IRC logs for #openttd on OFTC at 2023-09-05
⏴ go to previous day
00:38:05 *** gelignite has quit IRC (Quit: Stay safe!)
03:01:15 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:01:39 *** D-HUND is now known as debdog
04:19:18 *** tokai|noir has joined #openttd
04:19:19 *** ChanServ sets mode: +v tokai|noir
04:26:01 *** tokai has quit IRC (Ping timeout: 480 seconds)
04:33:06 *** Flygon has quit IRC (Read error: Connection reset by peer)
08:48:16 *** gelignite has joined #openttd
10:16:38 *** gelignite has quit IRC (Quit: Stay safe!)
10:23:12 <truebrain> Rust and its errors ... most of the time they are really good .. sometimes you get `^ the trait HostFunction<MyEnv, _, _, wasmer::externals::function::WithEnv> is not implemented for closure [closure@src/main.rs:36:70: 36:115]`
10:42:06 <peter1139> LordAro, hmm, might have to do an "amble" ride to cover some missing map squares.
11:10:19 *** linda[m]1 has left #openttd
12:16:25 <peter1139> There's a few squares that actually don't have any roads or tracks though :(
12:22:10 <LordAro> that is the tricky bit
12:22:19 <LordAro> it's rare that there's nothing at all
12:28:56 <peter1139> There's often a farm track that's closed off with a gate...
12:33:55 <LordAro> can you give an example square?
12:36:44 <peter1139> Actually that track/road across the middle might be passable.
12:37:56 <LordAro> ultimately it depends how willing you are to go through farmyards
12:38:01 <peter1139> It's not even remote, we just seem to have quite a lot of expanses with no small through roads.
12:38:27 <LordAro> the fact that they've got paths on the other end of the 'roads' suggests that there's nothing actually 'private'
12:38:42 <LordAro> so you *shouldn't* run into issues with farmers etc
12:38:48 <peter1139> It's one of the biggest complaints about here, every time they close off the minor roads for road works (HS2 has killed a few) we always have to ride out on the main roads.
12:39:34 <LordAro> we had one of those a couple years ago, they built a bypass around Leeds and cut off a road
12:39:49 <LordAro> when we first discovered it we had to clamber over a partially contructed road, rebar and all
12:40:07 <LordAro> luckily, they've since completed it and put in a lovely segregated cycle path all the way along the new road
12:40:49 <_glx_> isn't there a tacite rule saying if you open a gate, just close it behind you ?
12:41:21 <peter1139> Of course, but most gates are locked to keep people out.
12:42:08 <_glx_> it must be annoying for farmers too to have to carry keys for the gates
12:42:40 <peter1139> Hmm, these map squares don't line up with the OS grid unfortunately :)
12:43:01 <peter1139> OS maps are much better for showing right of ways.
12:47:21 <LordAro> the strava heatmap is useful too
13:00:29 <peter1139> Yeah I'll give it a go :)
13:01:05 <peter1139> Might be time to bring out the touring bike -- dynamo lights so doesn't matter if I get lost and stay out in the dark, and better tyres for this sort of stuff.
13:01:55 <peter1139> And being nearly twice as heavy doesn't matter when I don't have to keep up with everyone else.
13:10:33 *** gelignite has joined #openttd
14:10:47 <peter1139> And some very unfriendly looking signs on the fence :)
14:13:13 <peter1139> (I haven't been looking at this for this many hours, I just happened to switch back to the wrong tab which was still open :))
15:20:57 <LordAro> peter1139: jump over the hedge with your headunit
16:41:19 *** HerzogDeXtEr has joined #openttd
16:52:41 <brickblock19280> how is the order in the object order meant to work?
16:53:19 <brickblock19280> these are two different grfs which but the thing at the bottom is from the top grf
16:53:30 <brickblock19280> but has an extended id
17:57:27 <truebrain> okay, Rust + wasmtime means every TCP connection consumes roughly 30 KiB of RAM, and it can spawn at least 1000 connections per second ... and with this, every connection is assigned its own WASM instance, meaning that if something goes wrong, only that single connection is impacted
17:57:34 <truebrain> that is like 20x better than the Python alternative ..
17:58:02 <truebrain> let's see if I can port bananas-server to this ... that will take a few days of work 😛 But at least conceptually this works fine .. a lot less memory, a lot less CPU, a lot easier to work with ..
17:58:13 <truebrain> owh, and it only takes 110 lines of code .. lol ..
18:01:01 *** frosch123 has joined #openttd
18:01:01 <frosch123> brickblock19280: if you change/reload newgrf in-game, new objects are added at the end. start a fresh game to see the real order, which should be grf-order + grf-local id
18:01:40 <brickblock19280> I'll try it thx
18:38:25 <DorpsGek> - Update: Translations from eints (by translators)
18:57:16 *** Wormnest has joined #openttd
19:24:40 *** blindvt has joined #openttd
19:25:13 *** blindvt is now known as Guest1952
19:26:38 <Guest1952> hi folks! I'd like to propose a 2-liner patchlet to the game core, but i'm a bit lost where to mail it nowadays?
19:27:31 *** michi_cc[d] has joined #openttd
19:27:39 <Eddi|zuHause> open a pull request on github
19:29:21 <Guest1952> jikes. Oh how i love Microsoft. Resistance is futile i guess :-/ I'll open a pull request then. Thanks!
19:34:52 *** virtualrandomnumber has joined #openttd
19:35:24 *** virtualrandomnumber has quit IRC (Remote host closed the connection)
19:41:13 <andythenorth> I think we agreed some time ago that Microsoft are no longer the Borg?
19:41:17 <andythenorth> and are just another vendor
19:41:26 * andythenorth wonders if the memo got lost :)
19:50:43 <frosch123> they are still the borg, we just try to balance them against the klingons
19:51:16 <Guest1952> frosch123: *grin* well said
19:53:11 <truebrain> frosch123: So are klingons worse than borg?
19:53:28 <truebrain> Kinda depends in what area you define that statement .. hmmm
19:53:45 <frosch123> it does not matter, as long as they eat each other, we have a chance to hide from both
19:54:50 <Guest1952> As long as they continue to act according to their Embrace, Extend, Extinguish mindset, i see no reason to change my mind on them. But maybe that's just /me
19:58:03 <truebrain> 2010 called, they want their motto back
19:59:49 <truebrain> Hmm, tuning wasm engines is tricky
19:59:49 <truebrain> ... lot of moving parts to keep track off. Either running out of virtual memory or physical memory 😄
19:59:59 <truebrain> Guess I prefer the first ..
20:00:15 <truebrain> Odd thing to balance against 😄
20:01:29 <andythenorth> last I checked MS just want to sell a lot of Azure
20:02:13 <frosch123> does the uk get the special EU version of office?
20:02:47 <frosch123> office without teams, and windows with default browser setting
20:04:28 <truebrain> For some reason I was thinking about The Office, and was puzzled
20:05:54 <Guest1952> To be fair, there's one product by M$ that i liked: The original natural keyboard was really nice IMO. I just mention that to emphasise that not _everything_ they produce is complete and utter crap |-)
20:06:29 <truebrain> It is really like we went back to 2010 .. did I enter a time capsule? 😄
20:06:54 <frosch123> why 2010? browser war was 2000
20:07:21 <frosch123> now we are again debating about an OS forcing a browser
20:10:36 <_jgr_> Arguably OpenTTD itself is a time capsule from the mid 1990s 😛
20:10:50 <truebrain> A nightmare that doesn't leave me alone 😛
20:11:31 <andythenorth> can we debate OLE vs Java Applets?
20:11:46 <andythenorth> where do you stand in the RISC vs. CISC wars?
20:12:28 <andythenorth> or (I stole this), will next year Truly Be The Year of the Linux Desktop?
20:12:50 <_glx_> apple constantly jump from RISC to CISC to RISC
20:12:56 <truebrain> Isn't every year the year where the next year will be that year? 😄
20:13:11 <andythenorth> you get tonight's prize
20:13:17 <frosch123> yesterday some senior developer at work asked, whether we already use the log4j version with the fixed string interpolation
20:13:34 <andythenorth> ouch, frosch too soon
20:13:35 <truebrain> Only yesterday? Oof
20:13:39 <andythenorth> that was triggering TBH
20:13:40 <_glx_> that's a very fast reaction time
20:13:48 <andythenorth> I have memories of a lost December
20:14:03 <truebrain> Next he asks you if Citrix was updated
20:14:16 <truebrain> The answer is more often "no" than one would like
20:14:27 <frosch123> usually people complain that the log4j scanner still runs every week and consumes cpu and disk io. so it was kind of funny that someone asked whether it is fixed :p
20:15:08 <frosch123> i was tempted to reply "if you still have access to the corporate network, and you were not blocked, you already have the fixed version"
20:20:57 <Eddi|zuHause> <truebrain> Kinda depends in what area you define that statement .. hmmm <-- were there ever mirror universe borg?
20:21:49 <truebrain> Why is filling in a template so difficult for people .. I wonder how we can improve people not throwing it away ...
20:23:51 <_glx_> use dorpsgek to validate like translator issues ?
20:25:50 <truebrain> I am going to store that reply, nice 🙂
20:28:32 <truebrain> For issues you can force the template .. can you do the same for PRs, I wonder..
20:42:07 <peter1139> LordAro, Hmm, thought I might go from 12x12 to 14x14 or 15x15... got 17x17 :D
20:43:41 <peter1139> LordAro, looks like going to right to the very end of that lane managed to get both squares. And set of a burglar alarm.
20:51:47 <truebrain> Libcurl code also suggests that env variables should already work. Well, on mobile, can't test 🙂
20:55:58 <Guest1952> truebrain, filling out the template took 5 times the time to pull --rebase, see you pulled in curl, throwing away my old use-the-proxy-luke-code from a couple of yeas ago, write the 2 lines for curl and compiling. Sorry for that, i've filled out the template now, please accept my apologies
20:59:59 <_glx_> Guest1952: only https is supported IIRC
21:00:12 <Guest1952> truebrain: unfortunately the env vars do _not_ work out of the box. Maybe that's due to the curl_easy_ API usage. If i manually set the proxy then it works as expected
21:01:14 <andythenorth> Mr Weebl lives about 2 mins from me
21:01:23 <frosch123> i just tried, "https_proxy" works, "http_proxy" does not
21:01:53 <_glx_> expected I guess as all openttd.org are https
21:03:38 <truebrain> Libcurls code confirms that. Some libraries try both https and http for https connections, but libcurl does not
21:03:47 <truebrain> Guess that is not a weird choice
21:03:48 <Guest1952> _glx_: hm. Could be, i traditionally have http_proxy exported and IIRC (been ages) the proxy settings are passed to interested parties by my dhcpd
21:04:06 <truebrain> all_proxy should work too
21:05:02 <truebrain> And no_proxy normally is a list of hosts to ignore. Not a binary yes/no.
21:05:31 <frosch123> yes, both all_proxy and ALL_PROXY work
21:06:09 <truebrain> Libcurl stopped using uppercase http_proxy btw, as webservers were being stupid. For all others it does work. Funny
21:06:39 <truebrain> For those that like to read some more details 🙂
21:06:52 <Guest1952> ok so a proper "fix" would maybe be to mention the env var all_proxy and/or https_proxy in the manpage
21:06:58 <truebrain> frosch123: Tnx for confirming! I always assumed, never tested 🙂
21:07:57 <_glx_> you mean openttd manpage mention http_proxy ?
21:08:48 <Guest1952> or is the openttd manpage a debian thing? Then maybe put it in --help, would that be acceptable?
21:09:08 <Rubidium> yes, there is a manpage and it's reasonably up-to-date as well
21:09:24 <peter1139> Just doesn't get installed when I self build of course.
21:09:41 <peter1139> The manpage does not specify anything proxies though.
21:10:15 <_glx_> yeah it only lists command line args
21:10:40 <Rubidium> not sure how to handle an optional library in the manpage though
21:11:07 <peter1139> Well, could make it non-optional :-)
21:12:13 <Guest1952> If using libcurl, the usual https_proxy and all_proxy environment variables are used for HTTPS connections.
21:12:44 <truebrain> peter1139: Yes plz 😛
21:13:13 <Guest1952> that's all i'd have needed in the manpage it seems ;)
21:13:51 <peter1139> Guest1952, I guess you could rewrite 11264 to be a manpage update? :D
21:15:47 <Rubidium> it's a bit of a gray area, because should it also document that the usual SDL environment variables are used when using the SDL video driver? Same for SDL2, and probably others as well
21:16:48 <peter1139> Does Allegro have environment variables?
21:16:53 <_glx_> and the OTTD_ env vars ?
21:17:40 <Rubidium> at least (lib)fontconfig does as well
21:18:52 <frosch123> i didn't even know there were OTTD_ envs :p
21:19:00 <frosch123> but they look more like debug variables
21:19:11 <_glx_> they are debug variables
21:23:39 <truebrain> And I love them a lot
21:23:52 <truebrain> But pretty sure I am their #1 user 😄
21:24:09 <truebrain> (Helps testing our backends a lot)
21:27:36 <_glx_> or touching `/etc/hosts`
21:32:13 <truebrain> The latter doesn't allow changing ports, what I needed 😛
21:40:25 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
21:48:28 <peter1139> Hmm vec.resize() then std::copy(), or vec.reserve() then vec.insert() ...
21:49:07 <peter1139> resize causes unneeded default initialization I suppose.
21:58:15 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
continue to next day ⏵
