IRC logs for #openttd on OFTC at 2024-03-25
⏴ go to previous day
00:07:11 *** Wormnest_ has quit IRC (Ping timeout: 480 seconds)
00:43:20 *** Wormnest_ has joined #openttd
00:45:20 <reldred> Multi-cargo is quite nice
00:45:40 <reldred> Have been enjoying my time with multi-shark
01:37:31 *** Wormnest_ has quit IRC (Ping timeout: 480 seconds)
01:47:47 *** Wormnest_ has joined #openttd
01:48:18 *** tokai|noir has joined #openttd
01:48:18 *** ChanServ sets mode: +v tokai|noir
01:53:15 *** Wormnest_ has quit IRC (Quit: Leaving)
01:55:12 *** tokai has quit IRC (Ping timeout: 480 seconds)
03:25:15 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:27:08 <kamnet> reldred: haven't used it yet but I put it in my newgrf list for my birthday server
04:40:42 <DorpsGek> - Update: Translations from eints (by translators)
04:46:07 <rau117> JGRennisonviaGitHub: Btw how about the same, but for the road vehicles?
04:54:12 <emperorjake> There would still need to be the separation between passengers and other cargos since they use different station types
04:54:22 <emperorjake> but it would be nice for cargo trams and such
05:44:17 <wensimehrp> rau117: Bus with mail cabin?
05:46:40 <rau117> Well... better to say international bus with a luggage compartment
05:46:40 <rau117> And so, I asked it for all sorts of industrial trams with several trailers
05:47:40 <rau117> if the code works for ships, I think it would be (relatively) easy to add it for road.veh
05:57:41 <emperorjake> wensimehrp: Like I said, it wouldn't work for that, something would have to be done to the stations
07:11:48 <DorpsGek> - Add: summary for week 12 of 2024 (by OpenTTD Survey)
07:44:40 <kuhnovic> emperorjake: Personally i've always found it weird that RVs have separate station types for passengers and cargo. None of the other transport types have this distinction, and I've often found myself building the wrong type. Maybe it's just me 😊
07:52:25 <andythenorth> kuhnovic: I frequently build the wrong type
07:55:04 <emperorjake> Same, but not not much now with the new version of Fridaemon's roadstops that has different styles for bus and truck stops
08:03:55 <jfs> well idea to fix that then: make train station tile types that are only capable of loading/unloading specific cargo types/classes
08:04:20 <jfs> so if you don't build a coal station you can't load those hoppers
08:08:21 <reldred> that... could actually be kinda fun
08:08:54 <andythenorth> jfs: I am somewhat in favour of this idea
08:09:48 <jfs> the real challenge is that you may want to need station types to correspond to vehicle types, rather than cargo types
08:10:14 <andythenorth> would we need to teach the pathfinder 😛
08:10:31 <andythenorth> if a station is composed of different tiles with different cargos...
08:12:32 <jfs> now imagine if you could have vehicles that are capable of charging a fee if they get loaded with specific cargo types.,.. like sending a container flatcar to load coal: the coal has to be loaded into a hopper conforming to the shipping container standard, and then loaded.
08:12:32 <jfs> oh wait, that's refit costs.
08:13:50 <andythenorth> imagine the linkgraph for a cargo is between stations with specific handling for that cargo
08:14:06 <jfs> actually, not quite... since maybe you'd want to charge the fee every time the car gets loaded, and not just when it changes cargo type
08:14:49 <reldred> Could you all quit talking dirty
08:15:16 <jfs> or what about cargo unload fees... or what if they could be negative
08:15:33 <jfs> (negative cargo unload cost: passengers paying for a first class carriage)
08:16:00 <ahyangyi> I remember negative numbers and inflation were a fun combo
08:16:08 <reldred> Imagine with something like Steeltown or AXIS, having to keep expanding stations out to handle all the various complex cargo’s and their loading/unloading requirements
08:16:45 <reldred> And then PR8480 so Andy can’t cheat with overflow depots 😛
08:17:31 <jfs> build a container loading facility for the tracks, and then some facilities that can re-pack the cargo into container form
08:18:05 <andythenorth> I am just going to try to divert this into a resurection of YACD 😛
08:18:05 <jfs> have trains that carry empty containers back
08:18:25 <jfs> Yet Another Container Disaster?
08:18:51 <jfs> or is the D for something else?
08:21:50 <andythenorth> we do an intermediate demand creation approach
08:22:00 <andythenorth> YACD tried to connect sources / sinks
08:22:07 <andythenorth> cdist only uses the orders linkgraph
08:22:33 <jfs> ah okay cargo destinations, yeah
08:22:44 <andythenorth> YAYACD uses stations with specific cargo handling to create the linkgraph
08:22:51 <andythenorth> then the demand is exposed to grf and/or GS
08:23:14 <andythenorth> and is otherwise flat, by distance (on a curve) or random-weighted, per player preference
08:23:41 <andythenorth> the player has somewhat control by specifying the station cargo handling
08:24:08 <andythenorth> but there is the gameplay / goal element of the actual demand is determined by the game
08:24:44 <andythenorth> there is no way that reddit or steam are going to understand 'add a tile for this cargo' 😛
08:24:51 <andythenorth> way too complicated
08:25:46 <reldred> all the more reason to do it
08:27:49 <andythenorth> track / dock / road tiles are neutral, but non-track 'facility' tiles have to be added, one per cargo handled
08:28:01 <andythenorth> makes building RV stops in cities suck I guess
08:30:15 <jfs> can we make "post office" an industry instead of every home having their own?
08:30:29 <andythenorth> we could do that in grf
08:30:39 <andythenorth> houses stop generating mail
08:30:46 <andythenorth> post office industry counts population
08:31:04 <jfs> the main thing would be making cities always have a post office
08:31:11 <andythenorth> or....overly complicated edition....buildings write mail demand/generation to town register
08:31:21 <andythenorth> GS can place a post office in every town
08:31:31 <andythenorth> there's nothing in grf that could force it
08:31:46 <andythenorth> grf town industries work badly currently, without GS
08:32:06 <andythenorth> as the industry weights aren't set up to assume 1 or 2 industries per town
08:32:50 <jfs> can we make the Bank industry be a thing that levels up with the size of the city? and then have more dense commercial buildings be around the bank
08:33:08 <andythenorth> grf could do that
08:33:23 <jfs> levels up = changes visual design so it becomes more of a tower the bigger the city
08:33:40 <andythenorth> if an industry could influence zone on nearby tiles
08:33:53 <andythenorth> then (1) banks could produce high rise (2) steel mills could avoid it
08:34:30 <jfs> house property: "distance to industry that (produces/accepts) cargo X"
08:34:43 <reldred> We really need some better zoning controls
08:34:55 <reldred> rather than just radii/number of houses from centre
08:36:34 <jfs> tbh, that's another thing Transport Fever 2 actually does quite well, having zones within cities, and their shape and location are partly affected by the transport services offered
08:37:00 <andythenorth> TB had the strong concept of attaching GS individually to towns
08:37:13 <andythenorth> can't remember why it was appealing, but there was something neat about it
08:37:31 <jfs> towns having personalities could maybe be fun
08:37:32 <ahyangyi> Yeah, the very round zones are a bit comical when cities get large
08:38:02 <ahyangyi> Like the city planners have a strange sense of humor or something
08:38:03 <jfs> I guess that kind of GS could be called a mayor script
08:39:26 <ahyangyi> TBH we already have "personalities", just in the shallowest possible way: each town builds roads in one of the four different patterns
08:39:51 <ahyangyi> so anything that makes towns more different isn't a new direction we take, but enhancement of what we already have
08:40:32 <jfs> having scripts for towns could also allow different towns to offer different actions for companies to take, and respond differently to them
08:42:36 <jfs> town action: "eminent domain", costs a fortune, but lets you freely demolish houses and roads for a limited time
08:43:20 <jfs> (but if your company then doesn't start offering a good transport service within the town after a while, you're getting punished hard in some way)
08:43:57 <ahyangyi> What if you demolished all their houses, do you automatically get a good rating because there's no passenger to move any more
08:46:53 <ahyangyi> Anyways, what can a town do to a company?
08:47:08 <ahyangyi> I thought the worst they could do was to refuse dealing with them?
08:56:05 <jfs> eminent domain the company property back
08:56:58 <jfs> offer exclusive subsidies to competitors
09:05:59 <ahyangyi> I don't think eminent domain would be effective if the town could only grab what's within its borders, but exclusive subsidies sound fun
09:08:01 <ahyangyi> Otherwise I guess people will try to "feed" towns with very distant properties across the map for fun 😛
09:11:45 *** gelignite has joined #openttd
09:48:24 <merni> andythenorth: It'd be an option anyway
09:48:45 *** jeevadotnet has joined #openttd
09:48:45 <jeevadotnet> Hello, been playing this game for 30 years & 20 for OpenTTD. I want to build a trainset, is there perhaps a guide?
09:49:10 <jeevadotnet> South African Trains should be represented
09:49:21 *** gelignite has quit IRC (Quit: Stay safe!)
09:58:33 <ahyangyi> jeevadotnet: some of the popular newgrfs are also open source, using them as a starting point might be easier than from scratch
10:20:41 *** jinks has quit IRC (Remote host closed the connection)
10:41:14 <rau117> *Yes, yes... this is not the best place to suggestions…*
10:41:14 <rau117> How about finally disabling breakdowns by default? If there's one thing that can improve the lives of newbies, it's definitely disabled breakdowns.
10:43:20 <xarick> hmm... I twitch my nose
11:07:12 <peter1138> rau117: Multi-cargo ships works by using the articulated vehicles feature to create a chain of vehicle parts. Road Vehicles already have that...
11:09:52 <rau117> That is, there is an opportunity to create multi-refit road.vehs, but it is not used? Hmm…
11:10:50 <peter1138> You can build articulated road vehicles. They behave differently from non-articulated road vehicles of course.
11:30:21 <ahyangyi> they can only use drive-through stations
11:30:48 <Eddi|zuHause> and they can't overtake
11:31:20 <ahyangyi> Makes my decision to make all road vehicles articulated (for smooth turning) very dubious
11:31:56 <ahyangyi> I guess I should someday make a PR to add variables to enable smooth turning without fake articulation
11:32:21 <Eddi|zuHause> better invest in a PR to make smooth turning without variable
11:33:09 <ahyangyi> I guess grf creators still want control of the level of smoothness 🙂
11:33:52 <Eddi|zuHause> well, they can still use variables to undo the smooth turning :p
11:34:15 <Eddi|zuHause> but more seriously, they can just not provide the additional sprites
11:34:39 <ahyangyi> And road vehicles are inheriently unsymmetrical; turning right have different frames from turning left.
11:34:56 <ahyangyi> Well, the problem is that we don't have the variables in the first place
11:35:25 <ahyangyi> All existing variables return differences between the current wagon and another wagon, which is why they aren't useful in non-articulated vehicles
11:35:49 <Eddi|zuHause> yes, i know how that works. i invented most of that :)
11:36:43 <ahyangyi> They are so useful in most cases, but non-articulated road vehicles is the weird corner case >_<
11:36:46 <peter1138> Ideal solution is you don't need any variables.
11:37:01 <peter1138> But that's a long way off.
11:40:09 <Eddi|zuHause> you'd need to track differences in front and back position if you want to have a decision basis without articulated parts
11:40:12 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
11:40:23 <ahyangyi> Alternatively, allow articulated road vehicles to overtake and be overtaken and use dead end stops. But I remember seeing discussion before, and the conclusion seems to be that they are quite hard...?
11:40:44 <Eddi|zuHause> you have to completely redo overtaking
11:41:05 <Eddi|zuHause> and some way to drive backwards
11:41:13 <_jgr_> Articulated road vehicles overtaking/being overtaken on straight road segments is not really difficult
11:42:08 <Eddi|zuHause> overtaking has two problems: a) deciding whether it's safe to overtake, and b) articulated parts smoothly switching lanes
11:42:14 <_jgr_> Bay road stops are not doable because of the weird rotations that are done driving in and out of the stop
11:42:29 <ahyangyi> Eddi|zuHause: Oh good point
11:42:40 <ahyangyi> They make smooth turning even harder 😛
11:42:49 <ahyangyi> Now that they need to take care of overtaking as well
11:42:49 <Eddi|zuHause> a) is currently done by looking at the number of vehicles, which is why articulated parts break this
11:43:34 <LordAro> there are already bugs with overtaking (#9033, #10028), why would we want to introduce more? :p
11:43:40 <Eddi|zuHause> you can solve a) by introducing some sort of reservation system
11:45:00 <Eddi|zuHause> if you can reserve a stretch of the opposite lane long enough to complete the overtaking process, you can start overtaking
11:46:32 <ahyangyi> Can an overtaking vehicle break down, BTW?
11:48:18 <_jgr_> If the vehicle is about to break down it will not start overtaking, or will abort an overtake
11:50:26 <_jgr_> Hmm, actually, that might just be in my branch
11:50:30 <ahyangyi> Eddi|zuHause: Speaking of which, I kind of think the raw x/y/z/direction values within the grid might also be useful. Somewhere during the turn, the x/y differences might be the same but we might still want different visuals...
11:51:35 <Eddi|zuHause> ahyangyi: that may impose future limit of changing the movement patterns along the curve
11:53:20 <ahyangyi> Probably, but adding a caveat to the variable should be good enough? Even if we change movement patterns in the future, the worst thing that could happen would be an extremely smooth newgrf becomes only decently smooth
11:54:00 <peter1138> Correct solution is for the game to handle it all and not require any special variables.
11:54:28 <peter1138> Maybe I should update my patch that starts on that.
12:00:10 <ahyangyi> I am still not quite convinced... I feel the black box approach doesn't blend well with articulated vehicles, because each part can't predict how the game handles the other parts.
12:10:51 <ahyangyi> Yeah, perhaps I should come up with concrete examples to make this point
12:42:22 <Eddi|zuHause> you can't handle EVERY case with sprites. you'd need to switch to 3D models then
12:43:24 <peter1138> With enough sprites you can 😄
13:40:16 <alfagamma7> ahyangyi: The oldest government approved excuse
13:40:26 <alfagamma7> Never goes out of fashion
14:09:07 <peter1138> Hmm, read errors reported from my CD/DVD ROM drive.
14:09:23 <peter1138> Which is odd because there was no disc in it.
14:09:44 <_glx_> that could be why it fails to read 🙂
15:12:21 *** Flygon has quit IRC (Read error: Connection reset by peer)
15:26:19 *** Wormnest_ has joined #openttd
15:46:39 <__karma> Hey guys, I need some help with git, I clicked sync fork and now git created a commit with message "merge branch openttd master into my branch", is it ok to delete all the new commits from my fork, force push the changes and re-do the rebase?
15:58:43 <LordAro> it's "ok" in that you won't (or can't) break anything at our end
15:59:00 <LordAro> depending on what you're doing though, it might screw up your own branch
15:59:34 <peter1138> And always use branches for your own changes, avoids this 🙂
16:09:18 <__karma> Could you explain further? When I started I forked the project and created a new branch.
16:09:18 <__karma> Btw is it ok to open a PR 1 commit behind master? (git is quite confusing 🫠 )
16:11:14 <_glx_> so you did the right thing when starting (never work directly on master)
16:12:40 <_glx_> when you check the logs you see master commits with your commits on top and a merge commit ?
16:14:47 <__karma> It used to show my commits followed by the new commit from master and finally a merge commit on top, but I reset the head and forced up the changes, now it's back to what is was before I clicked sync fork button ig?
16:15:02 <ahyangyi> __karma: It's always OK to open a PR, the worst that could happen is that your PR can't be merged automatically
16:15:08 <__karma> (so it is still 1 commit behind master)
16:16:33 <__karma> Oh alright thx for the help :)
16:17:28 *** sparknf has joined #openttd
16:18:00 <ahyangyi> ( but you should probably play with PRs in your own repo instead of the main OpenTTD repo, until your are more or less clear about what's going on )
16:19:18 <sparknf> I am a new dev, and i wanted to start with openTTD, so i am now cloning the repo, any tips of what i should start with ? I want to rebuild the game, and get a understanding of it.
16:19:28 <_jgr_> When you're working with branches/PRs/etc with git, it's very helpful to use a tool which will show you what is going in graph form
16:20:05 <_jgr_> I use gitk mostly, but there are various others which do this
16:20:15 <talltyler> The first step is setting up your environment so you can compile OpenTTD — can’t do much without that 🙂
16:20:45 <ahyangyi> github has a "network graph" functionality as well -- is that the graph form you meant?
16:20:51 <sparknf> talltyler: Okay. I am brand new with everything, I only have some python experience, So i would love some guidence.
16:22:33 <_glx_> first question is which OS ?
16:22:35 <__karma> _jgr_: Oh alright I was using GitHub desktop app to do everything, I’m going to take a look at that
16:22:38 <talltyler> Start by following this guide, and if you get stuck ask here 🙂
16:24:57 <sparknf> how can i install the libraries on linux ? I dont see anything in the guide on HOW to do so.
16:25:24 <_glx_> ha on linux it's expectd you know how 🙂
16:25:36 <ahyangyi> Oh, you are using Linux. Then you probably just use your system package manager to install the libraries. For example, under Ubuntu they are typically called `libfoo-dev`
16:25:53 <ahyangyi> replace foo with the name of the library you are looking for
16:35:49 *** Leopold has quit IRC (Ping timeout: 480 seconds)
16:39:19 *** Leopold has joined #openttd
16:53:40 <talltyler> (No, I am not about to make a lot of spam 🙂 )
17:18:17 <xarick> uh... reading comprehension
17:21:57 <xarick> vehicle has 1 order, go to is selected...
17:22:32 <xarick> and i'm complaining that it is not cloning the orders of another vehicle
17:27:42 <__karma> Hey, does this look good or do I need to be more detailed?
17:32:25 <merni> __karma: It could do with being more high-level, as in, describing the effects of the change from the user PoV, as well as describing the change in terms of code
17:32:34 <merni> For example with a screenshot
17:33:01 <merni> But it's good enough to submit and edit later IMO
17:33:56 <__karma> Alright, I'll add a before and after screenshot then
17:37:14 <merni> xarick: re: 12121: the check is on "Vehicle 2", not "Vehicle 1". In your example save, both RV no. 2 and aircraft no. 2 have two orders each
17:38:01 <merni> Sorry, my comment on the issue should have said "more than or equal to two orders"
17:39:45 <merni> In the savegame you attached,
17:39:45 <merni> > Open Aircraft or Road Vehicle list.
17:39:45 <merni> > Click vehicle 2 from the list.
17:39:45 <merni> Vehicle 2 has two orders already and so quick creation is (intentionally) disabled
17:40:50 <xarick> oh, it has 2 orders, only 1 is station, quick goto is still on
17:41:23 <merni> But apparently even with 1 order it doesn't work
17:41:23 <_glx_> enabling of goto is done only if you have 0 or 1 orders
17:41:43 <merni> _glx_: Yes, but even with 1 order clicking on another vehicle doesn't seem to copy it
17:42:05 <_glx_> to copy you need to select the "end of order"
17:43:23 <merni> that doesn't seem to work either
17:44:13 <merni> the issue or the comment?
17:44:34 <xarick> the comment now says 2 orders
17:45:15 <merni> hm I should doenload RC3 first
17:45:28 <xarick> it needs to count 2 station orders for it to not activate quick goto
17:45:38 <xarick> it has 1, i should be more specific there i guess
17:45:52 <merni> but doesn't matter, it doesn't work even with one station order alone
17:48:00 *** HerzogDeXtEr has joined #openttd
17:49:43 <merni> the behaviour currently is:
17:49:43 <merni> - with 0 orders at all, goto is selected by default and clicking another vehicle copies orders (as expected)
17:49:43 <merni> - with 0 or 1 station order but at least one order in total, goto is selected by default but clicking another vehicle doesn't copy anything (even if you select --End of orders-- and then click). This is the bug
17:49:43 <merni> - with >=2 station orders goto is not selected by default (as expected)
17:51:54 <truebrain> Make sure this makes it into the issue if you don't mind 🙂
17:52:12 <_glx_> ha indeed it doesn't copy if there's any order
17:54:25 <xarick> so it's a mix of expected but inconsistent behaviour?
17:54:35 <xarick> or is it a me problem?
17:54:43 <_glx_> if (this->vehicle->GetNumOrders() != 0 && !share_order) return false;
17:55:14 <_glx_> I think it's to prevent accidental overwrite
17:55:55 <_glx_> unless you want to share orders (you pressed ctrl when clicking)
17:56:20 <merni> Why is overwriting OK when sharing but not copying?
17:56:23 <xarick> was hoping for the orders to be copied, without shared orders
17:57:01 <merni> xarick: Yeah imo that's the expected behaviour
17:58:29 <merni> Misclicking and overwriting the orders could be a big pain though especially if you have timetables
18:00:26 <peter1138> I have a patch somewhere that merges all identical orders into shared orders.
18:00:57 <peter1138> Mainly for testing performance between non-shared and shared orders.
18:01:04 <peter1138> It's not very useful :S
18:01:21 <merni> From gameplay pov probably not desirable unless it's something you explicitly do
18:03:16 <_glx_> BTW it's usually better to share orders than to copy
18:03:49 <_glx_> and yes missclick easily happen
18:03:54 <merni> But why does it make a difference in this case? If I'm overwriting existing orders, it doesn't really matter whether I'm replacing them by ordinary or shared orders
18:04:24 <_glx_> I won't count the number of times I have 'land info" cursor active
18:04:35 <merni> `* TODO: give a warning message */ `
18:04:39 <merni> Maybe we ought to do that
18:04:44 <_glx_> and this one is not the most dangerous
18:05:35 <xarick> if i click on a vehicle which is not in a station, could it at least copy the orders of the vehicle?
18:05:37 <merni> `We disallow copying orders of other vehicles if we already have at least one order entry ourself as it easily copies orders of vehicles within a station when we mean the station. `
18:05:37 <merni> Hmm, this is a good point
18:06:14 <xarick> actually... nevermind, better close it indeed
18:06:55 <xarick> the quick goto is the confusing one
18:07:15 <_glx_> quick goto just keeps the "goto" button active
18:07:19 <xarick> it should perhaps not be active if there's more than 0 orders
18:07:33 <merni> That makes sense to me tbh
18:07:45 <_glx_> and auto activates it if there's 0-1 orders in the list when you open it
18:08:06 <merni> I remember before discovering the quick goto setting
18:08:16 <merni> Go to > click station > Go to > click station .....
18:09:24 <_glx_> an order list needs at least 2 orders to be useful
18:09:39 <_glx_> so it's assumed you want to complete it
18:10:49 <merni> Honestly I think the warning message route for copying orders might be a good idea. Anyway when this copying is disabled it opens up the window of the vehicle you clicked, so instead it could open a popup "Overwrite existing orders and copy orders of Vehicle X? Yes/No"
18:11:22 <xarick> open a mini-menu instead
18:11:31 <merni> Otherwise it is not really discoverable (via the tooltip either) that you need to have 0 orders to copy
18:11:57 <xarick> copy orders of clicked vehicle, add goto station order, something else?
18:12:25 <xarick> "and don't ask me again"
18:13:14 <merni> "To disable cookies, untick every vehicle in this list of 1000 vehicles"
18:15:12 <xarick> when placing stations, holding ctrl opens a window
18:15:31 <xarick> needs something like that but for goto orders? 🙂
18:21:08 <merni> xarick: Why don't you PR it :p
18:21:38 <merni> Would it just have one option if the vehicle isn't in a station, though?
18:22:14 <merni> And how would it work if you click on the vehicle in the vehicle list window/depot window/whatever instead of on the map :P
18:59:59 <Eddi|zuHause> opening an extra menu on ctrl+click kinda defeats the point of ctrl being a shortcut... you can just select all the things ctrl can do from the order window
20:18:40 *** gelignite has joined #openttd
20:42:23 <__karma> Hey, for reviewers would you prefer me to add more commits to fix things brought up in the comments or to "redo" a commit (reset and force push a new commit)?
20:46:02 <frosch123> adding more commits implies "squash"-later
20:46:36 <frosch123> rebasing allows keeping the commits, which makes sense if they do different things (refactor first, fix second)
20:46:42 <_glx_> depending on complexity, multiple commits may help reviewing
20:47:05 <_glx_> (even if everything is squashed at the end)
20:49:08 <__karma> I see, at the moment, I need to change a functions name and to remove language txt files from my commit, I think it's better to redo the commit?
20:50:48 <Rubidium> just leave the txt files of languages other than english.txt untouched. The translator tool will automagically fix those (including comments)
20:51:03 <frosch123> if you edit the history, reviewers can easily see the changes: github has an extra diff button for that
20:51:19 <frosch123> it's just hard, if you also change the base-revision at the same time
20:55:58 <__karma> So it's better to make a new commit?
21:00:36 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:02:09 <frosch123> in case of #11345, it would be easier readable, if the addition of "get_def_cb" was a commit on its own. and the usage of it for service intervals was separate
21:02:20 <frosch123> one commit should do one thing, not two
21:03:35 <__karma> frosch123: I used git rebase to split the commit into 2 and I was going to force push it
21:03:39 <frosch123> currently there is one commit which does mass copy&paste changes to all the ini files, and hidden inbetween are functional changes by adding GetDefaultServiceInterval
21:03:55 <frosch123> yes, please force push, are not afraid of it
21:04:29 <__karma> Alright, what about the 3rd commit I need to revert the language text files, one more commit to do so or just redo the commit?
21:12:42 <truebrain> no clue what MacOS is doing there; also no way to debug it
21:41:07 <_glx_> `-Wl,-ld_classic` should be easy to add (because bumping minimum deployement is a clear no)
21:41:16 <_glx_> and testable via nightlies
21:41:52 <truebrain> random testing stuff via nightlies; that sounds like a recipe for success 😄
21:43:20 <_glx_> well we know at least an affected user
21:43:48 <_glx_> and getting nightlies via steam is simple
21:44:53 <_glx_> and I don't know if we have access to macOS12- users here 🙂
21:46:39 <_glx_> though it should be fixed with 15.1
21:49:26 <truebrain> yeah ... starting fee, 1 month
21:49:36 <andythenorth> hmm we joke about running the infra on my MBP
21:49:48 <andythenorth> but I have a spare intel MBP and I could open a port and give someone SSH on it
21:50:20 <andythenorth> my upstream isn't very quick
21:50:49 <_glx_> we would need something running xcode15, and something to run the resulting build on 12-
21:51:32 <andythenorth> I just upgraded from 12 😛
21:52:37 <_glx_> and RC2 was fine on 12 for you ?
21:55:14 <andythenorth> I didn't try RC2, RC1 was fine
21:56:03 <andythenorth> macOS 12 is dead, we shouldn't support it
21:57:32 <_glx_> well here it's 10.15.7 user complaining 🙂
21:58:25 <andythenorth> we did draft a sort of policy for mac support once
21:58:37 <andythenorth> I think it was 2 major versions, or something like that
21:58:57 <_glx_> yup CI change between RC1 and RC2
21:59:18 <truebrain> N - 2, yes. As that is what MacOS supports 😄 Tbh, it could also just be the answer: sorry, we don't support MacOS 10.15, as MacOS doesn't do it, and clearly they changed something which breaks on older versions
22:03:59 <_glx_> I surprised steam client still supports 10.15
22:04:57 <_glx_> (they require windows 10+ now)
22:06:01 <truebrain> Steam dropped 10.13 and 10.14 in February this year
22:06:09 <truebrain> (effectively dropping 32bit support)
22:06:34 <reldred> andythenorth: you shut your mouth, I'm still stuck on 12.6 😦
22:06:56 <reldred> (but at least I have four usb-c ports)
22:07:02 <_glx_> oh and RC2 or RC3 run for you ?
22:07:07 <truebrain> 10.15 is out of support since 1 year and 6 months 😄
22:09:33 <_glx_> (nightlies should have the issue too)
22:16:24 <xarick> I fail at squirrel classes
22:25:13 <andythenorth> squirrel classes are weird
22:25:26 <andythenorth> it's not clear they're meaningfully classes, they're just table slots of some kind
22:25:40 <andythenorth> and they have two different syntax, and neither is obviously correct
22:26:17 <andythenorth> squirrel is the most novelty language I have used, although my experience is not extensive
22:27:02 *** gelignite has quit IRC (Quit: Stay safe!)
22:31:41 <_jgr_> Actually having classes at all already puts Squirrel fairly high up as these things go
22:32:25 <_glx_> and table is perfectly fine to store class
22:33:52 <andythenorth> squirrel is like playdoh
22:34:39 <andythenorth> it seems to not really matter how you use it
22:34:45 <andythenorth> choices have no consequence 😛
22:35:47 <andythenorth> I am firmly indoctrinated by python though 😛
22:35:47 <andythenorth> `There should be one-- and preferably only one --obvious way to do it.
22:35:47 <andythenorth> Although that way may not be obvious at first unless you're Dutch.`
22:36:27 <andythenorth> the existence of python code golf suggests that usually there are many more than one ways to do it 😛
22:36:40 <peter1138> switches in python eh?
22:37:38 <andythenorth> they exist recently
22:37:49 <andythenorth> some syntax I forget
22:39:00 <andythenorth> was it ~lunch~ naptime
22:42:50 <andythenorth> yes it was naptime
22:44:52 <peter1138> `__cxx_global_var_init` is a nice place to crash.
22:44:58 <peter1138> Not actually started the real code yet.
22:54:29 <truebrain> hence Rb blaming the linker 😛
23:01:39 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:14:16 <_glx_> hmm do we really need xcode 15 stuff for 14.0 or was it only needed for 15.0?
23:39:14 <truebrain> zero guarantees how long release-14 branch compiles with older xcode 🙂 Very easy to sneak some backport PR in that breaks it
23:48:54 *** jinks has quit IRC (Remote host closed the connection)
23:56:31 *** Wormnest_ has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
