IRC logs for #openttd on OFTC at 2020-08-17
⏴ go to previous day
01:47:56 *** WormnestAndroid has quit IRC
01:48:08 *** WormnestAndroid has joined #openttd
02:08:23 <Gustavo6046> I have been toying with the concept of intertwining a city's road network with waterways. Did I do it right?
02:09:20 <Gustavo6046> I feel like I did one thing wrong, but I always feel that way lol
02:09:25 <Gustavo6046> (perhaps because I always do at least one thing wrong)
02:42:25 <Gustavo6046> You really have to add buoy orders manually?! ;-;
02:42:37 <Gustavo6046> Why not have the buoys be like pathnodes, so we can have a hierarchical-based pathfinding algorithm?
02:44:01 <Gustavo6046> Maybe keep track of what buoys can reach what buoys/docks (by searching for pathfindable waypoints in a manhattan radius of 30 tiles)
02:44:16 <Gustavo6046> Then when a ship wants to go from A to B, consider this high-level graph first
02:44:24 <Gustavo6046> before connecting each node on a per-tile basis
04:32:33 *** Gustavo6046 has joined #openttd
07:35:56 *** Guest5399 is now known as planetmaker
08:06:33 *** sla_ro|master has joined #openttd
09:29:49 *** andythenorth has joined #openttd
09:36:08 <Wolf01> Hmmm, I wanted to make some little scripts for fun, but I know well myself and I would end up creating entire applications and overcomplicate the starting idea
09:36:34 <Wolf01> Now I understand andythenorth :|
09:36:54 <andythenorth> think of the time you'll save though
09:37:07 <andythenorth> and you'd have a framework for creating scripts
09:37:57 <Wolf01> By not doing them, yes, because doing them is a waste of time anyway as they will be useless and I could already find alternatives
09:38:47 <andythenorth> the most best way to complete a project: 'invalid'
09:39:23 <andythenorth> my favourite ticket statuses
09:40:22 <Wolf01> I should tell this to boss when I'll get to work again, I wonder if I'll get to keep the job
09:50:06 <andythenorth> I am in the top 3 people for invalidating tickets :P
09:50:19 <andythenorth> I'm only not #1 because I am sometimes lazy and ask someone else to do it
10:55:06 <TrueBrain> right ... guess I should start working on the AWS migration again
10:55:15 <TrueBrain> the temperature is down enough that I guess focusing is possible again
10:58:22 * andythenorth needs to do AWS stuff
10:58:42 <andythenorth> but it's Monday, so I have to do socially-responsible capitalism first
10:59:02 <TrueBrain> "cloudfront: (cloudfront) Changed IDs for Distributions (will cause resource replacement)." <- owh boy, updating CDK will be slow
10:59:37 <andythenorth> also I should learn about cloudfront, and stop clicking the 'manual invalidation /*' option
11:00:12 <TrueBrain> only if you want to :)
11:00:35 <andythenorth> AWS have a politely worded hint that It's Very Silly
11:01:23 <TrueBrain> I automated it via cli commands in a GitHub Action :P
11:02:54 <andythenorth> I'm going to do something a bit similar
11:09:51 <TrueBrain> I am always scared to death to redeploy with the latest CDK .. no clue what breaks ...
11:09:56 <TrueBrain> it is going to redo most of the DNS entries
11:29:32 <TrueBrain> guess it is summer .. not many nightlies this month .. (2 so far :P)
11:31:22 *** frosch123 has joined #openttd
11:39:11 <LordAro> 2/16 isn't a good hit rate
11:39:16 <LordAro> someone should do some stuff
11:39:30 <TrueBrain> yeah, LordAro , do some stuff :P
11:44:01 <TrueBrain> changing the IAM policy on the ECS cluster .. what could possibly go wrong .. fingers crossed
11:45:29 <andythenorth> I have to set up a proper IAM policy
11:45:34 <andythenorth> I hate this kind of admin :)
11:45:39 <andythenorth> so the opposite of what I'm good at
11:45:56 <andythenorth> weekend project: draw Venn diagram of things andythenorth is good at, and everything else
11:46:38 <TrueBrain> updating all DNS records ... lets see if everything is still working in a bit :P
11:48:29 <TrueBrain> seems to be okay :P
11:53:23 <TrueBrain> meh; GitHub dependabot still cannot make a single PR for all dependency updates
11:53:33 <TrueBrain> I really do not like a PR per dependency ..
11:53:42 <TrueBrain> guess that only works if you keep up constantly or what-ever
11:53:50 <TrueBrain> requires a lot of automated testing to make that useful, I guess
11:55:19 <TrueBrain> andythenorth: why are you talking about "weekend projects" on a monday .. do you know what can happen between now and the weekend?!
11:56:14 <andythenorth> is this a question of semantics, and the unknowable future?
11:56:30 <andythenorth> or do you think I might get hit by a bus?
11:56:44 <TrueBrain> no no, nothing of that sorts
11:56:55 <TrueBrain> more that you will have 5 more of these "weekend projects" before the week is over
11:56:59 <TrueBrain> and then feel sad you cannot do them all
11:57:14 <FLHerne> We should just clone him
11:57:21 <TrueBrain> do we REALLY want to do that?
11:57:22 <andythenorth> isn't that like....the essence of life though?
11:57:25 <TrueBrain> can you really handle that FLHerne ?
11:57:29 <andythenorth> life is making choices
11:57:40 <andythenorth> mostly I choose to play Blitz, and regret it
11:57:55 <TrueBrain> I am trying REALLY hard to not start WoW ..
11:58:00 <TrueBrain> it is REALLY hard not to :(
11:58:04 <TrueBrain> I like writing REALLY like this
11:58:06 <FLHerne> But if you invest a few million dollars in this biotech company I'm about to start, in a few short years who knows what might be possible?!
11:58:11 <TrueBrain> it makes me feel REALLY important
11:58:24 <TrueBrain> FLHerne: I was not talking about the possibility of it, more the outcome
11:58:28 <TrueBrain> I mean .. more than one andythenorth ..
11:58:32 <andythenorth> I keep playing warcraft 1
11:58:33 <TrueBrain> can we really handle that?
11:58:40 <andythenorth> I stayed up waaaaaaaay too late
11:58:43 <TrueBrain> They will talk to each other ....
11:58:50 <andythenorth> I talk to myself, it's fine
11:58:58 <TrueBrain> and opening their forum PMs
11:59:03 <FLHerne> We'll end up with generated files where the line number overflows
11:59:05 <andythenorth> TrueBrain I think you're describing the current situation?
11:59:22 <FLHerne> Also, all the bugs will be closed for spurious reasons
11:59:30 <andythenorth> I really wanted to try a custom game in Warcraft 1 with no peons at the start
11:59:35 <TrueBrain> we will hit negative issues on GitHub :)
11:59:37 <andythenorth> but if peons are 0, you can't train any
11:59:55 <andythenorth> I did try 1 peon and 6 catapults and a raider
12:00:18 <TrueBrain> that's what she said
12:00:39 <andythenorth> do we keep doing that joke?
12:00:43 <andythenorth> I thought it was over
12:00:53 <TrueBrain> why would you think that?
12:01:10 <TrueBrain> at what point did anyone give any indication it would be over?
12:01:33 <andythenorth> there seems to be a mismatch
12:01:43 <andythenorth> between how I want the world to be, and how it is
12:01:46 <andythenorth> that's just me, right?
12:01:51 <andythenorth> nobody else feels that?
12:01:58 <TrueBrain> looks at US .. no, we all have this :P
12:05:03 <TrueBrain> so, where was I on my trello board ..
12:08:03 <TrueBrain> CDK is still deploying AWS .. after an update it can take such a long time ..
12:09:03 <TrueBrain> git commit -m "Fixed the bugs.txt"
12:14:03 <TrueBrain> okay .. so now it is time to upgrade the EC2 instances .. this causes some downtime, which is very unlikely anyone notices :)
12:21:28 <frosch123> wow.... turns out github wiki has no method to upload images, unless you do a local git-checkout
12:21:44 <TrueBrain> drag&drop it in there, not?
12:22:44 <TrueBrain> talking about wiki, where shall I document how BaNaNaS works? In the wiki of BaNaNaS repo, or in bananas-api? Or a new repo with commits? Any preferences?
12:22:59 <FLHerne> First one sounds sane
12:23:16 <FLHerne> I don't think anyone would ever find it in the others
12:23:28 <frosch123> TrueBrain: what is your audience?
12:23:41 <frosch123> users, uploaders, bananas devs?
12:23:48 <TrueBrain> documentation about the BaNaNaS YAML format, why the md5 stuff, how these repos are organised
12:23:53 <TrueBrain> so more the dev-side
12:24:06 <TrueBrain> users and uploaders should be taken care of by bananas-frontend-web
12:24:33 <TrueBrain> wiki vs git is mostly about tracking and approving changes
12:24:44 <TrueBrain> the repo itself is more: what repo do you open to find documentation
12:24:53 <TrueBrain> guess we can link to a single one from the rest, so not a biggy
12:26:34 <TrueBrain> I was considering making a bananas-docs, just to have a single place to store that all :P
12:28:40 <TrueBrain> okay, no documentation it is, tnx :) :P
12:29:03 <TrueBrain> that was not really a helpful reply ;)
12:29:43 <FLHerne> I already find the split of bananas-related repos confusing
12:29:52 <TrueBrain> that cannot be helped, really
12:29:54 <FLHerne> Adding another one would not be helpful, IMO
12:30:06 <TrueBrain> well, it would, if it explains how the other interconnect ofc
12:30:13 <TrueBrain> and link them all back to the documentation
12:30:16 <TrueBrain> so you can click through them
12:30:28 <TrueBrain> the problem I have, there is not a single repo where all the documentation should go
12:30:34 <TrueBrain> so that would result in splintered documentation
12:30:41 <TrueBrain> which is possibly more confusing
12:31:00 <TrueBrain> (but this is also exactly why I bring it up, as I need opinions :D)
12:31:17 <TrueBrain> BaNaNaS is a complex setup, basically :)
12:31:35 <TrueBrain> I wish you could group repositories in GitHub, that would be nice
12:34:06 <andythenorth> after I tidied it all up so nicely!
12:34:33 <TrueBrain> sorry, I really refuse to document this on the wiki
12:34:43 <andythenorth> oof, how rude :)
12:37:03 <TrueBrain> funny, you can have projects in the organisation now
12:37:06 <frosch123> TrueBrain: i consider bananas-api the master repo. the other bananas-* are secondary
12:37:10 <TrueBrain> allows you to make tickets cross repos :)
12:37:21 <TrueBrain> frosch123: not even bananas-server? :D
12:37:50 <frosch123> from the nml experience, people want to extend docs together with the code
12:37:53 <TrueBrain> I am fine by marking bananas-api as primary; in the end, it doesn't really matter, as long as I have a single place I can write down what we did and why and blablabla :P
12:38:05 <frosch123> so, it should be a doc/foobar.md within the bananas-api repo. no separate wiki
12:38:26 <TrueBrain> not even the GitHub wiki?
12:38:41 <frosch123> github wikis are a separate git repository
12:38:41 <LordAro> i dislike github wikis
12:38:49 <frosch123> so you cannot make a PR for both code and docs
12:38:52 <LordAro> they've always had a "tacked on" feel to them
12:38:57 <TrueBrain> frosch123: that is very true
12:39:07 <TrueBrain> LordAro: but perfect for documentation :)
12:39:28 <TrueBrain> frosch123: I cannot come up with a reason not to do it like that, so meh ..
12:39:38 <TrueBrain> well, the folder name .. should be "docs" in my opinion, not "doc" :D
12:39:44 <frosch123> LordAro: i just spent 4 days to extract ottd wiki into a git repository. and now learn that github wiki has no method to upload images for regular users
12:39:51 <frosch123> LordAro: so, yeah. github wiki is insane
12:40:12 <frosch123> TrueBrain: it's "docs" in ottd
12:40:37 <TrueBrain> okay, I will write some shit down; after that people can tell me what I missed :p
12:41:11 <TrueBrain> 2 out of the 4 EC2 instances replaced .. 2 more to go :)
12:42:27 <frosch123> TrueBrain: and yes, that image drag/drop only works for issues, not for the wiki
12:42:46 <frosch123> and the url will be disassociated from the wiki data, so a hell for all future migrations
12:42:48 <TrueBrain> frosch123: that -is- insane
12:42:57 <frosch123> it's essentially like an external url
12:43:07 *** Gustavo6046 has joined #openttd
12:43:07 <TrueBrain> I did not expect that tbfh
12:43:37 <frosch123> me neither. i was checking all those template-magic, that fails in gollum/github wiki
12:43:46 <frosch123> i did not expect it would fall flat on images :p
12:44:01 <frosch123> (though they work fine, if you add them via a git checkout)
12:45:00 <TrueBrain> the unexpected problems :(
12:45:15 <TrueBrain> I really do love how OpenTTD now runs on AWS ..
12:45:42 <TrueBrain> updating EC2 instances is like this: update the base image in the configuration of the auto-scale group. Scale up to twice the instances you had. Scale down to the former value
12:45:50 <TrueBrain> everything is taken care of automatically
12:46:09 <TrueBrain> when scaling up, the instances receive their configuration, and are added to the pool of what-ever they are for
12:46:29 <TrueBrain> when scaling down, they are first deregistered, it waits till all TCP connections are gone (or 5 minute timeout), and destroys the VMs
12:46:40 <TrueBrain> as the VMs don't contain any data, this works absolutely fine :)
12:46:45 <andythenorth> we use something like this for running automated tests
12:46:54 <TrueBrain> yeah, the bill is annoying :P But .. we still have plenty of credits :)
12:46:59 <andythenorth> moving prod to this is a whole other world, though
12:47:19 <andythenorth> "as the VMs contain lots of data, this is a problem" etc
12:47:23 <TrueBrain> well, I just upgrades all the machines in ~20 minutes, mostly because I forgot to press buttons when it was done with the next step :D
12:47:52 <TrueBrain> yeah .. VMs in AWS should really not contain any data :)
12:48:24 <TrueBrain> and now, lunch time!
12:49:41 <andythenorth> pls restore peter1138 from backups
12:49:47 <andythenorth> I miss lunch advice
12:50:01 <frosch123> btw. the .git dir for the wiki is 1 GB
12:54:12 <LordAro> TrueBrain: i think i might have said it before a long time ago, but i would like to keep the wiki around - i'm happy to try to do some work on making it work with the new infrastructure
12:54:38 <LordAro> i saw something recently that suggested we might be able to actually create a docker image without too much effort
12:54:51 <andythenorth> is mediawiki hosting a thing?
12:55:23 <TrueBrain> Not meant rude, but yes, you have said this for a few years now ;) but talk to frosch123 about this :)
12:55:34 <andythenorth> "Get The 20X FASTER MediaWiki Host Featuring High Performance Turbo Servers." definitely I'm convinced
12:56:21 <frosch123> LordAro: the problem is hosting the database
12:56:26 <frosch123> they are all expensive
12:57:17 <LordAro> and there's nothing magical that AWS provides, i guess?
12:57:28 <frosch123> so, if we keep mediawiki, we will probably keep a vps, and not do it in the cloud
12:58:23 <TrueBrain> And then we need to talk about who is going to maintain that :)
12:58:45 <andythenorth> can't we move it all to fandom?
12:59:10 <LordAro> ugly and full of ads? ew.
12:59:48 <TrueBrain> AWS supports EFS on ECS these days btw (read, shared storage for docker)
13:00:21 <frosch123> hmm, we could make a silly webapp for uploading images to github wiki :p
13:00:43 <TrueBrain> Haha, not super difficult tbh
13:01:04 <TrueBrain> Cannot believe they do not support it ..
13:02:02 <frosch123> oh, looks like gollum supports uploads...
13:02:11 <frosch123> let's try in my local installtion
13:14:59 <frosch123> well, you can upload files, but you cannot update them
13:15:10 <frosch123> you have to delete/rename the old one first
13:17:44 <frosch123> so you always rename files in your git checkout, before you add the new file with your edits?
13:19:42 <TrueBrain> Lol, that is one way of translating what I said. Not the most productive one, but okay :p but no, most systems don't allow you to overwrite files on upload
13:19:57 <TrueBrain> They either add a (1) or refuse it
13:20:43 <TrueBrain> For wikis you have the added issue you don't know where the image is used .. so that would be a bonus, not to be able to overwrite, not?
13:22:02 <frosch123> when people update images, they replace them with images from newer ottd releases, or (in the past) replaced them with more aggressive png compression
13:22:22 <frosch123> it would be fine, if there was a button "upload update for this upload"
13:22:48 <frosch123> anyway, i'll ignore the images for now (custom app is possible), and check out how much of the template magic fails
13:23:09 <TrueBrain> people do that? Funny :) Wouldn't expect that :)
13:23:31 <TrueBrain> and if we want, we can always make a PR in gollum I guess :)
13:23:33 <frosch123> current state is that gollum aborts rendering when it does not a single part of some template (which appears stupid)
13:23:38 <frosch123> and it fails with the inter-language links
13:23:45 <frosch123> so most ottd wiki pages currently render as "error" :p
13:23:58 <TrueBrain> which is an upgrade for most pages :D
13:26:42 <andythenorth> sounds like I need to leave the channel now :D
13:27:17 <frosch123> TrueBrain: "css" is close to "brexit"
13:28:16 <andythenorth> I have a theory about who ends up writing the css
13:29:10 <TrueBrain> ah; well, tnx for volunteering :)
13:29:26 *** DorpsGek_III has joined #openttd
13:29:26 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain opened pull request #60: Fix: convert f-strings without variable to strings https://git.io/JJFKu
13:31:04 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain updated pull request #60: Fix: convert f-strings without variable to strings https://git.io/JJFKu
13:31:25 <TrueBrain> cool, vscode inserted my credentials correctly when I wanted to push via the Terminal inside vscode .. that makes me happy :)
13:32:05 <DorpsGek_III> [OpenTTD/bananas-api] LordAro approved pull request #60: Fix: convert f-strings without variable to strings https://git.io/JJFKw
13:32:45 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain merged pull request #60: Fix: convert f-strings without variable to strings https://git.io/JJFKu
13:34:02 <TrueBrain> LordAro: anyway, to get back to mediawiki: I personally really do not care how we solve the wiki. Fact is, the current setup cannot survive :) So far frosch123 is the only one who actively has been busy looking for a solution :)
13:36:44 <andythenorth> unrelated, I'm assuming coop devzone is dead?
13:38:50 <frosch123> sometimes parts that were killed before start moving again in some weird way
14:31:13 *** GroovyNoodle has joined #openttd
14:41:37 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain opened pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFXM
14:41:50 <TrueBrain> very likely I forgot to write things down, but lets update the documentation when we get questions about stuff not documented
14:42:07 <TrueBrain> this should resolve my promise to LordAro to document it better :)
14:42:32 <TrueBrain> but do tell me if this isn't the case :)
14:44:44 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain updated pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFXM
14:45:39 <TrueBrain> FLHerne: if you have some time, I would love to know if documentation like this (start at introduction.md) helps out understand the amount of repos involved :)
14:51:03 <TrueBrain> guess pyup has to do, till GitHub can create a single PR for all updates ..
14:57:15 <DorpsGek_III> [OpenTTD/bananas-api] frosch123 commented on pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFM1
14:58:24 <DorpsGek_III> [OpenTTD/bananas-api] FLHerne commented on pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFMQ
14:58:30 <TrueBrain> frosch123: I really cannot remember any discussion around hyphens. Only around plural :P
14:58:37 <TrueBrain> I am fine with what-ever, so I will change it :)
14:59:01 <frosch123> i am not sure how helpful the paragraph on the byte-swapping of unique-ids is.
14:59:13 <frosch123> it may be more confusing than helpful :p
14:59:35 <frosch123> maybe just call it "human-readable" format (which may include byte-swapping)
15:01:57 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain updated pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFXM
15:05:10 <DorpsGek_III> [OpenTTD/bananas-api] frosch123 approved pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFDS
15:05:17 <frosch123> unless you want to wait for the customers
15:37:27 <TrueBrain> it is a beginning. I rather receive requests :)
15:37:42 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain merged pull request #61: Add: bunch of documentation of how BaNaNaS components work together https://git.io/JJFXM
15:39:16 *** supermop_Home has joined #openttd
15:41:59 <supermop_Home> weekend of no trains
15:42:13 <supermop_Home> only comical numbers of subarus
15:46:41 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain opened pull request #64: Fix: move text about documentation to top of README https://git.io/JJF94
15:46:44 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain opened pull request #26: Fix: convert f-strings without variable to strings https://git.io/JJF9B
15:46:51 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain opened pull request #38: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJF9R
15:46:53 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain opened pull request #9: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJF9u
15:47:49 <nielsm> f-strings sounds dirty
15:48:48 <TrueBrain> so does the language D I guess? :)
15:53:24 <_dp_> hm, weird, when I generate temperate 2030 game sometimes Dash(Diesel) is available and sometimes it isn't...
15:55:48 <DorpsGek_III> [OpenTTD/bananas-api] frosch123 approved pull request #64: Fix: move text about documentation to top of README https://git.io/JJFHI
15:56:26 <DorpsGek_III> [OpenTTD/bananas-server] frosch123 approved pull request #26: Fix: convert f-strings without variable to strings https://git.io/JJFHt
15:57:19 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 approved pull request #38: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJFHY
15:57:53 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] frosch123 approved pull request #9: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJFHG
15:58:27 <LordAro> _dp_: right on the edge of expiring, perhaps?
15:58:34 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain merged pull request #64: Fix: move text about documentation to top of README https://git.io/JJF94
15:58:45 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain merged pull request #26: Fix: convert f-strings without variable to strings https://git.io/JJF9B
15:58:55 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain merged pull request #38: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJF9R
15:59:01 <DorpsGek_III> [OpenTTD/bananas-frontend-cli] TrueBrain merged pull request #9: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJF9u
15:59:10 <TrueBrain> revenge revenge spam :)
15:59:40 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain opened pull request #27: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJFHu
15:59:51 <TrueBrain> and one more :) (last one, I promise)
16:00:12 <DorpsGek_III> [OpenTTD/bananas-server] frosch123 approved pull request #27: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJFH2
16:01:09 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain merged pull request #27: Add: hint in the README where to find more documentation about BaNaNaS https://git.io/JJFHu
16:01:37 <TrueBrain> right, that should resolve most of the "I do not understand BaNaNaS" questions :)
16:01:53 <_dp_> LordAro, yeah, looks like it
16:02:09 <_dp_> weird that it outlasts all other rail engines by 10 years though
16:02:19 <TrueBrain> who are you calling weird?
16:22:05 <FLHerne> TrueBrain: Nice, thanks
16:24:47 <_dp_> what does sort() in nml do?
16:24:55 <_dp_> sort(FEAT_TRAINS, [item_rail_EARLY_01, item_rail_EARLY_02, ...
16:25:23 <frosch123> it defines the order of trains in the purchase list
17:40:23 *** Wormnest has joined #openttd
18:37:54 *** gelignite has joined #openttd
18:52:02 *** Progman has joined #openttd
18:52:03 *** HerzogDeXtEr has joined #openttd
19:06:31 <Gustavo6046> Why do the locks look like antigravity water slopes?
19:08:08 <nielsm> because anything more advanced would be annoying to code
19:09:44 <Gustavo6046> I'm sure you can make building sprites look deeper than the actual tiles they're located in.
19:09:58 <Gustavo6046> "Anamorphism" is a stretch. It's.. just sprites. There are only four angles.
19:10:18 <Gustavo6046> the throughput of multiple ships?
19:10:30 <Gustavo6046> I think bottlenecking at locks could be a pretty nice balancing tool!
19:10:36 <Gustavo6046> Maybe make it opt-in in the Limitations menu.
19:19:48 <nielsm> all kinds of problems, multiple ships is one of them
19:21:27 <glx> they can pile up everywhere anyway
19:26:56 <Gustavo6046> nielsm: so no multi-staged locks?
19:27:06 <Gustavo6046> or whatever it's called
19:27:11 <Gustavo6046> one that goes up and down
19:27:16 <Gustavo6046> Why not just have a bunch of ships go in?
19:27:19 <Gustavo6046> Kinda like a shopping mall lift
19:27:33 <Gustavo6046> 65535 ships go in, lock goes up, 65535 ships go out
19:27:52 <nielsm> and then you'd want single-direction locks?
19:28:06 <Gustavo6046> we could have a more advanced lock
19:28:08 <nielsm> or ships waiting in fron of the lock if it's at the wrong height atm?
19:28:34 <Gustavo6046> a lift-lock, that starts in a ground canal, and you can add a level up (each level = exponential price), and each level must end in an aqueduct in one of four directions
19:28:59 <Gustavo6046> nielsm: single-direction?
19:29:10 <Gustavo6046> Also, waiting is fine
19:29:16 <Gustavo6046> kinda like train signals
19:29:19 <nielsm> like two locks side by side and one only goes up and one only down
19:29:40 <Gustavo6046> but then that would be two individual locks
19:30:03 <Gustavo6046> I did suggest at one point having a more hierarchical graph-based approach to ship pathfinding, where waypoints (buoys, docks, locks, etc) would be nodes in a graph, and the tile pathfinder would merely find and cache paths between those
19:30:32 <Gustavo6046> it would split cached paths into regions of 8x8 or something along these lines, so a single region change only needs an update in those particular path sections
19:30:35 <Gustavo6046> that would be efficient
19:30:40 <Gustavo6046> but I digress, that's on a more technical side
19:31:03 <Gustavo6046> the point is, with graphs, having "signals" internally to tell ships to wait would be trivial
19:31:27 <Gustavo6046> and really it would basically go unnoticed by the player, thanks to something called noclip (which all ships seem to have lol)
19:31:50 <Gustavo6046> nielsm: locks could be one-way, but then that would be opt-in
19:32:01 <Gustavo6046> dropdown -> "Both" "Up" "Down"
19:32:06 <Gustavo6046> I want to try to implement that
19:32:15 <Gustavo6046> on the other hand I don't think I can do any actual improvements to existing OpenTTD pathfinders
19:32:26 <Gustavo6046> and I don't want to implement a new tile-based pathfinder, just a high-level pathfinder that uses that
19:32:31 <Gustavo6046> nodes -> regions -> tiles
19:32:44 <Gustavo6046> regions can be updated individually without having to update the entire path
19:33:09 <Gustavo6046> also, are contributions accepted via patch submission, or GitHub pull requests, or?
19:33:32 <nielsm> if you take the approach that water pathfinding features (like bouys and docks) are part of a graph and you only pathfind on that graph
19:34:01 <Gustavo6046> that is the point
19:34:23 <nielsm> it would make sense that this graph is made by a simple nearest neighbors algorithm kind of thing, i.e. a bouy just searches for the nearest 3 neighbors or a max distance whichever is hit first
19:34:56 <Gustavo6046> nielsm: additionally, we could have those waypoints cache all reachable water tiles in a radius of up to 20 or something
19:35:23 <Gustavo6046> but we could add an upper limit to the length of a path between two nodes
19:35:31 <Gustavo6046> something between 30 and 50 or something
19:35:37 <Gustavo6046> have it be adjustable (and opt-out, of course!)
19:35:45 <Gustavo6046> and buoys would serve as pathnodes
19:35:53 <Gustavo6046> it wouldn't be a per-tile search based limit so
19:36:15 <Gustavo6046> also, I don't know how comfortable I am in writing my contributions in C++ in contrast to Rust, but I will try my best
19:36:48 *** tokai|noir has joined #openttd
19:36:48 *** ChanServ sets mode: +v tokai|noir
19:48:55 <andythenorth> rust conversion :P
19:50:37 <Gustavo6046> also, if I have a peninsula with rails on it, what is the best way to get ships across it (more efficient than going round)? Two spires with locks and an aqueduct?
19:52:46 <nielsm> I guess you could lower the inside part of the peninsula so you can build aqueduct over it
19:52:48 <Gustavo6046> Imagine underground aqueducts
19:53:22 <nielsm> lower the inside part to sea level
19:53:30 <Gustavo6046> But then how do I put rails
19:53:38 <nielsm> so you don't need two locks on each end
19:53:40 <Gustavo6046> the sea level part does not have water
19:53:46 <Gustavo6046> because it's sealed by terrain
19:53:51 <Gustavo6046> but this is sea level
19:54:33 <Gustavo6046> but I think I got it
19:54:40 <Gustavo6046> I have to lower the rails down but not let water get in
19:54:45 <Gustavo6046> and then I only need two locks, not four?
19:54:51 <Gustavo6046> so one in each end
19:55:13 <nielsm> since locks slow down ships significantly
19:55:19 <Gustavo6046> also there are already maglevs but I have this one crappy monorail line I'm a bit lazy to change lol
19:56:07 <Gustavo6046> I just found out.. it turns out those aren't my rails
19:56:33 <Gustavo6046> I wonder what is the new state-of-the-art for AI.
19:56:39 <Gustavo6046> I remember Admiral being cool
20:11:10 <Gustavo6046> Wolf01: we have one here in Porto Alegre
20:11:19 <Gustavo6046> but it's a more modern one so
20:56:58 *** b_jonas has joined #openttd
22:11:41 <andythenorth> oh frosch come back :P
22:25:13 <Eddi|zuHause> he always disappears way too early for a meaningful conversation :p
22:40:36 * andythenorth wonders if we need a format painter mode
22:40:58 <andythenorth> modifying appearance of individual trains is desirable
22:41:10 <andythenorth> modifying appearance of many trains at once is desirable
22:41:22 <andythenorth> this seems to point to some kind of painter tool, on a group basis
23:58:44 <andythenorth> 2px taller than UK Horse
continue to next day ⏵
