Config
Log for #openttd on 14th July 2020:
Times are UTC Toggle Colours
00:21:17  *** WormnestAndroid has quit IRC
00:21:57  *** WormnestAndroid has joined #openttd
00:39:53  *** Wormnest has quit IRC
00:40:38  *** WormnestAndroid has quit IRC
00:41:03  *** WormnestAndroid has joined #openttd
00:55:57  *** Wormnest has joined #openttd
01:01:54  *** b_jonas has quit IRC
01:17:02  *** keoz has quit IRC
01:49:00  *** Wormnest has quit IRC
01:56:47  *** glx has quit IRC
02:26:58  *** Wormnest has joined #openttd
02:39:20  *** supermop_Home has quit IRC
03:07:09  *** Wormnest has quit IRC
03:09:17  *** Flygon has joined #openttd
03:15:40  *** supermop_Home has joined #openttd
04:22:53  *** supermop_Home has quit IRC
07:15:39  *** WormnestAndroid has quit IRC
07:15:52  *** WormnestAndroid has joined #openttd
08:36:04  *** iSoSyS has joined #openttd
08:37:34  <DorpsGek_III> [OpenTTD/OpenTTD] ldpl updated pull request #8115: GS methods to let company have exclusive access to the industry https://git.io/JfZt7
09:09:17  *** Smedles has quit IRC
09:09:30  *** keoz has joined #openttd
09:17:40  *** gelignite has joined #openttd
09:40:30  *** gelignite has quit IRC
09:58:55  *** Laedek_ has joined #openttd
10:06:05  *** Laedek has quit IRC
10:21:19  *** mindlesstux6662 has quit IRC
10:22:50  *** mindlesstux6662 has joined #openttd
12:04:35  *** tokai has joined #openttd
12:04:36  *** ChanServ sets mode: +v tokai
12:11:33  *** tokai|noir has quit IRC
13:29:49  <DorpsGek_III> [OpenTTD/OpenTTD] SylvainDevidal opened issue #8272: Stopped vehicule destination should be red when mouse over https://git.io/JJ35w
13:33:00  *** glx has joined #openttd
13:33:00  *** ChanServ sets mode: +v glx
13:33:39  <DorpsGek_III> [OpenTTD/OpenTTD] ldpl opened pull request #8273: Center text and icons in the status bar vertically https://git.io/JJ355
13:36:40  *** nielsm has joined #openttd
14:17:42  <DorpsGek_III> [OpenTTD/OpenTTD] glx22 commented on issue #8271: Crash when creating a new group https://git.io/JJ3Y6
14:21:27  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #8271: Crash when creating a new group https://git.io/JJ3Y6
14:21:27  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro closed issue #8271: Crash when creating a new group https://git.io/JJ3Y6
14:21:43  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #7838: Crash in VehicleGroupWindow::GroupNameSorter https://git.io/JeKL5
14:21:43  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro closed issue #7838: Crash in VehicleGroupWindow::GroupNameSorter https://git.io/JeKL5
15:23:42  *** WormnestAndroid has quit IRC
15:24:56  *** WormnestAndroid has joined #openttd
16:01:35  *** WormnestAndroid has quit IRC
16:01:52  *** WormnestAndroid has joined #openttd
16:20:24  <DorpsGek_III> [OpenTTD/OpenTTD] glx22 dismissed a review for pull request #8192: Change: Always generate API files at build time https://git.io/JJvmQ
16:20:24  <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #8192: Change: Always generate API files at build time https://git.io/Jf1KV
16:21:31  <DorpsGek_III> [OpenTTD/OpenTTD] glx22 commented on pull request #8192: Change: Always generate API files at build time https://git.io/JJ3jQ
16:26:25  *** WormnestAndroid has quit IRC
16:27:06  *** WormnestAndroid has joined #openttd
16:29:21  *** supermop_Home has joined #openttd
16:29:27  <supermop_Home> hi
16:33:18  *** Progman has joined #openttd
16:34:48  *** Flygon has quit IRC
16:43:02  *** b_jonas has joined #openttd
16:50:01  *** gelignite has joined #openttd
16:58:51  *** HerzogDeXtEr has joined #openttd
17:01:39  *** frosch123 has joined #openttd
17:06:15  *** Wolf01 has joined #openttd
17:18:13  <supermop_Home> hi Wolf01
17:21:57  *** sla_ro|master has joined #openttd
17:27:51  <Wolf01> o/
17:28:18  <supermop_Home> too bad andy is not around - on sunday I went to the real steeltown
17:28:30  <supermop_Home> and saw some molten iron wagons
17:34:18  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 opened pull request #44: Add: [release-docs] support for cmake generated api files https://git.io/JJsfd
17:39:42  <Wolf01> <supermop_Home> too bad andy is not around <- Just say his name 3 times
17:45:29  <CornsMcGowan[m]> <_dp_ "Corns McGowan, do you know you c"> isn't that how it works normally
17:45:40  <CornsMcGowan[m]> nvm i see what you mean
17:47:05  <CornsMcGowan[m]> <_dp_ "Corns McGowan, or ctrl-click liv"> ah i see, weird it doesn't work in the group livery tab though
17:49:50  *** WormnestAndroid has quit IRC
17:49:53  *** WormnestAndroid has joined #openttd
17:59:07  <CornsMcGowan[m]> how do i shut down and restart a server while preserving company passwords?
17:59:33  <_dp_> you don't
17:59:39  <_dp_> it's a security "feature"
18:00:18  <b_jonas> I finally started a game with FIRS continental basic. FIRS seems fun.
18:01:39  <b_jonas> Although now I have to build some sort of feeder system that divides engineering supplies from a single port to three mines. Any hints for that?
18:02:57  <frosch123> low capacity vehicles
18:04:12  <CornsMcGowan[m]> ^
18:04:18  <CornsMcGowan[m]> or cargodist enabled for engg supplies
18:04:20  <b_jonas> frosch123: yeah. two of the mines are close to each other but at some distance from the port, so I'll probably have a long train transfer engineering supplies to a transfer station nearby, then feed the two mines from that with either a short train or trucks
18:04:35  <b_jonas> the third mine is close to the port
18:04:36  <CornsMcGowan[m]> anybody know if cargo payment rates are affected by daylength?
18:06:50  <CornsMcGowan[m]> a: no
18:21:41  <nielsm> have anyone suggested andy making the various supplies cargoes the "valuables" class so you can set cdist for them separately from the regular cargoes?
18:23:15  <frosch123> messes with vehicles though
18:23:26  <frosch123> armored vans
18:24:45  <frosch123> also, i think andy considered cdist broken for supplies
18:25:11  <frosch123> he wants distance-independent distruibuton, which cdist does not offer
18:26:28  <nielsm> hm true
18:26:58  <CornsMcGowan[m]> distance independent would be v nice
18:30:24  <_dp_> is it sending more closer or further?
18:33:13  *** JGR has joined #openttd
18:36:57  <nielsm> closer I think
19:11:32  <DorpsGek_III> [OpenTTD/CompileFarm] TrueBrain commented on pull request #44: Add: [release-docs] support for cmake generated api files https://git.io/JJstK
19:18:58  *** y2kboy23 has quit IRC
19:20:43  *** y2kboy23 has joined #openttd
19:25:03  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 updated pull request #44: Add: [release-docs] support for cmake generated api files https://git.io/JJsfd
19:26:01  *** frosch123 has quit IRC
19:26:51  *** y2kboy23 has quit IRC
19:28:38  *** y2kboy23 has joined #openttd
19:35:42  *** sla_ro|master has quit IRC
19:36:21  <DorpsGek_III> [OpenTTD/CompileFarm] TrueBrain approved pull request #44: Add: [release-docs] support for cmake generated api files https://git.io/JJsm0
19:41:53  <b_jonas> also, OpenTTD putting foundations under everything (tracks, roads, stations, depots, bridge heads, town buildings, industries) makes it so much easier to build on an uneven terrain than in TTD
20:05:02  *** JGR has quit IRC
20:43:07  *** nielsm has quit IRC
20:51:07  <DorpsGek_III> [OpenTTD/CompileFarm] LordAro merged pull request #44: Add: [release-docs] support for cmake generated api files https://git.io/JJsfd
20:51:08  <DorpsGek_III> [OpenTTD/CompileFarm] LordAro pushed 1 commits to master https://git.io/JJs3M
20:51:08  <DorpsGek_III>   - Add: [release-docs] support for cmake generated api files (by glx22)
20:51:21  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #8192: Change: Always generate API files at build time https://git.io/JJs3y
20:58:38  *** dihedral has quit IRC
21:17:13  *** dihedral has joined #openttd
21:27:16  *** openttdguest has joined #openttd
21:29:43  <openttdguest> hi what is the easiest way to get the full md5s of the network content without the game?
21:32:29  *** openttdguest has quit IRC
21:32:44  <LordAro> i suppose i could've linked them to the bananas repo
21:32:58  <Eddi|zuHause> why didn't you? :p
21:33:05  <LordAro> eh
21:33:44  <LordAro> also they waited less than 3 minutes
21:33:55  <LordAro> i'm not sure i could've typed the link out fast enough
21:35:25  <Eddi|zuHause> gotta get faster :) ... some people stay barely a minute :)
21:36:43  *** openttdguest has joined #openttd
21:36:50  <LordAro> i didn't see it immediately :p
21:39:50  <openttdguest> i got disconnect, how do I get full md5 of network content outside of game?
21:40:04  <supermop_Home> here's your chance LordAro
21:40:10  <LordAro> openttdguest: https://github.com/OpenTTD/BaNaNaS
21:40:33  <LordAro> though i guess you only get partial md5s
21:40:46  <LordAro> what are you trying to do with them?
21:40:52  <LordAro> supermop_Home: :p
21:42:06  *** openttdguest has quit IRC
21:43:55  *** Wolf01 has quit IRC
21:44:02  *** openttdguest has joined #openttd
21:44:03  <LordAro> supermop_Home: eh, i tried
21:44:18  <openttdguest> LordAro: do you know how to get full md5s?
21:44:36  <LordAro> i do not
21:44:45  <LordAro> i don't think there is a way, without downloading things first
21:45:36  *** openttdguest has quit IRC
21:45:51  <LordAro> well they actually quit that time
21:45:58  <LordAro> ¯\_(ツ)_/¯
21:46:36  <LordAro> TrueBrain: is "how bananas works" documented anywhere? i can't find anything in any of the bananas repos...
21:46:48  <glx> I think I'll merge #8192 after next nightly, to be sure everything is ok
21:57:25  *** HerzogDeXtEr has quit IRC
22:03:01  *** Progman has quit IRC
22:03:04  *** tycoondemon has quit IRC
22:23:42  *** WormnestAndroid has quit IRC
22:29:31  *** openttdguest has joined #openttd
22:29:33  *** openttdguest has left #openttd
22:34:38  <TrueBrain> LordAro : yeah, most repos have documentation in their readme
22:35:02  <TrueBrain> There is also a gist with the design
22:35:07  <LordAro> TrueBrain: they all have "how to run this", but none of them have "how this works"
22:35:09  *** openttdguest has joined #openttd
22:35:21  <TrueBrain> I will soon put that somewhere more permanent
22:35:26  <LordAro> :)
22:35:34  <openttdguest> anyone know how to get full md5sum outside of game?
22:36:09  <TrueBrain> LordAro : but there really is not much to it, so it is just a few lines explaining how things link up
22:36:20  <glx> there are many tools for that openttdguest
22:36:29  <TrueBrain> But, that should be in a repo, not a gist :)
22:36:54  <LordAro> TrueBrain: i was thinking something explaining the format of the files in the BaNaNaS repo too
22:37:07  <TrueBrain> Also in a gist :)
22:37:11  <LordAro> openttdguest: please explain in more detail what you are trying to achieve
22:37:46  <TrueBrain> Will put it on my todo. Anything else specific you would like to have documented?
22:37:48  <openttdguest> I would like to download network content outside the game
22:38:16  <LordAro> TrueBrain: a brief sentence on "why only partial md5s" as well :)
22:38:27  <TrueBrain> openttdguest : this is currently not supported till someone implements one of the issues we have open for this
22:38:34  <LordAro> because i couldn't answer ^ properly :)
22:38:47  <openttdguest> TrueBrain: is it supported how to get md5sum?
22:38:51  <TrueBrain> LordAro : will do!
22:39:23  <TrueBrain> openttdguest : no, currently this is not without faking a client or already knowing this
22:39:34  <openttdguest> I tried this: 09 00 01 01 00 02 42 76 49, but it doesn't work
22:40:04  <TrueBrain> https://github.com/OpenTTD/bananas-api/issues/29
22:40:13  <TrueBrain> Pull requests are welcome
22:40:33  *** gelignite has quit IRC
22:40:43  <TrueBrain> openttdguest : as LordAro mentioned, please state what problem you are trying to solve
22:41:32  <TrueBrain> LordAro : short answer btw is to make sure you can only easily download the latest version or you already need to know the md5sum
22:41:45  <TrueBrain> Call it a passphrase or something :)
22:44:28  <TrueBrain> We simply abuse the hash for that, basically :D But next month or so I will all make sure it is nicely written down in one of the repos :D
22:44:48  <openttdguest> TrueBrain: trying to download from browser, a CLIENT_INFO_LIST works (08 00 00 02 62 6d 08 18), but not a CLIENT_INFO_ID
22:45:42  <milek7> 24 bit is rather poor passphrase
22:45:48  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #8273: Center text and icons in the status bar vertically https://git.io/JJscn
22:47:03  <openttdguest> network content that is
22:47:19  <glx> as said it's not supported yet
22:47:52  <openttdguest> glx: I am trying to write a program to get the md5
22:48:53  <glx> I think you need to act like openttd
22:50:02  <TrueBrain> milek7 : it is 104 bits, but tnx for the nonconstructice remark
22:50:37  <TrueBrain> Hmm, no, 96 bits, oops, counting is difficult
22:50:50  <milek7> 24bit+counter is actually needed to download something
22:51:37  <TrueBrain> openttdguest : you confuse is a bit with the mention of 'from browser'. Via a browser you currently cannot download BaNaNaS content
22:52:16  <glx> (unless you know the complete url)
22:53:19  <openttdguest> glx: Yes, e.g., https://bananas.cdn.openttd.org/ai/41444d4c/4ccd92fb8f8f01045145be99a28e14a6/41444d4c-AdmiralAI-25.tar.gz, but I just need the md5sum
22:53:27  <TrueBrain> If you are faking a client, we don't really support that but you are free to do so within reason. See the OpenTTD client for an implementation for this
22:53:35  *** WormnestAndroid has joined #openttd
22:54:44  <TrueBrain> openttdguest : still we do not understand why you would need that. It is a technical detail, so what you are trying to do is unclear to us
22:55:25  <glx> IIRC openttd send a request to get the url, then it can download
22:55:32  <Timberwolf> I have been suckered into playing a FIRS Steeltown game as the only way to test that all of my various metal colours look good next to each other.
22:56:52  <glx> but everything is in openttd source
22:57:20  <openttdguest> glx: https://github.com/OpenTTD/OpenTTD/blob/master/src/network/network_content.cpp, for PACKET_CONTENT_CLIENT_INFO_ID, it seems like its 2 bytes packet size, 1 byte type, 2 bytes count, and 4 bytes id, but I do not get a response
22:57:53  <openttdguest> with type 0x0 packets, I get a response, but not type 0x1
22:58:16  <Timberwolf> Make that two games, I actually wanted FIRS 4 Steeltown not FIRS 3 Steeltown.
22:58:47  <TrueBrain> Seems we will never know what problem is being solved :D off to bed, night all
23:01:37  <openttdguest> TrueBrain: me? to download outside the game client
23:02:37  <_dp_> openttdguest, why download outside when you can download inside the game?
23:03:52  <milek7> openttdguest: what bytes you are sending?
23:04:23  <openttdguest> milek7: 09 00 01 01 00 02 42 76 49
23:06:50  <glx> https://github.com/OpenTTD/OpenTTD/blob/706c47265e3abf88341be8376532133bde582828/src/network/network_content.cpp#L336 you just need the 32bit id of content
23:09:24  <openttdguest> glx: for this newgrf, https://bananas.openttd.org/package/newgrf/49764202, it seems the 32bit id is 49 76 42 02, right?
23:09:38  <glx> yes
23:09:42  <TrueBrain> No
23:09:46  <glx> in reverse
23:09:55  <TrueBrain> Content-id is not the same as uniqueid
23:10:12  *** Eddi|zuHause has quit IRC
23:12:16  <openttdguest> TrueBrain: i see, where is the uniqueid?
23:12:30  <milek7> openttdguest: uniqueid is last 24bits of md5, and usually 0x00 at end
23:13:10  <milek7> eg. for querying admiralai -> 09 00 01 01 00 8e 14 a6 00
23:13:24  <TrueBrain> milek7 : shitty advise
23:13:44  <milek7> ?
23:13:51  <TrueBrain> openttdguest : type 2 allows you to query based on uniqueid, like the value you mentioned
23:14:11  <TrueBrain> Type 1 uses contentid, which you get from type 0 requests
23:14:50  <glx> with contentid you always get the latest release I think
23:15:00  <TrueBrain> With type2 too
23:15:39  <TrueBrain> Type2 packets also need the content-type btw
23:16:17  <TrueBrain> (As uniqueids are only unique within their category, strictly spoken)
23:17:48  <TrueBrain> But it all still depends on what you actually want to do, and why we keep asking this :)
23:18:02  <TrueBrain> Is it to download all content or just a specific version
23:18:13  <TrueBrain> Etc etc
23:18:20  <TrueBrain> Context is important :)
23:19:22  <TrueBrain> But normal client flow is: type 0 request, returns list of contentids, type 1 request for the ones the clients wants to download
23:19:34  <openttdguest> TrueBrain: is issue #29 (send download url via api) a program that inputs content id and outputs download url?
23:20:08  <TrueBrain> It adds the possibility of having download links on the webpage, yes
23:20:52  <openttdguest> TrueBrain: just for the newgame?
23:21:15  <glx> newgame means latest version
23:21:15  <TrueBrain> Yes. As by our terms of services, that is the only thing we offer
23:21:27  <TrueBrain> What glx says
23:22:01  <TrueBrain> It does not mean an in game new game :D
23:22:17  <openttdguest> TrueBrain: I am confused. type 1 requests use content id. Aren't I using content ids?
23:22:27  <TrueBrain> No, you use uniqueids
23:22:39  <TrueBrain> Contentids are only given from type0 requests
23:23:04  <openttdguest> TrueBrain: Do type 1 requests give md5sum?
23:23:07  <TrueBrain> And are unique to those two packets (well, two more places, but not important now)
23:23:48  <TrueBrain> openttdguest : only for the latest version of that content, yes. As does type2
23:24:13  <TrueBrain> (Ignoring some exceptions; nothing you will notice)
23:24:42  <glx> and you usually don't need to know the md5 to get a file
23:25:05  <glx> (unless when loading a savegame missing something)
23:25:10  <TrueBrain> But as we are flying blind here, it is hard to really help you
23:25:15  <openttdguest> TrueBrain: AdmiralAI I believe uses version 2 of the library while newgame is version 3, how does the game get the md5 for version 2?
23:25:36  <openttdguest> glx: Do you need the md5 to get it via http?
23:25:45  <TrueBrain> There is the exception, dependencies ;)
23:26:08  <TrueBrain> Type0 also shows older versions of content when it is a dependency of something else
23:27:35  <TrueBrain> openttdguest : okay, it becomes a tiny bit more clear what you try to do: never assemble the URL yourself. We change that often. Use the infrastructure in place to give you the complete URL. But LordAro pointed out to me that is not yet documented in a sane place how to do that
23:27:49  <TrueBrain> Mostly as nobody ever cared :D
23:28:00  <glx> :)
23:28:22  <glx> basically you request content list, then select ids in the list and ask for them
23:28:58  <TrueBrain> And because of historical reasons, another endpoint can give you the full url based on the contentid
23:29:50  <TrueBrain> openttdguest : you are the first I know that wants to do this, so basically we have nothing in place for it yet. The issue mentioned before is the proper solution I guess :)
23:30:14  <TrueBrain> Otherwise, follow exactly what the client does :)
23:30:21  <openttdguest> TrueBrain: looking at the code, does the game verify network content?
23:30:34  <TrueBrain> Define 'verify'
23:30:58  <openttdguest> TrueBrain: use a secure channel to verify file integrity
23:31:14  <TrueBrain> Still difficult to answer :D
23:31:30  <TrueBrain> It does not use an encrypted channel to protect in transit from spies
23:31:40  <TrueBrain> It uses TCP for integrity
23:32:56  <TrueBrain> So you can maninthemiddle the fuck out of it, but otherwise the integrity is there
23:33:55  <openttdguest> TrueBrain: is there anyway to tell if a file is completely downloaded?
23:34:08  <TrueBrain> Size, and in the end the checksum
23:34:23  <openttdguest> TrueBrain: using the partial checksum?
23:34:28  <TrueBrain> The md5sum is (partially) known
23:34:43  <TrueBrain> But the client does not do this actively to validate integrity
23:34:58  <openttdguest> TrueBrain: why not put the full checksum of newgame content on the website?
23:35:04  <glx> client trusts server and tcp
23:35:19  <TrueBrain> openttdguest : what problem are you solving?
23:36:36  <openttdguest> TrueBrain: well perhaps for the rare case of ambiguity of the first 8 bytes of the hash
23:37:31  <TrueBrain> The odds of a bit flip causing the same 8 bytes of md5sum is .... well, I did not run the math, but highly unlikely ;)
23:37:46  <TrueBrain> Not sure that is worth protection for :)
23:38:42  <TrueBrain> And there is no valid content with the same md5sumpartial (those uploads are denied)
23:39:22  <TrueBrain> Just remember this protocol is 15 years old. It hasn't been revised. But this part is the least of our issues :D
23:44:30  <TrueBrain> Anyway, I was off to get some sleep. It is 0140 here. Feel free to send an email if you want more info etc, but next month I will also update the docs on this!
23:44:44  *** openttdguest has quit IRC
23:45:14  <TrueBrain> Seems google translate doesn't know hi and goodbye :p

Powered by YARRSTE version: svn-trunk