IRC logs for #openttd on OFTC at 2019-11-23
⏴ go to previous day
00:11:19 *** supermop_work_ has quit IRC
00:11:49 *** supermop_work_ has joined #openttd
00:17:14 *** WormnestAndroid has quit IRC
00:17:52 *** WormnestAndroid has joined #openttd
01:08:31 *** supermop_work_ has quit IRC
01:09:24 *** supermop_work_ has joined #openttd
01:21:33 *** Wormnest has joined #openttd
01:38:12 *** Thedarkb-X40 has joined #openttd
01:44:36 *** supermop_work_ has quit IRC
01:45:31 *** supermop_work_ has joined #openttd
02:16:14 *** supermop_work_ has quit IRC
02:17:08 *** supermop_work_ has joined #openttd
02:47:21 *** supermop_work_ has quit IRC
02:49:18 *** supermop_work_ has joined #openttd
02:50:52 *** ChanServ sets mode: +v tokai
03:19:30 *** supermop_work_ has quit IRC
03:20:08 *** supermop_work_ has joined #openttd
03:50:20 *** supermop_work_ has quit IRC
03:50:46 *** supermop_work_ has joined #openttd
04:16:26 *** Wormnest has joined #openttd
04:20:59 *** supermop_work_ has quit IRC
04:21:34 *** supermop_work_ has joined #openttd
04:51:48 *** supermop_work_ has quit IRC
04:52:20 *** supermop_work_ has joined #openttd
05:22:34 *** supermop_work_ has quit IRC
05:23:55 *** supermop_work_ has joined #openttd
05:54:06 *** supermop_work_ has quit IRC
05:54:35 *** supermop_work_ has joined #openttd
06:24:48 *** supermop_work_ has quit IRC
06:27:01 *** supermop_work_ has joined #openttd
06:57:15 *** supermop_work_ has quit IRC
06:58:12 *** supermop_work_ has joined #openttd
07:28:24 *** supermop_work_ has quit IRC
07:29:03 *** supermop_work_ has joined #openttd
07:34:41 *** D-HUND is now known as debdog
07:54:31 *** Wormnest has joined #openttd
07:59:15 *** supermop_work_ has quit IRC
07:59:50 *** supermop_work_ has joined #openttd
08:26:16 *** andythenorth has joined #openttd
08:30:03 *** supermop_work_ has quit IRC
08:30:38 *** supermop_work_ has joined #openttd
08:53:52 *** sla_ro|master has joined #openttd
09:00:51 *** supermop_work_ has quit IRC
09:01:27 *** supermop_work_ has joined #openttd
09:01:51 <andythenorth> what's the version of a grf if it's not on a tag?
09:02:04 <andythenorth> the revision is trivially the rev count, we decided
09:02:13 <andythenorth> but is the version x.y.zM?
09:07:12 *** Progman has joined #openttd
09:07:52 <Wolf01> So... Transport Fever 2...
09:11:21 <Wolf01> andythenorth: do you have some LBG/metallic technic corrugated pipe? :P
09:11:41 <andythenorth> I think I do somewhere
09:11:56 <andythenorth> got boxes of random parts in a cupboard
09:12:24 <Wolf01> I only have 2L bits of it... I need some 6L and TLG doesn't want to sell them to me
09:14:28 <Wolf01> Yeah... but not for 2 pieces
09:15:50 <Wolf01> Hmmm, it seem I have 6 of them.. where?
09:18:49 <Wolf01> Rebrickable told me "there", but where is there?
09:28:16 <andythenorth> F1 racer has some in the engine?
09:28:24 <andythenorth> I recall because my kids smashed it up
09:31:39 *** supermop_work_ has quit IRC
09:31:59 <Wolf01> I don't like to cut tubes because I might want to rebuild sets
09:32:06 *** supermop_work_ has joined #openttd
09:48:32 *** gelignite has joined #openttd
09:50:27 *** frosch123 has joined #openttd
10:02:19 *** supermop_work_ has quit IRC
10:03:02 *** supermop_work_ has joined #openttd
10:04:34 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #7840: Fix #7836: check coherency of newgrf parameter min/max https://git.io/Jei0u
10:05:14 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #7000: Some NewGRF variables concerning railtypes https://git.io/Je2tX
10:05:42 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #7000: Some NewGRF variables concerning railtypes https://git.io/Jei0z
10:05:52 <Wolf01> Cat wrote " l.,,,,,,,,,,,,,,,,,,,,,,,,,"
10:08:53 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro dismissed a review for pull request #7286: Add #2155: newheightmapgame command https://git.io/fhN7O
10:24:13 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #7486: Fix: AI/GS settings with the flag SCRIPTCONFIG_RANDOM could be altered after loading from a savegame. https://git.io/JeiEL
10:33:15 *** supermop_work_ has quit IRC
10:33:44 *** supermop_work_ has joined #openttd
10:56:04 * andythenorth tried to run python processing pool with 4036 workers :P
11:03:56 *** supermop_work_ has quit IRC
11:05:24 *** supermop_work_ has joined #openttd
11:09:47 <TrueBrain> sadly there is no way to know how much it is costing, except for letting it run for a day or so to see
11:15:37 <andythenorth> TrueBrain: hurrah!
11:15:49 <andythenorth> and you got a subdomain!
11:16:04 <andythenorth> I don't much like the path to my AWS bucket :P
11:16:16 <andythenorth> but I had an idea for Bananas 2 publishing the docs :P
11:17:50 <TrueBrain> of course I have a subdomain .. even a valid certificate for it :P Otherwise it would all be a bit weird :D
11:17:56 <TrueBrain> but that was easier than I expected
11:18:00 <TrueBrain> (this is not hosted from an S3 btw)
11:18:18 <TrueBrain> strictly seen the website can, but having it in a docker makes it easier for people to work on it :)
11:18:30 <TrueBrain> NLB -> ALB -> ECS -> Docker instance
11:18:36 <TrueBrain> NLB -> ALB -> ECS -> EC2 -> Docker instance
11:18:43 <frosch123> so the certificate mails were actually legit?
11:18:47 <frosch123> i thought they were scam
11:19:09 <TrueBrain> frosch123: the comodo ones?
11:19:12 <TrueBrain> that is unrelated :)
11:19:39 <frosch123> well, there were two similar mails from different addresses
11:19:44 <frosch123> one with a zip attachment
11:20:01 <frosch123> i never got a certifiate myself, but it looked dubious to me :p
11:20:02 <TrueBrain> don't know what the PositiveSSL Wildcard is about tbh
11:20:11 <TrueBrain> owh, they are both Comodo
11:20:23 <TrueBrain> but yeah, they are legit; they are still used on the old infrastructure
11:20:39 * andythenorth had the idea to have Bananas 2 build grfs :P
11:21:09 <TrueBrain> but no, the certificate on the URL above is an AWS certificate
11:21:17 <TrueBrain> as doing letsencrypt was too complex for me :(
11:21:26 <TrueBrain> not that it matters .. should both be free
11:21:30 <TrueBrain> andythenorth: not a bad idea
11:21:46 <TrueBrain> I always wanted that all content on BaNaNaS also had their source attached
11:21:51 <TrueBrain> but I got flamed for that 15 years ago :P
11:22:04 <TrueBrain> turned out artists don't work like that, or something
11:22:19 <andythenorth> I don't actually mean that
11:22:24 <frosch123> they still do not work like that, andy is the only exception
11:22:49 <TrueBrain> that is something I learnt the hard way during those 15 years ..
11:23:08 <TrueBrain> as an engineer it is hard to understand that mindset; and I am sure they don't understand my mindset in this :D
11:23:14 <TrueBrain> it is not a complaint btw; just an observation
11:23:18 <andythenorth> frosch123 has that quote
11:23:31 <frosch123> the cliff stoll one?
11:23:52 <TrueBrain> andythenorth: but you can think about BaNaNaS2 all you want; I would just be happy to have a small up-to-date website where people can upload their content :)
11:23:56 <TrueBrain> that is all I ask for :D
11:24:17 <andythenorth> "artists think he's a great programmer, programmers think he's a great artist"
11:24:31 <andythenorth> it's quite brutal
11:25:01 <andythenorth> applies to me, V and pikka
11:25:44 <andythenorth> I arrived at the bananas 2 idea because of translations
11:25:58 <andythenorth> I don't want to give a random web service full commit access to my repos
11:26:23 <andythenorth> so I wondered about having the translator checkout my repo as a remote, then build it
11:26:51 <andythenorth> I'd never need the translations locally
11:27:13 <andythenorth> we could also ditch the eints database and just commit translations directly to git
11:27:25 <andythenorth> and we could build a rev automatically and release it whenever translations update
11:27:36 <andythenorth> instead of waiting months until the author does a release for new strings
11:29:52 <TrueBrain> right, time to destroy the AWS deployment again :)
11:30:22 <TrueBrain> tomorrow this will show me how much this would cost, give or take
11:30:23 <andythenorth> is it repeatable?
11:30:33 <TrueBrain> owh yeah, it is fully Infrastructure as Code
11:30:40 <andythenorth> cattle or pets? :P
11:30:41 <TrueBrain> so I can deploy this when-ever how-ever-often I want
11:31:17 <SpComb> fully destroyable apart from the bit that isn't
11:31:20 <TrueBrain> people doing pets should be killed
11:31:57 <TrueBrain> good thing about AWS, you can use CDK (or CloudFormation) .. that tracks the resources you created :)
11:31:59 <TrueBrain> so happy with that :D
11:32:08 *** Wormnest has joined #openttd
11:32:18 <andythenorth> we might have made a pet out of Ansible :P
11:34:17 <TrueBrain> I like nginx .. max memory usage was <1% .. I reserved 10% .. :D
11:34:32 <SpComb> also, NLB -> ALB? Are you using the lambda hacks to get a static UP or what?
11:35:13 <TrueBrain> well, I still have to add that hack, but yes
11:35:17 <TrueBrain> as I also have to hack IPv6 in there
11:35:24 <TrueBrain> as freaking NLB doesn't support IPv6 :(
11:35:39 *** supermop_work_ has quit IRC
11:35:42 <TrueBrain> so in reality it is (NLB + (2x EC2)) -> ALB
11:35:44 <TrueBrain> but who is counting
11:36:08 <SpComb> huh, what are the instances doing
11:36:44 <TrueBrain> no; you never NAT IPv6
11:36:51 <SpComb> if yoi just need static IPv4 for ALB you can use AWS Global Accelerator, but I think it's also IPv4-only
11:37:04 <TrueBrain> which is really shitty
11:37:18 <TrueBrain> so possibly I run IPv4 directly via ALB (and with GA)
11:37:29 <TrueBrain> but the IPv6 flow will be shitty (read: unmanaged)
11:37:52 *** supermop_work_ has joined #openttd
11:38:13 <TrueBrain> can't believe that in 2019 IPv6 is not considered commodity
11:38:22 <SpComb> lack of IPv6 for NLB is a little surprising
11:38:37 <SpComb> given that AWS has AFAIK the best VPC support for IPv6
11:38:40 <TrueBrain> main issue is that we have 2 services running on non-HTTP(S)
11:38:58 <TrueBrain> for that I sadly need the NLB to route into the ECS
11:39:21 <TrueBrain> lucky for me, they are both on different hostnames
11:40:12 <TrueBrain> in the early days that sounded like a good idea, not having them forced on the same IP .. they have been for years .. so I hope it also really works them not being on the same IP :P
11:40:44 <TrueBrain> now I come to think of it, it might be cheaper to just skip the NLB, and run both IPv4 and IPv6 for the TCP/UDP on the EC2
11:41:20 <SpComb> what are you using to LB the IPv6 traffic?
11:41:50 <TrueBrain> (as that is the only one that can do UDP LB wit Proxy Protocol .. can you believe that? No other piece of software does that .....)
11:42:07 <TrueBrain> HAProxy refuses to add proxy protocol support for UDP proxy
11:42:11 <TrueBrain> it is unbelievable :(
11:42:28 <TrueBrain> (I might be a bit frustrated how my experience is so far :P)
11:42:28 <SpComb> proxy protocol is pretty esoteric, never even wanted to use it
11:42:36 <andythenorth> hmm...why is Iron Horse grf so big filesize?
11:42:41 <andythenorth> that will be eating bandwidth :(
11:42:59 <andythenorth> sprites are only 4MB
11:43:03 <TrueBrain> SpComb: we need it, basically, as for the UDP we need to know the source IP
11:43:15 <TrueBrain> proxy protocol is used more and more in these setups
11:43:17 <andythenorth> oh wow 18MB of nfo
11:43:22 <TrueBrain> it is pretty nice tbh .. no stupid HTTP header and shit
11:44:21 <TrueBrain> our UDP is most special: you register to the master server, and the updater starts to query the source IP of how you registered to the masterserver + port in the registration packet
11:44:45 <TrueBrain> it is really difficult to bring this to the cloud, as I found out :)
11:45:41 <SpComb> for IPv4 yoi can avoid it because you have a proper network LB with source address preservation
11:45:41 <TrueBrain> but okay, it mostly means I have to write a lambda that updates the nginx configuration when ECS changes
11:45:52 <TrueBrain> yup .. NLB is exactly what I need
11:46:06 <TrueBrain> but ... 5+% of our users can't use IPv4 :(
11:46:22 <TrueBrain> we have a lot of IPv6 only users
11:46:31 <SpComb> hard to believe but ok
11:46:38 <TrueBrain> (well, CNAT on IPv4, but you can't host a server like that)
11:46:58 <TrueBrain> you do know large parts of Asia no longer support their users with IPv4, right? :)
11:46:59 <SpComb> does it let you host an IPv6 server? :D
11:47:02 <TrueBrain> (has been for a few years btw)
11:47:23 <TrueBrain> 10+% of our traffic is IPv6 btw
11:47:39 <TrueBrain> but I guess that happens if you support IPv6 for 12+ years now
11:47:46 <TrueBrain> and for some reason, AWS doesn't on their NLB :P
11:48:48 <SpComb> I wonder if you could hack IPVS to work in an AWS VPC without SNAT... maybe with your own EC2 backend instances and funny routing rules, but probably out of the question for ECS
11:49:17 <TrueBrain> ECS runs on EC2 in our case; at least till Fargate supports filesystems :P
11:49:50 <TrueBrain> and we also run NAT Instance (not NAT Gateway, as fuck that is expensive)
11:50:01 <TrueBrain> so if I change the NAT Instance to use IPVS, that might just work tbh
11:50:07 <TrueBrain> IPv6 is a bit more difficult
11:50:43 <TrueBrain> ECS runs on a local IP, so all traffic is sent back to the NAT Instance
11:50:48 <TrueBrain> IPVS picks it up from there
11:50:58 <TrueBrain> those routes are easier to configure, basically
11:51:05 <TrueBrain> IPv6 ... is more difficult
11:51:16 <TrueBrain> as the ECS EC2 instances will also get a public IPv6
11:51:21 <TrueBrain> (at least, if you follow spec)
11:51:30 <SpComb> doesn't it let you set up an instance route for IPv6?
11:51:48 <TrueBrain> it does; but it would break spec
11:51:50 <TrueBrain> not sure what the harm is
11:53:05 <TrueBrain> main issue for IPv4 and IPVS I guess would be that the public IP of the IPVS also needs to be on the ECS EC2 instances (otherwise the kernel will disregard the packet)
11:53:19 <TrueBrain> not difficult, just needs configuration, I guess
11:53:28 <TrueBrain> not a bad idea, need to think about it a bit, if it is doable :)
11:53:36 <TrueBrain> I just hope AWS announced IPv6 support for NLB next week :P
11:54:25 <TrueBrain> I still don't really understand how NLB works in AWS .. so little documentation how it works on a technical level :P
11:54:48 <SpComb> yeah, IPv6 UDP LB is certainly going to be painful on AWS one way or another until they bring official NLB support
11:55:10 <SpComb> time to upgrade the master server registration to use HTTP
11:55:15 <TrueBrain> the easiest solution by far is to not have the ECS in the private subnet, but in the public .. but that is just wrong
11:55:21 <TrueBrain> I honestly considered it
11:55:32 <TrueBrain> I even considered a small components that rewrites the UDP to HTTP on the edge
11:56:12 <TrueBrain> (as that would also mean strictly seen I could use API Gateway and Lambdas :P)
11:56:30 <SpComb> tbh I would prefer that over an UDP proxy, assuming that having the clients switch to HTTP was on the roadmap
11:56:45 <TrueBrain> OpenTTD has a roadmap? :P
11:57:06 <SpComb> up to your Product Owner :)
11:57:18 <TrueBrain> but yeah, it shouldn't be hard to change the master server registration to HTTP
11:57:23 <TrueBrain> we already use HTTP for downloads
11:57:31 <TrueBrain> I just wish we could make it HTTPS
11:57:34 <TrueBrain> but we have no SSL library
11:58:16 <SpComb> NSA is logging all the openttd multiplayer chats right now
11:58:30 <TrueBrain> well, the network protocol itself is also unencrypted yes, so yeah
11:58:50 <TrueBrain> my main issue with an UDP -> HTTP component btw is that it can easily be DoS'd
11:59:02 <TrueBrain> so it requires some engineering to prevent that
11:59:28 <TrueBrain> but if I do that for both the UDP and TCP, it would mean I don't need the NLB
11:59:50 <SpComb> more DoS:able than the master server itself?
12:00:02 <TrueBrain> and I guess I could put them in an Auto Scaling Group, and have a lambda to update DNS ..
12:01:56 <TrueBrain> okay, something to think about :)
12:02:10 <TrueBrain> never considered something we did 15 years ago to be difficult :P
12:03:42 <TrueBrain> would be nice if their Global Accelerator also supported IPv6, I guess :)
12:08:05 *** supermop_work_ has quit IRC
12:08:39 *** supermop_work_ has joined #openttd
12:22:25 *** HerzogDeXtEr has joined #openttd
12:30:24 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #7595: Possible fix for #7430: when train visits station, only reset time_since_pickup if has room to load https://git.io/Jeiuh
12:31:49 <andythenorth> Horse is currently 19MB
12:31:56 <andythenorth> it has one roster of vehicles
12:32:11 <andythenorth> I'm intending to do 5 rosters, so extrapolating gives 100MB
12:32:21 <andythenorth> OpenTTD is 7.5MB
12:32:27 <andythenorth> something is wrong in Gotham
12:38:53 *** supermop_work_ has quit IRC
12:40:09 *** supermop_work_ has joined #openttd
12:41:43 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro requested changes for pull request #7817: Feature: Topview screenshot (a.k.a minimap) https://git.io/Jeizk
12:45:32 <andythenorth> oh I forgot this one
13:10:23 *** supermop_work_ has quit IRC
13:10:59 *** supermop_work_ has joined #openttd
13:13:26 *** Wormnest has joined #openttd
13:20:29 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #7730: Change: Use vehicle model age for station rating calculation https://git.io/JeizR
13:27:57 *** Maarten has joined #openttd
13:37:39 *** HerzogDeXtEr1 has joined #openttd
13:41:12 *** supermop_work_ has quit IRC
13:41:59 *** supermop_work_ has joined #openttd
13:50:09 *** Maarten has joined #openttd
14:11:45 <DorpsGek_III> [OpenTTD/OpenTTD] Awizonosz opened issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
14:12:12 *** supermop_work_ has quit IRC
14:13:44 *** supermop_work_ has joined #openttd
14:15:15 <andythenorth> Wolf01: on a heavy wrecker / tow truck, what part shall I attach to the end of the winch line? :P
14:15:32 <andythenorth> hook? Or pin joiner?
14:17:29 <Wolf01> Metal hook if you have it
14:18:10 <andythenorth> unrealistic but the most boss?
14:18:36 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
14:19:04 <andythenorth> hmm looks like IRL has snap hooks, like giant carabiners
14:21:04 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro merged pull request #7840: Fix #7836: check coherency of newgrf parameter min/max https://git.io/JeKlG
14:21:05 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro closed issue #7836: Negativ Integer values for GRF Parameters crashes the game https://git.io/Jeozl
14:24:57 <Wolf01> I already spent over 200€ on it, and used parts from 3 mercedes arocs and other big sets
14:27:16 <LordAro> interesting, that save seems to cause an infinite loop
14:43:48 <DorpsGek_III> [OpenTTD/OpenTTD] Awizonosz commented on issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
14:43:58 *** supermop_work_ has quit IRC
14:44:39 *** supermop_work_ has joined #openttd
14:49:18 *** Thedarkb-X40 has joined #openttd
14:50:55 <LordAro> where are crash files put on Windows?
14:52:06 <glx> next to openttd.exe, or in documents
14:52:21 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
14:55:33 <glx> hmm the screenshot doesn't show a crash but a hang it seems
14:55:59 <glx> I think its (doesn't respond) in the title bar
14:56:02 <LordAro> certainly what i'm getting, but on a different date to them
15:01:35 <DorpsGek_III> [OpenTTD/website] glx22 commented on pull request #111: Add: 'Developer Files' section on download pages https://git.io/Jei2P
15:02:06 <DorpsGek_III> [OpenTTD/OpenTTD] Awizonosz commented on issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
15:02:43 <DorpsGek_III> [OpenTTD/website] LordAro approved pull request #111: Add: 'Developer Files' section on download pages https://git.io/Jei2M
15:03:00 <DorpsGek_III> [OpenTTD/website] LordAro merged pull request #111: Add: 'Developer Files' section on download pages https://git.io/Je2Eq
15:03:46 <glx> hmm old crashlog, useless because overflowed by news messages
15:05:00 <LordAro> oh, definitely unrelated then
15:07:52 <DorpsGek_III> [OpenTTD/nml] LordAro commented on pull request #59: Simplify pillow imports and version detection https://git.io/Jei2j
15:10:18 <DorpsGek_III> [OpenTTD/nml] LordAro approved pull request #60: Avoid duplicating rail/road/tramtype table code https://git.io/JeiaJ
15:10:33 <DorpsGek_III> [OpenTTD/nml] LordAro merged pull request #56: Add: regression workflow for PR and push to master https://git.io/JeVDh
15:11:10 <LordAro> glx: NML is missing review requirements on PRs, btw
15:12:51 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
15:14:26 <glx> I think review requirement is now added
15:14:53 *** supermop_work_ has quit IRC
15:17:03 *** supermop_work_ has joined #openttd
15:17:40 <DorpsGek_III> [OpenTTD/OpenTTD] Awizonosz commented on issue #7842: Crash on one my save in every version of OpenTTD https://git.io/Jeig0
15:21:31 <LordAro> interesting that it does eventually resolve itself though
15:21:37 <LordAro> so not actually an infinite loop
15:21:59 <glx> big map with many stations ?
15:22:26 <LordAro> well that's part of it, for sure
15:23:02 <glx> but I though linkgraph was threaded to not impact the game
15:23:26 <LordAro> it is, but something is waiting for the thread to complete still
15:23:27 <glx> well maybe if it takes really too long that may happen
15:32:02 <glx> hmm what happens if recalc interval is increased ?
15:33:10 <glx> I guess recalc happens too frequently for the data amount
15:33:12 <frosch123> yes, there is a setting to control how long the computation takes. the calculation runs in parallel, but it is merged at the deterministic timestamp into the gamestate
15:36:55 <glx> changing recalc_time and recalc_interval would probably help
15:47:16 *** supermop_work_ has quit IRC
15:47:45 *** supermop_work_ has joined #openttd
15:50:36 <glx> (ok it's a debug build but that's so slow)
15:51:16 <LordAro> glx: the same savegame? it was running at ~10fps for me
16:01:09 <LordAro> well that's something
16:01:29 *** Wormnest has joined #openttd
16:03:53 <glx> recalc_interval 4days, recalc_time 16 days
16:06:09 *** WormnestAndroid has quit IRC
16:17:56 *** supermop_work_ has quit IRC
16:18:37 *** supermop_work_ has joined #openttd
16:27:13 *** WormnestAndroid has joined #openttd
16:27:41 <Wolf01> Ok, enough lego for today, maybe tomorrow I'll build the defender
16:27:56 <FLHerne> LordAro: If you're ok with them, would you mind merging those PRs?
16:28:00 <FLHerne> I don't have write access
16:29:51 <glx> hmm it didn't like my force exit of the while loop in first pass
16:30:11 <LordAro> FLHerne: i know, i just generally don't merge PRs if i want someone else to look at them
16:30:18 <LordAro> not immediately, anyway
16:31:25 <FLHerne> LordAro: Ok, I wasn't sure to what extent "approved" meant "ok, do that" :P
16:32:07 <LordAro> for me, it means "as long as no one else finds/has any issue with it, it'll be merged soonish"
16:32:12 <glx> it just mean it's mergeable by someone with write access :)
16:32:56 <glx> and sometimes we are to fast on the trigger and regret it :)
16:38:08 <glx> just added the regression check as required before merge
16:45:38 <glx> (and of course all PRs need a rebase now)
16:48:48 *** supermop_work_ has quit IRC
16:49:27 *** supermop_work_ has joined #openttd
17:01:24 <frosch123> andythenorth: so, station rating. would you use a vehicle callback that allows to make some vehicles more favourable in station rating?
17:01:49 <supermop_Home> pacer reduces station rating?
17:03:07 <frosch123> basically #7730: people want to change the mechanics, but don't agree how. so the obvious solution would be to make the "speed of last vehicle" defined by the vehicle grf
17:03:36 <frosch123> which can then use other criterions than speed, like consist composition
17:07:24 *** WormnestAndroid has quit IRC
17:19:37 *** supermop_work_ has quit IRC
17:21:03 *** supermop_work_ has joined #openttd
17:31:23 *** WormnestAndroid has joined #openttd
17:38:44 *** tokai|noir has joined #openttd
17:38:44 *** ChanServ sets mode: +v tokai|noir
17:41:38 *** Wormnest has joined #openttd
17:50:16 *** Progman has joined #openttd
17:51:15 *** supermop_work_ has quit IRC
17:51:43 *** supermop_work_ has joined #openttd
18:21:55 *** supermop_work_ has quit IRC
18:23:10 *** supermop_work_ has joined #openttd
18:53:24 *** supermop_work_ has quit IRC
18:53:26 <DorpsGek_III> [OpenTTD/OpenTTD] telk5093 updated pull request #7817: Feature: Topview screenshot (a.k.a minimap) https://git.io/JegRL
18:54:41 *** supermop_work_ has joined #openttd
19:02:59 <andythenorth> frosch123: that would be an alternative way to implement things like 'luxury'
19:03:59 <frosch123> do you still use always-100% rating?
19:04:04 <andythenorth> no, I got bored of that
19:04:09 <andythenorth> it was fun for model train games
19:04:27 <andythenorth> it's not a wrong idea, I just did it enough
19:06:08 <andythenorth> it's probably better than messing around with a cargo payment bonus, which is my current solution to 'luxury'
19:10:04 <DorpsGek_III> [OpenTTD/nml] glx22 commented on pull request #65: Fix #57: Check coherency of GRF parameters limits https://git.io/Jei6H
19:11:05 <DorpsGek_III> [OpenTTD/nml] glx22 commented on pull request #65: Fix #57: Check coherency of GRF parameters limits https://git.io/Jei65
19:23:54 *** snail_UES_ has joined #openttd
19:28:45 <DorpsGek_III> [OpenTTD/nml] glx22 commented on pull request #65: Fix #57: Check coherency of GRF parameters limits https://git.io/JeiiO
19:34:04 <snail_UES_> I was reading in the IH thread about a “group liveries feature” developed by peter1138
19:34:14 <snail_UES_> is there any docs about how it works? or can anyone explain it?
19:35:08 <nielsm> you make vehicle groups using the regular grouping UI
19:35:28 <nielsm> then there is a button that opens the company livery window, where you can set livery for each group defined
19:36:14 <Eddi|zuHause> snail_UES_: i think the idea was that when you have multiple vehicles, you define one of them as a "master" vehicle, and the UI will then collapse those into one single entry on the purchase menu
19:37:06 <Eddi|zuHause> well, i might also be wrong :)
19:37:11 <snail_UES_> Eddi|zuHause: this sounds great...
19:37:33 <snail_UES_> I was waiting for something like that. But isn’t there any thread or documentation about it?
19:37:51 <Eddi|zuHause> snail_UES_: afaik: none of that is implemented (yet)
19:37:52 <snail_UES_> I’m still using refitting to cycle through liveries in my set, it’d be nice to use a less cumbersome method for my next one
19:38:30 <Eddi|zuHause> snail_UES_: what nielsm said, that is implemented.
19:38:56 <Eddi|zuHause> i.e. the player can now set company colours on vehicle group level
19:39:15 <snail_UES_> in the purchase window?
19:39:37 <snail_UES_> oh, so after you build your trains? in the roster of existing trains?
19:39:53 <snail_UES_> ah, so it’s not really what I was looking for
19:40:28 <Eddi|zuHause> afair, there is a company colour variable you can check in the varaction2, for visual differences
19:40:47 <snail_UES_> right, but that’s not exactly what I was looking for
19:40:51 <Eddi|zuHause> but that is probably not what you're looking for
19:41:01 <snail_UES_> I’d like to code different “versions” of the same engine as separate IDs, but then group them somewhat in the purchase window, to avoid cluttering
19:41:13 <Eddi|zuHause> yes. i would like that as well
19:41:46 <Eddi|zuHause> in my case, i want MUs with different amount of wagons
19:43:19 <Eddi|zuHause> i'm not 100% convinced, that the current working-proposal is really the best way to do that
19:43:21 <snail_UES_> you can code them as articulated engines and then add coaches to them, but it’s not exactly the same
19:43:31 <snail_UES_> especially when each car is powered
19:43:47 <Eddi|zuHause> yeah. but articulated vehicles can't be dual-headed
19:44:11 <Eddi|zuHause> and powered wagons are weird...
19:44:22 <Eddi|zuHause> especially price-balance-wise
19:44:22 <snail_UES_> right, you’d have to add the new coaches in the tail of the consist, then swap the graphics and the visual effects
19:44:36 <snail_UES_> ^^ what you just said is the point
19:45:32 <snail_UES_> in the current system there’d be little price difference between a “shorter” and a “longer” consist, if the “longer” is merely a “shorter” + a bunch of coaches
19:45:42 <Eddi|zuHause> snail_UES_: there's so much hacky newgrf stuff that has accumulated over the years, that can sort of do what you want, but not really...
19:45:48 <snail_UES_> and that’d be silly coz there’d be no incentive to build a “shorter” consist at all
19:46:14 <snail_UES_> yeah, pretty much everything is an afterthought :D
19:49:33 <Eddi|zuHause> i've been thinking of ways to get rid of this hacky/weird "livery refit using cargo subtypes" feature for years now
19:53:48 <snail_UES_> me too… but I couldn’t really find a workable solution, without cluttering the purchase list
19:54:22 <snail_UES_> and I don’t really want the same engine repeated, say, 8 times, with just different graphics and slightly different stats
19:54:39 <snail_UES_> OTOH, the current way of doing things doesn’t really work with auto replace
19:55:16 <snail_UES_> hence, I don’t use autoreplace in my games :p but this puts some people off
19:55:47 <Eddi|zuHause> with autoreplace, you have to be very strict about which subtypes you offer
19:56:06 <Eddi|zuHause> because the algorithm that tries to match them isn't very advanced
19:56:24 <snail_UES_> well, coding each subtype in a separate ID would work, right?
19:56:43 <Eddi|zuHause> and that's where the purchase list collapsing would come in
19:56:47 <frosch123> the autoreplace algorithm is pretty advanced :p
19:56:55 <frosch123> you just need to use the same strings for the same types
19:57:07 <Eddi|zuHause> frosch123: yes. :p
20:32:15 <andythenorth> oh snail left :P
20:32:19 <andythenorth> the group livery is done
20:32:30 <andythenorth> something about variants concept doesn't convince me
20:32:40 <andythenorth> it was my idea iirc, so I'm not being mean to anyone :P
20:58:17 <andythenorth> frosch123: got a patch then? o_O
20:59:03 *** andythenorth has joined #openttd
20:59:45 <frosch123> for station rating?
20:59:52 <frosch123> no, do you want one?
21:00:53 *** supermop_work_ has quit IRC
21:02:23 *** supermop_work_ has joined #openttd
21:05:26 <andythenorth> could be useful?
21:05:56 <andythenorth> I had an idea about cabooses
21:06:18 <frosch123> i wondered about cargo wagons
21:06:20 <andythenorth> ha ha, I also just had an idea about snowploughs :P
21:06:40 <frosch123> maybe specialised wagons are cooler than generic refittable wagons
21:06:42 <andythenorth> vehicles can't check snowline, right?
21:07:15 <frosch123> i think they can, but you can't use it for vehicle properties
21:07:28 <andythenorth> specialised > generic? o_O
21:07:47 <frosch123> car transporter > flat bed
21:07:59 <frosch123> hopper > open wagon
21:08:07 <Eddi|zuHause> frosch123: don't we have capacity and/or loading speed for those?
21:08:46 <frosch123> i just wonder whether station rating is a significant vehicle property
21:08:52 <frosch123> that allows for more vehicle variety
21:08:54 <Eddi|zuHause> trading capacity for autorefit, for example, would be a useful distinction between special and generic wagons
21:08:56 <andythenorth> there are some wagons e.g. refrigerated
21:09:06 <andythenorth> where it's impossible to actually give a bonus that's useful
21:09:26 <andythenorth> molten metal has the same issue, just hot not cold
21:09:37 <Eddi|zuHause> frosch123: no, because station rating would be the variable to optimize for, essentially reducing you to a single valid choice
21:10:00 <andythenorth> would it work like ships?
21:10:19 <frosch123> Eddi|zuHause: i don't think so
21:10:20 <andythenorth> rating's potentially interesting because it allows longer between pickup
21:10:22 <Eddi|zuHause> that ship bonus is silly
21:10:30 <frosch123> people do not go for larger/faster vehicles to please station rating
21:10:45 <andythenorth> and the original game only offered one size
21:11:48 <Eddi|zuHause> andythenorth: let me rephrase that: the rating decay for "not always a train waiting" is stupid, and the ship bonus is thus invalid
21:12:09 <Eddi|zuHause> ... as it is reinforcing a stupid concept
21:12:31 <andythenorth> I possibly agree, but it is what it is?
21:15:06 <frosch123> who is our vehicle property freak after V?
21:21:14 <supermop_work> no HST NG wagons andy?
21:21:30 *** snail_UES_ has joined #openttd
21:24:53 <supermop_work> what would a plow do with snowline information anyway?
21:25:24 <frosch123> you can raise and lower it like a the catenary thingie
21:26:27 <frosch123> that could potentially look awesome
21:26:32 <andythenorth> set higher station rating for plowing above snowline
21:26:44 <andythenorth> do I really want to animate a flanger? :P
21:26:54 <andythenorth> how about rotary?
21:27:08 <frosch123> as long as there is snow flying from it
21:28:37 <andythenorth> but a rotary would make more sense :P
21:28:58 <andythenorth> anyway station bonus for snowploughing :P
21:29:27 <frosch123> it would work if the station is above snow line
21:29:48 <andythenorth> in temperate, station bonus for running track maintenance machines
21:29:54 <andythenorth> and restaurant cars
21:32:36 *** supermop_work_ has quit IRC
21:32:59 *** supermop_work_ has joined #openttd
21:36:25 <snail_UES_> andythenorth: I’m completing the IH Italian translation
21:36:38 <snail_UES_> what the heck do you mean by “Lolz”? :p
21:37:17 <andythenorth> it's a temporary thing
21:37:27 <snail_UES_> I’ll leave it empty then
21:37:40 <andythenorth> translate it to 'shunter' or 'small shunter'
21:37:45 <andythenorth> I'll change it now in En
21:38:13 <andythenorth> it's not useful, it's just an eye candy / joke train
21:38:39 <snail_UES_> well, it’d be useful if we ever implemented that reversing patch someone did time ago
21:39:20 <andythenorth> oh I need to change the string name, that will break any translation you do sorry :P
21:40:02 <snail_UES_> anyway I’m reading your previous messages now, snowplowing would be a nice addition
21:40:14 <snail_UES_> it’s technically possible even now, but it’d be just eye candy...
21:40:38 <snail_UES_> improving station ratings for running snow plows (or maintenance vehicles) would be awesome though
21:40:43 <andythenorth> I've drawn a snowplough
21:41:17 <snail_UES_> but then we’d have to draw specific tracks, right?
21:41:26 <snail_UES_> right now I’m drawing snow-covered tracks for my set
21:41:59 <snail_UES_> we’d have to draw snow-covered tracks, and partially-snow-covered tracks to display immediately after a snowplow’s passage...
21:42:05 <supermop_work> my personal rating of a station goes down when im waiting for the subway late at night and instead a work train rolls through
21:43:12 <andythenorth> what if it's the vacuum train that hoovers up the dead rats and garbage?
21:44:02 <supermop_work> that one is very rare
21:44:11 <snail_UES_> no such thing in the NYC subway :D
21:44:26 <snail_UES_> we have the live rats that hoover up garbage instead
21:44:57 <supermop_work> also its so loud they usually won't run it when there are people waiting in the station
21:45:24 <snail_UES_> anyway, you put the snowplough as the head of a consist? I’ve rather seen dedicated consists, such as a single engine pushing a plow forward
21:45:45 <andythenorth> I've made it an engine so it can lead
21:46:50 *** arikover has joined #openttd
21:51:48 <snail_UES_> hi arikover, long time no see :)
22:03:14 *** supermop_work_ has quit IRC
22:03:38 *** supermop_work_ has joined #openttd
22:15:03 *** HerzogDeXtEr has joined #openttd
22:33:53 *** supermop_work_ has quit IRC
22:34:23 *** supermop_work_ has joined #openttd
23:00:58 <DorpsGek_III> [OpenTTD/OpenTTD] stormcone updated pull request #7800: Easier access for vehicles' group window. https://git.io/JeEtl
23:04:38 *** supermop_work_ has quit IRC
23:05:53 *** supermop_work_ has joined #openttd
23:25:14 *** andythenorth has left #openttd
23:38:37 *** supermop_work_ has quit IRC
23:39:51 *** supermop_work_ has joined #openttd
continue to next day ⏵
