IRC logs for #openttd on OFTC at 2020-02-04
⏴ go to previous day
00:14:17 *** WormnestAndroid has quit IRC
00:14:32 *** WormnestAndroid has joined #openttd
00:42:26 <Samu> 11 AIs with 5000 road vehicles, 4 more to go
02:17:15 *** snail_UES_ has joined #openttd
03:18:01 *** kelostone has joined #openttd
03:44:33 <DorpsGek_III> [OpenTTD/OpenTTD] James103 commented on issue #5213: Cloning vehicle with impossible orders silently fails to copy orders https://git.io/JvGsy
04:58:46 *** ChanServ sets mode: +v tokai
05:57:47 *** WormnestAndroid has quit IRC
05:59:22 *** WormnestAndroid has joined #openttd
06:35:24 *** WormnestAndroid has quit IRC
06:35:37 *** WormnestAndroid has joined #openttd
06:45:11 *** tokai|noir has joined #openttd
06:45:11 *** ChanServ sets mode: +v tokai|noir
08:07:06 *** andythenorth has joined #openttd
08:15:29 *** el3ktr4 has joined #openttd
08:34:14 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #5213: Cloning vehicle with impossible orders silently fails to copy orders https://git.io/JvGsy
09:36:45 *** el3ktr4 has joined #openttd
09:41:46 *** andythenorth has joined #openttd
09:59:42 *** el3ktr4 has joined #openttd
14:02:40 *** WormnestAndroid has quit IRC
14:02:52 *** WormnestAndroid has joined #openttd
14:14:27 <DorpsGek_III> [OpenTTD/OpenTTD] James103 opened issue #7972: Impossible orders silently fail to be copied and have their sharing settings fully protected. https://git.io/JvGub
14:33:15 *** tycoondemon has joined #openttd
14:56:37 *** snail_UES_ has joined #openttd
15:21:46 *** el3ktr4 has joined #openttd
15:29:09 *** el3ktr4 has joined #openttd
16:11:14 *** WormnestAndroid has quit IRC
16:12:59 *** WormnestAndroid has joined #openttd
16:13:30 *** stefino has joined #openttd
16:13:48 *** WormnestAndroid has quit IRC
16:16:57 <stefino> hi guys, I have a small problem with NML install on WIN. I need to install libraries but it seems I don't know how. I have started CMD, gone to NML folder where I have unziped NML master and then I'm writing pip3 install pillow or ply but nothing happen. "pip3"is not internal or external command
16:17:38 *** WormnestAndroid has joined #openttd
16:18:02 <stefino> ...operable program or batch file
16:19:15 <stefino> setup.py install works but in the end it writes that it needs to install the libraries first
16:20:01 <stefino> (I installed python 3.7.4 on WIN)
16:32:52 <planetmaker> stefino, pip3 is part of the python install, iirc
16:33:11 <planetmaker> maybe it's just called pip instead of pip3
16:33:24 <planetmaker> does calling python work for you?
16:34:26 <planetmaker> though it says for py >3.4 it's included
16:35:28 <LordAro> python -m pip <pip command> should Just Work(tm)
16:38:00 <planetmaker> he :) yeah. maybe better than asusming a pip being a primary command
16:40:00 <stefino> tried it with no result :/
16:40:12 <planetmaker> can you start python at all from cmd?
16:40:22 <planetmaker> "no result" is rather... non-descriptive :|
16:40:37 <planetmaker> python3 -m pip install ply
16:40:37 <planetmaker> 100% |████████████████████████████████| 51kB 875kB/s
16:41:50 <SpComb> openttdcoop banhammer?
16:43:36 *** Guest15453 has joined #openttd
16:46:25 *** Guest15453 is now known as pm
16:47:03 *** pm is now known as Guest15454
16:47:17 <stefino> I think that the problem is in python. In case of setup.py if I write python setup.py install, nothing happen. If I write only setup.py install, it works (but with the libraries erorr)
16:48:45 *** Guest15454 is now known as planetm-
16:57:38 <stefino> planetm-: I think that the problem is in python. In case of setup.py if I write python setup.py install, nothing happen. If I write only setup.py install, it works (but with the libraries erorr)
16:58:25 <planetm-> did you first install the libraries? If they are not installed, no point in trying anything nml
17:14:57 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 opened pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGwk
17:25:37 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh approved pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGwc
17:31:15 *** planetm- is now known as planetmaker
17:35:34 *** WormnestAndroid has quit IRC
17:39:41 *** WormnestAndroid has joined #openttd
17:41:17 *** WormnestAndroid has quit IRC
17:41:30 *** WormnestAndroid has joined #openttd
17:41:43 *** Terkhen has joined #openttd
17:41:43 *** ChanServ sets mode: +o Terkhen
17:42:13 *** V453000 has joined #openttd
17:42:13 *** XeryusTC has joined #openttd
17:42:43 *** planetma- has joined #openttd
17:43:43 *** fonsinchen has joined #openttd
17:43:53 *** Hazzard has joined #openttd
17:44:13 *** Hirundo has joined #openttd
17:44:20 *** planetmaker is now known as Guest15459
17:44:20 *** planetma- is now known as planetmaker
17:45:59 *** ^Spike^ has joined #openttd
18:11:28 *** y2kboy23 has joined #openttd
18:14:58 *** Progman has joined #openttd
18:27:34 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 dismissed a review for pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGwc
18:27:34 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGwk
18:28:24 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 commented on pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGrK
18:37:32 *** gelignite has joined #openttd
18:37:53 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh approved pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGrS
18:41:42 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 merged pull request #7973: Fix #7969: limit recursion during alias execution https://git.io/JvGwk
18:41:43 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 closed issue #7969: Console command "alias" can stack overflow and crash (bypassing crash handler) https://git.io/Jv3xF
19:02:03 *** Maarten has joined #openttd
19:29:19 * andythenorth learning to draw trains
19:35:55 *** frosch123 has joined #openttd
20:21:51 *** sla_ro|master has joined #openttd
21:10:00 <TrueBrain> frosch123: the bot was now missing permissions
21:10:07 <TrueBrain> I created a Team to compensate
21:12:04 <TrueBrain> GitHub Actions triggers failed for sure
21:12:13 <TrueBrain> not sure if eints failed; I never know if there was anything to do
21:12:23 <TrueBrain> but for sure DorpsGek had no write permission to anything, so it shouldn't have worked :D
21:12:58 <Samu> more than 10 seconds to advance a frame
21:13:42 <frosch123> ok, for eints we have two options
21:13:49 <frosch123> 1) set the group as "maintainer"
21:14:15 <frosch123> what do you prefer?
21:14:25 <TrueBrain> sorry, what problem are you trying to solve?
21:14:51 <frosch123> the repository is currently configured so that noone can push
21:14:55 <frosch123> but eints needs to push
21:15:08 <TrueBrain> so the first problem we had was that it couldn't even write to the repo
21:15:12 <TrueBrain> next indeed would be branch protection :D
21:15:25 <TrueBrain> I am in favour of branch protection
21:15:30 <TrueBrain> minimize the amount of rights the bot has
21:15:44 <TrueBrain> so setting him with maintainer access so it can only push, feels icky
21:16:44 <frosch123> ok, i added the team as exception
21:16:51 <frosch123> still not sure whether it will work
21:17:11 <frosch123> push may still get rejected because commit was not approved by anyone
21:17:17 <TrueBrain> we will find out tomorrow :P
21:18:12 <glx> only admin can push without approve if approve is required
21:18:49 <frosch123> it only mentions "status checks"
21:19:07 <frosch123> but the wording is all fishy
21:19:10 <TrueBrain> pretty sure that is not true glx; have you tested that?
21:19:14 <TrueBrain> well, I can test that right now honestly :)
21:20:05 <glx> for PRs I'm quite sure all checks must pass for a non admin to merge it
21:20:21 <TrueBrain> we are not talking about PRs :)
21:22:58 <TrueBrain> okay, as admin it is irrelevant what the checks are .. funny
21:23:14 <frosch123> there is a "include admins" setting
21:23:34 <frosch123> i checked it for all repos, but openttd, just for dorpsgek
21:23:45 <TrueBrain> remote: error: GH006: Protected branch update failed for refs/heads/master.
21:23:52 <TrueBrain> with that on, under my user
21:24:17 <TrueBrain> and I cannot overrule it now
21:24:40 <frosch123> not even with the "restrict who can push"?
21:24:53 <TrueBrain> but let me try it with DorpsGek token, that is a bit more fair
21:24:55 <frosch123> ok, so dorpsgek must be openttd repository admin
21:25:00 <frosch123> not organisation admin
21:25:06 <TrueBrain> that is too early to tell :)
21:25:18 <TrueBrain> let me do testing before we jump to conclusion based on some partials :)
21:25:47 <TrueBrain> that is, if I can remember how to push as someone else
21:28:23 <TrueBrain> nope, it still requires approval :)
21:28:37 <TrueBrain> so you are right glx. Just not sure you knew already, or that it was a guess :P
21:29:09 <glx> well it was partially a guess :)
21:29:18 <TrueBrain> frosch123: only as "admin" role I can overrule that
21:30:02 <glx> but it makes sense that a protected branch is really untouchable except by admins
21:30:04 <frosch123> yeah, but repo admin should be enough, no org admin needed
21:30:57 <TrueBrain> glx: it would be nice that if for bots they would allow exceptions tbh
21:31:01 <TrueBrain> frosch123: indeed, it is
21:31:58 <TrueBrain> but "Restrict who can push to matching branches" makes very little sense in this case tbh .. well, only if you ignore the approval part, I guess
21:32:02 <TrueBrain> they are a bit mutual exclusive of some sorts
21:33:23 <glx> I don't see anyone sane wanting to even think about using the two last rules
21:33:34 <TrueBrain> okay, the proper solution seems to be to make DorpsGek push to a branch, make a PR, and have an auto-merge-if-green action enabled
21:35:14 <frosch123> glx: "force pushes" are enabled for one eints branch :)
21:35:51 <glx> special case I guess, with limited access
21:35:57 <frosch123> the "eints with openttd interface" branch is always rebased onto the "pure eints" :)
21:36:27 <frosch123> TrueBrain: so, bots become repo admin for openttd?
21:36:37 <frosch123> done with testing alternatives?
21:36:54 <TrueBrain> yeah ... can't find another way
21:37:15 <TrueBrain> installing an auto-merge bot might be a bit too much hassle for now
21:37:29 <frosch123> the PR spam would be a bigger hassle
21:37:32 <glx> and DorpsGek is not a random bot from unknown source
21:37:57 <TrueBrain> glx: it is not about if DorpsGek should be allowed or not, it is about least privileges. We now have a bot with admin access. That is .. less optimal
21:38:19 <TrueBrain> always assume PATs get lost :)
21:38:29 <glx> but the PR option seems worse
21:38:44 <TrueBrain> it is what the rest of GitHub seems to be doing for these kind of jobs
21:38:55 <TrueBrain> so I would tend to say we are missing something :)
21:39:43 <TrueBrain> and there is a case to be made .. what if a language push made the CI fail. I guess that is a bit the idea behind forcing that flow
21:40:08 <glx> well CI can't fail if a language fails IIRC
21:40:29 <glx> it will warn but won't fail
21:41:03 <TrueBrain> nevertheless, that is I guess the idea of such workflow; force everything through the CI, don't allow things to sidestep it
21:41:09 <TrueBrain> unless you are really really high up in your permissions
21:41:20 <TrueBrain> that we completely abuse it ... well ..
21:42:08 <andythenorth> oh I looked away, now there is loads of chat :P
21:44:29 <TrueBrain> hmm, reading some more, "Restrict who can push to matching branches" might not do what we think it does .. I wonder if any non-admin can now merge PRs :D
21:45:04 <frosch123> someone will complain :)
21:45:58 <frosch123> but there is also the "only owner can approve" or something
21:46:21 <frosch123> hmm, approve is not merge
21:46:44 <frosch123> why didn't gh just do a spreadsheet with checkboxes for each team
21:46:59 <TrueBrain> hahaha, well, that is one solution :D
21:47:02 <frosch123> instead of prosa rules which add/revoke each other
21:48:46 <TrueBrain> "Restrict who can push to this branch is intended to exclude users or teams from pushing to important branches using any method, including merging their own pull requests onto the target branch. As we already saw in this article, merging a pull request is effectively the same as pushing commits to a branch. By enabling this setting you can prevent
21:48:46 <TrueBrain> a user from pushing commits to a branch while still allowing them to open pull requests."
21:48:54 <TrueBrain> okay, so that option is not doing what I expect it would do :D
21:49:08 <TrueBrain> but it literally doesn't allow anyone outside of that list to merge pull-requests :D
21:49:08 *** WormnestAndroid has quit IRC
21:49:12 <frosch123> interesting project... if gh does not have a public issue tracker, let's create one inside gh
21:50:03 *** WormnestAndroid has joined #openttd
21:50:41 <TrueBrain> so I think we want to disable that entry, as we already do this with Teams .. I *think* that should be enough :P
21:51:17 <michi_cc> nielsm: If you've not stuffed your Mac back into the closet, could you try reverting src/video/cocoa/event.mm to the version before 74b7f0a9aa7658857714db365853db8f6c8512bd?
21:51:40 <TrueBrain> frosch123: or possibly give Core Developers the "maintain" role
21:52:14 <TrueBrain> not sure .. very confusing entry, this "restrict push" button :D
21:52:26 <TrueBrain> but as you mentioned, someone will complain :)
21:52:35 <frosch123> TrueBrain: disable the setting sound better than "maintainer"
21:53:38 <TrueBrain> anyway, I am off for tonight .. this GitHub permissions is messing with my head :P
21:54:23 <nielsm> michi_cc: that one's before 1.8 isn't it?
21:54:32 <nielsm> I don't think anyone reported the issues for 1.8
21:55:00 <michi_cc> The commit was merged on 10.4.2018, i.e. after 1.8 release.
21:56:11 <michi_cc> It looks totally inconsequential, but it uses a different call for mouse cursor warping which might affect more in the event handling.
21:57:37 <nielsm> but I should try to do a proper bisect tomorrow
21:58:25 *** el3ktr4 has joined #openttd
21:59:51 <michi_cc> The amount of other commits affecting src\video\cocoa since 1.8 is almost zero if you subtract the color space changes, which were done in response to the slowdown problem, so clearly not the cause.
22:01:37 <michi_cc> The one other difference is the compiler environment, or more exactly the SDK and target versions used in the build. There are things OSX will do differently depending on the version the app was compiled for, which makes it a definite possibility as well.
22:02:22 <frosch123> andythenorth: all CC trains look the same. other than that, they are different
22:02:39 <andythenorth> it's very hard to draw steam trains :P
22:02:55 <andythenorth> they all have same CC kettle, tender, black chassis
22:02:57 <nielsm> oh right, 1.8 doesn't have fps window
22:03:56 <michi_cc> 74b7f0a9aa7658857714db365853db8f6c8512bd? is a one line add/remove change for event.mm, forward-porting should be hard :)
22:05:37 <DorpsGek_III> [OpenTTD/OpenTTD] LeandroSImmich opened issue #7974: Crash when CTRL+click to show a vehicle group that is collapsed https://git.io/JvGPA
22:05:50 <nielsm> but I'm not feeling any performance difference between 1.8.0, 1.9.3, and 1.10.0-beta2
22:06:10 <nielsm> so maybe I'm not fully equipped to test it anyway
22:08:35 <michi_cc> Same problem as me. It's not hard to run OSX virtualised on Windows, but there are exactly zero accelerated video drivers available, which makes any kind of performance testing pointless.
22:10:15 <nielsm> I'm not sure if we ever got a full confirmation whether the issue only appeared on os 10.14+ or not
22:10:27 <nielsm> but my machine can't install later than 10.13
22:10:48 <michi_cc> It's just two lines, manual editing should be possible :)
22:12:13 <milek7> i tried running it on kvm but performance was fine
22:13:49 <andythenorth> not exactly sure what we're trying to repro :)
22:13:53 <andythenorth> I've made the revert
22:14:02 <nielsm> still the weird slowness
22:14:07 <andythenorth> but there's no baseline peformances number, just weirdness
22:14:14 <nielsm> bad fps despite the gameloop and drawing times being good
22:16:47 <andythenorth> I've got nothing conclusive
22:16:57 <andythenorth> some of the issues are intermittent
22:16:58 <michi_cc> I don't really expect that change to do anything, but it is the only post-1.8 commit to even touch event handling.
22:18:08 <glx> hmm between 1.8 and 1.9 we moved the CF too IIRC
22:18:34 <nielsm> nope I'm not seeing any improvement with that change reverted
22:18:42 <LordAro> seems like it's almost certainly SDK related
22:19:44 <glx> I think most people in the issues use our provided builds
22:20:35 <nielsm> andythenorth: when you use the official build of 1.8 you don't have issues, right? what if you checkout 1.8 and build that yourself?
22:20:52 <nielsm> or can we try making the new cf build 1.8?
22:22:58 <nielsm> anyway, I'm off to bed, gn
22:23:53 <glx> and OSX version seems to matter too
22:25:22 <andythenorth> I'm not sure if 1.8.0 is information or not
22:25:30 <andythenorth> the reports aren't all using same save
22:25:36 <andythenorth> there's no fps window
22:25:40 <andythenorth> it's very subjective
22:26:32 <andythenorth> ffwd on an empty map is far faster on 1.8.0, but I'd have to time months to get a numeric value :P
22:27:03 <andythenorth> ffwd on master was briefly really fast on mac after peter decouple gui timing,
22:27:14 <andythenorth> then it got broken around the time kdtrees were added for stations
22:27:21 <andythenorth> but it wasn't necessarily causal
22:58:11 *** andythenorth has left #openttd
23:08:02 <LordAro> why does GH think that grfcodec was updated 2 hours ago?
23:08:57 *** supermop_work has joined #openttd
23:10:21 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro approved pull request #7859: Feature #7756: Allow server to supply a reason to kicked/banned clients https://git.io/JvG1l
23:10:44 <LordAro> aaand apparently i'm no longer authorised to merge PRs
23:15:15 *** frosch123 has joined #openttd
23:15:30 <frosch123> we have a test subject?
23:17:11 <frosch123> LordAro: try again?
23:17:34 <LordAro> frosch123: seems happier :)
23:17:43 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro closed issue #7756: Allow server to supply a reason to kicked / banned clients https://git.io/Jen01
23:17:43 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro merged pull request #7859: Feature #7756: Allow server to supply a reason to kicked/banned clients https://git.io/JeH2l
23:18:06 <frosch123> LordAro: i finally fixed grfcodec earlier, that was on the todo list for N months :)
23:19:36 <frosch123> no idea how it was originally imported, but it had random hashes for authors, and someone made a merge commit. so i reimported the repository
23:22:01 <glx> hmm peter has a cmake branch for grfcodec and I have an open PR against this branch
23:22:36 <frosch123> really, someone working on grfcodec? :o
23:23:05 <frosch123> i guess you need that rebase --onto thingie then
23:23:53 *** WormnestAndroid has quit IRC
23:24:34 <LordAro> should probably have all the CI gubbins set up with it as well
23:25:04 *** WormnestAndroid has joined #openttd
23:32:16 <frosch123> it is the least important tool, you can use a 7 year old release jsut fine
23:33:11 <glx> the cmake branch mostly ease compilation
23:42:56 <DorpsGek_III> [OpenTTD/OpenTTD] stormcone opened pull request #7975: Fix #7974: Crash when CTRL+click to show a vehicle group that is coll… https://git.io/JvGMs
23:50:14 <TrueBrain> LordAro: this time it was not me!! :p
23:50:50 <TrueBrain> I am so happy someone else broke it for once :)
23:51:27 <TrueBrain> (Really I am .. means someone else now also knows how it works :D)
continue to next day ⏵
