24:46 <Samu> you mean, keep the old behaviour just for the AIs, but create a new build behaviour for humans?
17:25:32 <planetmaker> AI can use libraries which can implement tasks many AI authors need.
17:25:41 <planetmaker> OpenTTD would not need to make a difference itself
17:28:54 <Samu> i see, hmm so there should be a notice about behaviour change somewhere in the noAI api
17:29:35 <planetmaker> though none of the calls themselves changes
17:30:43 <Samu> the return value of builddock, erm... let me find the correct name
17:31:01 <planetmaker> it still returns allowed or not allowed
17:31:08 <planetmaker> just more often allowed
17:32:44 <Samu> well, then there's nothing to do on the ai api side
17:32:48 <peter1138> Yay, cheap bike lights with dodgy USB sockets :/
17:33:21 <Samu> it's gonna be up to ai authors to adapt their code to the new behaviour
17:33:39 <Samu> as much as it saddens me, because most are inactive
17:33:56 <planetmaker> yes... that saddens me, too
17:35:00 <peter1138> I've seen savegames where AIs block their own stuff anyway...
17:36:16 <glx> most AIs start like it's a new game on loading
17:40:45 <Samu> will the pier tile require water?
17:41:04 <Samu> or will it build a canal if it's missing water?
17:42:23 <andythenorth> I would rather ease the building restrictions
17:42:52 <andythenorth> people will complain about MP griefing, but eh
17:44:00 <Samu> I got another project related to canal ownership on hold, didn't PR it because the changes are kinda massive
17:45:12 <Samu> in it, i made all water based constructions to automaticaly build a canal underneath the structures if it is missing water
17:45:44 <Samu> building in land is more fluid, but ultimately, ais in their current state, will suffer
17:46:03 <andythenorth> so a dock will place water under it?
17:46:24 <Samu> yes, but also set the owner of the water tile
17:46:37 <Samu> it's not just a single owner now
17:46:55 <Samu> dock owner, and canal owner
17:47:13 <Samu> and drive through stations
17:47:52 <andythenorth> I am +1 to easing building
17:49:36 <Samu> there's just an issue, a big one, about what to do these water canals if one removes the structure
17:50:01 <glx> andythenorth: canal around the land ?
17:50:06 <planetmaker> you build water, if required. You destroy water when it is removed
17:50:33 <Samu> but in the majority of cases, I don't want the water to be cleared
17:52:38 <glx> hmm looking at this screenshot I agree with nielsm, more space needed under aqueduct for road or tracks
17:53:08 <Samu> it's an issue about consistency, demolishing a ship depot built on canal tiles in current openttd version still keeps the canals
17:53:45 <Samu> i like this behaviour more than a bare land being created, but at the same time
17:54:01 <Samu> if the ship depot is the first structure I'm creating on land and I wanna remove it
17:54:11 <Samu> it's now gonna leave 2 water tiles behind
17:54:29 *** Progman has joined #openttd
17:54:45 <andythenorth> glx: canal around the land
17:55:09 <andythenorth> it's my older child
17:55:12 <andythenorth> he wasn't sure of money
17:55:28 * andythenorth builds airports on sea with canal tiles
17:55:47 <nielsm> glx yeah it's just an issue of "does the ship float on top of a 5 cm deep puddle of water?"
17:56:30 <planetmaker> lool @ screenshot
17:56:35 <nielsm> and similarly perhaps it should not be possible to tunnel directly below rivers/canals either
17:56:37 <planetmaker> but yes... I've built such already as well
17:56:54 <glx> for me it's more like can the vehicle under really go without touching
17:57:37 <glx> aqueduct seems thicker than other bridges
17:59:14 <glx> hmm for tunnels it's different, they can go deeper without anybody notice ;)
17:59:51 <nielsm> right, just assume that tunnel and any crossing below it fall some additional height
17:59:58 <glx> anything could happen between entry an exit
18:00:45 <glx> in ETS2 there's even tunnels making loops in scandinavia
18:02:59 <glx> but in the screenshot the aqueduct really feel very low
18:03:48 <Samu> raising 2 levels with locks alone require 3 * 2 tiles
18:05:08 <Samu> is this why you want foundations?
18:06:04 <glx> wow orudg.e is featured in this one
18:07:05 <Samu> i also have a patch about aqueducts requiring a min height of 2, let me dig
18:07:35 <glx> you sound like peter with your "I have a patch" :)
18:10:06 <Samu> makes it much much harder to raise 2 levels of land
18:10:24 <Samu> harder to build, too many constraints
18:11:52 <Samu> oh, i see it only applies to locks
18:12:34 <Samu> upper lock tile requires a bridge height of 2
18:12:43 <Samu> lower lock tile requires a briddge height of 3
18:16:06 <Samu> just the number of locks and space requires to build an aqueduct over a lock behind... it's bad, doesn't facilitate canal construction at all
18:28:57 <Samu> gonna try posting aircraft hangar issues a PR at a time
18:29:07 <Samu> which will trigger andythenorth anyway :|
18:29:33 <andythenorth> post one and see if it gets reviewed :)
18:30:51 <Samu> will decouple the fixes into single PR's
18:31:51 <glx> but yes one "feature" per PR is better
18:36:42 <LordAro> i have no issues with multiple fixes per pr, as long as they're in distinct commits
18:38:40 *** gelignite has joined #openttd
18:40:12 *** Gustavo6046 has joined #openttd
18:42:56 <peter1138> Deponia: The Complete Journey is free on HumbleBundle for a limited time.
18:44:21 <peter1138> That grammar though.
18:53:23 <TrueBrain> nielsm: it seems all Squirrel memory functions end up in a single file. And strictly seen we have a global telling which VM is active
18:53:33 <TrueBrain> would it be bad to abuse that knowledge? As that would make for a relative easy fix
18:53:37 <TrueBrain> just .. not really thread-safe :D
18:54:19 <TrueBrain> given there is a global to instance which instance is active, I doubt it
18:54:21 <TrueBrain> but I havent checked
18:55:58 <peter1138> I suppose if they have access to map or object info, threaded would probably not work well.
18:56:22 *** Wormnest has joined #openttd
18:57:22 <TrueBrain> it is more that Squirrel itself seems thread-safe. So adding code that breaks that, might do more harm
19:02:34 <TrueBrain> if I would be able to find what is new in Squirrel 3, it might be worth looking adding this to upstream, to fix properly
19:02:43 <TrueBrain> as it is not that difficult, just touching a lot of code, from what I can tell
19:05:46 <Samu> god damn dereference ...
19:06:13 <Samu> well, i'll test in release build
19:06:44 <LordAro> Samu: maybe you should fix that bug first :D
19:06:45 <TrueBrain> seems it is not that much different .. just some new things got added ..
19:07:16 <LordAro> TrueBrain: mm, i was doing some digging last night, there's not a lot that's changed
19:07:35 <TrueBrain> 'free' variables is the new thing in 3, it seems
19:07:37 <TrueBrain> what-ever that is :D
19:07:56 <LordAro> i couldn't find any sort of migration guide, other tham a few scattered forum posts about things not working with 3
19:08:00 <TrueBrain> might be worth updating to 3.1; then we can upstream a patch to lock memory to the vm
19:08:19 *** andythenorth has left #openttd
19:08:20 <TrueBrain> examples of what is not working?
19:08:57 <LordAro> i'll see if i can find it again
19:09:16 <TrueBrain> the file structure is near identical
19:09:19 <TrueBrain> seems the stdlib is bigger
19:12:13 <TrueBrain> parent, vargc and vargv are no longer keywords
19:12:47 <LordAro> TrueBrain: nope, dunno what i saw before, can't find it now :)
19:13:47 <peter1138> Samu, does #7099 fix aircraft that are already stuck, or just prevent them sticking in the first place?
19:14:59 <TrueBrain> sadly, the github repo starts a bit after 3.0 was released, it seems
19:15:12 <TrueBrain> the lexer shows some differences, but nothing that looks huge
19:16:35 <TrueBrain> was not expecting that name
19:16:39 <TrueBrain> took me WAY too long to find :D
19:17:38 <Samu> peter1138 just prevents them sticking in
19:17:54 <Samu> if they're stuck, they remain stuck
19:17:55 <TrueBrain> -removed 'vargv' and 'vargc' keywords
19:17:55 <TrueBrain> -now var args are passed as an array called vargv(as a paramter)
19:18:03 <TrueBrain> -removed 'parent' keyword
19:18:03 <TrueBrain> -added class getbase() built in method
19:18:38 <peter1138> Wonder if any scripts use them?
19:19:27 <peter1138> Hmm, v3 released 2011 :D
19:23:16 <TrueBrain> might be something for OpenTTD 2.0 :P
19:23:19 <TrueBrain> together with NRT? :D
19:28:05 *** frosch123 has joined #openttd
19:31:42 <Samu> how am i gonna explain that some of the changes only make sense together... :|
19:31:50 <Samu> i'm falling into this hole again
19:33:00 <Samu> you want to review them separately :|
19:33:42 <Samu> then it's rejected based on its own, the rest of the changes become meaningless
19:34:22 <Samu> I can already see all this rejected...
19:34:45 <peter1138> Explain it by itself.
19:36:51 <Eddi|zuHause> <peter1138> Deponia: The Complete Journey is free on HumbleBundle for a limited time. <- "unfortunately" i already have that..
19:37:14 <Eddi|zuHause> also, can that even still be called "complete"? i think it was only parts 1-3, not 4
19:37:28 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened pull request #7100: Fix #6574: Remove go to hangar orders when rebuilding airport https://git.io/fhoxm
19:38:54 <Samu> I am cancelling manual send to depot orders
19:39:03 <Samu> on its own, it's pointless
19:39:25 <Samu> the aircraft will still go flying around the airport
19:40:10 *** andythenorth has joined #openttd
19:42:38 <Samu> my other part of the code, would be making the airplane do automatic service at another airport than the next in the orders
19:43:26 <Samu> it was a manual order, I know, but it's cancelled
19:43:40 <Samu> allowing it to trigger automatic service
19:43:55 <Samu> it just requires my other part of the code there to actually start making sense
19:45:04 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7100: Fix #6574: Remove go to hangar orders when rebuilding airport https://git.io/fhoxp
19:47:11 <andythenorth> so I owe some reviews? o_O
19:47:37 <peter1138> You don't owe anything.
19:48:00 <andythenorth> you did the review
19:49:29 <peter1138> Oh that was HOURS ago.
19:49:34 <andythenorth> things move fast here
19:49:38 <andythenorth> keep up andythenorth
19:49:59 <peter1138> Who wants to watch me dev?
19:50:13 <andythenorth> you streaming? :P
19:50:28 <andythenorth> I watched nielsm play OpenTTD :P
19:51:34 <Samu> peter1138 by "remove" I meant mark it invalidated, my bad
19:51:58 *** Gustavo6046 has joined #openttd
19:52:03 <peter1138> Samu, so "RemoveOrderFromAllVehicles" doesn't remove orders?
19:52:18 <Samu> the function is named like that
19:52:31 <glx> hmm I think I have an idea for the deadlock, need to try it
19:52:31 <Samu> but it actually creates (Invalid Order) entry
19:53:26 <peter1138> The only thing is, what happens if you change your mind and replace the airport with one that does have a hangar?
19:53:51 <Samu> remains an invalid order
19:54:05 <andythenorth> isn't there some 30 day thing for that already?
19:54:26 <peter1138> andythenorth, for removed stations/depots, yes.
19:54:30 <DorpsGek_II> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7100: Fix #6574: Remove go to hangar orders when rebuilding airport https://git.io/fhop9
19:54:39 <andythenorth> building an airport without a hangar is removing a depot
19:54:46 <andythenorth> unless it's put back within the time limit
19:54:50 <andythenorth> otherwise order is invalid
19:54:55 <andythenorth> I assume I missed a memo :P
19:55:37 <Eddi|zuHause> andythenorth: i had the exact same thought
19:55:38 <andythenorth> Eddi|zuHause: feel like updating 7051? o_O
19:55:58 <peter1138> I wonder if it's treated differently because it's part of the airport.
19:56:00 <Eddi|zuHause> andythenorth: i only just noticed. maybe it should send out a notification?
19:56:06 <andythenorth> if only there was a bot
19:56:14 <Samu> hold on, im coming up with screenshots
19:56:31 <andythenorth> deliveroo, or should I cook?
19:56:54 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7100: Fix #6574: Remove go to hangar orders when rebuilding airport https://git.io/fhohJ
19:57:44 <andythenorth> so where do hangars exist?
19:58:01 <Eddi|zuHause> they're depot tiles
19:58:29 <andythenorth> but they don't trigger 'invalid order'?
19:58:45 <andythenorth> fix the root cause, replicate existing behaviour from elsewhere
19:58:53 <andythenorth> don't add new behaviour if not needed
19:59:45 *** Thedarkb-T60 has joined #openttd
19:59:46 * andythenorth cooks, deliveroo is expensive
20:00:17 <Samu> it's what happens when you remove a depot for a train, let me verify
20:00:34 <Eddi|zuHause> "your branch master is 54 commits behind origin/master" are you really moving this fast lately?
20:00:40 <andythenorth> supermop_work_: yo
20:02:26 <Samu> demolished train depot, it became Invalid Order immediately
20:02:52 <Samu> i don't really invalidate it on demolish, only when the rebuilt airport doesn't have hangar
20:02:52 <Eddi|zuHause> i hate patches with savegame bump, they're always conflicting when they're not merged quickly
20:03:35 <andythenorth> so 7051 changes train depot behaviour to be more forgiving
20:03:47 <andythenorth> I am +1 to not invalidating immediately
20:03:59 <andythenorth> it supports airport closure + replacement
20:04:19 <andythenorth> so waypoints and stations are invalidated on ~30 days
20:06:21 <peter1138> What about road vehicles and ships?
20:06:41 <peter1138> If it works for all types, then that's a better solution than Samu's.
20:09:07 <andythenorth> probably works for RVs and ships?
20:09:08 <Samu> i'm not sure what 7051 do about hangars, let me look
20:09:12 <andythenorth> I only tested trains :x
20:09:21 <andythenorth> code looks like it's generic?
20:09:53 <peter1138> May not be because, again, hangars are removed different from depots for other types.
20:10:46 <Eddi|zuHause> we need a more git friendly way of handling savegame bumps
20:11:20 <peter1138> It's mostly okay but yes, it'd be nice to not have to replicate numbers everywhere.
20:12:05 <peter1138> Maybe an enum of when a change is introduced. Then you just need to resolve the conflict in one place.
20:12:49 <Eddi|zuHause> yeah, that giant comment above SAVEGAME_VERSION might be turned into an enum
20:14:18 <nielsm> overkill solution: every savegame change has a GUID associated, every savegame contains a full list of all the GUIDs applicable to its data
20:16:08 <peter1138> Not sure what advantages that has.
20:16:51 <nielsm> orthogonal changes don't affect each other, doesn't need serialisation
20:18:33 <peter1138> Not sure how you'd handle savegame conversion.
20:19:32 <Eddi|zuHause> peter1138: the "before" check turns into "does not have GUID"
20:19:39 <peter1138> Or, indeed, handle future data changes.
20:20:12 <peter1138> It'd be neat, don't get me wrong. I'm all about the negativity :p
20:20:41 <peter1138> Hmm, how do I resize a window without changing its minimum size?
20:21:23 <Eddi|zuHause> peter1138: change its default size, reopen with new default size?
20:21:45 <andythenorth> did anyone make any PRs for Iron Horse? o_O
20:21:49 <andythenorth> eh I should move it to github
20:21:56 <andythenorth> then people draw all my sprites, right?
20:22:04 <LordAro> that's how open source works, yes
20:22:48 <Eddi|zuHause> that's how mafia works? sorry, i had to
20:23:28 <Samu> 7051 doesn't remove the hangar from the order after 30 days
20:23:30 <Eddi|zuHause> i'm unsure what the conflict in order_cmd means
20:23:35 <Samu> i replaced helidepot to heliport~
20:23:49 <Eddi|zuHause> need to look at both original commits
20:25:13 <Eddi|zuHause> and this saveload mess prevents me from handling this conflict individually
20:25:19 <Eddi|zuHause> gonna abort this
20:25:41 <Samu> hangars are a different speciment of depot
20:25:59 <Samu> can't be decoupled from their airport
20:26:45 <peter1138> Hmm, so much latency.
20:27:30 <LordAro> "Category: Open Transport Tycoon Deluxe" TRIGGERED
20:28:21 <Eddi|zuHause> i like how on twitch you can skip ads with f5
20:28:36 <peter1138> Yeah, I know, I emailed them about that a while ago and then forgot about it.
20:28:43 <andythenorth> I had to watch an ad
20:28:58 <peter1138> In fact I didn't even set it now, it was from when I last fiddled with twitch.
20:30:55 <LordAro> we're missing the top of your screen, btw
20:31:23 <peter1138> Looks alright for me.
20:31:24 <LordAro> ...i was scrolled down
20:31:32 <andythenorth> just give me all the train sprites for the group
20:31:36 <andythenorth> and a crayon tool
20:31:46 <LordAro> damn sticky headers confusing me about where the top of the page is
20:32:14 <peter1138> Hmm, wonder the list starts empty :/
20:32:20 <peter1138> Yeah, they're annoying
20:32:34 <peter1138> Oh, this is that 5000 ship savegame that used to be horribly slow.
20:32:50 <Eddi|zuHause> <andythenorth> can I? :O <-- yeah, twitch shows you an ad the first time you visit the site, but not the 2nd time. which is the reverse of youtube, btw.
20:33:21 <peter1138> Got rid of the checkboxes from the livery window, now has "default" in the dropdown list.
20:45:40 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened pull request #7101: Fix #6636: Airplanes could be sent to helicopter station depots https://git.io/fhKfk
20:48:27 <peter1138> Hmm, need to set initial scroll position.
20:48:47 <peter1138> And also, cook dinner.
20:52:33 <Gabda> it is quiet in the stream
20:53:18 <Gabda> are all 7 viewers from here? :)
20:57:53 <nielsm> constant buffering for me :(
20:58:13 <Gabda> Can I have a WIP tag on PR 7047 and 7025?
20:59:07 <Gabda> I am making a 3rd solution for the same problem
20:59:39 <Gabda> that was fast, thank you :)
21:01:33 <LordAro> nielsm: works for me, go for it
21:01:58 <DorpsGek_II> [OpenTTD/OpenTTD] nielsmh approved pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhKJZ
21:02:26 <LordAro> vast majority of it is for VS
21:02:36 <LordAro> so as long as someone's tested that :p
21:03:09 <Samu> an unexpected conflict :|
21:03:37 <Samu> ship cache conflicting in
21:04:31 *** Thedarkb-T60 has joined #openttd
21:05:49 <nielsm> peter1138: merge your own pr :)
21:06:20 *** Simons_Mith has joined #openttd
21:07:19 <nielsm> people are here and active, just not in this minute
21:07:52 <Simons_Mith> Was very busy late yesterday
21:08:59 <Samu> dest_tile is now SetDestTile
21:09:03 <Samu> a->SetDestTile(a->GetOrderStationLocation(destination));
21:10:12 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened pull request #7102: Fix: Direct the aircraft to the correct location of the hangar when s… https://git.io/fhKJN
21:11:24 <Samu> i think that one deserves a savegame to demonstrate the bug
21:14:59 <glx> I think I solved the deadlock, will PR
21:18:53 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7102: Fix: Direct the aircraft to the correct location of the hangar when s… https://git.io/fhKUR
21:20:06 <Simons_Mith> I was pondering the naming of depots.
21:20:16 <Simons_Mith> (Well it passes the time)
21:20:26 <peter1138> Ah harrr, now I'm back, but I've gotta eat :p
21:20:44 <Simons_Mith> I would have thought the first depot in a town would just be called 'depot'.
21:20:59 <Simons_Mith> There aren't any others, so there's no ambiguity,
21:21:33 <Simons_Mith> But in-game, for understandable reasons, our companies depandtically name every depot as RV/Train or whatever
21:22:44 <Simons_Mith> I was mulling putting custom code in to strip the redudant bit out, for the first depot in a town.
21:23:36 <Simons_Mith> Would mean that if you built a train depot first in one town and an RV epot first in aother, they'd both be called ABC Depot and DEF Depot.
21:23:55 <Simons_Mith> Little buit more verismilitude, little bit less consistency.
21:24:20 <Eddi|zuHause> <nielsm> have there been experiments with stations having capacity for waiting cargo? <-- the suggestion has been around for a bit, but i have no recollection of anyone actually trying it. main problem is, how do you handle it for people who don't have any station newgrfs loaded?
21:24:32 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 opened pull request #7103: Fix #7007: deadlock when launched with -n switch https://git.io/fhKUb
21:26:11 <nielsm> Eddi|zuHause: I suppose you'd have to supply some base "warehouse" expansions to stations, or just only enable limited capacity stations if any newgrf supporting it are loaded
21:28:48 <glx> maybe use the closer station name as reference
21:29:14 <Eddi|zuHause> nielsm: next thing is, people will want to disable the default station, if they want to introduce some stricter limits in their own station set
21:30:33 <glx> hmm there's a extra space in my commit message
21:30:40 <peter1138> Add a default capacity per station tile (maybe differ per type, as well)
21:31:30 <nielsm> roadside stops have lower capacity than bus stations e.g.
21:31:53 <Eddi|zuHause> well, i'm assuming here that the default station gets some generic "good for any cargo" limit, and the newgrfs can differentiate that by cargo type/class
21:32:29 <peter1138> Eddi|zuHause, yeah, I'd say so.
21:32:42 <peter1138> Though bus stops are obviously passenger only?
21:32:55 <LordAro> glx: it does indeed have an extra space :p
21:33:04 <LordAro> looks like all the webstuff suppresses it
21:34:47 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 updated pull request #7103: Fix #7007: deadlock when launched with -n switch https://git.io/fhKUb
21:34:49 <LordAro> oh, that's a very simple change
21:35:22 <glx> yes and I midified the comment a little while fixing the commit message
21:35:58 <glx> it's better with a "while"
21:37:23 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7103: Fix #7007: deadlock when launched with -n switch https://git.io/fhKTp
21:37:36 <LordAro> oh, the CI is borked again
21:38:27 <glx> at first I though it would be way harder to fix the deadlock, but was simple when the idea came :)
21:39:36 <glx> these can be very nasty to fix
21:41:29 <peter1138> Ok so... set a scrollbar position.
21:43:24 <Simons_Mith> BTW I generally take silence/no direct respose to mean 'nobody hates the idea'. Is that a reasonable assumption? Or is it wiser to canvas for an actual positive acknowledgement?
21:44:48 <glx> unless it's a very bad idea we rarely hate stuff
21:45:03 <glx> well not until we see the PR
21:45:28 <peter1138> We only hate Samu's stuff by default ;)
21:46:37 <Simons_Mith> Well now that pull requests are a thing it's much easier to get a suggestion into a hateable state.
21:47:17 <Simons_Mith> I think I'll start by resurrcting the copy refinements from that old thread.
21:47:24 <glx> like, if you were suggesting some kind of multi threading in game engine, the answer would be more direct :)
21:47:35 <Simons_Mith> I presume breaking it into reasonably bite size chunks is wise?
21:48:53 <glx> as long as each commit is buildable and works by itself
21:49:40 <Simons_Mith> What's your preference on putting currently unused 'hooks' in for future stuff?
21:50:19 <peter1138> Depends how soon you think it'll be used.
21:50:24 <glx> it's ok in preparation phase I think
21:50:40 <glx> like if it's used by next commit ;)
21:51:03 <peter1138> I once did it, in preparation for a commit later in the week. By the time I got there it'd been reverted, and that feature never happened.
21:51:54 <LordAro> it wouldn't have been merged nowadays ;)
21:56:14 <TrueBrain> my poor mailbox .. maybe I should unsubscribe from getting an email for any activity on any OpenTTD repo ..
21:56:52 <peter1138> Filter it into its own mailbox.
21:57:22 <TrueBrain> it already is; but I just mark them all as read now :P
21:57:57 <glx> hey I just get email for stuff I touched
21:58:29 <TrueBrain> why is Azure Pipelines triggering such a pain?
21:59:08 <glx> because it's the free version
22:00:14 <TrueBrain> its really weird .. sometimes things don't want to queue .. but the timing is highly irregular
22:01:26 <TrueBrain> lolz ... I named the new binary -docs-source.tar.xz
22:01:31 <TrueBrain> we also have -source.tar.xz
22:01:37 <TrueBrain> and the build system is now confused which I mean :D
22:02:04 *** Thedarkb-T60 has joined #openttd
22:02:07 <TrueBrain> (as it downloads all artifacts ending with -source.tar.xz)
22:02:56 <TrueBrain> how to fix that .. hmmm
22:04:53 <TrueBrain> :D That is cheating :P
22:04:57 <TrueBrain> itemPattern: 'bundles/openttd-*-source.tar.xz'
22:14:39 <peter1138> Sometimes, I wish I knew what I was doing...
22:15:26 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain opened pull request #24: Fix: [release-docs] name the source docs just '-docs.tar.xz' https://git.io/fhKLH
22:15:28 <peter1138> So I can select a group when opening the window, but when changing the selection with an existing window, it wipes out the list completely :/
22:15:36 <TrueBrain> peter1138: I just removed the -source part completely. As .. why not .. :D
22:15:44 <peter1138> TrueBrain, cheating is sometimes easiest :D
22:16:13 <TrueBrain> all other solutions I found were more terrible :P
22:16:20 <peter1138> TrueBrain, an update for the download descriptions too?
22:16:26 <TrueBrain> it already knows this one :)
22:16:30 <TrueBrain> I can remove the now unneeded, I guess
22:16:43 <DorpsGek_II> [OpenTTD/CompileFarm] PeterN approved pull request #24: Fix: [release-docs] name the source docs just '-docs.tar.xz' https://git.io/fhKLF
22:17:46 <peter1138> LordAro, ah yes, #7077 does work for VS, I do have VS here ;)
22:17:58 <DorpsGek_II> [OpenTTD/website] TrueBrain opened pull request #39: Remove: -docs-source.tar.xz was never in production https://git.io/fhKLA
22:18:13 <TrueBrain> right, it is indeed nicer to remove that entry :)
22:18:17 <DorpsGek_II> [OpenTTD/website] PeterN approved pull request #39: Remove: -docs-source.tar.xz was never in production https://git.io/fhKtv
22:18:31 <peter1138> You'll be wondering what it is for later on down the line...
22:18:37 <TrueBrain> I tried starting VS the other day .. it was telling me my license expired ... gave up after that :P
22:19:42 *** Thedarkb1-T60 has joined #openttd
22:20:35 <TrueBrain> funny, Visual Studio Code found the VS, and is now using the cl.exe :P
22:21:51 <DorpsGek_II> [OpenTTD/website] TrueBrain merged pull request #39: Remove: -docs-source.tar.xz was never in production https://git.io/fhKLA
22:22:49 * andythenorth played tanks for a bit
22:23:47 <TrueBrain> stop wasting your time on tanks!
22:24:04 <andythenorth> it's like 1 hour this week
22:24:13 <TrueBrain> GO KILL OTHER TANKS PLZ :D
22:24:14 <andythenorth> in 2018 I was playing like
22:24:22 <andythenorth> @calc 5 * 5 + 20
22:24:39 <andythenorth> 45 hours per day yes
22:24:49 <TrueBrain> that made the most sense to me too yes
22:24:57 <TrueBrain> playing tanks against yourself
22:25:52 <andythenorth> can GS trigger towns building stuff?
22:26:11 <andythenorth> also Samu's town growth ships thing
22:26:22 <andythenorth> regressions via GS :P
22:27:27 <DorpsGek_II> [OpenTTD/CompileFarm] TrueBrain merged pull request #24: Fix: [release-docs] name the source docs just '-docs.tar.xz' https://git.io/fhKLH
22:29:45 <peter1138> Why would this->owner be wrong when calling a function externally from a window, but correct within that window?
22:31:18 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain commented on issue #7043: Disconnected towns during world generation https://git.io/fhKqC
22:35:09 *** Thedarkb1-T60 has joined #openttd
22:39:43 <peter1138> Eh, same pointer value.
22:40:04 <peter1138> I'm using BringWindowToFrontById, and then casting the return value of that.
22:40:11 <LordAro> sounds like it's time for some watchpoints
22:41:24 <peter1138> I wonder if cgdb is still hideously slow with our debug builds.
22:41:44 <LordAro> it wasn't too bad for me the other day
22:43:12 <peter1138> this->owner has the same pointer value too... just a different value.
22:45:07 <peter1138> Yeah seems okay now.
22:45:12 <peter1138> Used to take about 3 minutes to start up.
22:45:37 <peter1138> Slows down a bit with a watch on :)
22:45:55 <peter1138> Okay maybe 5000 ships wasn't a good savegame to try it on.
22:47:09 <peter1138> Or at all. Add a watch makes it horrible slow even without leaving that function :/
22:49:35 <peter1138> Oh, it wasn't debug 3
22:50:46 <peter1138> Yeah, now it takes ages, woo.
22:53:43 <Samu> it was by pure coincidence
22:54:51 <Samu> it happened right in front of me
22:55:02 <Samu> and during my prevent ship stuff
22:55:14 <Samu> i know, but it's not derived from my stuff
22:55:58 <peter1138> Yeah but do you know the seed?
22:56:26 <Samu> towns generate in different places
22:56:29 <glx> the seed is useful, it determines all random calls
22:57:06 <Samu> erm, no, i mean the seed would only be useful for my prevent town grouth block thing build than the master
22:57:38 <Samu> i tried the same seed on a master build, and the towns are placed differently
22:57:58 <Samu> kind of expected, due to the different construction rules
22:58:41 <TrueBrain> including NoAI and NoGo docs :D
22:59:10 <Samu> i followed the code and found it deleting towns with 0 population as they are generated
22:59:17 <Samu> that's where I found it deleting the bridges
23:00:44 <Samu> a town was generated on some of the free space shown on that screenshot, then removed right away because it generated with 0, but took bridges of nearby towns with it
23:02:03 <peter1138> So this->owner is set to INVALID_COMPANY within FinishInitNested()
23:02:22 <peter1138> So why it is *not* INVALID_COMPANY when I refer to owner within the window? :/
23:02:32 <andythenorth> drawing pretendelino fronts :(
23:02:47 <LordAro> that right hand side does look very empty now
23:03:03 <TrueBrain> it does; suggestions are welcome
23:03:11 <LordAro> complete site redesign!
23:03:50 <andythenorth> TrueBrain: remove right hand column :P
23:03:52 <andythenorth> increase news space
23:03:59 <LordAro> andythenorth: except not virgin :p
23:04:56 <LordAro> they got east coast taken away from them
23:05:08 <andythenorth> TrueBrain: doing it
23:05:34 *** Thedarkb-T60 has joined #openttd
23:05:39 <peter1138> hardware watchpoint worked in debug-level 3
23:05:58 <TrueBrain> andythenorth: but but but .. the screenshots!
23:06:02 <TrueBrain> THINK ABOUT THE SCREENSHOTS!
23:06:14 <andythenorth> well put 4 screenshots there in a column then
23:06:32 <peter1138> One last crash and then it's ready.
23:06:57 <andythenorth> now I have to learn how the screenshots work :P
23:07:16 <andythenorth> we need to sample n?
23:07:34 * nielsm wants an annoyance-mode for rail signals where they only work within limited distance from a station (and have very significant maintenance cost per individual signal)
23:07:37 <andythenorth> LordAro: yeah nobody uses fully justified any more :)
23:07:50 <Samu> i better post it in the issue, chat a bit busy
23:07:51 <andythenorth> nielsm: force building signal boxes :P
23:07:51 <LordAro> andythenorth: also the header image
23:08:01 *** Thedarkb-T60 has joined #openttd
23:08:12 <LordAro> Samu: you should always post things relevant to issues... in the issues
23:08:30 <TrueBrain> andythenorth / LordAro: maybe leave it like this, and take longer to make it really better :P
23:08:34 <nielsm> semaphores only work when you have personnel stationed nearby to watch the trains go by and pull those levers!
23:08:48 <andythenorth> TrueBrain: I have learnt the hard way, don't make late changes on Friday, then ship
23:09:07 <LordAro> andythenorth: surely our equivalent is making late changes on Sunday? :p
23:09:12 <glx> Samu: yeah that should go to the issue, will help others wanting to work on it
23:09:27 <andythenorth> TrueBrain: I am happy to improve, but actually I am brain dead now
23:09:31 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro merged pull request #7103: Fix #7007: deadlock when launched with -n switch https://git.io/fhKUb
23:09:42 * LordAro googles "website designs"
23:10:13 <LordAro> "20 of the best website designs to inspire you in 2018"
23:10:14 <andythenorth> oof ad spam sorry
23:10:18 <TrueBrain> andythenorth: my point :) Lets just release this tomorrow :)
23:10:23 * andythenorth just uses bootstrap for everything
23:11:44 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on issue #7043: Disconnected towns during world generation https://git.io/fhK32
23:12:00 <andythenorth> can we use pointless photos from unsplash?
23:13:08 <TrueBrain> setting unrealistic expectations for the game are we? :P
23:13:15 <TrueBrain> I wonder if planetmaker already got the steam account claimed :D
23:14:30 <LordAro> they are very nice photos
23:14:58 <nielsm> works a bit better for simsig, with photos of the real locations simulated :)
23:15:06 <andythenorth> peter1138: what did you do :o
23:15:27 <andythenorth> used to go every yeare
23:15:32 <peter1138> andythenorth, I finished it.
23:17:15 <LordAro> i think i shall play some tanks
23:18:20 <andythenorth> who drew those icons :)
23:20:18 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened pull request #7104: Fix #5405: Determine the correct last and next destinations and then ensure these are within range of each other https://git.io/fhKsL
23:20:37 <andythenorth> oh the realism :)
23:22:26 <Samu> the next 5 changes are too difficult to post as different PR's
23:22:37 <Samu> they need to be together
23:22:53 <Samu> they depend on each other
23:23:21 <glx> if it forms a logical group of commits
23:23:26 <peter1138> Keep together but separate commits.
23:24:28 <andythenorth> peter1138: depot was slow to update livery after moving train to a group?
23:24:39 <andythenorth> [rest is pretty awesome, just saying what I see]
23:25:00 <andythenorth> seems like it's not marked dirty
23:25:02 <peter1138> Yeah I guess I need to handle that case.
23:25:06 <andythenorth> I click on it, it updates
23:28:42 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7077: Change: Always build baseset metadata, instead of including it in the repo. https://git.io/fhzn8
23:29:24 <peter1138> Be 'funny' if everything fails now ;p
23:29:42 <peter1138> Would be nice to do the same for the widgets/squirrel exports, too.
23:29:52 <peter1138> I partially got the squirrel exports stuff working.
23:30:05 <peter1138> They need changing because they do in-place modification.
23:31:56 <glx> but squirrel export already does that
23:32:11 <peter1138> I mean they replace src/ai/ai_instance.cpp, to do it properly it needs to build from src/ai/ai_instance.cpp.in -> src/ai/ai_instance.cpp
23:32:23 <peter1138> Then src/ai/ai_instance.cpp can be removed from the repo.
23:32:29 <peter1138> I have that bit working, actually.
23:32:50 <peter1138> But I'm not sure what a "blank" version of the file should actually contain, as it seems to rely on some of the autogenerated stuff already being there.
23:33:01 <glx> .sq are generated from nothing IIRC
23:33:14 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick commented on pull request #7104: Fix #5405: Determine the correct last and next destinations and then ensure these are within range of each other https://git.io/fhKGY
23:33:14 <peter1138> Yes, they are fine.
23:33:29 <peter1138> Just needs a README put in the directory because git can't store empty directories.
23:33:37 <peter1138> And the _controller one is a bit unique too.
23:33:43 *** oskari89 has joined #openttd
23:34:00 <glx> _controller is not generated
23:40:14 <andythenorth> now imagine it with ctrl-click? :D
23:40:33 <LordAro> i am absolutely terrible at this game
23:40:45 <LordAro> took me 4 games to actually hit anyone
23:40:51 <andythenorth> what are you playing?
23:40:54 <LordAro> and that didn't cause damage
23:41:00 <andythenorth> blitz or war thunder?
23:41:50 <andythenorth> WoT Blitz is multi-platform
23:42:31 <glx> I tried world of warships
23:42:34 <andythenorth> BAD FEATURE: is there an action 2 var for 'current company colour' (byte)
23:43:57 *** sim-al2 has joined #openttd
23:46:04 <supermop_work_> andythenorth: maybe
23:46:23 <supermop_work_> i recall something like that but may be misremembering
23:46:27 <peter1138> andythenorth, so it's working as you expected it?
23:46:34 <peter1138> Apart from the depot thing.
23:46:47 <andythenorth> I have....ideas :P
23:47:36 <andythenorth> let's try it on RVs :)
23:48:21 <TrueBrain> lol @ nielsm :) I like the "it isn't a bug, so I am have a feature request" approach there :D
23:48:40 <TrueBrain> well, time to drink more beer, I see
23:48:58 <andythenorth> yeah ok so RVs works nicely
23:49:03 <nielsm> that thing is really annoying :)
23:49:11 <andythenorth> can give trucks a trailer colour to match route or cargo
23:50:54 <nielsm> hmm wtf, it fails executing cscript
23:51:16 <nielsm> C:\Program Files (x86)\Microsoft Visual Studio\2019\Preview\MSBuild\Current\Bin\Microsoft.Common.CurrentVersion.targets(2817,7): error MSB4057: The target "CreateManifestResourceNames" does not exist in the project.
23:53:15 <nielsm> glx yes, that's why I've not sent in a feature request for that
23:53:57 <glx> I think it's been requested for visual studio code already
23:54:30 <glx> but the NULL one is weird
23:55:11 <glx> how can it fail to execute cscript ??? it's a windows tool
23:55:37 <nielsm> but it seems it was running the v141 toolset msbuild projects with the v142 toolset
23:55:47 <nielsm> so now installing the v141 toolset...
23:55:56 <nielsm> and may have to make updated projects for v142
23:58:22 <andythenorth> so how about...?
23:58:28 <andythenorth> 4 company colours? :)
23:58:32 <glx> it didn't suggest upgrade when opening vs141 solution ?
23:58:43 <andythenorth> there are some purples we could sacrifice, especially if it was flagged :P
23:58:45 <peter1138> andythenorth... no.
23:58:52 <nielsm> nope, it just loads it and shows (Visual Studio 2017) after every project name
23:58:56 <peter1138> You could do it with 32bpp, I suppose.
23:59:27 <andythenorth> no red wagons for my grey engine? :)
23:59:59 * andythenorth was trying to avoid suggesting 'wagon' and 'engine' options
continue to next day ⏵
