IRC logs for #openttd on OFTC at 2018-04-07
⏴ go to previous day
00:03:57 *** andythenorth has left #openttd
00:15:57 *** Mahjong2 has joined #openttd
00:29:26 *** andythenorth has joined #openttd
00:29:43 <andythenorth> how can I check if a consist has an odd number of vehicles?
00:29:52 <andythenorth> something % something
00:32:45 <Eddi|zuHause> you could use %2 or &1
00:33:39 <Eddi|zuHause> or if you're looking for whether you're the at the middle vehicle, use position == position_from_end
00:48:34 <Thedarkb> What kind of industrial station should I use with scrap metal on FIRS
01:45:42 <Eddi|zuHause> Pikka is the new DanMacK?
01:46:52 <Eddi|zuHause> it would fit from the timings :p
01:59:06 <Pikka> if it's something that needs patching in the API I guess it's a feature request for 1.8.1 then
02:11:07 *** supermop has joined #openttd
02:45:07 <supermop> hows tropical brutalism land?
02:46:03 *** ToBeFree has joined #openttd
02:46:52 <Pikka> they knocked it down to build a casino
02:49:20 *** Supercheese has joined #openttd
02:56:43 <Pikka> less than 20 years old when they pulled it down
02:59:12 *** Thedarkb-X40 has joined #openttd
04:06:56 *** CompuDesktop has joined #openttd
04:27:19 <supermop> whats your ai Pikka ?
04:27:52 <Pikka> trains are getting there
04:28:44 <supermop> is it up for testing?
04:30:02 <Pikka> once the trains and ships are done
04:30:15 <Pikka> the old version with just buses and planes is on bananana
04:30:45 <Pikka> currently wrestling with getting it to put the wagons behind the loco :P
04:30:58 <supermop> push only operation?
04:31:16 <Pikka> behind = in the same train as
04:32:24 *** muffindrake2 has joined #openttd
05:01:53 *** ToffeeYogurtPots_ has quit IRC
05:02:17 *** ToffeeYogurtPots_ has joined #openttd
05:03:47 *** muffindrake2 has joined #openttd
05:03:47 *** CompuDesktop has joined #openttd
05:03:47 *** Supercheese has joined #openttd
05:03:47 *** supermop has joined #openttd
05:03:47 *** sim-al2 has joined #openttd
05:03:47 *** mikegrb has joined #openttd
05:03:47 *** dustinm` has joined #openttd
05:03:47 *** nahkiss has joined #openttd
05:03:47 *** LANJesus has joined #openttd
05:03:47 *** Extrems has joined #openttd
05:03:47 *** HeyCitizen has joined #openttd
05:03:47 *** ccfreak2k has joined #openttd
05:03:47 *** nauticalnexus has joined #openttd
05:03:47 *** Maarten has joined #openttd
05:03:47 *** DarkSSHClone has joined #openttd
05:03:47 *** Tharbakim has joined #openttd
05:03:47 *** greeter has joined #openttd
05:03:47 *** TinoDidriksen has joined #openttd
05:03:47 *** techmagus has joined #openttd
05:03:47 *** Smedles_ has joined #openttd
05:05:32 <Pikka> also, ugggggggggghhhhhhh... just needed a -1 on AIVehicle.GetNumWagons(train) and all is well. AI builds a train.
05:06:09 *** HeyCitizen has joined #openttd
05:20:22 *** Supercheese has joined #openttd
07:49:26 *** andythenorth has joined #openttd
07:57:26 <Supercheese> when they crash they give jelly
08:05:02 <andythenorth> how many cars before they turn into 2 MUs?
08:05:15 <andythenorth> but 4 cars might look better
08:05:34 <Pikka> I don't... if people want two MUs, they can multi-head?
08:05:57 <Pikka> & give them poor enough TE that > 4 cars is a struggle for one unit :P
08:07:42 <andythenorth> oh redmine broke now :P
08:13:21 <Pikka> I'd tend towards 3 being 1
08:14:15 <andythenorth> that's what I'm about to do :)
08:42:41 <TrueBrain> oops ... migration hit the rate limit ... because I was doing too many other stuff ...
08:42:54 <TrueBrain> it is only at #3000 .. this will take 2 hours more :D
09:02:52 <andythenorth> 16 rows per pax vehicle
09:08:11 <andythenorth> I should automate something :P
09:10:35 <TrueBrain> automation is for the weak
09:10:40 <TrueBrain> real men do everything manual!
09:10:51 <andythenorth> what about manually writing the automation? o_O
09:12:33 <TrueBrain> only if it is done bit by bit
09:22:23 *** CompuDesktop is now known as Compu
09:24:30 *** sla_ro|master has joined #openttd
09:31:17 * andythenorth doesn't like writing 0s and 1s :(
09:48:56 <andythenorth> Pikka: so I did 8/8 railcars
09:49:06 <andythenorth> pls give me reasons not to do 4/8 xor 6/8
09:49:25 <andythenorth> 6/8 + 6/8 + 4/8 van = 1 tile train
09:49:37 <andythenorth> and the forums people seem to be obsessed with pacers
09:49:45 <Pikka> but who needs 1 tile trains? are station tiles in short supply?
09:50:19 <andythenorth> I'd draw 64 spriterows
09:50:24 <andythenorth> and never use them in my game
09:50:39 <andythenorth> but they'd look nice, right? :P
09:52:55 <Pikka> mmm maybe. can you draw anything reasonable at 4/8?
09:53:49 <Pikka> 8/8 is chibi enough for me :P
10:02:34 <andythenorth> but it's total pointless novelty
10:03:02 <andythenorth> and I have 180 vehicles to draw already :P
10:03:10 <Pikka> "come back" to it later
10:04:33 <andythenorth> I have spritesheets full of — views for 'come back to it later' :D
10:11:19 <TrueBrain> I remember what my issue was with Jenkins ... they explain how their plugins must be configured, but not how you should use the, ... which is a minefield :(
10:27:47 <peter1138> I never got auto-building to work yet.
10:28:00 <peter1138> Gotta log in and kick off a build manually :p
10:34:20 <TrueBrain> it turns out that what I want is very difficult :P I just want proper docker-in-docker support ..
10:36:38 <andythenorth> can you run a hypervisor in docker? :P
10:37:24 <TrueBrain> no; but you can connect to the docker tha tis running docker
10:37:56 <andythenorth> Turtles All The Way Down
10:38:01 <TrueBrain> but so I have in Configure Tools "Docker" with "Install automatically" .. yet .. it doesnt
10:41:03 <Wolf01> Mmmm I can't keep the windows open, it seem to live in a kennel... time to put on Sabaton to cover the noise
10:42:04 <andythenorth> whenever I've built an app with turtles in, it's sunk a company
10:42:15 <andythenorth> ok, I did it once
10:42:31 <andythenorth> dashboards that contained modules
10:42:38 <andythenorth> and one of the modules was 'dashboard'
10:43:26 <Wolf01> So I'm not the only one with that insane idea
10:43:46 <Wolf01> I wanted to propose it to my boss next week :P
10:44:56 <peter1138> andythenorth, the second awswer in that link is WTF
10:45:10 <peter1138> 'How would your users like it if they entered their payment details, clicked "Purchase"'
10:45:31 <peter1138> Do they think people are using CI to deploy to live services?!
10:46:17 <andythenorth> I once proposed that as a joke at work
10:46:27 <andythenorth> but green-light deploy is a whole thing
10:46:47 <andythenorth> @summon supermop
10:46:47 <DorpsGek> andythenorth: out of chalk
10:47:18 <TrueBrain> that first answer is BULLSHIT :P
10:47:24 <TrueBrain> Docker in Docker uses the Docker API
10:47:28 <TrueBrain> there is no concurrency issue
10:48:07 <TrueBrain> Docker in Docker doesn't mean running a dockerd in a Docker ... (as that doesnt really work)
10:48:18 <TrueBrain> someone is wrong on the interwebz :D
10:49:42 <TrueBrain> owh boy, they really wanted to run dockerd in a docker ..
10:49:47 <TrueBrain> that is just plain stupid and senseless
10:50:33 * andythenorth distracted TB sorry :(
10:52:31 <andythenorth> don't like it when devzone is dead :(
10:52:39 <andythenorth> if my SSD dies, I lose all my work, can't push
10:56:21 <TrueBrain> why is devzone dead?
10:56:33 <TrueBrain> and create a github project :P
10:56:35 <TrueBrain> private if youhave to
10:56:50 <andythenorth> can github do mercurial?
10:57:25 <TrueBrain> stop doing mercurial ffs
10:57:35 <andythenorth> can't, the compile farm requires it
10:57:45 <TrueBrain> things to fix, I hear ..
10:57:49 <andythenorth> I can't even do mercurial properly :P
10:58:30 <andythenorth> I am waiting for you to win
10:58:44 <andythenorth> then we can put grf-building-as-a-service on openttd CF :P
10:58:50 <andythenorth> and github the grfs
10:59:03 *** Progman has joined #openttd
10:59:05 <TrueBrain> create a Docker that can do that
10:59:06 <andythenorth> provide a docker image
10:59:12 <andythenorth> people can use the docker for dev as well
10:59:15 <andythenorth> if they have stupid platforms
10:59:29 <andythenorth> then we can stop building windows binary for nmlc
10:59:54 <TrueBrain> so make a Docker that builds grfs from nmls :)
11:00:35 * andythenorth finds Docker docs
11:01:14 <andythenorth> ok not today, but soon
11:01:17 <LordAro> docker? docker docker. docker.
11:01:23 <andythenorth> I need to learn docker for work I think
11:01:30 <andythenorth> we should maybe be using it
11:01:44 <andythenorth> what's it for? :P
11:02:02 <LordAro> y'know, if you wanted an alternative
11:02:15 <TrueBrain> you know, if you want to do the same but act like you are doing something else :P
11:02:18 * andythenorth is confused what docker is, because I've seen some very odd claims about it on the internet
11:02:35 <andythenorth> oh it's just jails
11:02:39 <andythenorth> we used to use jails
11:02:46 <andythenorth> then we dropped FreeBSD
11:02:54 <TrueBrain> jails only seperated filesystem, and a bit processes
11:03:01 <TrueBrain> docker namespaces nearly everything
11:03:10 <TrueBrain> and if you add cgroups, even resources can be tuned
11:03:12 <andythenorth> it's not immune to stupid people though?
11:03:20 <andythenorth> i.e. javascript developers
11:03:20 <TrueBrain> so a VM without a separate kernel
11:03:33 <andythenorth> how isolated is it?
11:03:40 <TrueBrain> lot of stupid people; not much any technique solves about that :D
11:03:50 <andythenorth> I've seen odd claims that docker containers are 'secure'
11:03:52 <TrueBrain> in principle, anything that runs in a Docker cannot escape to the host
11:03:56 <TrueBrain> but ... that is not completely true
11:04:12 <TrueBrain> Docker in the end runs as 'root' process on the host system
11:04:23 <TrueBrain> if you can influence how your docker is started, you can do anything
11:04:27 <TrueBrain> example, you can volume mount
11:04:28 <peter1138> Much like with kvm :p
11:04:36 <TrueBrain> much like any host/guest system
11:04:48 <TrueBrain> but here in a Dockerfile you define these things
11:04:52 <TrueBrain> so it can go unnoticed
11:04:54 <andythenorth> so Docker beat Vagrant
11:04:56 <peter1138> I remember early Linux containers, before lxc. I shut down a guest, and it shut down the whole machine.
11:04:59 <TrueBrain> Vagrant is dead, yes
11:05:04 <andythenorth> we were supposed to switch to Vagrant for dev
11:05:06 <andythenorth> but we hated the concept
11:05:08 <peter1138> Much like git won over mercurial.
11:05:16 <TrueBrain> dont use Vagrant ... it is only hurt
11:05:26 <TrueBrain> Docker For Windows is basically Vagrant btw
11:05:29 <andythenorth> every 2 weeks we have some painful build issue on local machines in dev
11:05:33 <TrueBrain> as it starts VirtualBox to start Ubuntu to run Docker :D
11:05:47 <andythenorth> so the 'solution' was to switch everything Vagrant, and have painful issues for every line of code
11:05:53 <andythenorth> and have to edit in emacs
11:06:03 <TrueBrain> yes; stupid idea :D (sorry, but it is)
11:06:04 <andythenorth> 'because it would be predictable'
11:06:04 <peter1138> I should probably read up on how to deploy my own docker stuff.
11:06:10 <TrueBrain> if you want reproducable builds, use Docker
11:06:25 <peter1138> Then again, currently I use kvm for full isolation anyway.
11:06:34 <andythenorth> but we're not sure whether we want reproducible containers, or full Cent OS images
11:07:05 <peter1138> But to deploy a generic kvm guest and then deploy a docker inside would be useful for reproducability, indeed.
11:07:21 <andythenorth> so how do I dev inside a docker container?
11:07:24 <TrueBrain> what I like about Docker is how I use it for the CF now .. somewhere else the steps are defined what should happen .. and everyone running that container the same stuff will happen .. so if something breaks .. it either breaks for all, or it is a local issue :D
11:07:27 <andythenorth> I have to switch all my tools to shell?
11:07:31 <TrueBrain> normally, you do not
11:07:32 <peter1138> So much faffing about configuring each kvm guest to be a specific machine.
11:07:33 <andythenorth> emacs and stuff?
11:07:41 <TrueBrain> you develop on your host, how ever you like
11:07:46 <andythenorth> or we just build dockers for deploy?
11:07:47 <TrueBrain> you use Docker to run what-ever you work on
11:07:49 <andythenorth> it's part of pipeline?
11:08:00 <TrueBrain> so, for example, I have used it to start Django with my website
11:08:02 <TrueBrain> have it autoloading
11:08:10 <TrueBrain> so I can edit on myhost system
11:08:16 <TrueBrain> and refresh my page
11:08:25 <andythenorth> sounds plausible
11:08:36 <TrueBrain> on linux, if you do 'ps aux' you see the docker processes
11:08:44 <TrueBrain> no; for most things it is near-host-speed
11:08:51 <TrueBrain> a few things are slow .. for example TCP port forwarding
11:08:55 <TrueBrain> (there is a proxy in between)
11:09:07 <TrueBrain> Docker is just a fancy shell around lxc, which is part of the kernel
11:09:09 <andythenorth> our local machines are faster than the production VMs anyway, going slower might be more realistic :P
11:09:22 <TrueBrain> it is nothing special tbh
11:10:30 <andythenorth> how do docker containers interact with orchestration (Ansible, etc)?
11:10:37 <andythenorth> are they frozen and never changed?
11:10:51 <TrueBrain> a docker image is build via a Dockerfile
11:11:01 <TrueBrain> and a bit depending on what you want, Dockers are not persistant
11:11:10 <TrueBrain> you mostly mount a volume from your host system in a docker
11:11:12 <TrueBrain> (to get files there)
11:11:17 <TrueBrain> or on build put those files in there already
11:11:30 <TrueBrain> there are ways of using Ansible with Docker .. but it is a bit 2 different worlds
11:11:48 <andythenorth> orchestration isn't needed if you have images?
11:12:20 <TrueBrain> depends on your goal :D No generic answer for that :)
11:13:34 <andythenorth> we use Ansible for (1) machine builds package upgrades (2) deploying / upgrading our apps (3) gathering facts our app instances in production
11:13:46 <andythenorth> probably (1) goes away with Docker
11:14:05 <TrueBrain> depending on your goal etc, you can replace your whole CI/CD with Docker
11:14:15 <TrueBrain> kubinetes (I think I spelled that correctly) is an example
11:14:40 <TrueBrain> CoreOS (which stricly seen runs rkt, not docker .. but they are similar in idea) has a lot written about stuff like that
11:15:18 * andythenorth will wait for something to win there :P
11:15:40 <andythenorth> we've adopted tech on the basis of blog posts before :P
11:15:46 <TrueBrain> lol ... I was wondering why my Jenkins didnt do what I expected ... I configured 2 things that do the same .. I liked one .. so I removed it .. and surprised it no longer works :D
11:15:57 <andythenorth> I now prefer to have "nothing" compared to "losing tech"
11:18:49 <TrueBrain> ha, this now works :D LOL! :D
11:18:54 <TrueBrain> okay .. this makes the CF a lot safer
11:19:08 <TrueBrain> every job spins up a new Docker which is the Jenkins agent, which spins up other dockers to do what-eer I tell it to do
11:20:00 <TrueBrain> now the next issue .. it assumes the docker host can write files it can read, Docker Pipeline in Jenkins
11:20:03 <TrueBrain> that is unexpected :P
11:22:48 *** HerzogDeXtEr has joined #openttd
11:25:59 <TrueBrain> anyway, now first some other job to preform ...
11:27:00 <TrueBrain> all issues arrived on the other side
11:27:08 <TrueBrain> guess it is time to put some other stuff in read-only
11:27:20 <TrueBrain> or shall I wait for frosch123 to arrive ...
11:35:33 <TrueBrain> well, after the restart, anyway :D
11:36:22 <TrueBrain> how to disallow svn commit .. hmm
11:36:25 <TrueBrain> we do that via ssh ..
11:38:27 *** andythenorth has left #openttd
11:40:32 <TrueBrain> there we go, svn read-only
11:40:51 <TrueBrain> now final conversion ..
11:42:54 <TrueBrain> backups with no purpose are no backups worth your while
11:43:33 <Wolf01> In every company I worked we had a lot of ability to fuck up things, badly
11:44:04 <TrueBrain> you always have a lot of ways to fuck things up
11:44:13 <TrueBrain> but last I read, making things read-only only makes that window smaller :D
11:46:06 <Wolf01> Ok, 1.5 hours of doing nothing could be enough... now I should do something
11:48:11 *** Wormnest has joined #openttd
11:53:09 <TrueBrain> okay .. think this is all how it should be ..
12:06:03 <peter1138> How do devs commit? :p
12:08:24 <TrueBrain> someone needs to go through the wiki and update all the links and insturctions there :)
12:08:41 <peter1138> Not really. We had git for years, and of course committing to that was a no-no (probably not even possible.)
12:11:10 <TrueBrain> we had git; we now have github
12:11:20 <TrueBrain> google tells you all about how github works, and how you can work with it
12:11:37 <TrueBrain> for any further instructions, frosch123 is your man :)
12:11:48 <peter1138> So we still use svn then.
12:12:08 <TrueBrain> svn is in read-only
12:12:52 <peter1138> This is far more awkward than it needs to be.
12:14:27 <peter1138> I'm guessing the answer to my question "How do devs commit?" is actually "instead of commiting to svn you now commit to github" or something like that. But you just tell me to Google it.
12:18:07 *** DorpsGek sets mode: +o TrueBrain
12:18:48 *** TrueBrain changes topic to "1.8.0 | https://github.com/OpenTTD/OpenTTD | Website: *.openttd.org (translator: translator, server list: servers, wiki: wiki, release info: finger) | Don't ask to ask, just ask | 'Latest' is not a valid version, 'Most recent' neither | English only | Logs: @logs | #openttd.dev if this channel is really spammy"
12:18:55 *** DorpsGek sets mode: -o TrueBrain
12:20:01 <TrueBrain> updated openttd.org frontpage, eints is read-only, flyspray has the correct notice, svn is read-only, http:// to vcs URLs point to github, IRC topics update .. hmm .. am I missing something ...
12:22:40 <TrueBrain> permissions on github should be okay too ..
12:22:46 <TrueBrain> that only leaves CI
12:25:26 <peter1138> Are the -Codechange: -Fix: etc commit messages still enforced?
12:27:10 <TrueBrain> hmm, pre-commit-hook .. hmm
12:32:52 <Pikka> how should I name AI trains?
12:33:29 <Pikka> buses are "Townname Number", planes are "Spirit of Townname", etc... trains are little industrial shunts to factories.
12:34:24 <Pikka> cutesy shunter names? Mercury, Sooty, etc? Hmm...
12:35:23 *** frosch123 has joined #openttd
12:38:19 <TrueBrain> frosch123: it is done
12:43:04 <frosch123> yay,work being done while i am asleep :)
12:47:19 <frosch123> i replaced the bug tracker links in the 1.8.0 and RC1 news items
12:48:25 <frosch123> also triggered a restart of the silly redmine vm
12:48:48 <TrueBrain> I already said to andy, he should make a docker to create grfs from nmls
12:49:42 <TrueBrain> right, lunch time; after that, more fiddling with Jenkins and Docker :)
12:50:01 <frosch123> so, did anyone figure out whether windows git can run python commit hooks, or only perl, or do we not care about windows folks?
12:50:47 <TrueBrain> what we can do, how-ever, is add to the CI that he does pre-commit validation
12:51:15 <frosch123> yes, i already assumed that pre-push commit hook is fine in python
13:01:40 <LordAro> i think the CI is probably the best way to go
13:01:51 <LordAro> assuming you want to continue enforcing commit messages
13:02:15 <frosch123> also indention and whitespace
13:03:03 <LordAro> identation is a bit harder, although checking for tabs instead of spaces seems reasonable
13:03:41 <frosch123> well, essentially the same as the svn hook did before
13:03:49 <frosch123> everything for the push check
13:04:07 <frosch123> if possible also checks for commit so people with usable dev environment get an earlier notice
13:05:16 <frosch123> not sure whether gh issues should get a prefix or not
13:05:20 <LordAro> shouldn't be too difficult to say "copy the contents of this folder into hooks directory"
13:05:26 <LordAro> or however git hooks actually work
13:05:36 <frosch123> "-Fix [gh#123]" vs "-Fix [#123]"
13:05:58 <frosch123> LordAro: it's easy on linux
13:06:01 <LordAro> you'll need to drop the [] if you want autolinking
13:06:21 <frosch123> you can just assume presence of python and stuff
13:09:15 <Eddi|zuHause> <andythenorth> %2 is easy <-- there's no reason why you shouldn't be capable of doing %3 or %4
13:11:10 <TrueBrain> frosch123: GH really like #123
13:11:14 <TrueBrain> it doesnt like GH#123
13:11:43 <TrueBrain> and if you can supply a Docker image which does pre-commit/post-commit validation, I can hook it up in the CI
13:12:15 <frosch123> TrueBrain: looks like i have no read access to the svn hook for c&p
13:12:24 <TrueBrain> haha, I had to disable SSH yes :D
13:14:28 <TrueBrain> frosch123: a tar.gz in your devs home folder
13:15:26 <TrueBrain> was a bit annoying .. you cannot make a subversion read-only if people can ssh to the box :D
13:16:17 <TrueBrain> not while keeping read mode
13:25:37 <Eddi|zuHause> but how does that work better than chmod a-w?
13:26:05 <TrueBrain> subversion is .. euh .... well .. weird; every Nth commit it makes a folder where he doesnt preserve the permissions set
13:26:17 <TrueBrain> so on regular intervals scripts run to fix permissions
13:26:23 <TrueBrain> and hoping I get them all is .. euh ..
13:26:29 <TrueBrain> well .. I rather disable ssh :)
13:27:03 <TrueBrain> the other issue is that subversion wants to write to some folders to read data
13:27:07 <TrueBrain> (lock-files, I always assumed)
13:27:34 <TrueBrain> rather safe than sorry :)
13:28:07 <Eddi|zuHause> which brings us back to how "ro" works then?
13:40:27 <Wolf01> TB: why not put SVN in read only by setting the permissions in the svn config?
13:40:49 <frosch123> it doesn't really matter
13:41:36 *** synchris has joined #openttd
13:44:09 <TrueBrain> Wolf01: we dont use svn protocol; we use svn+ssh
13:44:15 <TrueBrain> it ignores the authz file completely
13:44:21 <TrueBrain> (as it reads * = r :P)
13:44:34 <TrueBrain> and yeah, it really doesnt matter :)
13:44:41 <TrueBrain> OpenTTD runs in many many VMs
13:44:50 <TrueBrain> SVN is the only thing running on the VM I disabled SSH of
13:57:11 *** Thedarkb-X40 has joined #openttd
13:57:44 <TrueBrain> so badly documented, this pipeline stuff of Jenkins :( Snippets of examples .. that is it
14:06:45 *** HerzogDeXtEr has joined #openttd
14:10:51 *** Thedarkb has joined #openttd
14:16:05 *** HerzogDeXtEr has joined #openttd
14:20:40 <LANJesus> so.. poking around the admin interface i noticed the admin Poll command is supposed to send the update type as a uint8, but it sends the update type as a uint16 in the server Protocol packet
14:20:45 <LANJesus> what's up with that guys?
14:21:13 <peter1138> What's an admin interface?
14:24:00 <LANJesus> src/network/network_admin.cpp that stuff
14:24:47 *** andythenorth has joined #openttd
14:25:01 *** andythenorth has joined #openttd
14:25:14 <andythenorth> Pikka: QLD set is 'Caning It' no?
14:25:26 <LANJesus> specifically "AdminUpdateType type = (AdminUpdateType)p->Recv_uint8();"
14:25:52 <andythenorth> Brisvegas Nights?
14:26:06 <TrueBrain> LANJesus: I expect a pull request no later than tomorrow ;)
14:26:11 * andythenorth should go back to drawing mail cars eh :P
14:26:27 <LANJesus> TrueBrain: i can do that if you want, but it'd break things and change the admin protocol
14:26:36 <TrueBrain> or change the documentation? :D
14:26:46 <LANJesus> it'd break all the admin interfaces : P
14:26:55 <LANJesus> looks like that hasn't changed since 2014
14:26:56 <TrueBrain> documentation never broke anything! :P
14:27:10 <LANJesus> the documentation for this is severely lacking
14:27:21 <TrueBrain> it used to be fully documented :)
14:27:30 *** Wormnest_ has joined #openttd
14:27:36 <TrueBrain> but I am a bit lost .. you say it is wrong, but you cannot fix it :P I am missing apiece of the puzzle
14:27:44 <LANJesus> the code might be documented but the admin protocol is uh... lolz
14:27:56 <LANJesus> i can fix it but it'd break a bunch of things.
14:28:10 <LANJesus> probably better to fix when the protocol needs to be changed
14:28:43 <LANJesus> eg, server gives me a tomato, then tells me to stick a tomato in its potato socket later.
14:28:56 <TrueBrain> on the wiki, search for network protocol; that used to describe everything .. but I see it no longer does :)
14:29:00 <andythenorth> where is supermop? :P
14:29:12 <TrueBrain> andythenorth: you really expect us to answer that? :D
14:29:17 <peter1138> LANJesus, fix the documentation, IK think :-)
14:29:17 <LANJesus> i'm writing yet another admin interface ; )
14:29:35 <andythenorth> TrueBrain: maybe
14:29:44 <peter1138> I changed an interface the other day. I hope nothing breaks :p
14:30:05 <andythenorth> any kittens died?
14:30:33 <peter1138> Well the change is not live yet.
14:30:49 <frosch123> LANJesus: the documentation is in docs/admin_network.txt
14:31:00 <peter1138> But basically the interface only worked due to a bug in the framework. Now the framework is fixed, the interface had to be updated cos it was technically invalid.
14:31:02 <frosch123> there are also several implemenations in java, python, ...
14:31:40 <LANJesus> frosch123: didn't like them. i'm making my own with blackjack and hookers
14:31:54 <peter1138> LANJesus, we literally just changed to github, so yeah, not all the links have been updated yet :-)
14:32:19 <LANJesus> i've been working on this for about three weeks, off and on
14:32:34 <LANJesus> also working on my network coding chops
14:32:39 <LANJesus> and async programming. bleh
14:33:42 <LANJesus> example output: Chat Packet: {"ActionTypeCode":"Chat","DestinationTypeCode":"Broadcast","ClientId":1,"Message":"GREETINGS HUMAN","Money":0}
14:34:25 <TrueBrain> frosch123: I am a bit used to C these days
14:34:29 <TrueBrain> the - doesnt add anything for me
14:35:27 <peter1138> It was just to stop "stuff" type commit messages anyway ;p
14:35:42 <peter1138> But that guy left anyway.
14:36:00 <TrueBrain> that is true; it was literally created for 1 person
14:36:15 <LANJesus> are you guys familiar with squashing and rebasing?
14:36:29 <frosch123> yes, but now i want to reject "make cb123 do what my realism bs grf need" :p
14:36:40 <LANJesus> useful for intermediate commits on your local repo
14:36:50 <peter1138> frosch123, yeah it is useful to have a standard.
14:36:55 <LANJesus> when you want to merge back to the main branch
14:37:08 <peter1138> We've been using git for years, just not as the main repo.
14:37:23 <peter1138> Except for the strange people who settled on mercurial.
14:37:43 <TrueBrain> oeh, it seems I have something that looks like a CI working :D
14:37:51 <TrueBrain> just not a lot of control .. as this plugin really does everything
14:38:46 <LordAro> frosch123: i think i like C best as well
14:39:03 <LordAro> alberth & i dropped the dash for frct as well
14:39:12 <frosch123> so "fix" in front of all issues for the closing, hashes just as it
14:39:28 <frosch123> ah, frct, another project to eyeball for examples
14:39:32 <LANJesus> does github support # links to issues directly?
14:40:11 <frosch123> yes, and with certain keyword combinations it also closes them when merging to main branch
14:40:11 <LordAro> TrueBrain: frosch123: can't remember if i told you about openage - vast amounts of (python) CI checkers
14:40:28 <LordAro> thry go as far as checking the copyright in the files
14:40:31 <TrueBrain> put them in a docker and make us happy? :D
14:40:52 <frosch123> LordAro: i think openttd is the only project in the world doing tab indention
14:41:12 <peter1138> Was there an easy way to switch to the new repo is should I just do a new clone?
14:41:52 <andythenorth> Pikka: splitting mail and freight liveries on mail cars is stupid?
14:42:15 <andythenorth> too much drawing
14:42:15 <frosch123> peter1138: i pulled both, added tags to the latest identical revisions
14:42:35 <Pikka> and players will tend to use them for "mail-like" freight, I would think... they'd expect them to match
14:42:36 <frosch123> and then merged them into the deprecated branch
14:42:54 <frosch123> then i can rebase onto the deprecated head and the move over to the new branch
14:43:03 <TrueBrain> use git rebase --onto
14:43:13 <TrueBrain> allows you to give a total of 3 revisions
14:44:33 <andythenorth> also if they have a freight livery, they're just fast boxcars eh
14:44:33 <peter1138> Kinda pointless now.
14:44:44 <TrueBrain> frosch123: any problem if I add Jenkinsfile to the repo? :)
14:44:45 <LANJesus> whomp. i now have a fork of OpenTTD-Deprecated. w/e
14:45:50 <Wolf01> LANJesus: everyone does :P
14:47:31 <frosch123> hmm, how do i point the remote master branch to a different branch though...
14:48:40 <TrueBrain> get the branch local and push it to another place?
14:48:50 <frosch123> nah, i am converting my github fork
14:49:04 <TrueBrain> git remote rename origin deprecated
14:49:08 <frosch123> i guess i should just fork new
14:49:22 <TrueBrain> rename your current
14:49:25 <LANJesus> add new remote, merge it into existing ?
14:49:27 <TrueBrain> make a local git with both as remote
14:50:06 <TrueBrain> you have to fork both
14:50:06 <frosch123> yes, and i move the local master
14:50:14 <LANJesus> go to your fork, add the new repo as an additional remote
14:50:30 <LANJesus> rename your old one to "old" or something, call the new one origin
14:51:05 <LANJesus> meh, you'll figure it out. peace.
14:52:34 <TrueBrain> I would suggest: go to GitHub, rename your fork to OpenTTD-Deprecated, fork OpenTTD/OpenTTD, checkout the new fork of OpenTTD, add a remote to your forked OpenTTD-Deprecated
14:52:41 <TrueBrain> git rebase onto for the branches you want
14:52:47 <TrueBrain> and push to your forked OpenTTD
14:53:14 <TrueBrain> as I dont think GitHub otherwise will see that you changed fork-location
14:55:41 <peter1138> Meh, I don't appear to have a repo with anything else in it, so just doing a fresh clone anyway.
14:55:50 <peter1138> I had some stuff on my SSD which died ;(
14:56:05 <peter1138> I think I lost the patch which finished the game.
15:00:11 <TrueBrain> Bamboo ran out of memory; it didnt fit next to Jenkins :D
15:01:25 *** ZexaronS has joined #openttd
15:03:27 <LordAro> TrueBrain: maybe dorpsgek should add the translators email address
15:05:50 <TrueBrain> they tell me there is a visual pipeline editor in Jenkins
15:05:56 <TrueBrain> just all ways that MIGHT lead to it, I get errors
15:07:05 <TrueBrain> I make a file://, and it tells me: Saving Pipelines is not supported using http/https
15:07:08 <TrueBrain> yes .. I am using file ..
15:08:19 * peter1138 does the most important thing: updated my profile image on github :p
15:12:37 <TrueBrain> ah, found a way to get the editor .. but now I cannot see the resulting Jenkinsfile
15:16:45 *** supermop has joined #openttd
15:30:25 <andythenorth> TrueBrain is even noisier than me
15:30:35 <andythenorth> if he was here a lot, he would beat my channel line count
15:30:38 <Wolf01> How do I rebase via github?
15:31:07 <TrueBrain> Wolf01: check around 14:52 (currently 15:30)
15:31:13 <TrueBrain> still fits on my screen
15:34:28 *** synchris_ has joined #openttd
15:36:23 *** sla_ro|master has joined #openttd
15:41:27 *** Zexaron has joined #openttd
15:50:18 <TrueBrain> just wasted 15 minutes because I shadowed a variable ... and no indication of that .. ugh
15:50:20 <peter1138> Bah, how do I add the windows useful stuff to my project without making the project files change :p
15:52:28 <Wolf01> So I now have ottddepr into my ottd clone, must I run "git rebase --onto mybranch --root ottddepr/mybranch"?
15:53:37 <TrueBrain> best thing about git, you cannot go wrong if you just did a fresh clone
15:53:52 <TrueBrain> I didnt use --root btw
15:55:10 <TrueBrain> git rebase --onto mybranch ottddepr/master ottddepr/mybranch
15:55:23 <TrueBrain> is what I think you should do; but the git manual has pretty pictures
15:55:40 <TrueBrain> git rebase --onto master next topic
15:55:54 <TrueBrain> master is new master, next is old master, topic is HEAD
15:56:11 <TrueBrain> that takes the commits van old master .. HEAD
15:56:14 <TrueBrain> and puts it on new master
15:56:23 <TrueBrain> (you can also just cherry-pick btw)
15:56:39 <Wolf01> Hmm, fatal: needed a single revision
15:58:26 <Wolf01> I never understood this thing, I've done it in the past with SVN
15:58:51 <TrueBrain> if in doubt, checkout master from new OpenTTD repo, add remote to your old repo, and cherry-pick your commits
15:58:56 <TrueBrain> this works if the amount to cherry-pick is low
15:59:03 <TrueBrain> rebase --onto does the same, just bulky
16:00:00 <Wolf01> Oh, I must do it reversed, I might be a special kind of stupid
16:01:16 <frosch123> (the head of master-deprecated is the merge of both old and new head)
16:01:25 <Wolf01> Eh, I need to import my feature branches :P
16:02:22 <TrueBrain> frosch123: what have you done with your repo :P I assume you didnt rebase your feature branches yet? :D
16:02:36 <TrueBrain> behind: 27k commits, ahead: 27k commits :D
16:02:47 <frosch123> it contains both repositories
16:02:59 <frosch123> master is new master
16:03:04 <frosch123> master-deprecated is merge of both
16:03:12 <TrueBrain> how does merging help?
16:03:12 <frosch123> most stale branches are in the old branch
16:03:25 <frosch123> it links the chains
16:03:45 <TrueBrain> I understand what it does; but how does it help?
16:03:54 <frosch123> you can rebase on the old branch as normal, then switchover to the new one without any intermediate changes
16:04:05 <frosch123> TrueBrain: you do not need to lookup the hash in the new branch
16:04:18 <TrueBrain> for branches you dont want to rebase to HEAD?
16:05:00 <frosch123> yes, but i first want to rebase to HEAD of old branch
16:05:11 <frosch123> that should give you most automatic support
16:05:21 <TrueBrain> funny how git does so many shit for us :D
16:06:37 <Wolf01> Meh, if github explodes then it's my fault
16:11:46 <Wolf01> But must I create the branches first on the new repo?
16:12:02 <Wolf01> I can't select the branches I want as destination
16:12:22 <TrueBrain> I have no clue where you got stuck, so that is not really possible to answer :(
16:12:28 <Wolf01> Or maybe I'm still thinking in reverse
16:12:42 <TrueBrain> you made a new fork? you cloned that locally? you added the old fork as a remote?
16:12:52 <TrueBrain> so then something like this:
16:12:58 <TrueBrain> git checkout origin/master -b my-branch
16:13:10 <TrueBrain> git rebase --onto origin/master old/master old/my-branch
16:13:28 <TrueBrain> or what frosch123 says .. as that works too he says :D
16:15:05 <Wolf01> Shit, I have conflicts, I should sync the old branches
16:15:24 <TrueBrain> yes; the assumption is that all branches are up-to-date
16:18:51 <Wolf01> Meh, now I have 22k changes and I can't pull
16:19:08 <Wolf01> Reset doesn't seem to work
16:19:27 <Wolf01> Fresh clone in 3...2...1
16:19:46 <TrueBrain> first clone your deprecated fork
16:19:47 <frosch123> sounds like andy using hg
16:19:49 <TrueBrain> and get that up to speed :)
16:24:20 <Wolf01> I'll finish it tomorrow, maybe, now I must go
16:27:28 <TrueBrain> okay, CI finally triggers how I expect him to
16:34:22 <TrueBrain> 7 minutes and 34 seconds to do CI validation
16:34:36 <TrueBrain> I guess I have to install a shared ccache or something
16:37:20 <LordAro> that's not too bad really
16:37:52 <TrueBrain> frosch123: tomorrow I will first look into eints; but I think I need a git replacement for eintssvn.py
16:39:43 <LordAro> you going to bother doing a windows build at the same time?
16:39:53 <LordAro> or just nightlies, rather than PRs?
16:40:26 <TrueBrain> but .. Windows via Docker is ......
16:41:30 <LordAro> looked at appveyor at all?
16:41:45 <LordAro> they're the usual github windows builder
16:43:06 <TrueBrain> I started with Bamboo ... tried Jenkins now ... Travis doesnt do Windows ..
16:43:14 <TrueBrain> I guess I can skip on to the next fancy word :P
16:44:09 <TrueBrain> either way, I have a Windows Docker ready which kinda worked
16:44:15 <TrueBrain> biggest issue for Windows was our openttd-useful
16:45:15 <TrueBrain> but .. we are "special", so I had to fix up some stuff :P
16:45:42 <TrueBrain> like vspkg calls it lzma.lib, we call it liblzma.lib
16:46:26 <LordAro> sounds like the buildsystem needs tweaking :p
16:46:36 <TrueBrain> I was more thinking OpenTTD needs tweaking
16:46:46 <TrueBrain> maybe we should throw openttd-useful overboard, and use vspkg
16:47:48 <LordAro> that's what i was going for
16:47:49 *** andythenorth has joined #openttd
16:47:56 <TrueBrain> but currently I run Windows 10 Home, as that was sufficient for my needs .. but Windows Docker needs Windows 10 Pro :P
16:48:06 <LordAro> vspkg sounds like a good idea, pkg systems are always better
16:48:11 <TrueBrain> and the OpenTTD server is not the fastest (still runs spinning disks)
16:48:24 <andythenorth> we have a server? :o
16:48:26 <LordAro> or nuget or chocolatey or similar
16:48:47 <TrueBrain> so I was hoping AWS responded, but they seem to rather ignore emails :)
16:48:53 <andythenorth> like actual hardware? :o
16:49:01 <TrueBrain> yes ..... since 2004 ...
16:49:20 <andythenorth> getting rid of physical servers was a happy day for me
16:49:22 <TrueBrain> where else do you think stuff like the content service runs?
16:49:52 <TrueBrain> its fully virtualized, so meh
16:49:58 <TrueBrain> you just notice he doesnt have a lot of IOPS
16:50:04 <TrueBrain> and the machine is getting a bit old
16:50:12 <andythenorth> probably not worse than Rackspace public cloud
16:50:38 <andythenorth> public cloud, the product Rackspace most obviously wish they didn't have
16:52:52 <andythenorth> pikka how many engine liveries? o_O
16:54:35 <Pikka> can I say one? :P Maybe a few over time... "retro", "standard", "refurbished"? or are you thinking by cargo and that sort of business?
16:55:08 <andythenorth> I was hoping for one
16:55:23 <andythenorth> nice and consistent
16:55:33 <andythenorth> I have deleted some mail car liveries
16:55:43 <Pikka> although locos that stick around a long time might want updating, to match more modern locos that come along
16:55:55 <andythenorth> I just replace them with 'upgraded' model on new ID
16:55:58 <andythenorth> with black windows
16:56:01 <andythenorth> I want a new acronym also
16:56:12 <andythenorth> "just because we could doesn't mean we should"
16:57:29 <andythenorth> You Don't Have To Use All The Spec
16:58:07 <Pikka> because it was there (tm)?
17:00:43 <andythenorth> Because Andy Did
17:01:39 * andythenorth must to stop acronyms
17:09:39 <frosch123> was BANDIT created from BAD and ANDY?
17:16:01 <andythenorth> and was reversed acronymed :P
17:16:09 <andythenorth> I had forgotten BANDIT :P
17:16:48 <DorpsGek> andythenorth: Wolf01 was last seen in #openttd 52 minutes and 18 seconds ago: <Wolf01> Bye
17:17:03 <andythenorth> well who's going to redo the livery UI then? :P
17:31:40 *** tokai|noir has joined #openttd
17:31:40 *** ChanServ sets mode: +v tokai|noir
18:16:32 <andythenorth> and another 10 * 16
18:16:40 <andythenorth> not drawing 400 spriterows :P
18:23:04 *** Wacko1976 has joined #openttd
18:23:30 <Pikka> is AI automating playing the game? you can just watch it go.
18:23:45 <andythenorth> can it automate testing newgrf?
18:24:09 <TrueBrain> frosch123: there is not a real way to tell people svn.openttd.org is no longer the official source, except for shutting it down, I think .. possible just move it to svn-archive.openttd.org or something
18:24:27 <TrueBrain> the exposure is too low for anyone to pick up it changed
18:24:48 <TrueBrain> I will see if I can insert an echo in some hook to show a banner of some kind
18:24:51 <Pikka> it can automate testing how well it handles the newgrfs...
18:24:53 <TrueBrain> but even that is not really visible
18:24:54 <andythenorth> Pikka: have you automated crashing OpenTTD? o_O
18:24:56 <TrueBrain> something to consider :)
18:25:00 <Pikka> I haven't tried the trains with FIRS yet...
18:25:02 <andythenorth> I crash it a lot, it's work I could automate
18:25:21 <frosch123> TrueBrain: sounds like low priority :p
18:25:29 <Pikka> only with "default zoom screenshot"
18:25:41 <andythenorth> seems to hate the newgrf being changed under it :P
18:25:47 <frosch123> andythenorth: white an ai that tests whether newgrf do unexpected capacity changes when attaching wagons or refitting :p
18:25:47 <TrueBrain> frosch123: depending on your perspective, but ack :)
18:26:28 <andythenorth> frosch123: I would but I have to learn docker :(
18:26:40 <andythenorth> so I can stop using devzone bundles
18:27:06 <TrueBrain> or tell LordAro to do that for you andythenorth :)
18:27:17 <andythenorth> we need more contributors :D
18:27:26 <andythenorth> they're all playing minecraft or something
18:27:28 <frosch123> TrueBrain: how do you invoke the CI? for every commit, for every push? are there some envvars to figure out hashes before/after?
18:27:46 <TrueBrain> frosch123: atm, I have it on every PR that is being created
18:27:53 <TrueBrain> and on every PR accepted
18:28:26 <TrueBrain> atm a Docker starts with a shallow git checkout of that code
18:28:31 <TrueBrain> what are you looking for exactly here?
18:29:08 <frosch123> currently i have an "update" hook which gets the revision-range from a push and checks all commits for style
18:29:26 <frosch123> not sure how to do the same via the docker-ci
18:29:33 <frosch123> like, how to get the revision range
18:29:43 <TrueBrain> good question honestly
18:29:51 <TrueBrain> it is information that GitHub API sends out
18:29:58 <TrueBrain> but not information the Jenkins plugin returns
18:30:08 <TrueBrain> but if we assume all PRs are rebases
18:30:13 <TrueBrain> and we can just reject any PR that is not
18:31:18 <frosch123> should i take master..HEAD then?
18:31:40 <TrueBrain> I guess we can make a small CI that checks: is master in tree of HEAD
18:31:46 <TrueBrain> and abort the whole CI if it isnt
18:31:55 <TrueBrain> after that, master..HEAD is indeed the correct way
18:32:35 <TrueBrain> I guess that is fair too .. checking a PR that is not mergable is useless
18:33:15 <TrueBrain> so yeah, lets say that origin/master is available, and HEAD is what-ever we want to validate
18:33:32 <TrueBrain> and origin/master..HEAD is at least available
18:33:39 <TrueBrain> (older history might be missing)
18:35:16 <TrueBrain> now time for some dinner :)
18:40:42 <supermop> quite bored now that i gave up on car transporter
18:40:57 <supermop> maybe i should make some trains
18:43:05 <andythenorth> supermop: automate them :P
18:45:49 *** chomwitt has joined #openttd
19:06:17 <LordAro> frosch123: git rev-list origin..HEAD is a list of commits
19:06:51 <frosch123> that was not the question :)
19:17:30 <Pikka> my code in UKRS which disguises the last wagon of an AI train as a brakevan still works
19:17:49 <Pikka> there's bad features for ya
19:18:45 <andythenorth> Pikka Did, You Shouldn't
19:22:37 <LordAro> and have 22245 and 22153 different commits each, respectively.
19:36:46 <peter1138> Whew, that was... a ride.
19:38:29 <supermop> pikka, add code to disguise the brakevan of a human train as a AI wagon
19:44:25 <Eddi|zuHause> what happened to the idea to have different liveries for each player?
19:45:05 <supermop> hmm for some reason my browser is closing when trying to paste a copied url into a slack channel
19:48:36 <supermop> Eddi|zuHause: there is stuff in nml to do that
19:48:59 <supermop> i wonder who asked for it
19:49:05 <andythenorth> what if each player could choose a colour? o_O
19:49:09 <supermop> afaik no one has ever used it
19:49:24 <supermop> andythenorth: what if seinfeld was modarn
19:49:32 <LordAro> i've not got rebase --onto to work successfully yet
19:49:51 <LordAro> it seems to just remove commits and use the new tree
19:50:19 *** Supercheese has joined #openttd
19:50:28 <Eddi|zuHause> so what's the series of commands i should run to convert my svn checkout of trunk with the official dev branch?
20:10:43 * LordAro used a marginally manual `git cherry-pick oldfork/<branchname>~<n>..oldfork/<branchname>` instead
20:11:40 <Eddi|zuHause> i'm fairly sure i asked that before, but what does "yair" actually mean?
20:13:13 <LordAro> i was taking it as some form of "yeah"
20:14:44 <TrueBrain> so many solutions :P
20:15:39 <LordAro> TrueBrain: i looked at the jenkins, shouldn't it be doing a full build, instead of just with the latest changes?
20:16:08 <TrueBrain> I have no clue how to answer that question
20:16:21 <TrueBrain> do you like apples, or instead a sunny day
20:16:52 <LordAro> doesn't appear to have actually built ottd
20:17:01 <TrueBrain> no, PRs are not from scratch atm
20:17:11 <TrueBrain> not sure if it matters currently
20:17:30 <TrueBrain> only the first time a PR is built, is a clean build
20:17:41 <TrueBrain> the next push to the same PR reuses the build state
20:17:49 <TrueBrain> but that is by far the smallest problem I am having
20:17:51 <LordAro> i've gotten used to our work build system, which is really rather bad at incremental builds
20:18:10 <TrueBrain> OpenTTD's CI has been doing incrementals for years now
20:18:33 <LordAro> what are the other problems? :D
20:18:52 <TrueBrain> this grid stuff is pretty neat
20:18:57 <TrueBrain> but it is near impossible to configure
20:19:09 <TrueBrain> so it is a trade of no-configuration vs doing what I want
20:19:10 <LordAro> the blue ocean interface?
20:19:23 <TrueBrain> the project now shows the PRs, how it checked them, etc
20:19:38 <TrueBrain> its a special project type
20:23:23 <TrueBrain> the other issue I currently have that this supports 1 executor per host at best ... because I have to volume mount to the host on a fixed path
20:23:27 <TrueBrain> which is really really bad design
20:23:57 <TrueBrain> guess if I could volume mount to the current docker, I had less issues
20:25:32 *** supermop has joined #openttd
20:31:07 <peter1138> Come on then, who's going to be the first to commit to the new system? :p
20:31:54 <TrueBrain> maybe a bit more realstic? :D
20:32:02 <peter1138> Oh no! Not the R word!
20:32:57 <Eddi|zuHause> someone properly implement pikka's request so AI can choose to build semaphore/light signal?
20:33:50 <frosch123> someone could setup .gitattributes :p
20:33:53 <andythenorth> NRT won't merge anyway
20:34:02 <andythenorth> it's out of sync with trunk in a non-trivial way
20:34:19 <LordAro> remove all the hg/svn specific stuff from the buildsystem
20:34:45 <Eddi|zuHause> people chan pull into a hg repo from a git repo
20:34:58 <peter1138> Yeah but hg is madness
20:34:59 <Eddi|zuHause> so that would be really unwise
20:37:43 <LordAro> you can also make an svn checkout from a github repo, fwiw
20:37:44 <Eddi|zuHause> if it's possible, someone will do it
20:38:23 <Eddi|zuHause> but anyway, why would you remove something that works and needs next to no maintenance?
20:38:38 <TrueBrain> why would you keep it with the chance of it breaking?
20:38:39 <LordAro> because it requires some maintenance?
20:39:08 <TrueBrain> (still shocked OSX code is in OpenTTD :P)
20:39:17 <frosch123> we already have a findversion
20:39:24 <LordAro> besides, i think the version numbering stuff was changing anyway?
20:39:34 <LordAro> for nightlies, at least
20:40:19 <Eddi|zuHause> well, you can't really have incremental build numbers anymore
20:40:33 <TrueBrain> as long as you tag incremental
20:40:36 <TrueBrain> that value is incremental
20:40:53 <TrueBrain> just a shitload of conflicts with custom binaries
20:41:03 <Eddi|zuHause> yes, but is that reproduceable with separate repos?
20:41:17 <Rubidium> and with shallow clones?
20:41:40 <TrueBrain> I never said it was a good idea btw; but just the idea that git doesnt have incremental number is not true
20:41:51 <TrueBrain> the problem is that it is not a globally unique incremental value
20:42:09 <LordAro> it has an incremental number, not the incremental number
20:42:11 <TrueBrain> okay, I can now keep the volumes inside the docker .. so I can run multiple without the chance of breaking shit
20:42:19 <Eddi|zuHause> TrueBrain: yes, i was not questioning the existence of an incremental number, just the practicality
20:42:32 <TrueBrain> its even very practical for nightly builds
20:42:37 <TrueBrain> lot of companies use it for that
20:42:44 <TrueBrain> just you need to apoint a single truth
20:43:03 <LordAro> what does the D in DVCS stand for, anyway?
20:43:29 <TrueBrain> more the problem is, the only reason we have the incremental number is for network/savegame conflict resolving, and a bit of (mostly unused ) GRF stuff
20:43:34 <frosch123> hmm, was that a question? it had no emoji
20:43:56 <Eddi|zuHause> we need a blockchain!
20:43:57 <TrueBrain> OMG! WHO ATE THE EMOJI?!
20:44:09 <peter1138> When are we moving to Slack?
20:44:32 <frosch123> people are asking weird question like what does var42 bit 24 mean all the time...
20:44:32 * andythenorth has been wondering that
20:44:40 <andythenorth> did Slack actually win though?
20:44:44 <frosch123> how would i notice a rhetoric question?
20:44:45 <andythenorth> or did it win like Trello has won?
20:44:50 <TrueBrain> race is still going
20:44:54 <peter1138> I dunno, I've never used slack. Never intend to.
20:44:58 <TrueBrain> Discord is gaining a lot ...
20:44:59 <andythenorth> I haven't tried it, but I'm supposed to
20:45:08 <TrueBrain> HipChat is nearly dead .. stupid atlassian :(
20:45:31 <andythenorth> we are still on irc at work
20:45:34 <TrueBrain> and of course Skype ...... if you want to be sure your conversations are backuped up in some foreign country :P
20:45:43 <andythenorth> and if the updates work
20:45:44 <TrueBrain> I am happy we are no longer on IRC :D
20:45:52 <TrueBrain> I hate missing work conversations because my computer is not on
20:45:56 <frosch123> should we run a custom webex server?
20:46:06 <TrueBrain> YES! CUSTOM! CUSTOM! I WANT CUSTOM STUFF!
20:46:18 <frosch123> maybe andy should twitch-stream grf making
20:46:25 <andythenorth> TrueBrain: missing those conversations sounds ideal tbh
20:46:36 <TrueBrain> not if they are work related .. and not the bullshit channels
20:46:38 <LordAro> that seems like what email is for
20:46:39 <andythenorth> the idea of FOMO about work gives me ugh
20:46:42 <TrueBrain> the ones where people make choices :)
20:46:44 <peter1138> Yeah, at work we use email.
20:46:57 <andythenorth> if it's not on a ticket it didn't happen
20:47:00 <andythenorth> unless you phoned me
20:47:04 <TrueBrain> email ... havent used email for anything useful in months
20:47:13 <andythenorth> email is for password resets
20:47:21 <TrueBrain> yeah, and invite to parties
20:47:35 <frosch123> sounds useful then :p
20:47:37 <peter1138> Ooh when's the party?
20:47:46 <frosch123> tomorrow, your house
20:47:48 <TrueBrain> in my pants? RIGHT NOW
20:47:57 <TrueBrain> sorry frosch123, I took a different turn there :P
20:52:29 <LordAro> has anyone thought about minimum compiler versions yet?
20:53:47 <LordAro> src/language.h:108:8: error: ‘Collator’ does not name a type; did you mean ‘UCollator’?
20:55:02 <frosch123> anything that supports c++14
20:57:32 <LordAro> that appears to be gcc5, clang3.4
20:57:39 <LordAro> and msvc... as new as you can get
20:59:00 <TrueBrain> does OpenTTD support clang? :D
20:59:18 <LordAro> i have to fix warnings occasionally :p
20:59:38 <TrueBrain> hmmm .. guess that should be added to the CI too :)
20:59:59 <TrueBrain> currently it can only do GCC
21:00:05 <TrueBrain> maybe even different dockers .. would make it easier
21:00:09 <TrueBrain> now it iscalled linux-amd64
21:00:17 <TrueBrain> linux-amd64-gcc, linux-clang-gcc?
21:00:30 <TrueBrain> make yourself useful! :P
21:00:53 <frosch123> linux-clang-os9x-minmsvc?
21:01:38 <TrueBrain> sounds good to me LordAro
21:05:47 <andythenorth> all train roofs are ~same
21:06:06 <andythenorth> let's draw them once, not 720 times
21:06:54 <peter1138> When do we get VR support?
21:07:01 <TrueBrain> once you submit that patch of yours
21:07:15 <TrueBrain> owh, wait, it was on the SSD you lost, right? :P
21:13:56 *** Thedarkb has joined #openttd
21:15:23 <TrueBrain> hmm, with this new requirement OpenTTD-CF needs a bit of love :D
21:15:36 <TrueBrain> I can also remove the docker hub autobuild script shit
21:18:11 <LordAro> i'd imagine orudge can :p
21:18:56 <peter1138> Apparently I can, when I am logged in.
21:19:09 <frosch123> any other weird stickies?
21:20:09 <glx> there's also the option to add an URL rewrite
21:20:13 <frosch123> who even notices announcements?
21:21:16 <frosch123> TrueBrain: also the thread title :)
21:22:27 <TrueBrain> that was quick glx :P
21:22:37 <TrueBrain> this time it is for real :D
21:23:21 <TrueBrain> just as a general FYI (A PSA!): in Jenkinsfile you can add which CIs it should run
21:23:34 <TrueBrain> they run in parallel .. so dont go overboard .. maybe multiple stages are in order
21:23:51 <TrueBrain> they have to be PUBLIC docker files .. but soon I will make OpenTTD-CF auto-compile and publish :)
21:27:41 <frosch123> oi, this jenkins indeed looks way different than those i know
21:28:09 <TrueBrain> click Blue Ocean :)
21:29:04 <TrueBrain> any problem if I merge this?
21:29:51 <TrueBrain> I did get the comment message right btw, not?
21:29:59 <frosch123> yes, i checked that :)
21:30:09 <frosch123> only thing i knew about
21:30:13 <TrueBrain> okay, now I pushed it also validates that what I pushed is still valid
21:30:31 <peter1138> You committed it 14 days ago, woo
21:31:10 <TrueBrain> that were a lot of ammends :D
21:31:13 <TrueBrain> ammmmeeenndddddssss
21:31:24 <TrueBrain> funny that it picks that date
21:32:12 <TrueBrain> that is going to happen a lot btw
21:32:22 <TrueBrain> especially because we are not doing merge commits
21:32:35 <TrueBrain> (you also now dont see which PR caused this commit)
21:32:47 <TrueBrain> owh, you do if you click it
21:33:38 <TrueBrain> it is either this or a merge commit for every PR ..
21:35:20 <LordAro> merge commit for every pr isn't the worst thing in the world
21:35:28 <TrueBrain> just unneeded text :)
21:35:28 <LordAro> commits are cheap now :p
21:35:38 <TrueBrain> looking at other projects .. not many do merge commits
21:35:46 <TrueBrain> but a lot also use bots to condense pushes
21:35:49 <peter1138> No, I'm charging 50p per commit.
21:35:59 <frosch123> hmm, where does it show the pr?
21:36:05 <frosch123> also why is that important?
21:36:08 <TrueBrain> says master (#number)
21:36:22 <TrueBrain> always nice to jump back to the conversation leading up to the commit
21:36:25 <peter1138> Vaguely. It's the number, but doesn't actually say that that is a PR.
21:36:45 <frosch123> we could ammed the commit message?
21:37:32 <TrueBrain> you dont know the PR till you made the PR
21:37:43 <frosch123> yeah, a bit chicken/egg
21:38:01 <TrueBrain> but like I said, I only care about that number if I want to know what leads up to that commit
21:38:04 <TrueBrain> and the information is there
21:38:14 <peter1138> Woo, installed Blue Coean.
21:38:35 <peter1138> Separate link, weird :p
21:38:49 <TrueBrain> I am looking at CPython .. they make a PR without the number in the commit message
21:38:52 <TrueBrain> but when they merge, it is there
21:39:47 <frosch123> it is on github, but not in the repository
21:42:17 <TrueBrain> CPython does this .. but how .. hmm
21:42:23 <TrueBrain> nowhere in the PR is the PR number
21:42:25 <TrueBrain> but as soon as they merge
21:42:47 <frosch123> a bot adds it to all
21:43:04 <TrueBrain> yes, but how .. as the author merges
21:43:07 <TrueBrain> so when does the bot do this
21:44:19 <frosch123> it looks like people add "CLA signed" label
21:44:25 <frosch123> and then some bot does the merge
21:45:43 <TrueBrain> yeah, was just looing in the source
21:46:20 <TrueBrain> basically, they work around GitHub :D
21:47:48 <TrueBrain> they switched to GH-
21:47:54 <TrueBrain> their bot even reminds peopple about it :D
21:49:26 <TrueBrain> CLA is the permission thingy
21:51:09 <TrueBrain> they really have a few bots :D
21:52:09 <frosch123> so, should the ci amend all commit messages and merge after succesful CI?
21:54:36 <TrueBrain> still havent found where it happens
21:55:07 <TrueBrain> so many "fun" things in their code :P
21:55:10 <TrueBrain> randomizers and everything
21:57:26 <LordAro> frosch123: seems like a bad idea, imo
21:57:53 <LordAro> you also get commits signed by gpg keys, i'd imagine this would break such things
21:58:12 <TrueBrain> yet other big projects do this
22:00:27 <TrueBrain> and when you squash, it asks for the commit message
22:01:15 <TrueBrain> so that is just a matter of telling all devs to do that :D
22:03:51 <frosch123> the "CLA signed" label is actually from a bot
22:04:05 <TrueBrain> to indicate the person has signed the license agreement
22:04:15 <TrueBrain> all their bots have a theme :P
22:07:18 <TrueBrain> yes, it is part of their squashing
22:07:27 <peter1138> contributor license agreement?
22:07:32 *** muffindrake has joined #openttd
22:07:44 <TrueBrain> it seems that if you squash, by default it adds the (#123)
22:08:15 <frosch123> squashing is weird to me
22:08:18 <TrueBrain> that kinda forces small contributions :)
22:08:20 <frosch123> why would you do that?
22:08:25 <TrueBrain> it is really really useful tbh
22:08:32 <TrueBrain> it is like: I make 2 commits that are one thing
22:08:34 <TrueBrain> now I see a mistake
22:08:36 <TrueBrain> and I make a new commit
22:08:40 <TrueBrain> so you can see what I changed, as reviewer
22:08:46 <TrueBrain> I make a commit to fix it
22:08:54 <TrueBrain> then when we all agree it is perfect
22:09:04 <TrueBrain> amending is a bit evil
22:09:13 <TrueBrain> gerrit is one of the only systems I know that can display diffs of amends
22:09:19 <TrueBrain> (but gerrit is special :P)
22:13:22 *** Supercheese has joined #openttd
22:15:36 <TrueBrain> frosch123: the current method requires force pushes btw, which in general is also not that awsome to use in git :)
22:15:43 <TrueBrain> but you can just look over time what you like best
22:16:19 <TrueBrain> well, if you want the PR number in the commit message, I currently dont have a solution for the rebase tactic
22:16:40 <TrueBrain> as a bot cannot change the commit message in the PR, and I cannot find a hook that runs during the merge
22:19:22 <TrueBrain> well, happy with the result so far :)
22:21:18 <Thedarkb> How do I kick a player from the console?
22:21:29 <TrueBrain> with a bat, I imagine
22:21:38 <frosch123> listclients or something
22:21:46 <frosch123> then kick with the number
22:21:49 <TrueBrain> that doesnt hurt a physical person
22:22:21 <Thedarkb> I typed "rcon <password> kick <clientid>"
22:22:30 <andythenorth> specifically, public multiplayer :P
22:22:41 <andythenorth> that would only seem like a good idea to people who haven't met people
22:23:10 <Thedarkb> I'm playing in a semi public multiplayer
22:23:24 <TrueBrain> that sounds very .. dubious ..
22:23:31 <TrueBrain> only people who are N high?
22:23:36 <andythenorth> my kids keep trying to play public MP games
22:23:40 <Thedarkb> semi public = It's a server dedicated to a Discord I'm in.
22:23:41 <TrueBrain> only people who's IP ends with a 1?
22:24:09 <andythenorth> lots of MP casual games are predator's dream playground
22:24:22 <andythenorth> no filters, no checks
22:24:28 <andythenorth> unrestricted DMs
22:25:08 <TrueBrain> only twitter uses those :P
22:25:18 <andythenorth> and some casual games
22:25:23 <TrueBrain> I like how they are not calling it "private messages"
22:25:27 <Thedarkb> I used to play Quake when I was a kid and nobody predated on me.
22:25:30 <TrueBrain> I mean.. makes you wonder .. why direct, and not private ..
22:25:37 <TrueBrain> are they not .. private? *tinfoilhat*
22:25:51 <andythenorth> they're not private????? :o
22:25:59 <TrueBrain> that is the suggestion I take away from that :D
22:26:08 <andythenorth> you mean the platform might be data minging them?
22:26:11 <TrueBrain> I am sure enough privates go over direct messaging
22:26:13 <TrueBrain> but that is another issue
22:26:39 <andythenorth> anyway I keep banning my kids from games
22:26:43 <andythenorth> but it's a losing battle
22:26:55 <TrueBrain> better teach them to understand what is going on ;)
22:27:01 <andythenorth> that's what I figured
22:27:14 <TrueBrain> so it is time for "THE TALK"
22:27:25 <andythenorth> talk happened multiple times
22:27:35 <andythenorth> but then they go play games where other player nicks are 'fuck off' or 'big cock'
22:27:46 <Thedarkb> So they're playing with other kids then.
22:28:08 <andythenorth> I quit playing world of tanks due to the DMs
22:32:27 <peter1138> Does rcon still need extra quotes?
22:33:28 <TrueBrain> frosch123: #123 or GH-123 ?
22:34:25 <Thedarkb> I get "usage kick:<ip | client-id>"
22:34:54 <frosch123> rcon <pw> "kick id"
22:36:28 <ST2> always look better on buttons xD
22:36:42 <TrueBrain> frosch123: #6649 is a nice example what is very nice to have different commits for review, but squashing would be a lot better once accepted
22:47:39 <TrueBrain> right; time for some well deserved sleep :) Good night!
23:01:22 <supermop> i wonder if i can get some japanese construction worker coveralls sent here
23:11:33 <supermop> apparently i never had a 1.7.x version installed
23:13:06 *** sla_ro|master has joined #openttd
23:14:52 <supermop> hmmm no steeltown servers
23:20:58 *** Supercheese has joined #openttd
23:21:46 *** FLHerne has joined #openttd
23:30:09 *** Thedarkb has joined #openttd
continue to next day ⏵
