IRC logs for #openttd on OFTC at 2025-02-15
⏴ go to previous day
00:14:54 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
01:31:19 *** Flygon has quit IRC (Read error: Connection reset by peer)
02:21:48 *** gelignite is now known as Guest9169
02:21:51 *** gelignite has joined #openttd
02:29:09 *** Guest9169 has quit IRC (Ping timeout: 480 seconds)
03:45:59 *** Wormnest has quit IRC (Quit: Leaving)
03:50:49 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:27:24 *** gelignite has quit IRC (Quit: Stay safe!)
04:44:46 <DorpsGek> - Update: Translations from eints (by translators)
08:04:25 *** tokai has quit IRC (Quit: c('~' )o)
08:41:21 <truebrain> that moment the CI fails because there is a warning ... `Cache not found for keys`
08:47:39 <LordAro> that eslint config looks very "legacy"
08:48:14 <truebrain> I like that the new version was like: NO. Here is where I draw the line. UPPPGRRRAAADDDDEEEE
08:49:17 <LordAro> mm, and it's so much harder to write
08:49:29 <LordAro> and half the plugins don't support it
08:49:49 <LordAro> at least, that was the case when i tried it a few months ago
08:50:10 <truebrain> I just ran the conversion script and called it a day
08:50:58 <truebrain> wow, the linux arm runner makes things so much quicker
08:53:28 <LordAro> 4s for a complete flake8 run?
08:53:43 <LordAro> oh wait, 7 total. rounding :)
08:54:04 <truebrain> flake8 timing didn't change; why does 7s surprise you?
08:55:09 <LordAro> maybe github doesn't include the docker spin up time
08:55:23 <truebrain> it should; it just is really quick. Always has been, as far as I know 🙂
08:55:28 *** D-HUND is now known as debdog
08:55:30 <LordAro> gitlab's jobs aren't nearly that fast
08:55:37 <truebrain> no no no, gitlab is horrible
08:59:37 <truebrain> I was preparing myself to bump the Python in bananas-server to 3.11, only to find out it is already on 3.11
09:00:06 <truebrain> did that in Sept 2023 already
09:00:14 <truebrain> I thought bananas-api was the first one we did, a few months back 😄
09:01:25 <truebrain> `Dependabot encountered an error performing the update Error: (HTTP code 404) no such container - No such image`
09:01:27 <truebrain> well, that is just great
09:01:31 <truebrain> it can't find its own container
09:04:40 *** HerzogDeXtEr has joined #openttd
09:06:46 <peter1138> Well. I appear to still be in bed, not cycling...
09:09:02 <andythenorth> Except climbing not cycling
09:21:13 <LordAro> peter1138: it's very grey & dark outside
09:21:31 <LordAro> truebrain: just in time to upgrade to 3.13?
09:21:42 <truebrain> no, we decided on 3.11 a few weeks ago 🙂
09:22:25 <LordAro> as a minimum version, perhaps?
09:22:44 <truebrain> I do not plan to upgrade any of our software to newer
09:22:52 <truebrain> so not sure how you wage that "minimum" part of that sentence
09:23:30 <LordAro> minimum compatible version and what production runs on doesn't necessarily have to match
09:24:02 <truebrain> In case you didn't know, we pin all our versions and dependencies 🙂
09:24:05 <LordAro> but i guess that doesn't make a huge amount of sense for web services
09:24:18 <truebrain> I do not want to deal with some magic "owh, it is now a different version, and everything is burning" 😛
09:25:29 <truebrain> The difference between libraries and applications 🙂
09:26:03 <LordAro> i don't really account for libraries being different for different python versions
09:26:51 <truebrain> it is mostly about predictability 🙂 Sometimes nodes in our cluster crash, or a service crashes. It auto-restarts, but I really really want that to be the same version as has been running for a while 🙂
09:26:59 <truebrain> no midnight "omg, why is everything broken" calls for me please 😄
09:31:04 <truebrain> CodeQL does not appreciate `"([-a-z]*)/([0-9a-f]{8})/([0-9a-f]{8})$`
09:31:12 <truebrain> `Polynomial regular expression used on uncontrolled data `
09:33:27 <truebrain> It is not wrong, but also: meh
09:36:04 <peter1138> It is grey, dampish, and cold. But also I am getting fatter :(
09:37:43 <andythenorth> ok I have to get up
09:38:13 <peter1138> Right, what other stashes can I drop.
09:43:16 <truebrain> so there is a security issue with GitHub actions, which are fixed downstream, but they haven't updated their more upstream actions yet .. so I now have alerts about that, which I cannot resolve, because GitHub has to make a new release for some actions 🙂
09:43:45 <truebrain> Just funny to see how even they struggle with fixing things in such a way these time-gaps don't exist 🙂
09:48:23 <truebrain> Going to deploy a new dibridge; for IRC users: we will be right back 🙂
09:49:50 *** peter1138[d] has quit IRC (Remote host closed the connection)
09:49:50 *** _zephyris has quit IRC (Remote host closed the connection)
09:49:50 *** truebrain has quit IRC (Remote host closed the connection)
09:49:50 *** talltyler has quit IRC (Remote host closed the connection)
09:49:50 *** merni has quit IRC (Remote host closed the connection)
09:49:50 *** yiffgirl has quit IRC (Remote host closed the connection)
09:49:50 *** michi_cc has quit IRC (Remote host closed the connection)
09:49:50 *** florafex has quit IRC (Remote host closed the connection)
09:49:50 *** tionstav has quit IRC (Remote host closed the connection)
09:49:50 *** audigex has quit IRC (Remote host closed the connection)
09:49:50 *** belajalilija has quit IRC (Remote host closed the connection)
09:49:50 *** bohaska has quit IRC (Remote host closed the connection)
09:49:50 *** _jgr_ has quit IRC (Remote host closed the connection)
09:49:50 *** mnhebi has quit IRC (Remote host closed the connection)
09:49:50 *** frosch123 has quit IRC (Remote host closed the connection)
09:49:50 *** xarick has quit IRC (Remote host closed the connection)
09:49:50 *** jfs has quit IRC (Remote host closed the connection)
09:49:50 *** andythenorth has quit IRC (Remote host closed the connection)
09:49:50 *** audunm has quit IRC (Remote host closed the connection)
09:49:50 *** gwyd4016 has quit IRC (Remote host closed the connection)
09:49:50 *** johnfranklin has quit IRC (Remote host closed the connection)
09:49:50 *** wensimehrp has quit IRC (Remote host closed the connection)
09:49:50 *** DorpsGek_vi has quit IRC (Remote host closed the connection)
09:49:50 *** coobies has quit IRC (Remote host closed the connection)
09:49:50 *** _glx_ has quit IRC (Remote host closed the connection)
09:49:50 *** ahyangyi has quit IRC (Remote host closed the connection)
09:49:50 *** reldred has quit IRC (Remote host closed the connection)
09:49:50 *** rotterdxm has quit IRC (Remote host closed the connection)
09:49:50 *** emperorjake has quit IRC (Remote host closed the connection)
09:49:50 *** ufiby has quit IRC (Remote host closed the connection)
09:49:50 *** kuhnovic has quit IRC (Remote host closed the connection)
09:49:50 *** brickblock19280 has quit IRC (Remote host closed the connection)
09:49:50 *** xarothbrook has quit IRC (Remote host closed the connection)
09:49:58 *** DorpsGek_vi has joined #openttd
09:54:43 *** truebrain has joined #openttd
09:56:13 <truebrain> Now DorpsGek is going to be away for a moment 🙂
09:58:39 *** DorpsGek has joined #openttd
09:58:39 *** ChanServ sets mode: +o DorpsGek
09:59:56 <LordAro> truebrain: very scary
09:59:59 <LordAro> didn't know what to do
10:00:10 <truebrain> I looked in the mirror and I agree: very scary 🙂
10:04:12 <peter1138> Ah, peace and quiet for a second...
10:07:17 <peter1138> Hmm, clangd angry. Time to restart.
10:19:54 *** andythenorth has joined #openttd
10:25:34 <LordAro> truebrain: WT5(?) when :p
10:25:43 <truebrain> we once started on it
10:26:28 <truebrain> but both me and frosch kinda went AWOL a few in 😛 So who knows, maybe we pick it up again 🙂
10:28:26 <andythenorth> grf translations 😛 make it more interesting 😛
10:35:19 <LordAro> andythenorth: wouldn't expect it to be any different - it'd still have to go to a git repo somewher
10:35:40 <LordAro> unless you're thinking a complete rework of all translations and they're stored somewhere centrally
10:37:48 <truebrain> okay, deployment of game-coordinator failed horribly 😄
10:37:53 <truebrain> `TypeError: duplicate base class TimeoutError`
10:40:50 <truebrain> just happy I have this "Fail Deployment" button, which stopped the deployment of this new version and reverted it back 🙂 😄
10:43:23 <truebrain> they claim it is a drop-in replacement; let's see if that is true
10:55:26 <truebrain> okay, that multiplayer upgrade went .... difficult
10:55:31 <peter1138> I did a lazyweb search for "translator openttd" the other day and the new unfinished version of truebrain and frosch came up first.
10:55:35 <truebrain> I think everything is up on its feet again now, but pffff
10:55:44 <truebrain> peter1138: haha, that is priceless 🙂
10:56:06 <truebrain> hmm, no, something is still weird with the game-coordinator ... wth
10:56:43 <truebrain> `AttributeError: 'TransportSocket' object has no attribute 'close'`
10:57:40 <truebrain> `AttributeError: 'TransportSocket' object has no attribute 'send'`
10:57:53 <truebrain> I am not sure what happened there ...
10:59:19 <truebrain> File "/usr/local/lib/python3.11/asyncio/selector_events.py", line 902, in _call_connection_lost
10:59:29 <truebrain> I mean ... I sure hope that it isn't an issue INSIDE Python 3.11 😛
10:59:34 <LordAro> that's less than ideal
10:59:59 *** kuka_lie has joined #openttd
11:00:00 <truebrain> weirdly enough, everything does seem to work just fine
11:01:15 <LordAro> can _sock be set to the wrong type somehow?
11:01:51 <truebrain> the trace doesn't really tell me where in my code this was starting at ..
11:08:29 <truebrain> okay, it only happens during the connection-lost handler
11:11:37 <LordAro> which 3.11 release are you using? :p
11:12:03 <LordAro> probably not going to be fixed by 3.11.7 then
11:13:32 <truebrain> I think the issue is more that I do something on a connection-lost that is no longer allowed or something
11:14:31 <truebrain> no, my code-paths only do some minor internal cleanup
11:17:50 <truebrain> ``` These objects can be safely returned by APIs like
11:17:50 <truebrain> `transport.get_extra_info('socket')`. All potentially disruptive
11:17:50 <truebrain> operations (like "socket.close()") are banned.```
11:18:10 <truebrain> I am not 100% sure asyncio got that memo too 😛
11:19:07 <truebrain> okay, what happens is that the internal socket of `asyncio` is wrapped in a `TransportSocket`, which doesn't allow access to `send` and `close`
11:19:23 <truebrain> so I am guessing either myself or a library uses that socket to create a new selector
11:20:04 <truebrain> ```sock = writer.transport.get_extra_info("socket")
11:20:04 <truebrain> server = await asyncio.get_event_loop().create_connection(
11:20:04 <truebrain> lambda: GameProtocol(DetectGame(connected)),
11:20:12 <truebrain> I guess that would do it
11:33:56 <peter1138> Nowt to do with me guv.
11:35:46 <peter1138> I think that's the suggestion xarick was making ;-)
11:37:17 <truebrain> okay, the game-coordinator issue isn't an issue due to how I wrote the code. It still deserves fixing 🙂
11:37:30 <truebrain> basically, it fails when querying the server to see if it allows for direct connections
11:37:39 <truebrain> and if that fails, it falls back to STUN
11:37:45 <truebrain> which will always work, even for direct connections 🙂
11:38:59 <peter1138> I've been wondering, is it really an [out], to take a span that to be filled?
11:39:22 <LordAro> it's as much of one as a pointer "to be filled"
11:40:01 <peter1138> I guess it makes it clearer that you aren't expected to provide the shape :-)
11:44:10 <truebrain> LordAro: mind taking a look at #242 to see if you can think of a cleaner way in 2 minutes, or whether we just go with this?
11:47:37 <LordAro> truebrain: not immediately obviously, no
11:47:47 <LordAro> not that i've got a huge knowledge of asyncio :p
11:48:12 <LordAro> would recommend an issue as the (removed) deprecation notice recommends, see if anyone knows any better
11:52:19 <truebrain> I guess. I rather rewrite this GC in Go 😛
11:52:51 <truebrain> Current GC still has weird async issues we can't trace ..
11:53:09 <truebrain> So sometimes a server is kicked out wrongfully
11:56:03 <peter1138> Hmm, do we keep the co-authored-by: lines.
11:56:21 <LordAro> i think github does actually parse them
11:58:58 <_glx_> truebrain: Isn't that the one where the recommended solution is to use `@v<major>` but dependabot doesn't understand it's really the updated one ?
12:01:43 <peter1138> Hmm, Maybe 10063. Hmm.
12:03:20 <truebrain> _glx_: Nah, they didn't update their actions yet, but did update their octokit
12:14:01 <peter1138> Oh, didn't notice it failed :s
12:23:08 <xarick> 879 ticks to compute this table
12:29:42 <xarick> 949 ticks to compute all 11 tables
12:30:05 <xarick> hmm... i need to time the older versions
12:30:34 <truebrain> okay, so game-coordinatis running without errors now \o/
12:35:00 <xarick> DummyAI no longer errors out...
12:36:00 <xarick> nevermind, game was paused
12:42:33 <truebrain> frosch123: in case you read info@, I sent back an email asking for a bit more details.
12:59:06 <truebrain> 😮 that change doesn't actually work ... what did I do wrong ....
13:03:37 *** frosch123 has joined #openttd
13:03:37 <frosch123> That Korean bridge set is from 2013. I doubt they will find the author
13:05:30 <truebrain> Yeah .. I didn't want to go on the look for it, I have to admit. If you report an issue, at least give us concrete evidence to go on. This was, at least for me, to thin to act on
13:08:25 <truebrain> owh, wtf, another issue .. lovely!
13:11:08 <truebrain> Third time is the charm?
13:11:29 <truebrain> `User 'OpenTTD' does not have a verified primary email address.`
13:24:53 <_glx_> oh so many repo to update, then we can remove OpenTTD/actions/checkout
13:25:44 <truebrain> that ... is already removed?
13:31:47 <truebrain> `OpenTTD/actions/checkout-pull-request` is something different. That is an action that avoids fetching too many commits, while still fetching the full history
13:32:10 <truebrain> mostly needed for OpenTTD/OpenTTD, to avoid waiting a minute for a git clone to finish
13:33:40 <truebrain> but we use it in most repos for the commit checker
13:34:51 <_glx_> yes because we need to fetch commits to the parent commit
13:35:44 <truebrain> lol .. so `config.yml` does work in the OpenTTD/OpenTTD, but not if I put it in other repos .. wth? 😄
13:36:35 <truebrain> owh, when I copy it one-on-one, it does work .. wth ..
13:39:48 <truebrain> okay, two conditions were wrong .. it doesn't work if you do it without a template in general, and I made a typo in the configuration file
13:43:21 <xarick> selectable version to load is actually bugged
13:44:18 <xarick> but it's searching for the version which is highest that can load a v28 save
13:44:26 <xarick> that's not how it's supposed to work
13:51:19 <truebrain> okay ... I think I now updates all our software
13:51:22 <truebrain> that was long overdue 😛
13:51:43 <truebrain> and everything is Python 3.11 now \o/ 🙂
13:55:20 <truebrain> finally went cycling?
13:55:40 <peter1138> No, I went shopping.
13:55:48 <peter1138> Urgh, overflowing float, really?!
14:02:00 <peter1138> Alright, just make everything double? :o
14:08:52 <peter1138> Game: "Pressing shift always estimates the cost of doing something"
14:09:12 <peter1138> User; "Bug: pressing shift estimates the cost of doing something"
14:10:10 <truebrain> Developer: "Removes users"
14:10:28 <truebrain> it is the only way!
14:40:07 *** virtualrandomnumber has joined #openttd
14:40:30 *** virtualrandomnumber has quit IRC ()
14:42:41 *** talltyler has joined #openttd
14:42:41 <talltyler> Ah, the SkyNet solution
14:43:04 <truebrain> talltyler: when is the next meetup?
14:43:37 <talltyler> Dunno, I just (finally) quit my job that sent me to Europe a few years ago
14:44:17 <talltyler> I’m a contractor for a Scottish game studio now but don’t imagine they’ll fly me over any time soon
14:44:35 <talltyler> Maybe I’ll do a North America meetup and you’ll have to come here 😛
14:45:01 <truebrain> only if it is in Canada 😛
14:45:02 <talltyler> Or more likely, I’ll get a work visa and move to the UK…
14:45:31 <talltyler> I do not want to discuss politics though, you can infer my reasoning yourself 🙂
14:45:35 <truebrain> Also: gratz on the new job 🙂
14:45:52 *** xarothbrook has joined #openttd
14:45:52 <xarothbrook> Working on an unnamed game with a 6 in the title?
14:46:13 <talltyler> xarothbrook: I cannot confirm nor deny 😛
14:46:16 <truebrain> As a commit name showed a few days earlier, I went for a similar track. Just picked a studio a bit more West of yours 😛
14:47:15 <talltyler> Very cool! Never been but it looks like a beautiful place.
14:47:27 <truebrain> Totally unrelated, I booked a cruise to Iceland for August
14:47:30 <truebrain> really looking forward to it 🙂
14:49:52 <xarothbrook> talltyler: Gratz, also, Scotland is a very nice place... I've been meaning to go, and not just to try out how Irn Bru tastes like.
14:50:17 <Borg> I implemented Cargo RR balancing between 2 industries nearby
14:50:31 <Borg> works! :) kinda hacky.. but need to check if cargo is correctly split
14:51:25 <talltyler> I think JGRPP has a setting to evenly distribute cargo between all industries within the station catchment
14:51:53 <Borg> Im using vanilla OpenTTD. so I did it in GRF
14:52:03 <Borg> but of course it can work only with 2 industries
14:52:13 <talltyler> …how is that possible in GRF?
14:52:22 <Borg> with 3 or more .. 3th will not receive cargo
14:52:31 <Borg> talltyler: Opt out cargo acceptance magikery :D
14:52:53 <talltyler> How does the industry know another is within catchment?
14:53:07 <talltyler> *the same catchment
14:53:38 <Borg> talltyler: it doesnt care about catchmen.. every month (when mothly production call back kicks in) industry check for same industry type within 6 tiles.. if it founds one.. it activates RR cargo balancing
14:53:59 <Borg> I implemented it only for Power plants.. and I think only those will have RR cargo balancing
14:54:11 <Borg> for other industries I will leave cargo overflowing
14:54:48 <Borg> okey. it works nicely! just playing the game and monitoring 2 power plants.. cargo is evenly distributed :) not bad
14:54:52 <talltyler> Huh, seems like it would be easier to implement in OpenTTD itself 🙂
14:55:23 <talltyler> I’ve done some silly things with GRFs, then I realized it was easier to change the rules than find loopholes 😛
14:55:40 <Borg> maybe one day.. I will finaly manage to compile OpenTTD on my box
14:55:47 <Borg> but for now.. im happy with vanilla
14:56:10 <Borg> talltyler: its for my BSPI.grf.. Power Plant produce electricity.. and if you have lots of industries you need more Power Plants
14:56:25 <Borg> and I toyed with RR cargo balancing for a while at least for those :)
14:59:03 <Borg> I still need to lookup cargo distribution code in OpenTTD
14:59:24 <Borg> because im suprised that its being distributed to 2 industries on same tick..
14:59:38 <Borg> I tought it will go to 1 industry on one tick.. then to the other
15:02:38 <talltyler> I wouldn’t mind seeing cargo splitting as a feature in vanilla, but I think I’d want a more elegant user experience than a global setting to select “nearest industry to stations sign” or “all industries within station catchment”
15:02:55 <talltyler> Creative thinking required. 🙂
15:03:14 <Borg> I would really love to get EXTRA info in callback
15:04:05 <Borg> because for now.. it tries first industry it find. if its opt out.. tries next one.. and next one untill all are tried within catchment area.. or cargo is already distributed
15:04:12 <Borg> currently ci1 and ci2 is unused
15:04:38 <Borg> 18 lowest byte is Cargo ID
15:04:58 <Borg> so you can use higher byte for counter
15:05:13 <Borg> and so I could use smarter cargo spliting
15:07:17 <Borg> so cargo is not evenly distributed
15:12:31 <peter1138> > Variable GROVE_SEGMENTS is not used.
15:28:55 <andythenorth> oof unintended nap
15:29:10 <Borg> I miss NAND in VarAction2Advanced
15:38:38 <_glx_> it's not that hard, you just need to combine some AND and XOR
15:39:12 <_glx_> might require temp storage usage too
15:39:13 <peter1138> Borg, do you mean AND NOT?
15:54:23 <Borg> peter1138: no.. but I think its also not really NAND
15:54:50 <Rubidium> peter1138: but what about ORN and XORN? :D
15:55:08 <Borg> so 110 NAND 010 gives 100
15:55:13 <Borg> Rubidium: those are not really usefull
15:55:20 <Borg> but NAND can be very usefull
15:56:04 <Borg> need to lookup those logical operators :)
15:56:53 <peter1138> This is why we can't add them. Nobody knows ;-)
15:57:14 <_glx_> yeah your NAND example feels wrong 🙂
15:58:37 <Borg> it should be called then BCLR
15:58:55 <_glx_> for me result should be 101
15:59:15 <Borg> _glx_: yeah. perhaps.. as for NAND.. indeed
15:59:19 <Borg> thats not what I want :)
15:59:26 *** Wormnest has joined #openttd
16:00:20 <_glx_> 1 for anything other that 1,1
16:00:21 <Borg> NAND is bad name indeed... BCLR sounds much better
16:00:45 <_jgr_> Would you ever use that with something other than a constant?
16:01:14 <Borg> I can load flag bit into temporary storage
16:01:26 <Borg> and now.. clearing it.. is cumbersome :) need to negate it via XOR
16:01:32 <Borg> but first I need to confirm its 1
16:02:15 <_jgr_> OR, then XOR would be way to do that without needing to spill anything to temp storage
16:02:47 <_glx_> /2r flag, /2^0xFFFFFFFF, /2&value
16:03:07 <Borg> yeah.. that one will work too. but jgr's one looks prettier
16:03:38 <Borg> and easier to implement in grf
17:11:39 <peter1138> Okay, so I have the probability of land industries vs water industries.
17:11:56 <peter1138> And I have the proportion of land tiles...
17:12:51 <peter1138> Are those probabilities based on some assumption about how much land there is...
17:21:26 <peter1138> Hmm, yeah, default oil rigs only appear around the edges of maps, so scaling by amount of water everywhere... meh
17:38:03 <peter1138> Where are they meant to spawn if most of the edge of the map is water?
17:43:44 *** locosage has joined #openttd
17:43:44 <locosage> iirc at some point I was thinking about oil stuff and amount of water
17:43:56 <locosage> but ended up just increasing the range on distance setting
17:44:07 <locosage> as there doesn't seem to be any universal solution to it
17:51:33 <peter1138> Hmm, do oil rigs never spawn on generation, or have I broken something :D
17:54:19 <peter1138> Oil wells are spawning.
17:57:10 <peter1138> Initial generation probability seems to ignore date anyway.
17:59:30 <peter1138> Okay, creation probability is zero.
18:00:07 <peter1138> I mean, that means I don't need to care about scaling land/water industries... at least for vanilla.
18:00:35 <Borg> okey I scraped that idea of RR cargo distribution
18:00:49 <Borg> instead, I use adaptive Cargo opt out
18:01:15 <Borg> it will work better as I care only about suficient electricity production.. not true cargo distribution
18:02:24 <peter1138> Okay, 5120 industries. 4801 of them will be land industries, 318 of them will be water. Rounding error I guess.
18:02:46 <peter1138> Then I scale the land industries by amount of land, so that gets reduced to 1199. Water industries stay unscaled, so 318.
18:04:00 <Borg> Bonus is.. I can have more than 2 power plants in cluster :>
18:08:21 *** wensimehrp has joined #openttd
18:08:28 <xarick> I need #8430 limitations implemented
18:08:41 <wensimehrp> I thought you're gonna say hi 😦
18:08:53 <xarick> there is a "use case" now
18:11:43 <xarick> we can chose which version of scripts to run
18:11:53 <peter1138> OOf, overflow error.
18:12:31 <xarick> TrueBrain, the destroyer of features
18:14:15 <xarick> gonna look into `force_exact_match` see if I can do something
18:16:40 <xarick> 5 years later, I can understand TrueBrains viewpoint
18:17:09 <xarick> handling scripts was messy
18:17:15 <xarick> loading scripts and such
18:17:46 <xarick> messy code, so he makes it cleaner, but at the same time... kills features which is what annoys me
18:23:43 <peter1138> And you wonder why he has you on ignore...
18:25:57 <peter1138> Speaking of which, let's see what else I should close.
18:26:13 <truebrain> on a closing spree?
18:26:26 <truebrain> I closed 52 PRs today!
18:26:31 <truebrain> they were all from 1 person even!
18:28:03 <peter1138> Hmm, was "copy properties of an existing NewGRF entity" still a thing.
18:28:40 <truebrain> just a bot; not even more fancy than that 😛
18:28:43 <truebrain> (dependabot, to be more exact)
18:32:58 <peter1138> Hmm, what do I dinner?
18:33:35 <frosch123> Iirc in some places we scale coast/sea stuff linearly using the map width, instead of quadratic using map area
18:34:19 <xarick> subsidies should scale
18:34:51 <peter1138> Only for oil refinery and oil rig.
18:35:37 <xarick> would scale based on current era's engines speed/capacity or so
18:36:30 <peter1138> Objects also have some rules, but I didn't touch them. They already have a scalebywater flag, which complicates that.
18:38:18 <peter1138> For land industries, the scaling for coast is probably okay, if we assume the proportion of land/water is fairly equally distributied. Perhaps.
18:39:40 <xarick> so my shared list approach across multiple tables performs worse than computing it from scratch
18:40:16 <xarick> squirrel is quite interesting
18:42:43 <peter1138> Hmm, curry perhaps.
18:48:59 <andythenorth> wonder which is better
18:49:19 <andythenorth> variants sharing varact2 graphics chain, with a switch to check specific IDs
18:49:26 <andythenorth> or each variant has own varact2 chain
18:49:37 <andythenorth> probably potato / potato
18:49:54 <truebrain> _glx_: 40 linux wheels? Damn 😄
18:50:02 <truebrain> did not know there even were that many variations 😛
18:50:28 <peter1138> andythenorth, I have no real opinion other than "what the fuck was I thinking with variants"
18:51:27 <truebrain> (getch is get+catch 😄 )
18:55:06 <andythenorth> nothing wrong with variants like this
18:57:57 <peter1138> Mmm, hot food. Curry. Mmm.
18:58:01 <wensimehrp> peter1138: can a station tile read other station tiles' badges?
19:11:22 <peter1138> They should be able to but whether I have coded that yet is another matter.
19:12:13 <LordAro> truebrain: gonnagetsthemall
19:12:39 <peter1138> It's slightly awkward as it requires multiple parameters, so temp registers come into play.
20:05:59 <Rubidium> peter1138: I'm tempted to approve #13515 (improved tree placement), but I'm wondering whether the PR's name should be made more specific (especially for the changelog). Something like 'Change: Make tree placement at world generation look more organic"
20:14:44 *** Borg has quit IRC (Quit: leaving)
20:15:40 <peter1138> Okay, well, rename it :)
20:39:47 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
20:56:52 <truebrain> lol @ ply: `it is maintained as a mature library` .. last modified date? 2 years ago 😄
20:57:07 <truebrain> some ambitions died 🙂
20:57:33 <peter1138> Well, someone will probably complain that the number of trees isn't exactly the same.
21:01:36 <locosage> 13515 looks overly complex for what it achieves
21:01:59 <locosage> feels like just some wiggly function on distance and angle could've done the same
21:05:01 *** yiffgirl has joined #openttd
21:05:02 <yiffgirl> PeterNviaGitHub: just as i finished making my visualizer lol
21:05:02 <yiffgirl> it's actually quite fascinating. i'm not quite sure how it works, but `s`, `t`, and `d` are essentially "which side of line N is this point on".
21:05:02 <yiffgirl> as long as the point is within the triangle, they all share the same sign, but as soon as you cross one line, you're on the other side and thus at least one of the three don't match
21:05:43 <xarick> dP is the master of algorithms
21:07:04 <yiffgirl> eg: `s = (side2width * distanceFromVertex2.y) - (side2height * distanceFromVertex2.x)`
21:07:04 <yiffgirl> i'm not sure why that works but it's so interesting
21:08:51 *** johnfranklin has joined #openttd
21:08:51 <johnfranklin> I overslept and missed dinner
21:09:32 <truebrain> never too late for dinner!
21:10:41 <peter1138> If you miss dinner, then the next one is supper.
21:10:57 <peter1138> And yes, predictably all the experts come out after something is merged.
21:11:23 <yiffgirl> i'm just happy it's in
21:11:38 <yiffgirl> no more ugly diamonds!
21:13:54 <frosch123> yiffgirl: In case you really want to know: The sign of the determinant tells the orientation of the vectors
21:14:14 <yiffgirl> frosch123: the determinant?
21:15:33 <truebrain> yiffgirl: it is pretty 😄
21:15:57 <peter1138> Thanks for contributing.
21:16:57 <truebrain> what else will you be working on yiffgirl ? 😄
21:18:03 <yiffgirl> truebrain: there was that tree spreading pr.
21:21:33 <yiffgirl> other than that my general plan is to work on anything tagged stale that i think looks cool
21:21:33 <yiffgirl> probably more work on my newgrfs before that though
21:26:25 <truebrain> (I like faster regression runs 😛 )
21:28:00 <peter1138> Bumping the version number should be fine. I only bumped it to 3.16 because that was the highest version in the source at the time.
21:28:19 <peter1138> That's in the wrong place.
21:28:40 <peter1138> That's in the right place.
21:35:26 <peter1138> Let's hope the CI doesn't complain now :D
21:45:43 <truebrain> peter1138: CI likes it just fine 🙂
21:47:46 <truebrain> su has been typing for the last 10 minutes now
21:48:04 <yiffgirl> hm. my last comment on that spreading pr [slightly edited for clarity]:
21:48:04 <yiffgirl> > my intent was to make this forest spreading a new spreading setting rather than relying on the "improved" placement, but i'm realizing that doing so would involve changing the spreading options from
21:48:04 <yiffgirl> > - spread only in rainforests
21:48:05 <yiffgirl> > - spread everywhere
21:48:09 <yiffgirl> > - spread only in rainforests
21:48:11 <yiffgirl> > - spread only when near many other trees
21:48:11 <yiffgirl> > - spread only in rainforests, near many other trees
21:48:13 <yiffgirl> > - spread everywhere
21:48:15 <yiffgirl> > which is a bit more complexity than i was bargaining for. i'd like your feedback before i go ahead with it.
21:48:15 <yiffgirl> i think that probably still needs addressing? or maybe it's just bikeshedding and i shouldn't bother with making spreading interact with rainforests
21:49:46 <andythenorth> wave form collapse 😛
21:53:23 <frosch123> yiffgirl: "Spreading in rainforest" is related to original TTD and the lumbermill industry. It can only be built in rainforest and cuts actual trees
21:58:49 <yiffgirl> i think then that i should go with
21:58:49 <yiffgirl> - spread only in rainforests
21:58:49 <yiffgirl> - spread only when near other trees
21:58:51 <yiffgirl> and have the "only in rainforests" setting use the current spread mechanism
22:10:15 <peter1138> What if there is no new option, and "spread everywhere" means "spread everywhere, biased to near other trees"?
22:14:25 <peter1138> Hmm, why does trying to overbuild stations (or otherwise modify) someone elses station not give the usual "... owned by" error.
22:15:30 <peter1138> The owned by error only appears if you try to destroy or remove it.
22:19:20 <peter1138> Sped up Windows CI tests by ~ 10 seconds or so? Or just a coincidence...
22:25:47 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:32:25 <xarick> is it possible to use references in squirrel
22:39:22 *** kuka_lie has quit IRC (Quit: leaving)
22:46:35 <yiffgirl> peter1138: funny you should say that,
22:47:23 <peter1138> Sorry I don't remember things from 9 months ago...
22:47:51 <yiffgirl> not a problem, just funny
22:48:32 <peter1138> Well, diamonds, eh?
22:50:29 <yiffgirl> if i recall right people seemed divided on if it should be a new thing or if it should overwrite an old thing, even on the original pr.
22:53:26 <peter1138> Looking at Zorg's last post.
22:53:50 <peter1138> The shape is best in the second picture, although perhaps too many trees.
22:54:11 <peter1138> The JGRPP picture is... weird?
22:54:50 <peter1138> Hmm, I think I can split off the error-company changes first.
22:55:29 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:56:39 <peter1138> But maybe the idea of "trees: low, med, high" is viable.
22:56:59 <peter1138> But then maybe not.
22:58:35 <yiffgirl> plant a new tree only if tree count <= some number, perhaps?
22:59:39 <yiffgirl> the jgr picture seems like it was just uniform but less dense
22:59:44 <peter1138> After 3 years my map is filled with trees :D
22:59:57 <peter1138> The uniformness is not nice.
23:01:40 <yiffgirl> trees turning the entire map into a completely uniform mass has been my biggest issue with the game for a long while.
23:03:37 <yiffgirl> which is maybe a bit silly of me that that's what irks me the most, but still
23:07:45 <peter1138> Hmm, why is the tree growth algorithm not changable in game...
23:08:05 <peter1138> Or... does that only affect world gen. That would make sense.
23:22:13 <peter1138> Maybe something like water flooding...
23:34:38 <peter1138> If you simply limit growth, then tree clumps "migrate" a bit.
23:47:36 <peter1138> Oh, that's why my trees are growing so fast.
23:48:07 <peter1138> It's set to 120, so my "after 3 years" was actually "after 30 years"
23:49:01 <peter1138> This needs an algorithm with built in hysteresis.
continue to next day ⏵