IRC logs for #openttd on OFTC at 2021-04-10
⏴ go to previous day
00:11:20 <DorpsGek> [OpenTTD/OpenTTD] perezdidac opened issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
00:27:40 <spnda> By now, you've really gotten the hang of writing list filters haven't you
00:28:36 <didac> hahah yeah I've gone too far
00:28:37 <didac> filter is my middle name
00:51:43 *** didac has quit IRC (Remote host closed the connection)
01:05:16 <DorpsGek> [OpenTTD/OpenTTD] glx22 opened pull request #8985: Fix e0561dbde: [MinGW] use ofstring(wchar_t*) as ofstring(wstring) doesn't exist https://git.io/JOfWe
01:26:39 *** spnda has quit IRC (Remote host closed the connection)
01:55:25 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler opened pull request #8986: Fix: Missing 'Town names:' colon in map gen GUI https://git.io/JOfRa
02:28:20 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8891: Introduce snow/desert-coverage and custom terrain type, and move "maximum map height" to settings only https://git.io/JOfzd
02:33:08 *** iSoSyS has quit IRC (Remote host closed the connection)
02:37:08 *** Wuzzy has quit IRC (Quit: Wuzzy)
02:43:44 *** Wormnest has quit IRC (Quit: Leaving)
04:37:45 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
04:42:09 *** supermop_Home has quit IRC (Remote host closed the connection)
05:01:36 *** y2kboy23 has joined #openttd
05:26:29 *** riis has quit IRC (Remote host closed the connection)
05:37:41 *** snail_UES_ has quit IRC (Quit: snail_UES_)
06:32:20 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8986: Fix: Missing 'Town names:' colon in map gen GUI https://git.io/JOfA1
06:38:20 <nielsm> morning, how's the outlook for 1.11.1 release soon? any major outstanding issues?
06:38:29 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JOfxP
06:38:33 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JOfxX
06:39:34 *** Progman has joined #openttd
06:40:51 <LordAro> nielsm: the GL(?) related crashes would be good
06:45:06 <nielsm> also again a situation where being able to release a version that's 100% network compatible with a previous one, and just has "client fixes" would be nice
06:46:27 <nielsm> that could also allow releasing test betas trying different approaches to solving an issue and have many players test it, without impacting their abilito to play with others
06:47:31 <DorpsGek> [OpenTTD/OpenTTD] James103 opened issue #8987: Loading JGRPP savegames in official OpenTTD gives "too new" instead of "modified version" error https://git.io/JOfhZ
06:49:34 <DorpsGek> [OpenTTD/OpenTTD] nielsmh approved pull request #8985: Fix e0561dbde: [MinGW] use ofstring(wchar_t*) as ofstring(wstring) doesn't exist https://git.io/JOfh6
06:52:18 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #8986: Fix: Missing 'Town names:' colon in map gen GUI https://git.io/JOfhA
07:30:28 *** andythenorth_ has joined #openttd
07:43:00 *** Progman has quit IRC (Remote host closed the connection)
07:54:25 *** WormnestAndroid has quit IRC (Remote host closed the connection)
07:54:38 *** WormnestAndroid has joined #openttd
07:56:51 *** didac has quit IRC (Remote host closed the connection)
08:09:00 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
08:09:13 *** WormnestAndroid has joined #openttd
08:28:39 <DorpsGek> [OpenTTD/OpenTTD] nielsmh requested changes for pull request #8006: Codechange: Increase scrollbar length limit to 32bits https://git.io/JOJON
08:47:43 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 opened issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
08:56:11 <andythenorth_> this is...interesting
08:56:12 <andythenorth_> { 0x44442202, 0x44440111 }, // UKRS addons modifies UKRS
08:56:26 <andythenorth_> specific support for specific newgrfs? :o
08:57:22 <peter1138> This is backwards compatibility when the enginepool stuff was done.
08:59:11 <peter1138> A comment would have been useful :p
09:00:26 <peter1138> Basically the enginepool stuff broke anything that deliberately modified another set's vehicles, so it's a bit of specialness to allow that.
09:00:47 <TrueBrain> amazed you still remember that :o
09:01:23 <peter1138> It's amazing, because I didn't even write that particular bit of code but forgot that I hadn't ;)
09:01:34 <andythenorth_> it actually increments with cargo aging
09:01:41 <andythenorth_> it doesn't reset when cargo is loaded
09:01:57 <andythenorth_> also cargo packet age is statistical
09:02:11 <andythenorth_> so I might need to rethink a tiny bit of this idea
09:02:26 <peter1138> TrueBrain, but yes. I don't remember what I was doing yesterday...
09:03:58 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8989: Fix 70bc55cfd6e: snow line height was set while calculating desert line https://git.io/JOJcB
09:04:04 <peter1138> Well before covid I could at least say I had been out, or was still out, on the bike.
09:04:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8891: Introduce snow/desert-coverage and custom terrain type, and move "maximum map height" to settings only https://git.io/JOJcE
09:04:11 *** FLHerne has joined #openttd
09:05:07 <TrueBrain> LordAro: I plan on spending some time on the bugs we have open next week; doesn't really mean anything in terms of when-to-release, but .. yeah
09:05:20 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain merged pull request #47: Add: translate Discussion events into IRC messages https://git.io/JYHGz
09:05:26 *** sla_ro|master has joined #openttd
09:05:43 <TrueBrain> I now first have the pleasure of merging bump-PRs
09:06:02 <DorpsGek> [OpenTTD/OpenTTD] PeterN approved pull request #8989: Fix 70bc55cfd6e: snow line height was set while calculating desert line https://git.io/JOJcy
09:06:14 <peter1138> I assume it compiles...
09:06:27 <TrueBrain> peter1138: I assume too .. never tested it
09:06:49 <DorpsGek> [OpenTTD/bananas-api] TrueBrain updated pull request #84: Scheduled monthly dependency update for April https://git.io/JYov3
09:06:49 <TrueBrain> I mean, it is clearly wrong, so ... YOLO :P
09:06:52 <andythenorth_> now how do I read my own var in nml?
09:06:54 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #19: Scheduled monthly dependency update for April https://git.io/JYovg
09:07:07 <LordAro> you'd hope it would compile, really
09:07:26 <LordAro> i'm not sure what you'd have to do to make the previous valid but the new invalid
09:07:36 <LordAro> overridden assignment operator, perhaps?
09:07:50 <TrueBrain> do we want to know if it is? :P
09:09:13 <LordAro> not sure if it could return different types though
09:10:01 <DorpsGek> [OpenTTD/bananas-api] TrueBrain updated pull request #84: Scheduled monthly dependency update for April https://git.io/JYov3
09:10:24 <andythenorth_> no fricking idea how to reset the counter on cargo load though
09:10:37 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain updated pull request #19: Scheduled monthly dependency update for April https://git.io/JYovg
09:10:41 <TrueBrain> okay, chardet can be upgraded this month, it is just idna that still isn't possible
09:10:43 <TrueBrain> because of requests
09:10:53 <TrueBrain> seems we are slowly getting out of this pip dependency hellhole
09:11:18 *** Progman has joined #openttd
09:13:01 <DorpsGek> [OpenTTD/bananas-frontend-web] TrueBrain updated pull request #55: Scheduled monthly dependency update for April https://git.io/JYovM
09:13:35 <DorpsGek> [OpenTTD/bananas-server] TrueBrain updated pull request #45: Scheduled monthly dependency update for April https://git.io/JYoff
09:13:50 <DorpsGek> [OpenTTD/master-server] TrueBrain updated pull request #28: Scheduled monthly dependency update for April https://git.io/JYoJ0
09:13:56 <andythenorth_> ok so let's play text adventure
09:14:03 <andythenorth_> you are in a train coach with no toilets
09:14:21 <LordAro> peter1138: how far do i want to cycle today?
09:14:28 <DorpsGek> [OpenTTD/master-server-web] TrueBrain updated pull request #21: Scheduled monthly dependency update for April https://git.io/JYoJS
09:14:39 <peter1138> You, probably lots. Me, 0.
09:14:49 <andythenorth_> what var do you wish to check?
09:14:50 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain updated pull request #46: Scheduled monthly dependency update for April https://git.io/JYofc
09:14:57 <TrueBrain> only had to manually fix 3 PRs this time, wauw
09:14:59 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #8989: Fix 70bc55cfd6e: snow line height was set while calculating desert line https://git.io/JOJcB
09:15:07 <andythenorth_> - time since cargo last loaded?
09:15:19 <peter1138> andythenorth_, Twin Kingdom Valley. Except it's graphical.
09:15:25 <andythenorth_> - average age of cargo, compared to average age of cargo when loaded?
09:15:35 <andythenorth_> - days since last station visit?
09:15:51 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #8985: Fix e0561dbde: [MinGW] use ofstring(wchar_t*) as ofstring(wstring) doesn't exist https://git.io/JOfWe
09:15:51 <peter1138> Actually you can turn the graphics off.
09:16:04 * andythenorth_ looks up "Twin Kingdom Valley"
09:16:21 <andythenorth_> peter1138 does it feature trains without toilets?
09:16:42 <peter1138> It features a cabin without toilets
09:17:09 <LordAro> peter1138: i have a 125k route plotted
09:17:34 <peter1138> I have failed in life.
09:19:35 <andythenorth_> I'm basically doing a sort of 'vehicle range' function here
09:20:01 <andythenorth_> but relying entirely on running costs and not trying to force anything beyond that
09:20:29 <DorpsGek> [OpenTTD/OpenTTD] nielsmh opened pull request #8990: Fix: [Win32] Font glyphs of certain widths broke https://git.io/JOJWj
09:20:54 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain approved pull request #19: Scheduled monthly dependency update for April https://git.io/JOJlf
09:21:07 <andythenorth_> measuring the cargo age could have really unexpected effects
09:21:15 <peter1138> andythenorth_, back when conserving memory really was necessary.
09:21:16 <DorpsGek> [OpenTTD/bananas-api] TrueBrain approved pull request #84: Scheduled monthly dependency update for April https://git.io/JOJlL
09:21:19 <LordAro> nielsm: how long's that been broken?
09:21:28 <andythenorth_> cargo age is an average of the age of the cargo packets on the vehicle
09:21:29 <DorpsGek> [OpenTTD/bananas-frontend-web] TrueBrain approved pull request #55: Scheduled monthly dependency update for April https://git.io/JOJlq
09:21:42 <DorpsGek> [OpenTTD/bananas-server] TrueBrain approved pull request #45: Scheduled monthly dependency update for April https://git.io/JOJlm
09:21:49 <DorpsGek> [OpenTTD/master-server] TrueBrain approved pull request #28: Scheduled monthly dependency update for April https://git.io/JOJlO
09:21:54 <DorpsGek> [OpenTTD/master-server-web] TrueBrain approved pull request #21: Scheduled monthly dependency update for April https://git.io/JOJls
09:21:59 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain approved pull request #46: Scheduled monthly dependency update for April https://git.io/JOJln
09:22:07 <nielsm> LordAro: since the win32 native font rendering was introduced
09:22:29 <TrueBrain> another 2009 fix incoming? :P
09:22:31 <TrueBrain> 1.11 is full of those :D
09:22:46 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8990: Fix: [Win32] Font glyphs of certain widths broke https://git.io/JOJlE
09:23:24 <LordAro> hmm, should #8934 be backported?
09:23:33 <LordAro> i'd be tempted to say no..
09:23:34 <TrueBrain> not even a reference to the commit where it broke?! :(
09:23:38 <andythenorth_> DaysInTransit could have very wide range, if cdist
09:23:41 <FLHerne> andythenorth_: Age since loading onto the vehicle, or since origin?
09:23:56 <TrueBrain> LordAro: I agree, no
09:24:14 <andythenorth_> FLHerne for the idea I have I want to measure aging on the current journey leg
09:24:22 <andythenorth_> DaysInTransit is an average, so that could get interesting
09:24:24 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8975: Change: do not disable NewGRF window apply button if dev tools are enabled https://git.io/JOJl1
09:24:26 <LordAro> TrueBrain: in which case...
09:24:26 <FLHerne> The latter would be an issue for journeys that are "all the way down the map on a nice express train, then branch line"
09:24:38 <TrueBrain> (I already removed the label :P)
09:24:45 <andythenorth_> yeah since origin doesn't work, there's already a var for that
09:24:53 <andythenorth_> I would be using it if viable :)
09:25:11 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8986: Fix: Missing 'Town names:' colon in map gen GUI https://git.io/JOJlb
09:25:11 <andythenorth_> if half your cargo packets are age 0 and half are age 255
09:25:15 <DorpsGek> [OpenTTD/bananas-frontend-cli] TrueBrain merged pull request #19: Scheduled monthly dependency update for April https://git.io/JYovg
09:25:19 <andythenorth_> age will be 128
09:25:28 <DorpsGek> [OpenTTD/bananas-frontend-web] TrueBrain merged pull request #55: Scheduled monthly dependency update for April https://git.io/JYovM
09:25:33 <FLHerne> Also, your tail makes my client colour your name wrong, and it's confusing :p
09:25:39 <DorpsGek> [OpenTTD/bananas-server] TrueBrain merged pull request #45: Scheduled monthly dependency update for April https://git.io/JYoff
09:25:41 <andythenorth_> why do I have a tail?
09:25:42 <FLHerne> Now you're blue like Eddi when you should be pink
09:25:47 <DorpsGek> [OpenTTD/master-server] TrueBrain merged pull request #28: Scheduled monthly dependency update for April https://git.io/JYoJ0
09:25:47 *** andythenorth_ is now known as andythenorth
09:25:54 <DorpsGek> [OpenTTD/master-server-web] TrueBrain merged pull request #21: Scheduled monthly dependency update for April https://git.io/JYoJS
09:26:00 <DorpsGek> [OpenTTD/DorpsGek] TrueBrain merged pull request #46: Scheduled monthly dependency update for April https://git.io/JYofc
09:26:05 <andythenorth> is blue ignore, or is pink ignore? :P
09:27:07 * FLHerne is an equal-opportunity ignorer, and makes strong efforts not to discriminate based on the colour of one's nick :p
09:28:14 <FLHerne> Difference in age is just due to loading times?
09:29:30 <andythenorth> difference in age is due to time spent on vehicles
09:32:19 <FLHerne> But if it's per-leg, how can there be a difference in time on vehicles except loading times
09:32:25 <FLHerne> oh wait there are multiple stops
09:32:57 <andythenorth> there are a few fiddly aspects to it
09:33:20 <andythenorth> my current patch just increments a counter every time cargo aging is called
09:33:28 <andythenorth> and needs to reset it when cargo is loaded or something
09:33:37 <andythenorth> or the train visits a station
09:33:53 <andythenorth> visiting a station would be tediously easy to exploit
09:34:28 <andythenorth> I think resetting when any cargo is loaded / unloaded would be the best
09:34:47 <andythenorth> any choice has edge cases though
09:35:26 <peter1138> Bah at when you find a nice looking font but the baseline is slightly off.
09:35:46 <andythenorth> e.g. cargo time on vehicle could be messed up by trains going off to service
09:35:49 <nielsm> are you doing a thing where you effectively keep track of "actual age" and "vehicle-requested age" separately?
09:36:15 <andythenorth> nielsm not sure :) Can you expand?
09:37:18 <nielsm> if you increment a new counter every time cargo aging is called, then that counter essentially measures how much the cargo has actually aged without taking vehicle effects into account, doesn't it?
09:39:24 <andythenorth> currently it increments when the vehicle ages cargo
09:39:32 <DorpsGek> [OpenTTD/OpenTTD] nielsmh merged pull request #8990: Fix: [Win32] Font glyphs of certain widths broke https://git.io/JOJWj
09:39:47 <andythenorth> if I've understood correctly
09:41:11 *** DorpsGek` has joined #openttd
09:41:11 *** DorpsGek is now known as Guest1063
09:41:11 *** ChanServ sets mode: +o DorpsGek`
09:41:11 *** DorpsGek` is now known as DorpsGek
09:41:11 *** DorpsGek` is now known as DorpsGek
09:41:14 *** Guest1063 has quit IRC (Remote host closed the connection)
09:42:27 <DorpsGek> [OpenTTD/bananas-api] TrueBrain merged pull request #84: Scheduled monthly dependency update for April https://git.io/JYov3
09:42:45 <andythenorth> hmm, maybe I should write a gist of what I'm trying to do
09:42:56 <andythenorth> mine don't come out as neatly as TrueBrain's though :|
09:43:11 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
09:43:13 <peter1138> Write a gist of what you want to achieve
09:44:27 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
09:44:30 <TrueBrain> it works, but not to this channel yet ..
09:45:21 <peter1138> andythenorth, I'm not sure why you'd be putting cargo aging stuff on the vehicle, though. It's the cargo that ages...
09:45:36 <andythenorth> the vehicle handles returning the current age
09:45:54 <andythenorth> it's an average of the cargo packets in the vehiclecargolist or so
09:46:05 <peter1138> The cargo age comes from the cargo packets.
09:46:18 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8991: Change: [DorpsGek] also announce Discussion creation/comments from GitHub https://git.io/JOJRE
09:46:21 <andythenorth> yes, the cargo packet sums and divides them
09:46:38 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8991: Change: [DorpsGek] also announce Discussion creation/comments from GitHub https://git.io/JOJR2
09:46:42 <andythenorth> I closed the code to write a gist, but eh, implementation isn't the interesting bit
09:46:51 <andythenorth> except...OpenTTD facts are what they are
09:46:59 <peter1138> Sums and divides? What?
09:47:12 <andythenorth> hang on I'll find the line
09:47:20 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8991: Change: [DorpsGek] also announce Discussion creation/comments from GitHub https://git.io/JOJRE
09:47:50 <andythenorth> cargo packet DaysInTransit()
09:48:00 <andythenorth> L255 cargopacket.h
09:48:37 <andythenorth> which the vehicle handles as 80+ var 0x3F
09:48:51 <andythenorth> code is clear, my explanation might be piss poor
09:49:05 <peter1138> But that's only a NewGRF var. That isn't used for the actual payment function.
09:49:13 <andythenorth> no I'm leaving the payment function well alone
09:49:18 <andythenorth> messing with payment...no
09:49:19 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
09:49:21 <andythenorth> I don't think it can be done
09:49:32 <andythenorth> it will mess with transfer payments
09:49:33 <TrueBrain> cool, you can now sort Discussion replies on amount of votes or new/old
09:49:40 <peter1138> I don't even know why you'd need to access the cargo age in varactions.
09:52:33 <FLHerne> peter1138: He's written a grf that makes the running costs very high if the cargo is too old
09:52:44 <FLHerne> there's a screenshot ^ up there somewhere
09:53:14 <FLHerne> Because if the cargo is too old it might get ants?
09:55:40 <peter1138> Let's make it realistic. Passengers pay for a ticket before they get on the train... ;)
09:56:15 <Wolf01> Couldn't you just change the cargo aging factor in relation of vehicle age?
09:56:41 <andythenorth> I would advise not changing the cargo aging factor
09:56:55 <andythenorth> I've tested it, I think it's unwise
09:58:03 <andythenorth> peter1138 the 'why' is because running metro on all routes is boring
09:58:10 <andythenorth> but that's what Iron Horse implies you should do
09:58:54 <Wolf01> The problem of TT is that it doesn't work at all like R-life, so you shouldn't even apply R-life rules
10:00:01 <Wolf01> I as passenger pay the same even if the train coach is rotting away or it takes 2 hours more because whatever happened on the trip, I would really like if I could pay at the end of the trip, so I could ask money instead of paying
10:00:41 <andythenorth> RL rules are highly variable anyway
10:01:15 <andythenorth> in some systems you get some money back for delays
10:01:18 <Wolf01> Yes, but usually transport companies don't sell cargo
10:01:27 <andythenorth> in other systems you don't pay at all for some legs of the trip
10:01:38 <andythenorth> not trying to model reality here
10:01:53 <andythenorth> oof I think I played way too much tank game
10:02:50 <Wolf01> But the majority if the transport methods work with "you know how much to transport, and estimated time and pay accordingly in advance"
10:03:33 <Wolf01> You may consider others as edge cases
10:05:35 <andythenorth> yeah we don't do that
10:08:00 <andythenorth> oof I don't know
10:08:06 *** andythenorth has quit IRC (Quit: andythenorth)
10:10:05 *** Rubidium has joined #openttd
10:10:05 *** ChanServ sets mode: +o Rubidium
10:11:09 <TrueBrain> we used the name Rubidium too often lately .. it seems that summons him :p
10:13:08 <Rubidium> sorry for being away for so long
10:13:35 <TrueBrain> Rubidium: how rude!
10:13:35 <LordAro> omg, it's the big cheese
10:13:52 <Rubidium> oh darn, I'm still an op in here? I guess Celestar's proposal never made it
10:14:10 <TrueBrain> You expect us to take care of these things while you are away?
10:14:37 <TrueBrain> how is our lost son doing these days? :)
10:15:08 <Rubidium> still working with trains
10:15:52 <TrueBrain> haha, really? Nice :D
10:17:22 <Rubidium> yup. Doing the software development for doing measurements of the rail infrastructure
10:17:43 *** Synck has quit IRC (Remote host closed the connection)
10:17:45 <TrueBrain> still making sure we can take the train :) Nice :) Still enjoying it too?
10:19:26 <Rubidium> as long as there is no politics/management involved, definitely. Though that politics/management you probably got everywhere. Nevertheless "debugging" systems from suppliers based on the output of the measurement systems can be fun, just like hunting desyncs
10:19:34 <LordAro> i know someone working in that area, anything to do with Altran?
10:20:12 <TrueBrain> so what brings you back to this corner of the internet? :P
10:20:17 <TrueBrain> was it our livestream?! :D
10:21:02 <TrueBrain> frosch123: can I deploy eints (with the Romanian thing) to production? Or does that need anything else before we do that?
10:21:44 <LordAro> TrueBrain: there's some comments in #8936
10:21:54 <Rubidium> it wasn't the livestream but rather it being fools day and thinking about doing releases, though the livestream gave a little extra push
10:22:24 <Rubidium> I don't known Altran, if its the company in Helmond then definitely nope
10:22:35 <TrueBrain> LordAro: that doesn't help me which is the order shit should happen in :P
10:23:10 <TrueBrain> Rubidium: volunteering to do the next release? How nice of you!
10:24:01 <Rubidium> I already did r10k, r20k and r25k (if you look at the second SVN numbering)
10:24:41 <TrueBrain> wtf, reddit server is not the most popular atm
10:25:09 <TrueBrain> okay, all software is redeployed again, except for eints .. "seems to be working"
10:25:10 <LordAro> have we still got more clients than servers?
10:25:22 <TrueBrain> There are 342 clients, 442 IPv4 servers and 101 IPv6 servers in this list.
10:25:27 <TrueBrain> no, we are back to the old ways of doing this
10:25:56 *** sla_ro|master has quit IRC ()
10:28:01 <FLHerne> and someone's covered a continent in roads for no clear reason
10:28:48 <TrueBrain> so I guess this weekend we try to get the issues below 100 again? :P
10:33:22 <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJaW
10:33:32 * LordAro goes outside for a bit
10:34:43 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8987: Loading JGRPP savegames in official OpenTTD gives "too new" instead of "modified version" error https://git.io/JOfhZ
10:34:46 <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJaW
10:37:05 <TrueBrain> LordAro: "dest_tile" contains the industry id? Wuth?
10:37:45 <TrueBrain> better use i->index , I would think
10:38:16 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJaj
10:38:18 <peter1138> Seems wrong for vehicles, as they move.
10:38:31 <peter1138> Okay, maybe not for a disaster :p
10:39:35 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJVZ
10:40:16 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJV8
10:40:18 <TrueBrain> nah, seems I confused vehicles with industries
10:40:45 <TrueBrain> so the change should never have been made :P
10:41:05 <LordAro> disaster vehicles do an awful lot of "variable reuse"
10:41:13 <LordAro> they use orders as as state machine, for instance
10:41:20 <TrueBrain> make them into a scripting language, you say? :P
10:41:54 <TrueBrain> owh, right, I was going to enjoy some lunch ..
10:43:15 <peter1138> Unless industries move, can't it always just use the tile?
10:52:12 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
10:55:53 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #8975: Change: do not disable NewGRF window apply button if dev tools are enabled https://git.io/JYhMY
10:58:16 <DorpsGek> [OpenTTD/OpenTTD] JGRennison opened pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOJrr
11:01:03 <TrueBrain> I like how the draw cursor is slowly morphing into something completely different from where it started :D
11:03:17 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOJoC
11:10:09 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8968: Arrow key camera panning not consistent with framerate https://git.io/JYAK6
11:13:23 <peter1138> Yeah, I always have smooth scrolling on, so never noticed it.
11:13:41 <peter1138> Btw, I think the credits window scrolls too slowly now (my fault probably)
11:14:05 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8970: Ramsomware Protection preventing NewGRF download https://git.io/JYxe3
11:14:10 <michi_cc> So, Rb is back here, peter approves PR. If a Bjarne now suddenly turns up, the world really did a 720 :P
11:14:11 <peter1138> You have to be REALLY bored to get all the through it.
11:14:26 <peter1138> I did that accidentally. I forgot I wasn't going to.
11:14:36 <TrueBrain> peter1138: I feel special now :)
11:14:50 <TrueBrain> and scrolling .. I want to fix all the scrolling in the game, but it requires AA in the blitter :P
11:14:55 <TrueBrain> not sure yet how to approach that problem :)
11:15:09 <TrueBrain> I want to be able to draw stuff on subpixels
11:15:33 <TrueBrain> should allow smooth vehicle movement, smooth news movement, smooth text scrolling, etc
11:16:29 <michi_cc> Vehicle movement isn't even on pixels for 2/4x zoom, still some space left before sub-pixels :)
11:17:04 <TrueBrain> the vehicles themselves are on subpixels
11:17:11 <TrueBrain> but the blittering doesn't
11:17:22 <TrueBrain> so .. I think we have to increase the resolution blittering happens on
11:17:51 <TrueBrain> which is mostly code in the Viewport
11:18:11 <Rubidium> just switch between drawing two versions of the pixels quickly, that'll give the illusion of AA, right?
11:18:20 <TrueBrain> which .. I want to refactor first anyway, as it is a class without being called a class .. and that would allow releasing the game-state-lock during viewport loops ..
11:18:39 <TrueBrain> Rubidium: lol, that would work. I was more thinking about pixel blending on 32bpp :)
11:18:49 <TrueBrain> well, colour blending
11:19:10 <TrueBrain> but either way .. we need to know if something is on x=1.0 or 1.2 or 1.5 or 1.8 ... :)
11:19:15 <TrueBrain> information currently not available :P
11:19:50 <peter1138> Ah, temporal antialiasing.
11:22:42 <DorpsGek> [OpenTTD/OpenTTD] JGRennison updated pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOJrr
11:25:00 <peter1138> Ah, the about window scrolling is by pixels, and not scale depending on zoom level.
11:25:04 <FLHerne> TrueBrain: The comment in that saveload PR is a link to existing OTTD code
11:25:09 <FLHerne> (which apparently doesn't work)
11:25:10 <peter1138> So 1x is fine, 2x and 4x are tediously slow.
11:25:17 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
11:25:42 <TrueBrain> FLHerne: sorry? What PR? "that saveload" is not ringing any bells :)
11:26:27 <TrueBrain> hmm .. tempted to add a "low priority" label for bugs
11:26:31 <FLHerne> TrueBrain: The "For future patchpack creators: please follow a system like JGRPP, where the version is masked with 0x8000" text
11:26:45 <FLHerne> is in saveload.h as it exists now
11:26:49 <TrueBrain> FLHerne: linking the PR also works :P
11:27:10 <FLHerne> so it seems a bit odd to decline to support it when it's officially recommended :p
11:27:11 <TrueBrain> FLHerne: the fact that we mention if they can says that they aren't :P That is why I am reluctant :)
11:27:25 <TrueBrain> but .. typetype that in the ticket please
11:28:33 <TrueBrain> fuck it, I made "priority: low" label
11:31:18 <michi_cc> TrueBrain: Do you know https://obs.ninja ? Looks like it might be something to do non-Discord livestreams in the future.
11:32:27 <TrueBrain> michi_cc: you mean have everyone stream their webcam directly to OBS?
11:33:37 <TrueBrain> we will have to test it; main reason I picked Discord as it syncs up voice and video for us
11:33:48 <TrueBrain> when doing it via OBS I had the issue they could be out of sync :P
11:34:01 <TrueBrain> (mainly as OBS insists on a 1MB buffer)
11:34:09 <TrueBrain> so if they solved that ... :D
11:34:28 <michi_cc> Hmm, FAQ of obs.ninja seems to indicate that it is basically WebRTC from your browser to OBS.
11:34:37 <TrueBrain> owh, they use a browser to show it
11:34:43 <TrueBrain> that could have a lower buffer for sure
11:34:47 <TrueBrain> but yeah, worth testing
11:34:52 <DorpsGek> [OpenTTD/OpenTTD] FLHerne commented on issue #8987: Loading JGRPP savegames in official OpenTTD gives "too new" instead of "modified version" error https://git.io/JOfhZ
11:34:53 <TrueBrain> Discord is annoying as fuck :P
11:44:17 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
11:45:28 <DorpsGek> [OpenTTD/OpenTTD] PeterN opened pull request #8994: Fix: Adjust scrolling interval of credits to account for text line height https://git.io/JOJMJ
11:47:50 <TrueBrain> smart solution peter1138 :)
11:50:38 <TrueBrain> ugh, most bugs are hard to pick up :D
11:51:21 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
11:53:30 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
11:53:35 <Wolf01> angrythenorth still andy? :P
11:54:51 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
11:55:05 <TrueBrain> very odd ... only the OSK has \ in it
11:57:59 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8887: Emergency netsave on server restart connection failure saves title game https://git.io/JYvCc
11:59:07 *** frosch123 has joined #openttd
12:01:55 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #8961: Add: Larger size renderings of the logo/icon https://git.io/JOJyG
12:04:42 <frosch123> TrueBrain: who's on stream next week? :p (reminder from last night)
12:04:48 <frosch123> TrueBrain: i'll figure out eints today
12:05:49 <Wolf01> who's on stream next week? <- me as viewer for sure :P
12:06:02 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8911: Suggestion: Replace target of the order without recreating it, and/or copy order's specs. https://git.io/JYlBY
12:06:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #8911: Suggestion: Replace target of the order without recreating it, and/or copy order's specs. https://git.io/JYlBY
12:06:13 <TrueBrain> Wolf01: sorry, instead I removed the "next stream next week" :P
12:07:25 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8994: Fix: Adjust scrolling interval of credits to account for text line height https://git.io/JOJSU
12:08:22 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8874: It is possible to run the `rescannewgrf` command while NewGRFs are being scanned. https://git.io/JmrJo
12:10:08 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOJSa
12:12:43 <TrueBrain> right .. I was hoping fixing a few of these bugs would be easier
12:13:30 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 opened issue #8995: Installer have not option to import original Transport Tycoon Deluxe files / Original Transport Tycoon Deluxe files not detected https://git.io/JOJ9k
12:15:16 <DorpsGek> [OpenTTD/OpenTTD] FLHerne opened pull request #8996: Fix #8987: Display 'saved with patchpack' error when failing to load JGRPP savegames https://git.io/JOJ98
12:15:44 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on issue #8995: Installer have not option to import original Transport Tycoon Deluxe files / Original Transport Tycoon Deluxe files not detected https://git.io/JOJ9k
12:16:06 <nielsm> I guess that feature of the installer got lost in the switch to cpack
12:16:19 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8987: Loading JGRPP savegames in official OpenTTD gives "too new" instead of "modified version" error https://git.io/JOfhZ
12:18:59 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8996: Fix #8987: Display 'saved with patchpack' error when failing to load JGRPP savegames https://git.io/JOJ9j
12:19:16 <DorpsGek> [OpenTTD/OpenTTD] FLHerne updated pull request #8996: Fix #8987: Display 'saved with patchpack' error when failing to load JGRPP savegames https://git.io/JOJ98
12:20:01 <TrueBrain> was pretty deliberate to no longer do that
12:21:21 <FLHerne> TrueBrain: It's not clear to me if we should recommend that 0x8000 means "patchpack" or specifically "patchpack using JGR's extended versioning"
12:21:52 <FLHerne> I mean, to some extent the latter solves our versioning problems
12:22:04 <TrueBrain> FLHerne: this comes back to that it is not sufficiently described what is expected here :D
12:22:27 <FLHerne> because it already has a system for advertising which savegame changes are actually made
12:22:27 <TrueBrain> which is a bit the core of my issue with this :)
12:22:56 <FLHerne> Well, you wrote the original comment :p
12:23:13 <FLHerne> So I assume you know what it was meant to mean...
12:23:17 <TrueBrain> who said I was perfect? :D
12:23:36 <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJaW
12:23:59 <TrueBrain> I just know it is a bit messy atm .. and I know someone already looked into backporting the extended versioning stuff from JGRPP
12:24:27 <TrueBrain> my suggestion would be for someone to work that out a bit more and see if we can find some common ground to work from .. as mentioning WHICH patchpack created the savegame would be a huge boost in user-friendliness, in my book :P
12:24:42 <TrueBrain> always the balance between "the easy solution for now" and the "proper solution for the future"
12:24:46 <TrueBrain> dunno how to balance it here :P
12:25:25 <TrueBrain> either way, if you want this to be about the JGRPP extended savegame stuff, please also update that in the comment above. As it is not only mentioning another chunk :D
12:25:41 <TrueBrain> or at least sync the language, let me put it that way
12:25:52 <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJHj
12:26:15 <TrueBrain> SL_JGRPP_EXTENDED_SAVEGAME ?
12:26:32 <TrueBrain> I dunno FLHerne :) Maybe others have ideas here :)
12:26:47 <TrueBrain> to be clear, I have nothing against your solution; just silently hoping a bit someone picks this up for realz :D
12:27:09 <DorpsGek> [OpenTTD/OpenTTD] FLHerne commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJQq
12:28:38 <FLHerne> TrueBrain: I thought that was what your comment said already :p
12:29:19 <TrueBrain> FLHerne: well, we use different language, so that is not helping :)
12:29:28 <FLHerne> please follow a system like JGRPP, where blah and and the true version is stored in its own chunk with feature toggles.
12:29:32 <frosch123> FLHerne: TrueBrain: i think you can read the gamelog (even for otherwise unknown savegame versions) to give a versionstring for the openttd that saved a savegame
12:29:41 <FLHerne> ^isn't supposed to recommend that patchpack makers use it?
12:29:48 <FLHerne> I like your proposed name better
12:30:29 <TrueBrain> I would say, sync up the comments to read the same, and I am not against it :)
12:30:41 <frosch123> if the message would say "savegame was created with XYZ and can not be loaded with this version of OpentTD" you can avoid all that newer/older trouble
12:31:04 <TrueBrain> frosch123: sounds like a very solid solution
12:31:25 <TrueBrain> "savegame was created with OpenTTD 1.11 and cannot be loaded .."
12:31:28 <TrueBrain> if you load the game in 1.10
12:31:35 <TrueBrain> would help a lot of people
12:34:01 <DorpsGek> [OpenTTD/OpenTTD] FLHerne commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJ7f
12:34:12 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8995: Installer have not option to import original Transport Tycoon Deluxe files / Original Transport Tycoon Deluxe files not detected https://git.io/JOJ9k
12:34:15 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #8995: Installer have not option to import original Transport Tycoon Deluxe files / Original Transport Tycoon Deluxe files not detected https://git.io/JOJ9k
12:36:42 <DorpsGek> [OpenTTD/OpenTTD] FLHerne commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJ7c
12:38:19 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJ7D
12:39:21 <TrueBrain> sorry FLHerne for confusing you; I should have used a bit more words on GitHub :)
12:40:01 *** virtualrandomnumber has joined #openttd
12:40:57 *** virtualrandomnumber has quit IRC ()
12:40:58 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJ5O
12:44:21 <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #8992: Fix #8956: Industry disaster news messages showed the wrong location https://git.io/JOJaW
12:44:24 <DorpsGek> [OpenTTD/OpenTTD] LordAro closed issue #8956: Disasters: wrong location shown in factory/refinery destroyed message https://git.io/JY7L2
12:48:05 *** FLHerne has quit IRC (Ping timeout: 480 seconds)
12:49:33 *** FLHerne has joined #openttd
12:52:23 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJFG
12:54:08 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8995: Installer have not option to import original Transport Tycoon Deluxe files / Original Transport Tycoon Deluxe files not detected https://git.io/JOJ9k
12:54:42 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #8996: Fix #8987: Display 'modified version' error when failing to load JGRPP savegames https://git.io/JOJFP
12:55:26 *** HerzogDeXtEr has joined #openttd
12:56:10 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on issue #8987: Loading JGRPP savegames in official OpenTTD gives "too new" instead of "modified version" error https://git.io/JOfhZ
12:56:25 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJbq
12:56:39 <TrueBrain> michi_cc: ^^ maybe you know the answer to my mentioned limitation :)
12:57:49 <peter1138> Right, breakfast/lunch time.
12:58:35 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8987: Loading JGRPP savegames in official OpenTTD gives "too new" instead of "modified version" error https://git.io/JOfhZ
12:58:55 <LordAro> peter1138: seems you were right about "50", haven't gone anywhere, so i'll do a "quick" 50k loop later
12:58:58 <LordAro> bike is clean now though
13:04:34 <peter1138> I'm a bad influence.
13:05:31 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJN2
13:06:17 <LordAro> peter1138: nah, i should have started earlier
13:06:34 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
13:07:14 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJNS
13:08:01 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJNA
13:08:43 <TrueBrain> michi_cc: "next present"?
13:08:46 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJAT
13:09:07 <michi_cc> TrueBrain: Whatever OS call causes a buffer flip/blit to front buffer.
13:09:23 <TrueBrain> currently we only set it during AllocateContext
13:09:30 <TrueBrain> so asking for a game restart is far easier to implement :P
13:09:36 <michi_cc> SwapBuffers for Windows
13:09:37 <TrueBrain> mostly as I don't know what SDL is doing :D
13:09:48 <TrueBrain> otherwise we need runtime code to check if the setting has changed
13:09:54 <TrueBrain> not sure if that is worth it :)
13:11:38 <peter1138> It bugs me that I never fixed this. There are two dropdown lists off the main toolbar that have blank spaces in, except the blank space is provided by a string (in a sequential list of strings) that is simply blank.
13:12:00 <milek7> vsync should be enabled by default imo
13:12:47 <peter1138> In other places we explicitly build the dropdown list and use a different dropdown list item which draws a clear clear unselectable separator.
13:13:00 <michi_cc> But I think OTTD takeing more than 16ms for the draw part would probably be a bigger problem anyway.
13:14:03 <TrueBrain> oof, game options are not aligned
13:14:09 <TrueBrain> which shows when adding yet another option
13:14:13 <TrueBrain> labels and buttons are .. off :D
13:15:19 <peter1138> milek7, vsync slows the game down substantially for me.
13:16:17 <TrueBrain> I opened some AAA games, none have vsync enabled when not in fullscreen
13:16:23 <peter1138> Oh that was unfair. Fullscreen on left monitor != fullscreen on right monitor.
13:16:46 <peter1138> 1920x1080 vs 2560x1440!
13:18:06 *** Progman has quit IRC (Remote host closed the connection)
13:18:22 <peter1138> Still substantially slower.
13:19:10 <peter1138> Would be nice if the framerate window could show more history.
13:19:45 <TrueBrain> you know where the source code is :P :P :P :D
13:19:49 <TrueBrain> (And I agree with you)
13:20:01 <peter1138> It's almost like the gameloop gets stuck at 60Hz, even though it's not.
13:20:25 <peter1138> I don't see any massively different reading to explain the slowdown :/
13:22:30 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJbq
13:22:32 <peter1138> Game is paused, graphics frame rate is down to 51fps
13:22:47 <peter1138> Open "View output frame rate graph" and it goes back up to 60fps.
13:23:06 <peter1138> (Or any of the graphs)
13:23:23 <peter1138> I wonder if my GPU is throttling based on workload.
13:27:18 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
13:29:06 <milek7> maybe it shouldn't try to sleep with vsync enabled
13:30:32 <milek7> maybe just add 'vsync' as another option to 'Display refresh rate'?
13:30:54 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJbq
13:31:12 *** Gustavo6046 has joined #openttd
13:31:35 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #8986: Fix: Missing 'Town names:' colon in map gen GUI https://git.io/JOJj9
13:32:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOJjN
13:34:16 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8986: Fix: Missing 'Town names:' colon in map gen GUI https://git.io/JOUea
13:34:29 <peter1138> I'll try commenting out std::thread_thread::sleep_for()
13:36:42 <TrueBrain> too bad WSL doesn't proxy OpenGL properly (yet), so I cannot easily debug this with you :)
13:38:21 *** jottyfan has joined #openttd
13:38:31 <milek7> vsync doesn't necessarily block on swapbuffers, it might block on any future GL call
13:38:42 <milek7> maybe it's blocked by GL while holding lock or something
13:49:54 <TrueBrain> peter1138: interesting ...
13:50:38 <peter1138> They are a security-specialist type though, so a little biased.
13:50:57 <TrueBrain> yeah, initial I also considered UPnP a security issue
13:50:59 <TrueBrain> but some more thinking
13:51:12 <TrueBrain> thinking you can be "secure" by not allowing incoming connections
13:51:18 <TrueBrain> is not really understanding how TCP works
13:51:23 <TrueBrain> which .. NAT punching proofs just fine
13:51:33 <TrueBrain> I mean .. I send 2 connects to each other, and it magically makes a two-sided stream
13:51:59 <TrueBrain> but my initial response too came from the feeling: if you can connect from outside to inside, that is a security risk
13:52:01 <TrueBrain> but really .. it is not
13:52:04 <TrueBrain> talking to the Internet is :D
13:53:20 <TrueBrain> and instead of finishing my lobby code, I am looking through bugs .. these bugs are much harder than normal :(
13:53:42 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JOUJE
13:54:41 <milek7> UPnP is almost useless
13:55:16 <milek7> even if router would support it, it will fail at stacked NATs anyway
14:00:54 <peter1138> Hmm, so the colour of the Graphics frame rate value in the Frame rate window is dependent on the Display refresh rate value, regardless of the VSync setting.
14:01:25 <milek7> in my experience typical setup is: operator throws everybody behind NAT -> router does another NAT for local network -> user is annoyed by weak wifi signal in another room, and buys another router (which is configured as router by default) -> they have triple-NAT :P
14:03:39 <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOUTk
14:04:12 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8972: I can't build tram tracks in an existing road tunnel without demolishing the tunnel https://git.io/JYpXL
14:07:13 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8997: Feature: allow a toggle to enable/disable vsync https://git.io/JOUTi
14:07:17 <TrueBrain> nice catch peter1138
14:08:48 <TrueBrain> just no clue how to solve it properly :)
14:09:59 <FLHerne> milek7: At least in the UK, ISP routers come with UPnP enabled and I've never seen carrier-level NAT
14:11:46 <FLHerne> Certainly that's the case for my last two internet plans, with the largest and third-largest ISPs
14:12:00 <FLHerne> between them that's over half the UK market
14:12:53 <FLHerne> TrueBrain: yeah, the most securee router is an empty box with ports on each end
14:15:45 <milek7> curious why they didn't use truewiki :P
14:16:42 <TrueBrain> tell them they should!
14:16:44 <FLHerne> TrueBrain: That's less secure, you can plug it in backwards :p
14:17:15 <TrueBrain> still only allows a one-way connection :P
14:17:38 <TrueBrain> it btw is a serious awesome product, especially as it reproduces a two-way protocol for many protocols on both sides
14:20:56 <TrueBrain> funny, minimap is not scaling to Interface zoom :D
14:21:00 <TrueBrain> did not expect that :P
14:23:04 <Rubidium> I remember the minimap not being small enough...
14:23:47 <TrueBrain> "In my days ..." :D :D
14:23:49 *** supermop_Home_ has joined #openttd
14:27:02 <Rubidium> it was introduced in the current major version, so it's not that old ;)
14:27:20 <DorpsGek> [OpenTTD/OpenTTD] JGRennison updated pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOJrr
14:27:59 <frosch123> eints is sooo mean...
14:28:32 <frosch123> there is a default filesize limit for uploading language files, which is set to 100k (too low for ottd)
14:28:52 <frosch123> eints can warn about that, but it fails hard if truncating at 100k results in invalid utf-8
14:29:09 <frosch123> turns out romanian.txt just had a utf-8 sequence at 100k, so it failed hard
14:29:18 <frosch123> and romanian is what i tried to upload...
14:33:35 <TrueBrain> Lol .. what are the odds :p
14:56:11 *** Progman has joined #openttd
15:20:25 <EER> If only I were smart enough to run the pre-commit checks BEFORE sending a PR -_-'
15:23:59 <DorpsGek> [OpenTTD/OpenTTD] frosch123 updated pull request #8936: Add: Plural support for Romanian translations https://git.io/JYX84
15:24:52 <DorpsGek> - Update: Translations from eints (by translators)
15:32:44 *** snail_UES_ has joined #openttd
15:33:31 <DorpsGek> [OpenTTD/OpenTTD] frosch123 merged pull request #8936: Add: Plural support for Romanian translations https://git.io/JYX84
15:35:12 <frosch123> lol, noone checked the build logs before, everyone stopped at the commit-checker :p
15:35:45 <frosch123> ah, nevermind, it was not noticeable before
15:39:11 *** Speedy` has quit IRC (Read error: No route to host)
15:41:27 <frosch123> no, i can't read :(
15:44:40 <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on issue #8981: Train never finishes entering depot, blocks itself from leaving depot. https://git.io/JOv93
15:48:41 <DorpsGek> - Update: Translations from eints (by translators)
15:51:51 *** andythenorth has joined #openttd
15:56:10 *** Beerbelott has joined #openttd
15:58:35 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
16:00:56 <frosch123> does someone want to save the nightly by approving #8999? :)
16:00:59 <Beerbelott> Hello. I am experiencing errors related to communication with the master server from my dedicated servers. Is this something known?
16:04:11 *** Wormnest has joined #openttd
16:05:59 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8995: Installer have not option to import original Transport Tycoon Deluxe files / Original Transport Tycoon Deluxe files not detected https://git.io/JOJ9k
16:11:21 <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOU0s
16:13:37 <peter1138> Beerbelott, nobody else has mentioned anything.
16:22:16 <Beerbelott> peter1138 From a player there, it seems to happen every couple months
16:23:03 <peter1138> Okay. What happens every couple of months?
16:23:04 <Beerbelott> 'cause actually it's not only the MS, but the whole connectivity
16:23:09 <Beerbelott> either IPv4 or IPv6
16:23:51 <DorpsGek> [OpenTTD/eints] frosch123 opened pull request #41: Improve error checking when uploading translations https://git.io/JOUuB
16:25:23 <glx> Beerbelott: is there a message with the disconnect ?
16:26:13 <glx> anyway master server is only used for the server list, once connected to a server it's not used
16:26:18 <TrueBrain> frosch123: why is there a max filesize in the first place?
16:26:59 <frosch123> probably so you cannot OOM eints by uploading 300 GB
16:27:39 <TrueBrain> doesn't bottle already limit that? At least, assuming you run any proxy in front of this (which you should) already limits the max :P
16:27:45 <TrueBrain> just a bit uncommon :)
16:28:27 <DorpsGek> [OpenTTD/eints] TrueBrain approved pull request #41: Improve error checking when uploading translations https://git.io/JOUzY
16:28:35 *** WormnestAndroid has quit IRC (Remote host closed the connection)
16:28:59 *** WormnestAndroid has joined #openttd
16:29:00 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
16:30:07 *** WormnestAndroid has joined #openttd
16:31:46 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #8993: Fix thread safety issues around the cursor and cursor sprites in the OpenGL backends https://git.io/JOJrr
16:31:53 <Beerbelott> glx It says nothing comes fro mthe server and then I drop. The server logs merely people leave
16:32:38 <Rubidium> I wonder whether OOM-ing eints is the problem, or pushing 300 GB worth of language file to the git repository. That would probably be painful
16:32:42 <glx> people leaving your server has nothing to do with master server
16:32:56 <Beerbelott> I ma running a test server on which I test the connection every second
16:33:48 <Beerbelott> glx Lemme clarify things: connectivity problem with the MS has just shed a light on something bigger. Connectivity drops from time to time.
16:34:15 <TrueBrain> Beerbelott: if players are being dropped and MS connection is failing, it sounds like you have a connection problem
16:34:16 <glx> I think it's on your side only
16:34:18 <Beerbelott> Someone playing on the serevr said it happend every couple moths in-game or so. This is new.
16:35:48 <Beerbelott> TrueBrain Yup. Anything done to the network stack for 1.11.0 to investigate?
16:35:58 <TrueBrain> it suggests it is your system
16:36:24 <nielsm> I don't think there were any changes to networking on 1.11 at all?
16:37:50 <nielsm> okay there were a few minor fixes, mostly UI related things
16:37:52 <andythenorth> eh what am I hallucinating?
16:37:55 <andythenorth> Rubidium is in the channel :o
16:38:08 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
16:38:44 <TrueBrain> I am slightly disappointed he didn't make a PR yet :P
16:39:11 *** WormnestAndroid has joined #openttd
16:39:48 <andythenorth> I spend a few hours in discord and THIS is what happens :P
16:39:54 *** sla_ro|master has joined #openttd
16:40:00 <andythenorth> if dalestan shows up, then what?
16:41:02 <TrueBrain> its funny, everyone has a different name to use for this
16:41:15 <TrueBrain> I read Celestar, Bjarni en now Dalestan today
16:41:28 *** Flygon has quit IRC (Read error: Connection reset by peer)
16:42:06 <frosch123> TrueBrain: yesterday you spent way too little time in how to pronounce rondje
16:42:09 *** jottyfan has quit IRC (Quit: jottyfan)
16:42:22 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
16:42:22 <frosch123> you should have dragged it for 5 minutes, and make everyone say it correctly :)
16:42:36 <TrueBrain> I was amazed how good it sounded already
16:42:43 <TrueBrain> I cannot punish that
16:42:44 <Rubidium> and don't forget misremembering r10k
16:43:00 <TrueBrain> I was afraid of the "corrections" :P
16:43:38 <TrueBrain> no, but I told everyone that pretty clearly :)
16:44:03 <Rubidium> 10k was somewhere in Leiden, 20k might have needed to be more westerly because of our friend from the real mountain
16:44:50 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
16:45:07 <TrueBrain> Rubidium: was the 10k cake in Leiden?
16:45:12 <TrueBrain> I only remember boekabart making a cake
16:45:16 <TrueBrain> which made me laugh :P
16:45:27 <TrueBrain> ugh ... shows how good my memory was :D
16:45:53 <TrueBrain> we went to Germany twice
16:45:55 <Eddi|zuHause> i wasn't in leiden, and all the partys i was at were in braunschweig
16:46:02 <TrueBrain> once we even flew in some crazy person
16:46:30 <frosch123> and made them quite smoking
16:46:56 <TrueBrain> so, okay, we told a lie on stream
16:47:25 <TrueBrain> I will just let Timberwolf tell any part of OpenTTD's history
16:48:23 <TrueBrain> but more important, we should bribe pm again to host a party when ever we can do that again :)
16:48:40 <frosch123> we can just show up :p
16:52:57 <frosch123> does someone know the difference between "हिन्दी" and "हिंदी" ?
16:53:18 <frosch123> it's probably political, so i cannot just pick one
16:54:03 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
16:55:11 <andythenorth> I always pick dalestan because newgrf :)
16:55:33 <andythenorth> and because I've never seen anyone walk the line so close up to abuse of community members without quite crossing it :P
16:56:38 <glx> frosch123: indeed google translate doesn't help here :)
16:59:54 <peter1138> train_cmd.cpp:2685 is where it marks a train as stuck when entering a depot. So I wonder what cases that section fixes.
17:00:24 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
17:01:41 <frosch123> glx: hmm, i had both from wiki pages, but now i no longer find the other one
17:02:21 <Beerbelott> Might there be limits per IP address enforced on the MS?
17:03:52 <glx> as said, MS only updates the server list, it has nothing to do with player leaving
17:04:40 <Beerbelott> I'm talking about a TCP connection to the game port failing while server advertisement is not compelte
17:04:48 <glx> if player leave when you MS can't reply, it's your connection
17:05:23 <nielsm> Beerbelott that log just shows me that your server loses internet connectivity
17:05:33 <Beerbelott> nielsm How do I do that? I'm talking server advertisement fro mthe game
17:05:38 <TrueBrain> Beerbelott: no, server advertisement is normally completely separate from the rest of the game. So advertising should have zero influence on the game itself
17:05:58 <TrueBrain> it is an "extra", basically :)
17:06:02 <nielsm> connected clients don't care at all about your advertisement to the master server
17:06:28 <nielsm> and the server doesn't care about its advertisement to the master server failing, it can still continue running and talk to existing connections and accept new connections all the same
17:06:36 <TrueBrain> but, there can be other (local) relations between advertisement starting and clients being dropped
17:06:36 <Beerbelott> nielsm At the same time I got a process in the container probing on ::1 3982 which tells me nothing aswers. Are we blaming the localhost interface now? :D
17:06:56 <nielsm> but if your master server advertisement fails and at a near point in time all clients also drop, that kind of points to your internet connectivity failing
17:07:14 <nielsm> Beerbelott I don't care that you can talk to other processes running on the same hardware
17:07:27 <TrueBrain> Beerbelott: this connect to ::1, is that done from within the Docker context?
17:07:31 <Beerbelott> nielsm ::1 3982 fails at times mentioned in the log. This is not an infra problem.
17:07:32 <nielsm> have a host somewhere else on the internet talk to your hosting computer
17:07:33 <TrueBrain> so not from the Docker host, but inside the container?
17:07:53 <Beerbelott> TrueBrain Yup. Inside the container, ::1 3982. Cannot do closer.
17:08:16 <TrueBrain> where I have seen similar issues, is when the docker messes up iptables
17:09:01 <TrueBrain> weren't you also the one reporting shutdown issues?
17:09:02 <Beerbelott> No external routing there. Indise a container, talking to a process there on the loopback interface.
17:09:21 <Beerbelott> Shutdown? Doesn't ring a bell
17:10:00 <TrueBrain> what is your Dockerfile?
17:10:07 <Beerbelott> Is there a way to do server advertisement manually, deactivating it inside the cam's conf?
17:10:12 <TrueBrain> I might be able to reproduce it that way
17:10:22 <TrueBrain> it is either enabled or disabled
17:10:26 <TrueBrain> there is nothing in between
17:10:43 <andythenorth> frosch123 crude test of measuring time-since-loaded (not very accurately)
17:10:44 <TrueBrain> and really, I cannot imagine the advertisement being the issue, it is just what makes the issue visible, I think
17:11:41 <nielsm> yes the server advertisement failing is a symptom, not a cause
17:11:54 <Rubidium> I think there is a problem with the IPv6 firewall configuration
17:12:05 <TrueBrain> the IPv6 master server announcement took 20 seconds
17:12:13 <nielsm> as I said, make something on the internet, a different computer on a different network, that continually talks to your hosting computer or ideally the openttd process inside your docket, and log the results
17:12:47 <nielsm> it will almost certainly show that your openttd process stops responding to packets at intervals
17:14:01 <TrueBrain> yeah, something odd with UDP going on .. the master servers sends one packet, waits 5 seconds, sends another
17:14:01 <Rubidium> getting the session key works for both IPv4 and IPv6, so the IPv6 routing should be okay (unless it is intermittently wrong; are you hosted by OVH in Frankfurt by any chance?), but I'm not seeing the UDP queries from the master server for IPv6 arrive, which means the IPv6 part of the advertising fails even though the IPv4 part might have succeeded
17:14:04 <TrueBrain> but 3 of them arrive at the same time
17:14:22 <Beerbelott> nielsm I showed that a process probing ::1 on TCP/3982 lost connectivity during the MS announcement
17:14:26 <TrueBrain> Rubidium: those 3 IPv6s are the master server queries :)
17:14:35 <nielsm> Beerbelott I don't care about your ::1
17:14:43 <Beerbelott> nielsm I can see that
17:15:17 <Beerbelott> Knowing that TCP/3982 listens on all interfaces, it means the game port doesn't answer. Cue disconnections?
17:15:28 <peter1138> Hmm, is there an easy test for a train entering (rather than leaving) a depot?
17:15:58 <TrueBrain> Beerbelott: but again, what Dockerfile are you running? Allows me to look with you
17:16:09 <TrueBrain> as currently you two are running circles :)
17:17:43 <Beerbelott> TrueBrain WI'm on it already: we'll tackle that in DM (o:
17:22:42 <TrueBrain> okay, so what is really odd
17:22:50 <TrueBrain> is the delay between receiving keys via IPv4 and IPv6
17:22:56 <TrueBrain> that should be near instant at the same time
17:23:57 <TrueBrain> so I wonder if there is something suspending the OpenTTD process ..
17:24:01 <TrueBrain> an AV of some kind?
17:24:25 <peter1138> Blocking on the socket?
17:24:41 <TrueBrain> you also see that the IPv4 queries from other clients stop for 20 seconds
17:24:53 <TrueBrain> so from this brief log, it seems like the server is suspended for 20 seconds :P
17:25:06 <TrueBrain> but the fact that it does catch up with 3 IPv6 packets are are 100% sure never send on the same time
17:25:11 <TrueBrain> suggests it is something weird
17:25:22 <TrueBrain> as they are UDP, they should have been dropped
17:26:55 <TrueBrain> so the question is .. does OpenTTD go zombie for 20 seconds
17:27:01 <TrueBrain> or is some external influence causing that
17:27:15 <Beerbelott> I had 2 probes: one on the host probing the IPv6 address, one in the container probing ::1
17:27:33 <Beerbelott> Both showed the exact same behaviour
17:27:37 <TrueBrain> what if you probe from inside the container to something external?
17:28:19 <Beerbelott> Gimme a couple minutes
17:28:33 <TrueBrain> in the meantime, what OS?
17:28:58 <TrueBrain> (the Docker host :P)
17:29:41 <TrueBrain> and can you do a "ps aux | grep openttd" in those 20 seconds?
17:29:52 <nielsm> okay yes so, if network_udp.cpp:564 hangs (send packet on UDP master socket while holding _network_udp_mutex), then that can cause NetworkBackgroundUDPLoop() to hang (line network_udp.cpp line 649), and since that is called from NetworkBackgroundLoop() (network.cpp line 840) that can cause all network send/receive to block
17:30:02 <nielsm> I guess nobody is ever expecting a UDP send operation to hang
17:30:18 <TrueBrain> networking with a VERY small send-buffer?
17:30:19 <Beerbelott> Probed ipv6.google.com on TCP/443: no sign of trouble during the exact same stages (MS advertisement after map generation)
17:30:22 <nielsm> but maybe if the kernel network stack is overloaded and outgoing buffers are full or something?
17:31:28 <Beerbelott> TrueBrain I did even better that monitoring processes, I monitored listening sockets, and they did not withdraw at any point
17:32:23 <TrueBrain> but, what OS are you running?
17:33:45 <DorpsGek> [OpenTTD/OpenTTD] PeterN opened pull request #9000: Fix [#8981]: Don't attempt to re-reserve path if already entering/entered depot. https://git.io/JOU6O
17:36:51 <peter1138> Oh, dodgy commit message :p
17:37:33 <TrueBrain> Beerbelott: from within the Docker context, what does "sysctl net.ipv4.udp_rmem_min" and "sysctl net.ipv4.udp_wmem_min" return?
17:37:34 <DorpsGek> [OpenTTD/OpenTTD] PeterN updated pull request #9000: Fix [#8981]: Don't attempt to re-reserve path if already entering/entered depot. https://git.io/JOU6O
17:38:23 <Beerbelott> TrueBrain Debian-like
17:39:09 <Beerbelott> net.ipv4.udp_rmem_min = 4096; net.ipv4.udp_wmem_min = 4096
17:39:33 <TrueBrain> and "sysctl net.ipv4.udp_mem" ?
17:40:21 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
17:40:58 <Beerbelott> TrueBrain I don't have that
17:41:00 <TrueBrain> LordAro: exactly what I initially tried, but failed horrible on non-english :D
17:41:11 <TrueBrain> Beerbelott: ah, kernel differences .. euh .. what was the generic one ..
17:41:36 <Beerbelott> Searching for that ^^
17:41:58 <TrueBrain> this is too long ago for me :P
17:42:24 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
17:43:56 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
17:43:59 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
17:44:39 <DorpsGek> [OpenTTD/OpenTTD] Limyx826 commented on issue #8988: The current fonts missing some of the characters used in the texts for this language https://git.io/JOJG7
17:44:53 <Beerbelott> TrueBrain Your query is valid... It just isn't available from the container. Might be related to capabilities. From the host I got net.ipv4.udp_mem = 187596 250129 375192
17:45:15 <TrueBrain> a bit on the small side, but shouldn't be any real issue
17:45:28 <Beerbelott> No tuning involved
17:45:41 <TrueBrain> really odd, that is mostly what I can say about this :)
17:46:13 <Beerbelott> I guess I'll stop most of the servers and only run with one.
17:47:02 <TrueBrain> if it doesn't happen with fewer servers
17:47:11 <TrueBrain> that does help confirm what nielsm mentioned
17:47:30 <Beerbelott> TrueBrain I would say evenmore: that's odd :P
17:48:14 <Beerbelott> The thing with network stack overloaded and buffers?
17:48:27 <DorpsGek> [OpenTTD/OpenTTD] nielsmh opened pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUPv
17:48:55 <Beerbelott> Ah it seems I may be behind this 9001, jusging from its title :P
17:49:22 <nielsm> can we make "over 9000" jokes now?
17:49:36 <TrueBrain> the idea is that it avoids "hanging" the game .. but I still wonder how the hell a send() could be blocking
17:49:44 <TrueBrain> send() -> UDP send()
17:50:08 <nielsm> yeah the only way a UDP send could block is if the OS has something really weird going on
17:50:22 <nielsm> usually everyone thinks of UDP as instant fire and forget
17:50:47 <TrueBrain> so that PR might solve clients being dropped .. for sure it will still be terribly annoying for the host :D
17:50:51 <nielsm> and even if it blocked, it would be for a fractions of a second
17:52:44 <Rubidium> Beerbelott: can you run the same test as before, but with "openttd -d sprite=3 -D" instead of just "openttd -D" (put the -d sprite=3 before the -D)
17:52:46 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUPi
17:53:17 <Rubidium> that should show whether the GameLoop gets blocked or not
17:54:24 <peter1138> nielsm, won't it still block on the UDP send?
17:54:54 <Rubidium> nielsm: how does this help if any of the replies of a query get blocked?
17:55:13 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUP5
17:55:38 <nielsm> yes it will still block the UDP send, but at least it won't block the main loop and lose clients connected to the game
17:56:00 <TrueBrain> well, it only solves the issue of send() done in the MS advertisement thread
17:56:12 <TrueBrain> if an UDP send() triggered by the mainloop hangs ... :)
17:56:58 <TrueBrain> sadly, several Receive_NNN functions trigger a send()
17:57:29 <TrueBrain> so I guess it depends on what send() blocks, if this solves Beerbelott 's issue :)
17:58:42 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUX2
17:59:59 <TrueBrain> I am lost in JGR's reply :P
18:00:30 <TrueBrain> anyway, in the upside, if I ever get time to finish my STUN/lobby solution, we will no longer use UDP for MS :P
18:01:48 <Rubidium> it was win9x that was having problems with making/accepting a lot of TCP connections, right? So maybe just go to TCP
18:02:05 <TrueBrain> that is the plan, yes
18:02:08 <TrueBrain> which has .. some challenges
18:02:15 <TrueBrain> but one I am trying to face :)
18:02:33 <DorpsGek> [OpenTTD/OpenTTD] nielsmh updated pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUPv
18:02:34 <TrueBrain> (having 10k open TCP connections is not trivial for a server to handle properly :P)
18:04:00 <DorpsGek> [OpenTTD/OpenTTD] nielsmh updated pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUPv
18:04:10 <TrueBrain> nielsm: awh, I was about to comment you liked spam :)
18:04:41 <nielsm> also changed to modulo 32 instead to allow the compiler to make it FAST
18:05:04 <nielsm> (and closer to 1 second between spam)
18:05:11 <TrueBrain> which is a joy .. as a lot of network code hasn't been touched how you left it :P
18:07:22 <Rubidium> that's a lie, I've seen a lot of changes
18:07:33 <Rubidium> (like NULL -> nullptr)
18:07:46 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOU1D
18:08:09 <TrueBrain> Rubidium: yes darling, you are correct :P
18:09:53 *** jottyfan has joined #openttd
18:12:25 <DorpsGek> [OpenTTD/OpenTTD] stormcone commented on pull request #8982: Feature: text filter for the industry directory https://git.io/JOUMB
18:12:31 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUM0
18:13:05 <TrueBrain> we should help Beerbelott test this change, see if that really is what is causing his problem :)
18:13:25 <TrueBrain> the communication with the MS is so much lower than any of the other communication
18:13:39 <TrueBrain> it is hard to imagine that send() being the blocker :(
18:13:48 <Rubidium> TrueBrain: looks tricky to get functioning and especially tested
18:14:20 <TrueBrain> Rubidium: so far, 5 out of the 6 people that tested it, could NAT punch
18:14:28 <TrueBrain> but I guess that is 5 people more than without
18:14:32 <nielsm> I am in fact running a modified server right now (my "1.11.0b" patchset) which is accepting clients normally
18:14:45 <nielsm> it's not too hard setting up at least, if you have a build environment
18:20:53 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8966: Feature: News/advice setting to warn if no depot order in vehicle schedule https://git.io/JOUDa
18:23:19 <peter1138> I imagine that if that UDP send blocks, something else could easily block later anyway.
18:23:36 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
18:24:23 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9001: Fix: If a thread is blocking on UDP send, it can block the entire main loop https://git.io/JOUDA
18:25:31 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
18:25:34 <DorpsGek> [OpenTTD/OpenTTD] perezdidac closed issue #8983: Editbox losing focus triggers Editbox's OnEditboxChanged https://git.io/JOf39
18:33:56 <Beerbelott> TrueBrain I am going to run that openttd -d sprite=3 -D test now
18:34:32 <DorpsGek> [OpenTTD/bananas-frontend-cli] erenes updated pull request #22: Fix: Use urllib to combine urls (#5) https://git.io/JYQbM
18:37:10 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on issue #8978: OpenTTD downloading from OneDrive freezes at startup https://git.io/JYhyh
18:42:50 <Rubidium> there are no sprite compacting messages, so the GameLoop is not running at all so UDP send() blocking seems likely
18:47:27 <Beerbelott> Just an idea on one QoL improvement: Since I had to struggle through code to realise that dedicated servers were started with net=6 debug level, wouldn't it be nice to drop a log line telling about any >0 debug level set, for any facility?
18:53:58 *** didac has quit IRC (Remote host closed the connection)
18:54:01 *** gelignite has joined #openttd
18:56:19 <Beerbelott> TrueBrain Oh wow. Just realised I actually *did* some sysctl tuning on that machine. Don't recall when... years ago.
18:58:05 <Beerbelott> TrueBrain For the changed things the only ones which I guess could be of interest is a change to rp_filter and tcp_synack_retries, as they apply to incoming connections
18:58:12 <nielsm> Beerbelott: if you want to try my patch on your server I can help you build a 1.11.0 compatible version with it applied
18:59:16 <Beerbelott> nielsm I can pull than a build from a branch. It is native to the way I build OpenTTD over here. (o:
18:59:48 <nielsm> if you want clients to be on your server you will probably need to be 1.11.0 compatible ;)
18:59:53 <Beerbelott> nielsm Trying that right now and keeping you posted
19:00:10 <DorpsGek> [OpenTTD/OpenTTD] michicc opened pull request #9002: How to break cargo ageing and cargodist in one go. https://git.io/JOUQn
19:00:36 <michi_cc> andythenorth: #9002 is just for you :)
19:01:08 <frosch123> LordAro: it's over 9000!
19:01:09 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8982: Feature: text filter for the industry directory https://git.io/JOUQ8
19:01:38 <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8982: Feature: text filter for the industry directory https://git.io/JOfI5
19:02:17 <Beerbelott> Yup. Too late for a celebration. Gotta wait for 10k
19:02:32 <nielsm> if you build that it should report as 1.11.0 on the network and show internal version as 1.11.0.2
19:02:39 <Beerbelott> nielsm I know all about it. On it.
19:03:25 <Beerbelott> Based on your PR, I will pull your udp-blocking branch
19:04:12 <michi_cc> andythenorth: It's not really that crackpot, except that it might insta-kill all existing cargodist savegames.
19:05:58 <Beerbelott> nielsm: Oh I see what you did in nielsm/udp-blocking-fake1.11.0. Don't worry, I am already overriding the version number when I wanna run changes that may not be canonical (not the case in our tests)
19:06:12 <Beerbelott> I will be using udp-blocking
19:07:04 <andythenorth> michi_cc seems 'fine' o_O
19:07:27 <andythenorth> presumably we can then have newgrf station tiles that modify the station aging rate? :D
19:08:11 <nielsm> andythenorth yes but they will also need to have a capacity so only the first 200 fish that arrived can be refrigerated
19:08:15 <michi_cc> andythenorth: Yes, but that probably needs permanent storage (or something like userbits) for stations first. Doing a callback on all stations tiles might not be very performant.
19:08:52 <michi_cc> I'm hoping my pitch is convincing enough that somebody feels triggered enough to produce colourful diagrams.
19:09:38 <andythenorth> I do think YOLO explains it all nicely
19:10:59 <andythenorth> I am waiting for frosch123 to spot the deliberate errors in my gist patch :P
19:11:17 <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8794: Feature: Show rail/road/tram NewGRF name in Land Area Information window https://git.io/JthCX
19:11:20 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOUQn
19:11:29 <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8794: Feature: Show rail/road/tram NewGRF name in Land Area Information window https://git.io/JOU79
19:14:08 <LordAro> looks possibly like something to do with the script compatibility script copying rules
19:14:09 <Beerbelott> nielsm Shall I run OpenTTD with any arguments, like "-d sprite=3"?
19:16:34 <peter1138> Dinner was good, I think I'll do another cycle ride like this morning's...
19:16:46 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOU5B
19:16:52 <LordAro> peter1138: i got rained on :(
19:17:16 <peter1138> Timberwolf, no it's necessary
19:17:23 <nielsm> Beerbelott not needed, the DEBUG I added is level 0 for network
19:17:28 <andythenorth> it's about right
19:17:32 <FLHerne> Timberwolf: Probably, but it is interesting
19:17:40 <nielsm> so it will always spam you if the lock is hit
19:17:47 <FLHerne> Only thing I'd criticise is the road centrelines
19:18:00 <LordAro> Timberwolf: lol those drains
19:18:15 <FLHerne> dashes should be longer and wider-spaced, probably
19:18:37 <LordAro> Timberwolf: does it have a RHD variant? :p
19:18:38 <peter1138> More like a cattle-grid.
19:18:52 <peter1138> At least they're oriented safely.
19:18:54 <LordAro> they are about half the width of a vehicle...
19:19:47 <peter1138> Okay, can we identify the cars?
19:20:05 <peter1138> I mean in Timberwolf's graphs.
19:20:06 <andythenorth> didn't even look )
19:20:48 <peter1138> The estate looks a bit like a Volvo.
19:21:03 <peter1138> The rear is a bit too slanty.
19:21:17 <Timberwolf> The Motorail photo has (I think) a Renault 16, an original Range Rover, and then I run out of resolution.
19:21:23 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOU5d
19:23:30 <andythenorth> so motorail in Horse, I just give the wagons pax capacity, ignore realism yes?
19:23:56 <andythenorth> as we don't have signed running costs :P
19:23:59 <Timberwolf> Estate is not a Volvo. Maxi is *close*, (and excusable at that accuracy), it's the same model as the tan and golden-brown car though.
19:24:48 <andythenorth> is there a Saab!?
19:25:18 <Timberwolf> Aw, I didn't want to go through the pain of making it year-dependent but now you're tempting me...
19:25:57 <Timberwolf> Motorail pic I think Saab 99 is plausible for the car in front of the Range Rover.
19:27:26 <DorpsGek> [OpenTTD/OpenTTD] spnda opened pull request #9003: Change: Query for servers on opening multiplayer window https://git.io/JOUdX
19:27:40 <spnda> idk I opened that PR just because. Can be closed if it just is unnecessary.
19:28:17 <DorpsGek> [OpenTTD/OpenTTD] spnda updated pull request #9003: Change: Query for servers on opening multiplayer window https://git.io/JOUdX
19:28:54 <Timberwolf> From front - Triumph Dolomit, Jaguar XJ6, Austin Princess, unsure (possibly Skoda of ome kind), 2CV, unknown, unknown, Mk1 Granada.
19:29:02 <DorpsGek> [OpenTTD/OpenTTD] ldpl commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOUd7
19:29:27 <andythenorth> Timberwolf I knew there was a reason I didn't do motorail in Horse yet
19:33:13 <Beerbelott> Oh my bad... I forgot to remove the additional debug level from TrueBrain...
19:34:55 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOUFK
19:36:52 <peter1138> Yeah so is it just hitting that lock in normal situations anyway?
19:38:27 <glx> didac: could you paste the corresponding output from output tab and not the summary from error list ?
19:38:41 <glx> there should have more details
19:38:46 <nielsm> it would need to hold the lock for nearly 1 second constantly, or hit really unlucky situations
19:40:29 <andythenorth> michi_cc 9002 will be so horrible with pax :)
19:40:44 <andythenorth> it's impossible to clear stations in any big city, so the waiting will be ... brutal
19:41:00 <Wolf01> andythenorth: started a new game... got confused by food and wood :S
19:41:09 <michi_cc> andythenorth: It only ages cargo that is in-transit, not if it wasn't transported yet.
19:41:31 <Wolf01> I didn't understand why a farm would produce wood
19:42:17 <andythenorth> michi_cc (I didn't read PR code yet sorry) does it age at standard 185 rate?
19:42:20 <michi_cc> So you "just" have to make sure that transferring pax/cargo will get moved away in time. Or you need vehicles with very low ageing to still get some profit in the end.
19:42:43 <michi_cc> No, 370 for pax/mail and 740 for all other cargoes.
19:43:20 <michi_cc> And I even added a NewGRF cargo prop to change it :)
19:44:13 <andythenorth> I am hoping to _remove_ non-standard cargo aging from my vehicle sets
19:44:18 <andythenorth> it has weird effects
19:44:48 <andythenorth> it seems to show up as reduced profit on subsequent legs due to cargo being already aged
19:45:06 <andythenorth> which is...tricky...to understand and diagnose
19:45:43 <peter1138> That should happen regardless of tweaking the interval.
19:46:20 <Beerbelott> Btw, I shut down all but 1 server on the side of the test one. So conditions shall have improved drastically if that had an impact on the situation... Initial advertisement still smh fail
19:46:30 <peter1138> It's pretty much by design that if you reduce the payment due to cargo aging, then ... as the cargo is aged... the payment is... reduced...
19:49:37 <andythenorth> peter1138 yes, but messing about with vehicles aging rates gets results that are ... complicated
19:50:19 <michi_cc> It might make your slowly ageing long-distance coaches better.
19:50:35 <Beerbelott> Rubidium I can integrate that on my local code repo if you wish I give it a run?
19:50:47 <andythenorth> based on literally...days now...of testing
19:50:56 <andythenorth> I would leave cargo aging well alone for vehicles
19:52:35 <nielsm> Rubidium: maybe even add the OS socket number (file number) to the debug? to be able to distinguish between the "client", "network" and "master" sockets
19:54:41 <Beerbelott> Building & launching now
20:02:44 <Beerbelott> Ah this is the shutdown issue you mentioned earlier?
20:06:16 <TrueBrain> I think it describes the exact same issue ;)
20:09:28 <Beerbelott> MMmh nvm. master.openttd.org:3978 is trying to reach the server on TCP/3982 or UDP/3982, right?
20:10:27 <Beerbelott> peter1138 Yeah that would have been enough for the address part, but maybe not for the data amount?
20:14:19 <Rubidium> so, at :25 is says "advertising" and at :35 it says "sending 30 bytes", followed by "advertising" in the same second. The 30 bytes sending is before the actual "sendto", so it is not actually the UDP sending of the message but something that happens before that
20:15:05 <Rubidium> having said that, due to the blocked messages I assume it is already in SendPacket quite soon
20:16:39 <Rubidium> so it's probably the address resolving that takes a while
20:19:06 <Rubidium> Beerbelott: does nslookup master.openttd.org respond immediately, or does it take a while to get something?
20:21:16 <Beerbelott> No name resolution delay
20:24:47 <TrueBrain> nice find Rubidium .. so what is it doing earlier that can lock up ..
20:26:47 <Beerbelott> peter1138: Also, message has to be dispatched on a potentially blocking send ;)
20:28:37 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOUQn
20:29:50 <Rubidium> TrueBrain: that's going to require a lot of debug statements I fear, or maybe profiling?
20:32:24 <TrueBrain> yeah .. but without being able to reproduce, that is hard as fuck :P
20:32:47 <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #9000: Fix #8981: Don't attempt to re-reserve path if already entering/entered depot. https://git.io/JOUhw
20:36:33 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #8994: Fix: Adjust scrolling interval of credits to account for text line height https://git.io/JOJMJ
20:41:13 <nielsm> btw feel free to close my PR on that mutex lock if it turns out I was completely off track :)
20:47:31 <Rubidium> for now it's mostly confirming whether the delay is in the resolve
20:47:49 <TrueBrain> wauw michi_cc , I see someone got out of bed on the wrong side .. lol .. I hope you don't take it personally, as I really like that you are just trying things to see how it looks / goes :D
20:47:51 <TrueBrain> that is really cool :)
20:53:12 <_dp_> if you mean me ofc I'm not gonna be happy about pr that destroys years of work on balancing and play testing game modes for no apparent reason and no analysis whatsoever
20:53:41 <TrueBrain> yeah, but the way you worded that ..... you could be a bit more friendly :)
20:54:49 <glx> so yes it fails earlier with assert enabled
20:55:13 <TrueBrain> so happy you make those backtraced readable for us glx :D
20:55:19 <TrueBrain> I tried it myself the other day ..
20:55:23 <TrueBrain> took me ... over an hour :P
20:55:28 <_dp_> Xaroth, it worked fine for about a decade
20:56:22 <_dp_> the "less" I'm talking about 99.9% of players aren't even good enough to notice :p
20:56:35 <glx> it's easy, get the exe, the pdb and the dmp in the same place, checkout right source somewhere, open dmp
20:57:24 *** snail_UES_ has quit IRC (Quit: snail_UES_)
20:57:24 <TrueBrain> with opengl, I first had to find the right DLLs :P
20:58:09 *** Samu has quit IRC (Quit: Leaving)
20:58:18 <glx> VS download dlls by itself
20:58:25 <TrueBrain> only if it can find them :P
20:58:39 <TrueBrain> nvidia dlls are not part of that :D
20:59:25 <TrueBrain> either way, really appreciate you doing it :)
20:59:43 <peter1138> Yeah, I don't think VS Code has any of those features.
21:00:58 <glx> vscode can probably debug locally
21:01:09 <glx> vs can even debug on remote
21:01:21 <TrueBrain> I used MSVC for this :D
21:02:09 <Beerbelott> Rubidium I will do that a tad later tonight
21:02:24 <glx> and I finally found the right intellisense settings for mingw
21:03:38 <glx> they are wrong in the default mingw configuration for cmake
21:04:33 <FLHerne> How do I tell if OTTD is using OpenGL?
21:05:13 <TrueBrain> Game Options, Hardware Acceleration
21:05:15 <TrueBrain> if it is one, it is on
21:05:50 <nielsm> nah if you have forced the regular video driver via -v option or config file that option does nothing
21:06:03 <nielsm> same if you have forced an opengl variant video driver, it also does nothing
21:07:43 <Beerbelott> Rubidium btw patch line 11 had a space. Git complained (o:
21:08:58 <_dp_> TrueBrain, I think I was friendly enough for how I feel about this kind of shit
21:09:09 <TrueBrain> that doesn't make it friendly for people to try stuff
21:09:14 <TrueBrain> you can also just say you dislike it and state your reasons
21:09:26 <TrueBrain> this is short to bullying
21:09:58 <_dp_> to state constructing reasons I need some substance
21:10:18 <_dp_> all I see so far is that pr having to motivation or analysis whatsoever
21:10:32 <_dp_> I don't even understand what the problem is
21:10:36 <_dp_> it worked fine for years
21:10:38 <Xaroth> then kindly ask for a motivation?
21:10:41 <glx> with correct version and intellisense mode set to windows-clang-x64 it works better (correct defines are set at least)
21:11:04 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOTJQ
21:13:42 <glx> TrueBrain: 'crash8967-2.dmp' (Minidump) : Chargé '*C:\Program Files\NVIDIA Corporation\CoProcManager\nvd3d9wrapx.dll'. Binaire correspondant introuvable.
21:13:42 <glx> 'crash8967-2.dmp' (Minidump) : Chargé '*C:\Program Files\NVIDIA Corporation\CoProcManager\nvdxgiwrapx.dll'. Binaire correspondant introuvable.
21:13:42 <glx> "can't find corresponding binary" doesn't prevent loading the dmp to get the trace ;)
21:14:08 <TrueBrain> for me it couldn't continue the trace after the missing DLLs
21:16:24 <glx> and it didn't find "C:\Windows\System32\ig4icd64.dll" (intel driver)
21:16:52 <glx> so I fear something weird happens with dual GPU thing
21:18:51 <_dp_> Xaroth, what part of "If it's the latter can you please state what's the actual problem and" is not kind enough?
21:19:02 <glx> btw it seems I fixed my FTTH issue, I just put a 120mm fan blowing on the ONU, and everything works fine
21:19:07 <_dp_> I'm not a Shakespeare, that's all the english I know besides technicall stuff
21:19:14 <Xaroth> sorry, I stopped reading after `Is this a troll or are you actually serious?`
21:20:04 <_dp_> there were troll prs before
21:20:20 <Xaroth> from a core developer?
21:20:20 <_dp_> how am I supposed to know with all the fuss about cargo aging lately?
21:20:34 <peter1138> I amused by the idea that it is currently balanced, because everyone else says it's not.
21:20:45 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOTUj
21:21:29 <_dp_> peter1138, with enough effort you can make it work
21:22:16 <_dp_> balancing, configuring, whatever
21:22:18 <peter1138> Effort working around the fact it's unbalanced?
21:22:37 *** EER has quit IRC (Remote host closed the connection)
21:23:39 <peter1138> So it's balanced after you've balanced it. But nobody else can try balancing it a different way.
21:23:41 <_dp_> peter1138, unbalanced what? if you talk about vanilla it has no goal, what are you even balancing for?
21:24:19 <FLHerne> Xaroth: Did you see the original PR title?
21:24:26 <peter1138> "The game is broken, fix it" *gives something a go* "NO! NOT LIKE THAT!"
21:24:59 <FLHerne> Xaroth: I don't think it's wrong to ask if a PR titled "How to break cargo ageing and cargodist in one go." is a troll, especially when it does
21:25:12 <_dp_> peter1138, I can do 10 prs in a day like that. no motivation, no analysis, just random code changes
21:25:25 <peter1138> FLHerne, it's more ... an experiment, I would think.
21:25:53 *** glx is now known as Guest1106
21:26:03 <FLHerne> I guess the idea is people can play with it and consider whether the changes are an improvement, or too far the other way
21:26:08 <michi_cc> FLHerne: The original title was a troll to andy. I did change it shortly afterwards.
21:27:24 <FLHerne> _dp_: Anyway, various cargo aging concerns were discussed on IRC in the last week-ish, mostly andythenorth trying to hack different mechanics into his grfs to make up for the stock mechanics not being strong enough
21:27:57 <FLHerne> Including an insane patch to allow vehicle running costs to increase based on cargo age
21:28:56 <_dp_> it was thousands messages with tons nonsense, do I really need to read them all?
21:29:08 <_dp_> github should contain full info
21:29:25 <peter1138> I didn't realise _dp_ was OpenTTD's project manager.
21:29:48 <FLHerne> Eh, I think at this point this is "here's another idea, let's see if it helps" rather than something to be merged anytime soon
21:29:57 <Xaroth> peter1138: do you have any chairs, I seem to have fallen off of mine.
21:30:06 <FLHerne> It's a better idea than the other patch
21:30:11 *** Guest1106 has quit IRC (Ping timeout: 480 seconds)
21:30:38 <_dp_> When I skip on pr description I'm told to fix it
21:30:44 <_dp_> When I ask I'm told to fuck off
21:31:10 <peter1138> Only one person used the f-word. Twice nice.
21:33:07 <FLHerne> _dp_: It really does just come down to "andy thinks cargo aging isn't strong enough, so make it stronger and see how it breaks everyone's games"
21:33:59 <FLHerne> It's marked "Draft", detailed assessment of whether it improves things will have to wait on people testing it
21:34:46 <FLHerne> If anyone seriously considered merging it as it stands, I'd agree with you 100%
21:35:40 <_dp_> peter1138, I didn't know this chat was G-rated :p
21:37:33 *** erle- has quit IRC (Quit: Leaving)
21:41:47 <_dp_> FLHerne, well then, I guess it found a first victim
21:42:02 <_dp_> and that's why I asked, to understand how seriously should I take it
21:42:46 *** gelignite has quit IRC (Quit: Stay safe!)
21:43:20 <DorpsGek> [OpenTTD/OpenTTD] andythenorth commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOTIR
21:43:42 <_dp_> andy whined for over a week about it, could've been enough for emergency backport for all I know :p
21:47:27 <peter1138> Hmm, somehow reduce frequency of viewport-based sounds, scaled by area of viewport?
21:47:36 <andythenorth> FLHerne could you...not misrepresent what I say please
21:48:13 <peter1138> Leave the misrepresentation to andy.
21:48:23 <andythenorth> I can misrepresent myself quite well
21:48:27 <andythenorth> been doing it a while
21:49:38 <DorpsGek> [OpenTTD/bananas-api] erenes opened issue #86: Error when uploading a newgrf to bananas-api running in docker https://git.io/JOTIh
21:50:26 <Rubidium> nielsm, TrueBrain: the problem with the slow "advertising" is due to getaddrinfo taking 5 seconds (i.e. waiting for the default timeout) for some reason in Beerbelott's container (he confirmed that a lookup in the container takes 5 seconds). Solving that issue should "solve" the problem for OpenTTD too; no idea what the cause of that is. Maybe triggering the resolve of the address earlier helps?
21:50:46 <peter1138> Maybe it is enough to just include the volume scaling.
21:51:06 <TrueBrain> Rubidium: hmm .. honestly, that is a local issue, and not something OpenTTD should try to prevent, I think?
21:51:30 <peter1138> Why do OpenTTD need to resolve it?
21:51:31 <FLHerne> andythenorth: I try, but presumably I got it wrong
21:51:37 <TrueBrain> Docker DNS timeouts are not uncommon btw ..
21:51:48 <andythenorth> TL;DR I think cargo aging is not a very useful vehicle balancing property, even though SOME VERY IMPORTANT AND FAMOUS FORUM PEOPLE AGREED IT SINCE 2013 ALREADY
21:51:56 <FLHerne> So yes, if you could misrepresent yourself that would be nice :-)
21:51:58 <TrueBrain> still don't know why, but from time to time, I have to change my resolve to 1.1.1.1 to fix it for me locally :)
21:52:28 <andythenorth> if you mess with cargo aging (1) you have to mes with it a lot to see the effect on 'small' maps (you know, like the ones we outdated people play :)
21:52:49 <andythenorth> (2) it will fuck up your cargodist network because you'll have really aged packets turning up sometimes as 'surprise!'
21:52:50 *** sla_ro|master has quit IRC ()
21:53:21 <Rubidium> peter1138: resolve as in "fix" the issue, or resolve as in "get the IP address for the master server"?
21:53:24 <peter1138> If they are really aged, what's the issue?
21:53:42 <andythenorth> really weird inexplicable changes in vehicle profit year on year
21:53:49 <andythenorth> which go away when I stop nerfing cargo aging
21:54:21 <andythenorth> set cargo aging back to 185, instead of some vehicles being 185 and some 32
21:54:24 <andythenorth> issue disappears
21:54:38 <peter1138> Rubidium, hah! The latter, but if it's that, then yes of course. Though normally you'd resolve and do other things and then come back once it's resolved.
21:54:47 <TrueBrain> Rubidium: but happy you found the root-cause, that is at least a clear explanation :)
21:55:35 <peter1138> I wonder why a 5 second pause in the UDP stuff causes the TCP clients to drop off.
21:55:46 <TrueBrain> the whole mainloop grinded to a halt
21:55:47 <Rubidium> peter1138: yeah, though the resolve happens in the bit of the code that sends the packet which has a massive lock around it :(
21:55:56 <Rubidium> and that prevents the main loop from going
21:56:05 <TrueBrain> rewrite the network code to be more async you say? :P
21:56:06 <Rubidium> and it resolves for IPv4 and IPv6
21:56:22 <peter1138> await this->Resolve(() => send())
21:56:56 <peter1138> PR #9005: Rewrite OpenTTD in C#.
21:56:57 <Rubidium> and once the 10 seconds have passed, it is time for the retry... so it continues blocking the main loop until it times out after a few tries
21:56:57 <TrueBrain> Rubidium: what we could do, is what we use for the content-server connection, is have the send-packets queue up
21:57:03 <TrueBrain> and send them once the connection is established
21:58:02 <TrueBrain> on the other hand .. this is now the first user that has this issue in 17 years ..
21:58:10 <TrueBrain> owh, no, network we did later
21:58:20 <peter1138> How many don't notice?
21:58:30 <Rubidium> the real PITA is that the resolve needs to be done for each socket, so you first "create" the connection and then you can resolve the IP address
21:59:03 <peter1138> How many did notice but we blamed their connection :D
21:59:06 <TrueBrain> and caching resolve results doesn't really help, as TTLs are low enough that by the next time you need it .. it is expired :D
21:59:16 <TrueBrain> peter1138: well, strictly seen still true :)
21:59:20 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOTLA
21:59:22 <TrueBrain> a DNS that takes 5 seconds to reply is .. euh .. special :)
21:59:49 <TrueBrain> but I have the same issue myself in docker from time to time, and there is a lot of software not behaving nicely :P
22:00:08 <TrueBrain> most just say: cannot resolve, because they refuse to wait a bit longer
22:00:27 <Rubidium> man resolv.conf says "... the default is RES_TIMEOUT (currently 5, ..."
22:01:28 <nielsm> why is it even doing a dns lookup at that time
22:02:10 <TrueBrain> nielsm: how else to connect to the master server? :)
22:02:26 <nielsm> you could assume the master server doesn't change address very often
22:02:28 <peter1138> Can you "just" release the lock while doing the DNS lookup?
22:02:35 <TrueBrain> nielsm: that is a wrong assumption :)
22:02:41 <TrueBrain> there are days the IPs change hourly
22:02:48 <TrueBrain> there are weeks it doesn't change at all
22:03:28 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOTtV
22:03:33 <nielsm> but yes at least resolve the address before taking the socket lock, you don't need the lock to do an address resolution
22:03:54 <nielsm> the lock is really only there because we have some internal state on an object wrapping the socket
22:05:04 <TrueBrain> maybe we should just pin this for now, and see how the lobby-server goes
22:05:06 <andythenorth> nice discord suggestion (someone made it here also a few days ago) just hide all profit display stuff everywhere
22:05:09 <TrueBrain> as that removes most of this code anyway
22:05:20 <TrueBrain> and I assume Beerbelott wants to resolve his DNS issues anyway :)
22:05:41 <TrueBrain> (and the TCP stuff already does resolving (and connecting) async)
22:07:00 <DorpsGek> [OpenTTD/bananas-frontend-cli] erenes opened pull request #23: Combined PR to fix issues #3, #4, #5 and PR #20 https://git.io/JOTtp
22:07:23 *** snail_UES_ has joined #openttd
22:08:09 <andythenorth> how about a lang that just blanks all the money strings? :)
22:08:21 <andythenorth> I am aware this is not a sustainable suggestion
22:08:51 <LordAro> i did start on some stuff that just disabled money
22:09:14 <andythenorth> if there was a way to do it without....fragmenting the UI and simulation into two games
22:09:18 <andythenorth> it might be really nice
22:09:53 <andythenorth> ModelTrainPixelTycoon
22:10:05 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9003: Change: Query for servers on opening multiplayer window https://git.io/JOTqG
22:10:06 <LordAro> i just hacked the command handler
22:10:21 <LordAro> so it never returned false for money reasons
22:10:35 <LordAro> or however it works, it was weeks ago, i don't recall :p
22:10:43 <andythenorth> weeks are like years
22:11:04 <andythenorth> would I play sandbox mode though? o_O
22:11:19 <andythenorth> goals like silicon valley are really more fun
22:11:44 <DorpsGek> [OpenTTD/OpenTTD] nielsmh merged pull request #9000: Fix #8981: Don't attempt to re-reserve path if already entering/entered depot. https://git.io/JOU6O
22:11:47 <DorpsGek> [OpenTTD/OpenTTD] nielsmh closed issue #8981: Train never finishes entering depot, blocks itself from leaving depot. https://git.io/JOv93
22:11:49 <andythenorth> and the money...(1) it's kind of easy (2) it's kind of easy (3) if it's not easy, I reset the grf costs / income :P
22:12:24 <andythenorth> oof give trains a happiness rating
22:12:29 <andythenorth> and let cargo age modify that
22:12:31 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOUQn
22:12:38 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #9002: Change cargo age to have more effect on cargo payment. https://git.io/JOTqa
22:12:41 <andythenorth> just try and make trains happy
22:12:52 <andythenorth> might get a bit like having children...very service oriented
22:13:01 <andythenorth> what was that pet thing?
22:13:12 <LordAro> andythenorth: can newgrfs vary stuff depending on how much money is being made? (i.e. vehicle profit or whatever)
22:13:22 <andythenorth> LordAro 'stuff'?
22:13:38 <LordAro> because that's basically running into the same issues that daylength does
22:13:42 <LordAro> andythenorth: i have no idea
22:13:49 <LordAro> performance or something
22:14:02 <andythenorth> that's a wonderfully broad question :D
22:14:03 <michi_cc> andythenorth: BTW, with 'not redefining PAX/MAIL' I meant 'not reusing the slots for something else'.
22:14:06 <LordAro> can you make a train that goes faster depending on how much money it's made?
22:14:11 <andythenorth> michi_cc oops sorry
22:14:15 <andythenorth> LordAro you can yes
22:14:29 <andythenorth> there are profit vars for this year and last year in 80+ vars
22:14:38 <andythenorth> and cb 36 can change vehicle properties like speed
22:14:48 <andythenorth> V453000 uses it for trains that have random moods in NUTS
22:14:56 <andythenorth> they change mood in depot or station I think
22:15:18 <andythenorth> this is where the 'red train goes faster' joke happens also
22:15:23 <andythenorth> can check company colour
22:15:25 <TrueBrain> Rubidium / nielsm : yeah, all TCP connections are made via a TCPConnector which threads the resolve & connect, so assuming I manage to remove all the UDP stuff (which I plan to), this problem should resolve itself soon :)
22:15:33 <TrueBrain> (had to double check they all use TCPConnecter)
22:16:04 <LordAro> would need to come up with a solution for that (perhaps just "yes i have made all the money" if asked for
22:16:06 <TrueBrain> unintentional fixing shit :D
22:16:17 <andythenorth> otherwise it might not be an accurate name
22:16:38 <andythenorth> LordAro just break those grfs in sandbox, problem solved :)
22:16:44 <andythenorth> or tell them profit is £0
22:17:07 <andythenorth> we put way too much privilege on grfs due to a couple of very noisy legacy forum people
22:17:12 <LordAro> like i said, basically the same issues as daylength
22:17:52 <andythenorth> I think it's a similar idea to groundhog year / daylength
22:18:01 <andythenorth> those lie to the game about time passing
22:18:05 <andythenorth> this would lie about money
22:18:33 <andythenorth> bankruptcy in MP if negative profit happens or something ?
22:20:07 <_dp_> michi_cc, btw, what I'm concerned the most is aging on stations, don't see any benefit in doing so
22:20:15 <_dp_> unclamping profit is just whatever
22:21:18 <_dp_> unless it does something silly on charts, 1/(x+1) sounds kinda arbitrary
22:22:14 <michi_cc> You could call it "realistic", it's a bit strange that if you transfer cargo, it doesn't matter how long it sits around in-between. Kinda makes having "slow" and "fast" cargoes pointless.
22:22:57 <andythenorth> hmm GS has deity mode?
22:23:02 <andythenorth> how does that deal with money?
22:24:05 <_dp_> there are some nice things you can do with station storage, aging on stations will greatly discourage that kind of play
22:24:28 <_dp_> and any transfers in general tbh
22:24:43 <_dp_> and they kinda suck already
22:26:04 <andythenorth> test it, see what it's like
22:26:14 <andythenorth> has it got a wasm build?
22:26:18 <michi_cc> 1/(x+1) (well, 31/(x/16+1) to be exact) has a) an +inf asymptotic limit of 0, and b) a slope at the origin that is somewhat similar to the -2 slope of the existing third payment section. Other than that, it is arbitrary.
22:26:51 <andythenorth> wasm multiplayer when?
22:27:24 <_dp_> it takes eternity to really test stuff like that properly
22:27:47 <_dp_> want to make sure I'm not just going just to waste my time first
22:28:17 <DorpsGek> [OpenTTD/OpenTTD] Booeraque opened issue #9005: I love your game and really hope you will develop it much better! <3 https://git.io/JOTmp
22:29:27 <andythenorth> _dp_ will it really matter? genuine question
22:30:42 <DorpsGek> [OpenTTD/OpenTTD] rexxars commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOTYI
22:31:07 <peter1138> Changing "Show vehicle-type specific liveries" mid-game does not appear to have an effect.
22:31:24 <peter1138> I guess cached colour map is not being invalidated.
22:31:54 <andythenorth> no effect at all, or just some vehicles are cached?
22:31:55 <_dp_> andythenorth, station aging? who knows, for average player probably not, they don't keep cargo on stations for long anyway
22:32:03 <andythenorth> company colour caching is slow to invalidate sometimes
22:32:08 <_dp_> for citymania it will be a disaster if added without an option
22:32:10 <peter1138> No effect on vehicles on the map.
22:32:14 <_dp_> just because in changes economy
22:32:32 <_dp_> and affects some important strategies
22:32:32 <peter1138> andythenorth, it was originally invalidated instantly.
22:34:30 <DorpsGek> [OpenTTD/OpenTTD] James103 commented on issue #9005: I love your game and really hope you will develop it much better! <3 https://git.io/JOTmp
22:34:40 <michi_cc> _dp_: Not really about cargo ageing specifically, but you're basically arguing that a large part of the game can never be touched without a disable setting because any change might affect some important strategy.
22:34:51 <andythenorth> peter1138 I see it now and then, but it's really hard to repro, so didn't make a bug report. Also because I am not entirely a good person :P
22:35:09 <andythenorth> it's not frequent enough in general gameplay to piss in anyone's cornflakes
22:35:22 <andythenorth> and I probably see it more because I change colours a lot for grf dev
22:35:36 <andythenorth> I guess what you're seeing is different :P
22:36:25 <michi_cc> _dp_: It's not a wrong opinion when having a specific goal/gameplay style in mind, but it also places limits on everybody else.
22:37:08 <peter1138> Hmm, it invalidated the livery window...
22:37:43 <_dp_> michi_cc, I realize that, but that's what is what I do and what's important to me
22:37:50 <_dp_> and it worked fine for years
22:38:13 <_dp_> also in a way it's a part of savegame compatibility that openttd is so keen about
22:38:24 <_dp_> what's the point of loading old saves if the game changed?
22:38:41 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOTY5
22:39:19 <_dp_> also what I'm opposed to the most is not any change but a change without a good reason
22:40:15 <_dp_> when station catchment was changed it also affected citymania quite significantly but I found a ways to counteract it because I understood that it was a good change
22:40:24 <michi_cc> We might somewhat guarantee that loading old games is possible, but definitely don't guarantee no variances in gameplay whatsoever.
22:43:01 <peter1138> Alright, what's the way to pretend my build is 1.11.0? :D
22:43:26 <_dp_> michi_cc, yeah, but where is the line? what's the point of loading original ttd saves in factorio? :p
22:43:40 <DorpsGek> [OpenTTD/OpenTTD] spnda commented on pull request #9003: Change: Query for servers on opening multiplayer window https://git.io/JOTOc
22:43:50 <TrueBrain> peter1138: cmake -DGENERATE_OTTDREV=1 -P cmake/scripts/FindVersion.cmake
22:44:09 <TrueBrain> it should, if nobody fucked it up, overwrite any detection :D
22:44:36 <_dp_> a lot of ttdcoop stuff is quite sensitive to changes for example, if you load it and it breaks there wasn't much point in having compatibility
22:44:45 <TrueBrain> nevermind .. it doesn't work in OpenTTD that way .. it does for all other repos as of late :P
22:44:48 <TrueBrain> sorry peter1138 , ignore me :)
22:44:53 <TrueBrain> .git wins from .ottdrev :P
22:45:57 <peter1138> git tag --delete, git tag ;)
22:46:59 <michi_cc> Just make sure to not push back to upstream :)
22:47:33 <peter1138> I haven't done it, it feels wrong, and yes, dangerous :)
22:47:54 <nielsm> peter1138: change GetNetworkRevisionString() to lie and then make a tag that consists of only numerals and points
22:48:50 <frosch123> we should probably split the "network version" from other version strings in the game
22:49:20 <frosch123> then citymania client could say "citymania" on intro screen, in gamelog and everywhere, and still cheat to be multiplayer compatbile
22:49:39 <frosch123> would make it easier for bug reporting
22:51:28 <_dp_> frosch123, cmclient follows vanilla releases, using different versioning will only add to players confusion
22:51:48 <_dp_> and it already can cheat whatever it needs to and put "citymania" wherever it can
22:53:05 <TrueBrain> OpenTTD 1.10.3 - Citymania edition
22:53:09 <_dp_> yeah, but that doesn't require any changes from vanilla
22:54:04 <glx> didac: 'Error copying file "D:/dev/OpenTTD/bin/ai/compat_1.12.nut" to "D:/dev/OpenTTD/build/ai/compat_1.12.nut".' was the missing info
22:54:22 <didac> yep I see that, it's trying to copy the same file to itself
22:54:46 <peter1138> bin to build isn't the same
22:54:59 <didac> well no, the file doesn't exist
22:55:11 <didac> oh wait different path, I read it wrong
22:55:31 <didac> compat_1.12.nut doesn't exist in OpenTTD/bin/ai
22:56:13 <didac> I might be really behind from master in that branch, lemme fix it...
22:56:38 <DorpsGek> [OpenTTD/OpenTTD] rexxars commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOT3k
22:57:00 <glx> and all these files are explicitely added in CMakeLists.txt
22:59:19 <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOT3W
22:59:31 <glx> and same error for bin/game
23:00:38 <glx> these 2 files were added after 1.11 branching
23:01:34 *** Gustavo6046 has joined #openttd
23:03:02 <FLHerne> ^cheerfully incompetent bug reporting :D
23:04:37 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:04:47 <DorpsGek> [OpenTTD/OpenTTD] PeterN opened pull request #9006: Fix: Invalid cached vehicle colourmaps when changing liveries setting. https://git.io/JOT37
23:05:13 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
23:07:01 <glx> FLHerne: yeah, steam users
23:07:32 <glx> still waiting for the same data in #8979
23:08:21 <glx> they just copy paste the message on crash dialog without reading the extra info in the box
23:13:01 *** FLHerne has quit IRC (Ping timeout: 480 seconds)
23:16:54 <andythenorth> I can add 2 kinds of fishing grounds
23:18:56 *** nielsm has quit IRC (Ping timeout: 480 seconds)
23:19:53 <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8794: Feature: Show rail/road/tram NewGRF name in Land Area Information window https://git.io/JthCX
23:25:32 *** andythenorth has quit IRC (Quit: andythenorth)
23:27:35 <DorpsGek> [OpenTTD/OpenTTD] michicc opened pull request #9007: Fix: Check for a validly mapped OpenGL screen buffer during driver init. https://git.io/JOTZc
23:34:03 <didac> +glx: I think so, now it's all good :D
23:38:22 <DorpsGek> [OpenTTD/OpenTTD] rexxars commented on pull request #9004: Feature: Reload NewGRF on SIGUSR1 signal https://git.io/JOTnT
23:56:46 <DorpsGek> [OpenTTD/OpenTTD] James103 opened issue #9008: OpenTTD accepts negative / too-high starting years via `-t` command line parameter https://git.io/JOTcQ
continue to next day ⏵
