IRC logs for #openttd on OFTC at 2026-10-25
⏴ go to previous day
00:02:59 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
00:03:12 *** MinchinWeb[m] has joined #openttd
00:06:51 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
00:07:07 *** MinchinWeb[m] has joined #openttd
00:55:28 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
00:55:51 *** MinchinWeb[m] has joined #openttd
00:55:56 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
00:56:13 *** MinchinWeb[m] has joined #openttd
01:04:23 *** MinchinWeb[m] has quit IRC (Remote host closed the connection)
01:04:56 *** MinchinWeb[m] has joined #openttd
01:08:59 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
01:09:47 *** MinchinWeb[m] has joined #openttd
01:13:49 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
01:14:54 *** MinchinWeb[m] has joined #openttd
01:47:10 *** audigex has joined #openttd
01:47:10 <audigex> The dog appears to be drunk
01:47:23 <audigex> peter1138: You could spill your drink on your laptop
02:03:35 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
02:03:57 *** MinchinWeb[m] has joined #openttd
02:04:05 *** herms22 has quit IRC (Quit: bye)
02:07:16 *** herms22 has joined #openttd
02:23:29 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
02:23:31 *** WormnestAndroid has joined #openttd
02:24:12 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
02:24:14 *** WormnestAndroid has joined #openttd
02:24:52 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
02:24:53 *** WormnestAndroid has joined #openttd
02:40:33 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
02:41:28 *** MinchinWeb[m] has joined #openttd
03:57:28 <peter1138[d]> talltyler: Could do with some announcement stuff, I don't have the powers here.
04:41:43 *** Phileman has joined #openttd
04:45:16 *** Philemon has quit IRC (Ping timeout: 480 seconds)
05:01:01 *** ChanServ sets mode: +v tokai
05:08:03 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
05:50:16 *** Izzybee has joined #openttd
06:00:02 *** Izzybee has quit IRC (Ping timeout: 480 seconds)
06:03:37 *** Izzybee has joined #openttd
06:24:07 *** Izzybee has quit IRC (Ping timeout: 480 seconds)
06:32:54 *** Izzybee has joined #openttd
07:21:43 *** kaibaneddy has joined #openttd
08:23:04 *** will_marshall_ has joined #openttd
08:24:00 <kaibaneddy> must be the coffee
08:29:26 <will_marshall_> it has this effect
09:11:06 <peter1138> A wild Darkvater appeared in t' forums.
09:17:08 <peter1138> A wild Sunshine Motors appeared, too.
09:18:14 <andythenorth> doubles and triples
09:22:02 *** gelignite has joined #openttd
09:44:38 <peter1138> andythenorth, no, I was only on pints, not spirits.
11:06:38 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
11:06:54 *** WormnestAndroid has joined #openttd
11:19:35 <LordAro> just looked at the kitchen clock and thought I'd start making lunch
11:28:43 *** vim6790 has joined #openttd
11:32:42 <will_marshall_> Power cut, might be time for a family trip to mcd's
11:32:49 <will_marshall_> oh except the power is out at mcd's too
11:33:12 <will_marshall_> heating some beans on the camping stove
12:02:28 <peter1138[d]> Walking along the canal, I've gone past 4 locks. Sounds like a good crypto key.
12:48:17 *** Izzybee has quit IRC (Ping timeout: 480 seconds)
13:07:44 *** Izzybee has joined #openttd
13:19:04 *** Izzy is now known as Guest3876
13:19:49 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
13:21:02 *** MinchinWeb[m] has joined #openttd
13:22:59 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
13:23:02 *** Izzybee has quit IRC (Ping timeout: 480 seconds)
13:23:32 *** MinchinWeb[m] has joined #openttd
13:26:37 *** Guest3876 has quit IRC (Ping timeout: 480 seconds)
13:28:58 *** Izzybee has joined #openttd
13:45:19 <rito12_13> stanleys12viaGitHub: slopley
14:26:47 <_glx_> rito12_13: I would have suggested the same code without needing AI 😉
14:55:05 *** Twofish has joined #openttd
14:56:55 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
14:57:00 *** MinchinWeb[m] has joined #openttd
14:57:19 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
14:57:37 *** MinchinWeb[m] has joined #openttd
15:17:58 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
15:19:43 *** MinchinWeb[m] has joined #openttd
15:23:44 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
15:24:11 *** MinchinWeb[m] has joined #openttd
15:30:44 <rito12_13> stanleys hasn't contributed to the project yet, so maybe AI helped them understand things better, I don't know
15:32:24 <andythenorth> curious if action 2 `remove_tmp_location` is expensive
15:52:37 *** wallabra has joined #openttd
16:01:40 *** wallabra has joined #openttd
16:05:29 <_glx_> doesn't look very expensive
16:06:13 *** Smedles has joined #openttd
16:09:30 *** wallabra has joined #openttd
16:10:52 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
16:11:25 *** MinchinWeb[m] has joined #openttd
16:17:05 <andythenorth> ```ncalls tottime percall cumtime percall filename:lineno(function)
16:17:05 <andythenorth> 33694410/2387 9.381 0.000 9.383 0.004 ../pypy3/lib/pypy3.12/site-packages/nml/actions/action2.py:137(remove_tmp_location)```
16:17:05 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
16:17:18 *** MinchinWeb[m] has joined #openttd
16:17:22 <andythenorth> not sure what that means though, blind profiiing here 😛
16:17:22 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
16:17:40 <andythenorth> rather you than me
16:18:53 *** MinchinWeb[m] has joined #openttd
16:18:59 <_glx_> percall are imperceptible 🙂
16:19:58 <andythenorth> tottime is in a profiling run, which is wildly slower
16:20:28 <andythenorth> about 7x slower 😛
16:20:45 <_glx_> tottime is the whole run, not just the function
16:31:04 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
16:32:20 *** MinchinWeb[m] has joined #openttd
16:36:25 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
16:37:22 *** MinchinWeb[m] has joined #openttd
16:45:03 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
16:45:03 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
16:45:21 *** MinchinWeb[m] has joined #openttd
16:45:30 *** MinchinWeb[m] has quit IRC (Remote host closed the connection)
16:45:46 *** MinchinWeb[m] has joined #openttd
16:46:14 *** Twofish has quit IRC (Remote host closed the connection)
16:46:35 *** Twofish has joined #openttd
16:51:18 *** wallabra has joined #openttd
16:56:31 *** WormnestAndroid has joined #openttd
16:58:23 *** wallabra has joined #openttd
17:13:04 *** MinchinWeb[m] has quit IRC (Remote host closed the connection)
17:13:21 *** MinchinWeb[m] has joined #openttd
17:17:24 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
17:18:49 *** MinchinWeb[m] has joined #openttd
17:26:58 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
17:27:11 *** MinchinWeb[m] has joined #openttd
17:31:13 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
17:32:41 *** MinchinWeb[m] has joined #openttd
17:44:55 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
17:45:33 *** MinchinWeb[m] has joined #openttd
18:04:03 <andythenorth> trying to apply the diff 🙂
18:13:05 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
18:15:45 *** MinchinWeb[m] has joined #openttd
18:16:24 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
18:17:50 <andythenorth> ```33694410/2387 8.356 0.000 8.357 0.004 .../pypy3/lib/pypy3.12/site-packages/nml/actions/action2.py:137(remove_tmp_location)```
18:18:26 <andythenorth> only one run though, could just be jitter?
18:20:09 <andythenorth> "I should script repeated runs or something"
18:24:55 *** MinchinWeb[m] has joined #openttd
18:25:18 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
18:25:51 *** MinchinWeb[m] has joined #openttd
18:27:14 <andythenorth> second run though
18:27:14 <andythenorth> ```33694410/2387 8.309 0.000 8.310 0.003 ../pypy3/lib/pypy3.12/site-packages/nml/actions/action2.py:137(remove_tmp_location)```
18:33:53 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
18:34:07 *** MinchinWeb[m] has joined #openttd
18:38:51 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
18:39:09 *** MinchinWeb[m] has joined #openttd
19:01:32 <_glx_> so `self.tmp_locations` is actually too small to see an effect with `if location in self.tmp_locations:` (for list it's O(n) but O(1) for set)
19:02:14 <_glx_> (well it's a 1s difference)
19:47:50 *** deku_shrub has joined #openttd
19:51:37 <peter1138> Step 1: don't use AI.
19:56:59 <andythenorth> "we have Claude at home"
19:59:50 <andythenorth> sorry deku_shrub the project doesn't accept AI contributions
20:15:58 <deku_shrub> I am aware, hence sharing it as a fork for discussion
20:18:19 <will_marshall_> It warrants as much discussion as you put effort into it.
20:22:56 <deku_shrub> Quite a lot actually
20:29:38 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
20:31:30 *** WormnestAndroid has joined #openttd
20:33:36 <LordAro> trouble is it's impossible to tell the difference
20:34:20 <LordAro> as the code is (for want of any better term) not useful to us, a video demonstrating the feature might be interesting
20:34:47 <peter1138> There is a video if you can stand listening to AI-generated voice-overs.
20:38:54 <_zephyris> The changes are +9,862/-205 lines. Reviewing nearly 10k lines of changes is a multi-week full-time effort - for a volunteer project, you're not going to find anyone willing to do that
20:42:13 <deku_shrub> peter1138: That is me, not an AI
20:44:53 <peter1138> Huh, then something about your recording set up adds the characteristic AI-voice-wishywashybuzziness.
20:46:32 <talltyler> I haven’t looked at the diff (currently sat on a train using my phone) but I’m suspicious of any 10k lines-of-code change. I spent 13 months rewriting how time works in OpenTTD, and that was only about 2k lines.
20:47:03 <andythenorth> nobody is going to review claude contribs
20:47:12 <LordAro> as a feature, i quite like it
20:47:18 <andythenorth> I'm here reviewing a GPT migration script for a grf, it's painful
20:50:36 <will_marshall_> As a feature it's fairly nauseating but I am not feeling 100% right now. The zoom is fine but the fly-to is crying out for an easing function.
20:51:37 <deku_shrub> will_marshall_: It's based on the algorithm that Google maps uses
20:52:08 <deku_shrub> peter1138: Was just recording on my laptop
20:56:12 <will_marshall_> In terms of solving the stated problem - I think it probably does give the user more context about the origin and destination of the camera move, but I would probably evaluate some less invasive methods to achieve the same thing. For instance, during camera pans we could show the smallmap with the current view extents in one corner of the screen and keep it presented for a few seconds after the
20:56:12 <will_marshall_> move. In my head I'm imagining that smooth zoom transitions came first and the zoom-out during the camera pan came second (and serves to justify the first feature).
20:58:01 <deku_shrub> Nah I did the zoom first, then I found it opened up the fly-to feature.
20:58:01 <deku_shrub> More navigation elements could be shown during the fly sure, but I was unhappy with the current pan-to behaviour as-is
20:58:37 <will_marshall_> That's what I said, innit?
20:59:52 <deku_shrub> I am aware this will be uphill because use of AI, but I thought some nice base quality of life stuff rather than e.g. full 3d conversion mods might be more worth engaging with
21:00:59 <peter1138> It's not uphill. It's an unscalable wall. You used AI, end of.
21:02:17 <peter1138> Enjoy your fork, I suppose.
21:03:07 <will_marshall_> Congratulations, it's a fork! You now have to care for and feed the thing for at least 18 years until it's all grown up.
21:03:49 <will_marshall_> There is a bloke down the street selling hessian sacks labelled "Archive this Repository" in the Github settings.
21:23:09 <andythenorth> I guess someone will put a giant EverythingAI fork together
21:24:55 <will_marshall_> What will happen first: the frontier labs go out of business or someone manages to have a big license-violation jamboree and makes a thin clone of OpenTTD with the serial numbers filed off and shoves it onto Steam?
21:26:34 <andythenorth> Steam review system must be drowning in junk?
21:26:44 <andythenorth> and license violations
21:26:47 <will_marshall_> What review system it seems to be just an open orifice through which slop flows?
21:27:21 <andythenorth> my AI is doing well porting my grf
21:33:55 <andythenorth> repeatable migration pipelines are my #2 use case for actually using AI
21:34:10 <will_marshall_> from my pov all AI use cases are #2 hayoooooo
21:34:25 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
21:34:27 <andythenorth> I am not a programmer, I just want to make stuff
21:34:45 <will_marshall_> I have AI use cases they're all just really really boring and workaday.
21:35:27 <will_marshall_> I can get everything that I need an LLM for done with a tiny little model on my spare gaming PC. 8gb critters.
21:35:27 <andythenorth> mine are basically "discovering new python modules", "avoiding javascript", "writing migration scripts" and "I will never learn bash"
21:35:54 <andythenorth> if everything to do with local LLM wasn't currently baffling, I'd run one locally
21:36:12 <will_marshall_> just install LM studio and click around it's actually easy for the babiest of brains
21:36:21 <mnhebi> andythenorth: you got 128gb of VRAM mate?
21:36:38 <will_marshall_> he's got a fairly sporty macbook if i recall
21:36:43 <will_marshall_> they have unified memory
21:36:49 <mnhebi> cause you'd be running a spellchecker capable LLM on anything less..
21:37:18 <mnhebi> and even the quants that fit 128gb need a lot of handholding..
21:37:44 <will_marshall_> I don't know how good frontier models have to be for you to say that the local models aren't usable at all.
21:37:58 <will_marshall_> But I don't want to find out because I cannot be compelled to talk about the stupid frontier models.
21:38:01 <mnhebi> they are usable, but in very specific use cases and it with a lot of handholding
21:38:13 <mnhebi> you basically got a frontier level or near frontier AI doing the handholding
21:38:19 <andythenorth> audio AI kills it
21:38:21 <will_marshall_> They seem to do everything I want just fine, but I'm giving them baby jobs.
21:38:43 <mnhebi> I mean yeah if you give them line by line edits
21:39:03 <will_marshall_> I am not vibe coding with them.
21:39:29 *** WormnestAndroid has joined #openttd
21:39:42 <will_marshall_> I fancy dinner.
21:40:07 <andythenorth> local dinner LLM
21:40:16 <andythenorth> vs. frontier dinner
21:40:18 <will_marshall_> my local LLM says it's glue pizza again 🙁
21:41:55 <will_marshall_> I already did that when I started nodding off at 4pm.
21:43:26 <talltyler> You must be hungry if I’ve had dinner first in US Eastern 🙂
21:44:04 <will_marshall_> Hangry, even. Had a bit of a nothing day between the extended powercut and two walks of the dog.
21:44:09 <peter1138> Any yellow-name fancy posting in the openttd-release channel on Discord?
21:44:48 <talltyler> Do I trust myself to post from my phone…? 🙂
21:45:37 <will_marshall_> The hard part is making the snazzy cover image? You'd have to end that tradition.
21:46:09 <reldred> Just post ‘new OpenTTD release check the website yourself I’m at the pub’
21:46:55 <talltyler> I’m on a train, even, on brand for OpenTTD
21:47:15 <will_marshall_> No, no. No trains until tomorrow.
21:47:19 <peter1138> reldred, I couldn't go tonight. That would be too much.
21:47:49 <reldred> I never did get around to doing tubes yesterday
21:48:09 <reldred> None of my chores got done
21:49:07 <talltyler> Someone should make a cover image featuring the helicopter which now lasts forever 🙂
21:50:08 <peter1138> talltyler, emergency title game, and it's just a helipad with lots of Guru X2s
21:51:29 <talltyler> Don’t let me forget to PR the new title game once we branch for RC1, the competition ended a while ago 🙂
21:52:28 <talltyler> Back to Temperate climate, I’ll miss Toyland 😛
21:55:35 *** Flygon has quit IRC (Read error: Connection reset by peer)
21:56:31 *** jfs_ has quit IRC (Ping timeout: 480 seconds)
22:01:44 <will_marshall_> rito12_13: oh hell yeah
22:02:27 <peter1138> Neat, I can remove my non-working stash.
22:11:48 <rito12_13> It crashes when a GS uses translations though
22:12:25 <LordAro> well that's their fault
22:26:16 *** Wolf01 has quit IRC (Ping timeout: 480 seconds)
23:27:57 <_glx_> and of course he decided to replace the generic `Array<Titem>` solution with a specific `Array<SQInteger>` one
23:29:09 <LordAro> I'm not sure a function that returns an array is particularly desirable either
23:29:55 <_glx_> as I understand it it's to be able to directly pass the return to a function requiring an array as parameter
23:35:06 <_glx_> anyway the function name is misleading
continue to next day ⏵
