IRC logs for #openttd on OFTC at 2020-07-15
#openttd / 2020 / 07 
            

00:03:01  *** Progman has quit IRC



00:03:04  *** tycoondemon has quit IRC



00:23:42  *** WormnestAndroid has quit IRC



00:29:31  *** openttdguest has joined #openttd



00:29:33  *** openttdguest has left #openttd



00:34:38  <TrueBrain> LordAro : yeah, most repos have documentation in their readme



00:35:02  <TrueBrain> There is also a gist with the design



00:35:07  <LordAro> TrueBrain: they all have "how to run this", but none of them have "how this works"



00:35:09  *** openttdguest has joined #openttd



00:35:21  <TrueBrain> I will soon put that somewhere more permanent



00:35:26  <LordAro> :)



00:35:34  <openttdguest> anyone know how to get full md5sum outside of game?



00:36:09  <TrueBrain> LordAro : but there really is not much to it, so it is just a few lines explaining how things link up



00:36:20  <glx> there are many tools for that openttdguest



00:36:29  <TrueBrain> But, that should be in a repo, not a gist :)



00:36:54  <LordAro> TrueBrain: i was thinking something explaining the format of the files in the BaNaNaS repo too



00:37:07  <TrueBrain> Also in a gist :)



00:37:11  <LordAro> openttdguest: please explain in more detail what you are trying to achieve



00:37:46  <TrueBrain> Will put it on my todo. Anything else specific you would like to have documented?



00:37:48  <openttdguest> I would like to download network content outside the game



00:38:16  <LordAro> TrueBrain: a brief sentence on "why only partial md5s" as well :)



00:38:26  <TrueBrain> openttdguest : this is currently not supported till someone implements one of the issues we have open for this



00:38:34  <LordAro> because i couldn't answer ^ properly :)



00:38:47  <openttdguest> TrueBrain: is it supported how to get md5sum?



00:38:51  <TrueBrain> LordAro : will do!



00:39:23  <TrueBrain> openttdguest : no, currently this is not without faking a client or already knowing this



00:39:34  <openttdguest> I tried this: 09 00 01 01 00 02 42 76 49, but it doesn't work



00:40:03  <TrueBrain> https://github.com/OpenTTD/bananas-api/issues/29



00:40:13  <TrueBrain> Pull requests are welcome



00:40:33  *** gelignite has quit IRC



00:40:43  <TrueBrain> openttdguest : as LordAro mentioned, please state what problem you are trying to solve



00:41:32  <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



00:41:45  <TrueBrain> Call it a passphrase or something :)



00:44:27  <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



00:44:48  <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



00:45:42  <milek7> 24 bit is rather poor passphrase



00:45:48  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #8273: Center text and icons in the status bar vertically https://git.io/JJscn



00:47:03  <openttdguest> network content that is



00:47:19  <glx> as said it's not supported yet



00:47:52  <openttdguest> glx: I am trying to write a program to get the md5



00:48:53  <glx> I think you need to act like openttd



00:50:02  <TrueBrain> milek7 : it is 104 bits, but tnx for the nonconstructice remark



00:50:37  <TrueBrain> Hmm, no, 96 bits, oops, counting is difficult



00:50:50  <milek7> 24bit+counter is actually needed to download something



00:51:37  <TrueBrain> openttdguest : you confuse is a bit with the mention of 'from browser'. Via a browser you currently cannot download BaNaNaS content



00:52:16  <glx> (unless you know the complete url)



00:53:19  <openttdguest> glx: Yes, e.g., https://bananas.cdn.openttd.org/ai/41444d4c/4ccd92fb8f8f01045145be99a28e14a6/41444d4c-AdmiralAI-25.tar.gz, but I just need the md5sum



00:53:27  <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



00:53:35  *** WormnestAndroid has joined #openttd



00:54:44  <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



00:55:25  <glx> IIRC openttd send a request to get the url, then it can download



00:55:32  <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.



00:56:52  <glx> but everything is in openttd source



00:57:20  <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



00:57:53  <openttdguest> with type 0x0 packets, I get a response, but not type 0x1



00:58:16  <Timberwolf> Make that two games, I actually wanted FIRS 4 Steeltown not FIRS 3 Steeltown.



00:58:47  <TrueBrain> Seems we will never know what problem is being solved :D off to bed, night all



01:01:37  <openttdguest> TrueBrain: me? to download outside the game client



01:02:37  <_dp_> openttdguest, why download outside when you can download inside the game?



01:03:52  <milek7> openttdguest: what bytes you are sending?



01:04:23  <openttdguest> milek7: 09 00 01 01 00 02 42 76 49



01:06:50  <glx> https://github.com/OpenTTD/OpenTTD/blob/706c47265e3abf88341be8376532133bde582828/src/network/network_content.cpp#L336 you just need the 32bit id of content



01:09:24  <openttdguest> glx: for this newgrf, https://bananas.openttd.org/package/newgrf/49764202, it seems the 32bit id is 49 76 42 02, right?



01:09:38  <glx> yes



01:09:42  <TrueBrain> No



01:09:46  <glx> in reverse



01:09:55  <TrueBrain> Content-id is not the same as uniqueid



01:10:12  *** Eddi|zuHause has quit IRC



01:12:16  <openttdguest> TrueBrain: i see, where is the uniqueid?



01:12:30  <milek7> openttdguest: uniqueid is last 24bits of md5, and usually 0x00 at end



01:13:10  <milek7> eg. for querying admiralai -> 09 00 01 01 00 8e 14 a6 00



01:13:24  <TrueBrain> milek7 : shitty advise



01:13:44  <milek7> ?



01:13:51  <TrueBrain> openttdguest : type 2 allows you to query based on uniqueid, like the value you mentioned



01:14:11  <TrueBrain> Type 1 uses contentid, which you get from type 0 requests



01:14:50  <glx> with contentid you always get the latest release I think



01:15:00  <TrueBrain> With type2 too



01:15:39  <TrueBrain> Type2 packets also need the content-type btw



01:16:17  <TrueBrain> (As uniqueids are only unique within their category, strictly spoken)



01:17:48  <TrueBrain> But it all still depends on what you actually want to do, and why we keep asking this :)



01:18:02  <TrueBrain> Is it to download all content or just a specific version



01:18:13  <TrueBrain> Etc etc



01:18:20  <TrueBrain> Context is important :)



01:19:22  <TrueBrain> But normal client flow is: type 0 request, returns list of contentids, type 1 request for the ones the clients wants to download



01:19:34  <openttdguest> TrueBrain: is issue #29 (send download url via api) a program that inputs content id and outputs download url?



01:20:08  <TrueBrain> It adds the possibility of having download links on the webpage, yes



01:20:52  <openttdguest> TrueBrain: just for the newgame?



01:21:15  <glx> newgame means latest version



01:21:15  <TrueBrain> Yes. As by our terms of services, that is the only thing we offer



01:21:27  <TrueBrain> What glx says



01:22:01  <TrueBrain> It does not mean an in game new game :D



01:22:17  <openttdguest> TrueBrain: I am confused. type 1 requests use content id. Aren't I using content ids?



01:22:27  <TrueBrain> No, you use uniqueids



01:22:39  <TrueBrain> Contentids are only given from type0 requests



01:23:04  <openttdguest> TrueBrain: Do type 1 requests give md5sum?



01:23:07  <TrueBrain> And are unique to those two packets (well, two more places, but not important now)



01:23:48  <TrueBrain> openttdguest : only for the latest version of that content, yes. As does type2



01:24:13  <TrueBrain> (Ignoring some exceptions; nothing you will notice)



01:24:42  <glx> and you usually don't need to know the md5 to get a file



01:25:05  <glx> (unless when loading a savegame missing something)



01:25:10  <TrueBrain> But as we are flying blind here, it is hard to really help you



01:25:15  <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?



01:25:36  <openttdguest> glx: Do you need the md5 to get it via http?



01:25:45  <TrueBrain> There is the exception, dependencies ;)



01:26:08  <TrueBrain> Type0 also shows older versions of content when it is a dependency of something else



01:27:35  <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



01:27:49  <TrueBrain> Mostly as nobody ever cared :D



01:28:00  <glx> :)



01:28:22  <glx> basically you request content list, then select ids in the list and ask for them



01:28:58  <TrueBrain> And because of historical reasons, another endpoint can give you the full url based on the contentid



01:29:50  <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 :)



01:30:14  <TrueBrain> Otherwise, follow exactly what the client does :)



01:30:21  <openttdguest> TrueBrain: looking at the code, does the game verify network content?



01:30:34  <TrueBrain> Define 'verify'



01:30:58  <openttdguest> TrueBrain: use a secure channel to verify file integrity



01:31:14  <TrueBrain> Still difficult to answer :D



01:31:30  <TrueBrain> It does not use an encrypted channel to protect in transit from spies



01:31:40  <TrueBrain> It uses TCP for integrity



01:32:56  <TrueBrain> So you can maninthemiddle the fuck out of it, but otherwise the integrity is there



01:33:55  <openttdguest> TrueBrain: is there anyway to tell if a file is completely downloaded?



01:34:08  <TrueBrain> Size, and in the end the checksum



01:34:23  <openttdguest> TrueBrain: using the partial checksum?



01:34:28  <TrueBrain> The md5sum is (partially) known



01:34:43  <TrueBrain> But the client does not do this actively to validate integrity



01:34:58  <openttdguest> TrueBrain: why not put the full checksum of newgame content on the website?



01:35:04  <glx> client trusts server and tcp



01:35:19  <TrueBrain> openttdguest : what problem are you solving?



01:36:36  <openttdguest> TrueBrain: well perhaps for the rare case of ambiguity of the first 8 bytes of the hash



01:37:31  <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 ;)



01:37:46  <TrueBrain> Not sure that is worth protection for :)



01:38:42  <TrueBrain> And there is no valid content with the same md5sumpartial (those uploads are denied)



01:39:22  <TrueBrain> Just remember this protocol is 15 years old. It hasn't been revised. But this part is the least of our issues :D



01:44:30  <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!



01:44:44  *** openttdguest has quit IRC



01:45:14  <TrueBrain> Seems google translate doesn't know hi and goodbye :p



02:00:55  <b_jonas> gray fish on the gray background of the station window is a bit hard to see



02:27:25  *** WormnestAndroid has quit IRC



02:27:58  *** WormnestAndroid has joined #openttd



02:41:48  *** Eddi|zuHause has joined #openttd



02:53:40  *** b_jonas has quit IRC



03:52:23  *** Flygon has joined #openttd



04:34:19  *** glx has quit IRC



04:50:03  *** Smedles has joined #openttd



06:15:06  <CornsMcGowan[m]> fish?



07:34:12  *** sla_ro|master has joined #openttd



07:37:52  <Eddi|zuHause> <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 ;) <-- i'm pretty sure if that were possible, it would implode the complete algorithm



08:01:57  *** sla_ro|master has quit IRC



08:29:00  *** snail_UES_ has quit IRC



08:50:54  <TrueBrain> https://www.reddit.com/r/openttd/comments/hrbni4/why_a_developer_of_openttd_should_release_openttd/



08:52:52  <Eddi|zuHause> that url is misleading



08:57:32  <dwfreed> reddit slugging sucks



09:05:38  *** nielsm has joined #openttd



09:06:30  <LordAro> lol



09:12:51  <Eddi|zuHause> https://www.reddit.com/r/openttd/comments/hrbni4/btw_you_can_write_just_about_anything_here



09:17:36  *** HerzogDeXtEr has joined #openttd



09:30:32  *** cHawk- has quit IRC



09:46:01  *** iSoSyS has joined #openttd



10:16:51  *** cHawk- has joined #openttd



10:32:57  *** WormnestAndroid has quit IRC



10:33:10  *** WormnestAndroid has joined #openttd



11:06:09  <CornsMcGowan[m]> whats slugging



11:06:18  <CornsMcGowan[m]> also wild idea: discontinuous landscape



12:02:48  <TrueBrain> The comments are funny, lot of (unfounded) speculation :D



12:09:10  <Eddi|zuHause> turns out i'm terrible at tropico



12:09:15  *** iSoSyS has quit IRC



12:10:04  <Eddi|zuHause> particularly, making money and keeping the superpowers happy...



12:15:58  <LordAro> TrueBrain: someone is wrong on the internet :o



12:33:26  *** tokai|noir has joined #openttd



12:33:26  *** ChanServ sets mode: +v tokai|noir



12:40:13  *** tokai has quit IRC



13:01:46  *** HerzogDeXtEr1 has joined #openttd



13:07:37  *** HerzogDeXtEr has quit IRC



13:26:40  *** snail_UES_ has joined #openttd



13:33:23  <supermop_Home> hello



14:00:40  <DorpsGek_III> [OpenTTD/OpenTTD] Kuhnovic opened pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JJsQJ



14:39:29  *** tokai has joined #openttd



14:39:29  *** ChanServ sets mode: +v tokai



14:46:22  *** tokai|noir has quit IRC



14:52:59  *** gelignite has joined #openttd



14:54:45  *** glx has joined #openttd



14:54:45  *** ChanServ sets mode: +v glx



14:57:01  <DorpsGek_III> [OpenTTD/OpenTTD] Kuhnovic commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JJs5Z



15:04:40  <DorpsGek_III> [OpenTTD/OpenTTD] Kuhnovic updated pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JJsQJ



15:07:03  <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JJs5S



15:23:24  <DorpsGek_III> [OpenTTD/OpenTTD] Kuhnovic updated pull request #8274: Feature: option to auto remove signals when in the way during rail co… https://git.io/JJsQJ



16:06:43  *** gelignite has quit IRC



16:16:37  *** WormnestAndroid has quit IRC



16:17:15  *** WormnestAndroid has joined #openttd



16:58:52  <DorpsGek_III> [OpenTTD/OpenTTD] glx22 merged pull request #8273: Center text and icons in the status bar vertically https://git.io/JJ355



18:07:37  *** gelignite has joined #openttd



18:12:58  *** cHawk- has quit IRC



18:42:53  *** Flygon has quit IRC



18:49:46  *** frosch123 has joined #openttd



19:04:09  *** skrzyp2 is now known as skrzyp



19:04:56  *** Progman has joined #openttd



19:05:43  *** Wolf01 has joined #openttd



19:20:20  *** cHawk- has joined #openttd



20:51:50  *** cHawk_ has joined #openttd



20:57:57  *** cHawk- has quit IRC



21:10:28  <glx> ah I broke compile farm



21:10:47  *** sla_ro|master has joined #openttd



21:26:39  *** frosch123 has quit IRC



21:39:10  <Wolf01> @seen andythenorth



21:39:10  <DorpsGek> Wolf01: andythenorth was last seen in #openttd 3 days, 23 hours, 12 minutes, and 18 seconds ago: <andythenorth> :)



21:39:14  <Wolf01> Hmmm



21:39:23  <Wolf01> Suspicious



21:43:02  <supermop_Home> indeed



21:43:16  <supermop_Home> who am I going to tell about all the steelmills I saw



21:43:45  <supermop_Home> wait



21:44:15  <supermop_Home> my timestamps just changed by an hour



21:46:12  <supermop_Home> https://imgur.com/yOXbZOf



21:46:48  <supermop_Home> it's certainly 3:46 here, not 2:46



21:47:16  <Wolf01> Aahah



21:47:31  <Wolf01> NTP poison



21:52:47  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 opened pull request #45: Fix 091287d: [release-docs] cmake requires a compiler https://git.io/JJGIx



22:01:10  <DorpsGek_III> [OpenTTD/CompileFarm] TrueBrain approved pull request #45: Fix 091287d: [release-docs] cmake requires a compiler https://git.io/JJGLc



22:01:18  <TrueBrain> I am sure you have a reason for that :P YOLO! :D



22:03:14  *** Eddi|zuHause2 has joined #openttd



22:03:34  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 merged pull request #45: Fix 091287d: [release-docs] cmake requires a compiler https://git.io/JJGIx



22:03:35  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 pushed 1 commits to master https://git.io/JJGL2



22:03:35  <DorpsGek_III>   - Fix 091287d: [release-docs] cmake requires a compiler (#45) (by glx22)



22:04:56  <glx> hmm DorpsGek_III is noisy on this repo :)



22:07:52  *** Eddi|zuHause has quit IRC



22:14:54  <glx> ok let's try a rerun of failed nightly jobs



22:15:35  <michi_cc> Wolf01: https://www.tt-forums.net/viewtopic.php?p=1234185#p1234185 :(



22:16:05  <glx> should be safe as manifest was fixed the last time



22:16:37  *** WormnestAndroid has quit IRC



22:16:56  <Wolf01> Meh, ragequits, what do they solve?



22:17:08  <glx> ok failed again :(



22:17:08  *** WormnestAndroid has joined #openttd



22:19:11  <glx> oups my fault, it's not easy to do compilefarm stuff



22:42:19  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 opened pull request #46: Fix 3775a12: [release-docs] use default clang https://git.io/JJGql



22:42:28  <glx> same player try again :)



22:56:04  *** Eddi|zuHause2 is now known as Eddi|zuHause



23:08:11  *** Speeder has joined #openttd



23:18:12  *** nielsm has quit IRC



23:18:22  *** WormnestAndroid has quit IRC



23:18:29  *** WormnestAndroid has joined #openttd



23:33:24  <DorpsGek_III> [OpenTTD/CompileFarm] TrueBrain approved pull request #46: Fix 3775a12: [release-docs] use default clang https://git.io/JJG3k



23:57:44  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 merged pull request #46: Fix 3775a12: [release-docs] use default clang https://git.io/JJGql



23:57:45  <DorpsGek_III> [OpenTTD/CompileFarm] glx22 pushed 1 commits to master https://git.io/JJGsY



23:57:45  <DorpsGek_III>   - Fix 3775a12: [release-docs] use default clang (#46) (by glx22)