IRC logs for #openttd on OFTC at 2021-07-15
⏴ go to previous day
01:54:17 *** snail_UES_ has joined #openttd
02:00:23 *** WormnestAndroid has quit IRC (Read error: No route to host)
02:00:35 *** WormnestAndroid has joined #openttd
02:12:53 <DorpsGek> [OpenTTD/OpenTTD] bootmii opened issue #9442: [Bug]: FON raster fonts are not loaded or rendered correctly in Windows https://git.io/JWvF3
02:19:04 *** ChanServ sets mode: +v tokai
02:21:18 *** Wuzzy has quit IRC (Remote host closed the connection)
02:22:08 *** Wormnest has quit IRC (Quit: Leaving)
02:25:41 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
02:38:57 *** debdog has quit IRC (Ping timeout: 481 seconds)
02:45:52 *** tokai|noir has joined #openttd
02:45:52 *** ChanServ sets mode: +v tokai|noir
02:52:35 *** tokai has quit IRC (Ping timeout: 480 seconds)
06:02:40 *** andythenorth has joined #openttd
06:23:41 *** snail_UES_ has quit IRC (Quit: snail_UES_)
06:40:35 *** andythenorth has quit IRC (Quit: andythenorth)
06:49:07 *** andythenorth has joined #openttd
07:12:25 *** andythenorth has quit IRC (Quit: andythenorth)
07:27:45 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
07:28:36 *** WormnestAndroid has joined #openttd
07:36:40 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
07:38:22 *** WormnestAndroid has joined #openttd
08:04:07 <DorpsGek> [OpenTTD/OpenTTD] PeterN commented on issue #9442: [Bug]: FON raster fonts are not loaded or rendered correctly in Windows https://git.io/JWvF3
08:22:42 *** Progman has joined #openttd
08:27:26 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
08:31:52 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
08:42:47 *** andythenorth has joined #openttd
08:48:07 <andythenorth> TrueBrain are you gainfully employed yet?
08:54:29 <TrueBrain> is there any other type of employment? :)
08:55:05 <TrueBrain> also gainful, I hope :P
08:55:07 <peter1138> Oh, I finally found which window had left a tooltip open...
09:27:42 <peter1138> Hmm, does Freetype support rendering raster fonts?
09:29:45 *** andythenorth has quit IRC (Quit: andythenorth)
09:31:29 <peter1138> Alas, Freetype isn't used on Windows...
09:33:01 <Rubidium> "sorry, Windows' font rendering doesn't support its own fonts"
09:38:11 <peter1138> I didn't see an equivalent call for raster fonts when I did a quick search. I suspect it requires more of GDI, rendering text to a DC handle etc...
09:45:40 <LordAro> i'm not sure i follow
09:47:01 <peter1138> Do you ever have mirrors on your bicycle/person?
09:53:06 <LordAro> i ride with some people who probably should though
09:53:16 <LordAro> i'm happy enough turning my head
09:53:35 *** D-HUND is now known as debdog
10:01:07 *** virtualrandomnumber has joined #openttd
10:02:08 *** virtualrandomnumber has quit IRC (Remote host closed the connection)
12:05:28 *** snail_UES_ has joined #openttd
12:16:42 <TrueBrain> ha, I can do STUN connects with 2 Game Coordinator instances, where the server is connected to the first, and the client to the second :D
12:16:49 <TrueBrain> that required some inter-communication to get working :)
12:54:29 *** Progman has quit IRC (Remote host closed the connection)
13:28:15 *** y2kboy23 has quit IRC (Ping timeout: 480 seconds)
13:31:24 *** y2kboy23 has joined #openttd
13:34:44 *** gelignite has joined #openttd
13:36:08 *** WormnestAndroid has quit IRC (Remote host closed the connection)
13:36:21 *** WormnestAndroid has joined #openttd
14:16:50 *** roadt__ has joined #openttd
14:21:22 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
14:21:24 <TrueBrain> found one bug during testing .. not a bad outcome :)
14:23:53 *** roadt_ has quit IRC (Ping timeout: 480 seconds)
14:26:28 <DorpsGek> [OpenTTD/py-protocol] TrueBrain opened pull request #10: Feature: support for Game Coordinator protocol 3 https://git.io/JWBAe
14:26:59 <DorpsGek> [OpenTTD/py-protocol] TrueBrain updated pull request #10: Feature: support for Game Coordinator protocol 3 https://git.io/JWBAe
14:30:51 <DorpsGek> [OpenTTD/py-protocol] TrueBrain merged pull request #10: Feature: support for Game Coordinator protocol 3 https://git.io/JWBAe
14:33:20 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #18: Feature: support for Game Coordinator protocol 3 https://git.io/JWRk1
14:33:31 <TrueBrain> that required ... a lot of code :P
14:33:33 <TrueBrain> but I already knew that :D
14:34:33 <TrueBrain> still needs TURN support, but that isn't that bad, in terms of LoC
14:44:29 <orudge> Stunning work, TrueBrain :D
14:46:42 <orudge> It could be a real head-turner I'm sure
14:47:49 <TrueBrain> Let me TURN this conversation off
14:47:58 <TrueBrain> the puns are STUNning me .. I cannot handle it
14:48:48 <orudge> To be fair, that's pretty much all I've got. I won't be transmitting any more.
14:49:04 <TrueBrain> sorry, what did you say? There was some packetloss on the line
14:52:13 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain updated pull request #18: Feature: support for Game Coordinator protocol 3 https://git.io/JWRk1
14:53:15 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #18: Feature: support for Game Coordinator protocol 3 https://git.io/JWRk1
14:54:07 <TrueBrain> there is a memory leak in the game-coordinator, I am pretty sure of that :D
14:54:21 <TrueBrain> it currently crashes every 24h or so, because it goes OOM :P
15:12:08 *** Wormnest has joined #openttd
15:18:59 <TrueBrain> oof, cannot deploy the new code, as it seems AWS increased the wrong limit
15:19:04 <TrueBrain> despite the whole ticket talking about the right one :P
15:20:39 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #19: Change: [Actions] also deploy the STUN server upon merging / releasing https://git.io/JW0s9
15:21:08 <Timberwolf> I liked the OOM we had in an API I built a few years ago.
15:21:17 <Timberwolf> A constant amount of memory leaks per request.
15:21:43 <TrueBrain> I think in this case it is not cleaning up disconnected servers or something
15:21:44 <Timberwolf> It was deployed on two instances, behind a round-robin load balancer, with both instances brought up and provisioned at exactly the same time.
15:22:15 <Timberwolf> Guess when they both ran out of memory and crashed :)
15:22:44 <TrueBrain> it is not far off in our deployment currently ;)
15:23:07 <Timberwolf> Yep. Well, close enough that we went from "everything is fine" to "wait, we just lost both instances?" within a monitoring window.
15:23:12 <TrueBrain> but sometimes one of them crashes for unrelated reasons earlier .. so that offsets things again :D
15:23:27 <TrueBrain> just after each deployment that offset resets :D
15:25:03 <TrueBrain> these are two instances, and this is the "max memory" between the two
15:25:16 <TrueBrain> so you can see one crashed ~24h before the second :)
15:25:44 <TrueBrain> but now first I have to wait for AWS to fix their shit :P
15:31:20 <glx> I guess I'll have some tests to do (when it will be testable)
15:38:26 <TrueBrain> wow, AWS already replied .. they are never this quick with service limit requests :P
15:38:55 <TrueBrain> just nice to see that they don't take hours to fix mistakes on their side :)
16:28:29 *** HerzogDeXtEr has joined #openttd
16:35:54 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWE1W
16:55:11 *** andythenorth has joined #openttd
16:57:43 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #19: Change: [Actions] also deploy the STUN server upon merging / releasing https://git.io/JW0s9
17:15:36 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWu3o
17:17:12 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
17:19:55 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWunU
17:20:25 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
17:21:50 <TrueBrain> Connection type: Behind NAT
17:24:07 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWuWt
17:24:16 <TrueBrain> glx: if you wouldn't mind, all yours to test :)
17:24:27 <TrueBrain> IPv4+IPv6 should now work just fine :)
17:24:32 <TrueBrain> famous last words ofc
17:29:22 <glx> first step, hosts change is disabled :)
17:42:04 <glx> default hosts, IPv4+IPv6 (forwared IPv4, closed IPv6) : Behind NAT
17:43:45 <glx> opened IPv6, still Behind NAT
17:45:03 <TrueBrain> that is a bit odd :P
17:45:04 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
17:45:40 <TrueBrain> I am curious if it does work with master?
17:47:10 *** gelignite has quit IRC (Quit: Stay safe!)
17:50:31 <glx> IPv6 only (opened): Behind NAT
17:52:30 <glx> invite code is the same on all tests (which is expected IIRC)
17:59:44 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:03:44 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #20: Fix: direct-ips detected via IPv6 were stored incorrectly https://git.io/JWu1u
18:03:55 <TrueBrain> I see no reason why it wouldn't work on AWS, as it does work locally just fine
18:03:58 <TrueBrain> will check it out later tonight
18:04:04 <TrueBrain> but information if master does work would be helpful :)
18:22:53 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWu7N
18:48:52 *** gelignite has joined #openttd
18:49:38 <DorpsGek> - Update: Translations from eints (by translators)
19:07:35 <glx> <TrueBrain> but information if master does work would be helpful :) <-- lot of files to recompile ;)
19:15:46 <glx> TrueBrain: master, staging, IPv6 only (closed) : can't connect; IPv6 only (open): Public
19:20:58 <glx> won't test IPv4 only, because it's a pain to edit hosts :)
19:21:48 <glx> and I know it will be Public, (forwarding works)
19:28:42 <Rubidium> IPv4 only seems to work too
19:31:56 *** NGC3982 has quit IRC (Remote host closed the connection)
19:38:26 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9443: Feature: allow the use of STUN to connect client and server together https://git.io/JWZe2
19:40:14 <TrueBrain> okay, so that is really weird .. hmm ..
19:40:19 <TrueBrain> owh, I have an idea what it might be
19:40:38 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #20: Fix: direct-ips detected via IPv6 were stored incorrectly https://git.io/JWu1u
19:41:55 <glx> STUN priority over direct ?
19:43:57 *** NGC3982 has joined #openttd
19:44:18 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #21: Fix: direct-ip detection wasn't working for coordinator protocol v3 https://git.io/JWzZy
19:44:20 <TrueBrain> using the wrong port ;)
19:45:16 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain updated pull request #21: Fix: direct-ip detection wasn't working for coordinator protocol v3 https://git.io/JWzZy
19:45:39 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain updated pull request #21: Fix: direct-ip detection wasn't working for coordinator protocol v3 https://git.io/JWzZy
19:45:51 *** NGC3982 has joined #openttd
19:46:02 <glx> lol I was expecting the second update :)
19:48:53 <FLHerne> V453000: congrats on being in the Steam Deck demo video :D
19:49:16 <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #21: Fix: direct-ip detection wasn't working for coordinator protocol v3 https://git.io/JWzZy
19:52:33 <TrueBrain> glx: try it out now :)
19:58:00 <DorpsGek> [OpenTTD/master-server] TrueBrain opened pull request #44: Fix: entries on the gc-stream are now well defined https://git.io/JWzWM
19:58:46 <glx> closed -> behind NAT, open -> Public
19:59:08 <DorpsGek> [OpenTTD/master-server] TrueBrain updated pull request #44: Fix: entries on the gc-stream are now well defined https://git.io/JWzWM
19:59:19 <TrueBrain> for both IPv4 and IPv6?
19:59:23 *** NGC3982 has quit IRC (Remote host closed the connection)
19:59:47 <glx> let me change private.cfg :)
20:00:07 <glx> or maybe it's doable in console, so I don't have the restart openttd
20:00:10 <TrueBrain> I am mostly curious about the dualstack :)
20:01:20 *** gelignite has quit IRC (Quit: Stay safe!)
20:03:12 <glx> now I think about the earlier tests, IPv4 should say public, as it's forwarded
20:04:48 <glx> dual, with IPv6 open -> Public
20:05:36 <TrueBrain> can you leave a dualstack server open for a bit of time?
20:05:40 <TrueBrain> want to check if the API shows it correctly :D
20:06:12 <TrueBrain> hmm .. I think the web cannot show invite-code data yet .. lol
20:06:20 <TrueBrain> okay, I need to fix that first
20:06:23 <TrueBrain> will do that in a bit :)
20:07:22 *** NGC3982 has joined #openttd
20:08:59 <glx> server running, dual stack, IPv6 closed
20:09:17 <TrueBrain> sorry, I meant dualstack, as in: IPv4 and IPv6 open :)
20:09:55 <TrueBrain> I want to know if it lists both IPs correctly
20:11:35 <TrueBrain> I -think- it works :P
20:12:04 <TrueBrain> now to fix the frontend to show invite-code-based servers too :D
20:13:28 <DorpsGek> [OpenTTD/master-server] TrueBrain merged pull request #44: Fix: entries on the gc-stream are now well defined https://git.io/JWzWM
20:21:08 *** EER has quit IRC (Remote host closed the connection)
20:23:13 <glx> I have 2 weird lines in the log "dbg: [net] Client 5 made an error and has been disconnected: ndbg: [net] Closed client connection 5"
20:23:37 <glx> something is broken in the display
20:25:46 <glx> oh no, I know, it's threads
20:27:33 <Rubidium> so Debug needs to some locking?
20:28:33 <DorpsGek> [OpenTTD/master-server] TrueBrain opened pull request #45: Fix: allow invite-code as server-id in the API https://git.io/JWzuZ
20:28:59 <glx> but it's weird only "n" from the error string is shown
20:30:32 <glx> (I haven't really checked, but the only valid explication for the weird lines is threading)
20:31:01 <DorpsGek> [OpenTTD/master-server] TrueBrain updated pull request #45: Fix: allow invite-code as server-id in the API https://git.io/JWzuZ
20:31:12 <DorpsGek> [OpenTTD/master-server-web] TrueBrain opened pull request #32: Add: show invite-code if possible, over direct-ips https://git.io/JWzuH
20:31:24 <Rubidium> for Windows the same static buffer to write (fputws) is used, so that might explain a weird string getting printed
20:31:56 <Rubidium> hmm, never mind... it's not static
20:32:56 <glx> and GetString doesn't use a static buffer
20:33:16 <glx> so something happens inside Debug
20:33:30 <DorpsGek> [OpenTTD/master-server] TrueBrain merged pull request #45: Fix: allow invite-code as server-id in the API https://git.io/JWzuZ
20:34:05 <glx> fmt::format using shared buffer maybe ?
20:36:04 *** andythenorth has quit IRC (Quit: andythenorth)
20:39:07 <TrueBrain> glx: tnx, both IPs are indeed registered correctly :)
20:39:29 <TrueBrain> now let me fix that it shows up on the website .. invite-code only is enough :P
20:39:56 <DorpsGek> [OpenTTD/master-server-web] TrueBrain merged pull request #32: Add: show invite-code if possible, instead of direct-ips https://git.io/JWzuH
20:40:33 <TrueBrain> I think that is everything, besides a lot of testing ofc
20:40:43 <TrueBrain> I have noticed that "restart" on a server makes it unlikely clients reconnect
20:40:51 <TrueBrain> as they are too quick, and the invite-code is not registered yet :P
20:41:08 <TrueBrain> but I think this was already an issue, that if you reconnect too quick, the server wasn't there yet
20:41:30 <michi_cc> glx: Writing to a file (console) handle should be thread-safe on Windows, but I don't think for the console it is specified how the characters are sequenced.
20:42:05 <glx> yeah, but I'm more worried about the missing chars :)
20:42:45 <glx> have a string displayed in the middle of another happens, but both should be full drawn
20:44:00 <glx> I recognise my invite code :)
20:47:18 <Rubidium> glx: or might the string being put into that format be "broken"?
20:47:30 *** Wuzzy has quit IRC (Remote host closed the connection)
20:47:49 <Rubidium> e.g. a string of 4 characters with 3 '\0' or so
20:49:24 <glx> it's one of STR_NETWORK_ERROR_XXX
20:53:08 <Rubidium> well, I was more thinking about it reusing the memory the string of GetString was stored in for something else while doing the Debug formatting stuff
20:53:10 <glx> maybe "STR_NETWORK_ERROR_CLIENT_WRONG_REVISION :n° de version/révision incorrect"
20:53:42 <glx> well GetString in theory should allocate a new buffer
20:54:35 <glx> and the buffer is converted to std::string for return
20:54:54 <Rubidium> that sounds plausible; can you see what a Debug(net, 0, "[{}]", GetString(STR_NETWORK_ERROR_CLIENT_WRONG_REVISION)); does for you?
20:57:14 *** andythenorth has joined #openttd
20:57:28 *** Wormnest has quit IRC (Quit: Leaving)
20:58:56 <Rubidium> it seems to work fine for me (non-Windows at the moment)
21:00:10 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:08:24 <glx> converting as a console app to check something
21:09:22 <glx> in powershell it displays "dbg: [net] [n� de version/r�vision incorrect]"
21:11:11 *** andythenorth has quit IRC (Quit: andythenorth)
21:14:08 *** Samu has quit IRC (Quit: Leaving)
21:17:09 <glx> hmm but my windows should use utf8 (same for powershell)
21:22:07 *** Progman has joined #openttd
21:26:58 <glx> oh there's convert_to_fs() before printing
21:27:47 <glx> and an outdated * @param console_cp convert to the console encoding instead of the normal system encoding.
21:44:10 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
21:44:42 *** WormnestAndroid has joined #openttd
21:47:55 *** snail_UES_ has quit IRC (Quit: snail_UES_)
22:13:55 <peter1138> As if Factorio is being used as promo material for this new Steam device...
22:17:15 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
22:18:04 *** WormnestAndroid has joined #openttd
22:23:57 <glx> haha "-q" switch can't work, on broken save it calls GetString() before the language packs are loaded
23:41:13 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
continue to next day ⏵
