IRC logs for #openttd on OFTC at 2023-10-10
⏴ go to previous day
01:07:46 *** Wormnest has quit IRC (Quit: Leaving)
02:23:00 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:13:04 *** pm is now known as Guest2788
03:18:46 *** Guest2670 has quit IRC (Ping timeout: 480 seconds)
05:19:33 *** keikoz has quit IRC (Ping timeout: 480 seconds)
06:08:41 <peter1138> Hmm, in my compile_commands.json, everything that is `-DWITH_...` is listed as `-DW ITH_...`
06:28:10 <peter1138> Even more than just that.
06:28:21 <peter1138> Everything W in the terminal has a space after it. Huh.
07:22:46 *** gelignite has joined #openttd
09:01:16 *** godbed is now known as debdog
09:10:38 *** gelignite has quit IRC (Quit: Stay safe!)
09:36:23 <pickpacket> LordAro: removing the top __init__.py worked fine. Apparently it was unnecessary as you suggested :) I've also commented a lot more about *why* this change is needed and what it means for releasing new versions
09:36:50 <pickpacket> I'll make a PR and we'll take it from there :)
09:46:22 <pickpacket> alfagamma7: already? it's 11:46 here
09:46:34 <pickpacket> alfagamma7: are you in the UK?
09:46:50 <peter1138> It's 10:46 in the UK, so not lunch time.
09:49:29 <pickpacket> peter1138: that's my point. Seemed very early, even though I consider 11:00 to be an okay lunch time
09:50:09 <peter1138> Instead of assuming they are in a country where it is lunch time (given they said it's lunch time) you assume they are in the UK where it's not lunch time? Hmm.
09:50:52 <peter1138> NML_VERSION previously came from git, now it's hard-coded.
09:51:08 <pickpacket> peter1138: no. I assume that everyone in here is in the UK apart from myself :D Because that's the impression I've had
09:51:17 <pickpacket> peter1138: yes. I explain why in the comment
09:51:22 <peter1138> Okay. Most people here are not in the UK :)
09:53:12 <peter1138> "generally bad practice" could do with a citation.
09:55:48 <pickpacket> peter1138: You mean you have a citation claiming it's good practice? ;)
09:56:53 <pickpacket> just imagine if you did an "apt update" but couldn't get info on which versions of which packages are available because that's calculated at install time...
09:57:32 <pickpacket> (yes, I get that it's metadata in PyPI that can be added alongside, but still)
09:58:13 <pickpacket> anyway, I can see if I can find anything that supports my view of it :D Sadly it doesn't really matter because setup.py won't be able to do it in the future anyway
09:58:18 <peter1138> That isn't entirely an equivalent analogy.
09:59:31 <pickpacket> and I really think it's a quality-of-life thing for devs to not have to worry about it. Which is why I'm hoping it's possible to make a github action that can do it instead. But I don't know enough about GH to know if it's possible
10:01:23 <peter1138> That doesn't seem likely. Releases are tags, tags do not have commits (they just refer to a particular commit.)
10:01:54 <pickpacket> And github actions can’t be triggered by tags? 😕
10:03:26 <peter1138> That means only github can build the release.
10:03:31 <peter1138> That doesn't seem ideal.
10:04:26 <peter1138> You could provide a `setup.cfg.in` template, and then use the Makefile to process that and set NML_VERSION from git tags, and output it as `setup.cfg`.
10:09:20 <peter1138> That would probably conflict with how you are envisaging this gets installed. Maybe it just needs to be manual. Bit of a pain.
10:09:48 <peter1138> (I've never installed it, just run it from ./nmlc, which you are also removing.)
10:12:02 <peter1138> With your changes the regression test doesn't run.
10:18:02 <peter1138> (How do you even build?)
10:26:32 <_glx_> Nml version is not only determined in setup, but also at runtime in not installed IIRC
10:41:49 <pickpacket> peter1138: I install with pip and run nmlc
10:42:13 <LordAro> `pip install -e .` iirc ?
10:44:41 <peter1138> pip is a pain when you otherwise use system-provided packages.
10:45:48 <pickpacket> peter1138: yes. But afaik nml isn't available as a system-provided package. It is installable by pip, though
10:46:28 <peter1138> Only with `pip install -e . --break-system-packages` which is not all that nice.
10:46:38 <peter1138> (Or you set up the venv stuff)
10:46:57 <LordAro> setting up the venv stuff is good practice
10:47:28 <LordAro> `python -m venv venv && . ./venv/bin/activate`
10:47:50 <LordAro> can even use --system-site-packages for slightly less duplication
10:48:03 <pickpacket> with a venv you can know exactly what your dependencies are, if you've sort of installed them willy nilly when needed during dev and don't remember :D
10:50:48 <pickpacket> peter1138: I could put the nmlc file back if that makes you feel better :D
10:52:41 <_glx_> Windows standalone needs it
10:52:55 <pickpacket> hm. You mentioned the Makefile, and I'll have to figure out what the install stage should look like there
10:53:19 <peter1138> It's not about me feeling better, it's broken as is.
10:53:39 <peter1138> You may need to pay attention to the workflows too, they refer to setup.py
10:53:55 <pickpacket> yeah, I see that now
10:54:20 <pickpacket> with this change it's quite possible that the workflows need to be changed too
10:54:46 <pickpacket> "make install" and "make extension" don't work without setup.py
11:01:15 <_glx_> Extension is important (speed improvement)
11:04:18 <pickpacket> I could probably use some help with the Makefile things
11:05:54 <pickpacket> and obviously I can't do anything about workflow stuff 🤷
11:07:06 <truebrain> "obviously" .. does that mean you are incapable of learning new things, or how do I need to read that? 😄 🙂
11:08:09 <truebrain> (wow, that sentence sounds a lot harsher than intended .. the smileys are important, let's keep it at that 😛 )
11:12:32 <pickpacket> truebrain: hahaha! No, I just assumed I needed permissions in the repo for it :D Just now found that the workflow files are in the code
11:12:53 <truebrain> it even runs in your fork if you want to 🙂
11:13:25 <truebrain> btw, just to push your work a bit: setup.cfg is also going towards deprecation, and everything is done in pyproject.toml these days; might be worth just skipping setup.cfg
11:13:38 <pickpacket> that said learning new things takes a lot of time for me, calendar wise, because life
11:13:48 <pickpacket> truebrain: crap-a-doodle-do...
11:14:10 <pickpacket> I'll have to re-write all my personal python projects... *sigh*
11:14:30 <truebrain> "deprecation" in Python means: we will still run it, we just rather have that you don't
11:14:33 <truebrain> so you don't have to 🙂
11:14:55 <truebrain> but pyproject.toml makes it a bit easier to do dynamic version detection, like git-versions etc
11:15:09 <pickpacket> when it comes to setup.py it's more like "WARNING: this is deprecated and will be removed in the future"
11:15:27 <truebrain> yeah .. "removed" ... I doubt any tooling actually is going to 😛
11:15:47 <pickpacket> you know what? Let's just skip the setup.cfg thing
11:16:00 <pickpacket> I need to have a look at pyproject.toml anyway
11:17:06 <truebrain> but there are alternatives
11:42:12 *** virtualrandomnumber has joined #openttd
11:42:27 *** virtualrandomnumber has quit IRC ()
11:51:02 <LordAro> truebrain: idk, they've actually managed to remove distutils
11:55:47 <truebrain> how many years did it take? 😄
11:56:07 <truebrain> setup.py is so embedded in so many projects .. I think there will be some mad words when they do remove it, no matter when. Not sure that is a bad thing btw.
11:56:19 <truebrain> just when it comes to Python, I gave up what their "deprecated" actually means 😄
11:57:23 <_glx_> It's like deprecated stuff in windows API
11:58:20 <_glx_> The only "deprecated" that really matters is when it's in Apple documentation
11:58:49 <_glx_> Because they actually remove stuff
13:40:43 <Flygon> The moment Civilization II stops working due to a missing API, is the moment I know the Windows legacy is dead.
13:41:19 <Flygon> (...granted, that game already needs a good 32-bit compat patch because it has some 16-bit subroutines inside a mostly 32-bit app for whatever reason.)
13:41:57 <Flygon> I just wish it had more than 7 player support.
13:42:05 <Flygon> Unfortunately, not a very patchable thing :D
13:42:27 <Flygon> (also I kinda think Alpha Centauri is the best one sorry :D)
13:46:49 <peter1138> Freeciv step 1: Choose SDL, gtk or qt client...
14:33:33 *** HerzogDeXtEr has joined #openttd
15:48:48 *** limyx826 has joined #openttd
15:48:48 <limyx826> Is the baseset icons interface size determined by fixed value, e.g. 1x, 2x, 3x etc. or it can be set as any value, e.g. 1.2x, 1.5x, 2.7x etc. within the options?
15:49:29 <peter1138> Fixed at 1x, 2x or 4x.
15:58:23 <LordAro> `assert errors.__contains__(expected_errors[x][1]) and errors.__contains__("E001")`
15:59:33 *** Limyx826_ has joined #openttd
16:01:47 <LordAro> along with `from __future__ import print_function` in a different file
16:02:12 <LordAro> which is something i've not seen in a long time
16:03:31 *** Limyx826_ has quit IRC (Remote host closed the connection)
16:05:10 *** limyx826_ has joined #openttd
16:29:37 <peter1138> Android 14... dare I?
16:35:51 <Rubidium> I'd suggest installing it next Friday ;)
16:36:22 <peter1138> What happens next Friday?
16:39:52 <Rubidium> if it fails, then you can at least blame it on the thirteenth
16:40:40 <peter1138> Uh, that's this Friday.
16:41:38 <Rubidium> hmm... I guess I used the wrong wording then
16:44:33 <LordAro> this Nday vs next Nday is often ambiguous
17:17:17 *** Flygon has quit IRC (Read error: Connection reset by peer)
17:45:16 <peter1138> Uh, std::vector is empty, but std::find(begin, end, x) segfaults on null-pointer access o_O
18:33:18 *** Wormnest has joined #openttd
20:25:12 *** DorpsGek_vi[1] has joined #openttd
20:33:10 *** limyx826 has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** johnfranklin has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** beauxgus has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** peter1138[d] has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** michi_cc[d] has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** brickblock19280 has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** kamnet has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** belajalilija has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** emperorjake has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** georgevb has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** talltyler has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** _jgr_ has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** _pruple has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** alfagamma7 has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** _glx_ has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** _zephyris has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** andythenorth has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** truebrain has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** sinas128 has quit IRC (Ping timeout: 480 seconds)
20:33:10 *** DorpsGek_vi has quit IRC (Ping timeout: 480 seconds)
20:33:36 <peter1138> Was it something I say?
20:50:00 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:04:05 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:13:35 *** DorpsGek_vi has joined #openttd
21:24:06 *** DorpsGek_vi has quit IRC (Ping timeout: 480 seconds)
21:30:15 *** DorpsGek_vi has joined #openttd
21:32:35 *** truebrain has joined #openttd
21:32:35 <truebrain> well, then some nodes crashed on AWS .. lol
21:35:30 <truebrain> and now it echo's on Discord; nice 🙂
21:46:02 *** ufo-piloot has quit IRC (Remote host closed the connection)
21:46:23 *** ufo-piloot has joined #openttd
21:50:41 <truebrain> right, seems everything is on its feet again .. seems the machine ran out of memory, and crashed, cascading again .. I thought I had that fixed, but clearly I just extended the lifetime 😄
21:52:05 *** DorpsGek_vi[1] has quit IRC (Ping timeout: 480 seconds)
22:21:04 *** tokai|noir has joined #openttd
22:21:04 *** ChanServ sets mode: +v tokai|noir
22:27:43 *** tokai has quit IRC (Ping timeout: 480 seconds)
22:49:35 *** esselfe has quit IRC (Ping timeout: 480 seconds)
23:13:04 *** esselfe has joined #openttd
23:30:28 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
continue to next day ⏵