IRC logs for #openttd on OFTC at 2020-04-25
⏴ go to previous day
00:04:34 <TrueBrain> yeah, IPv6 also works :D
00:05:38 <DorpsGek_III> [OpenTTD/nml] LordAro commented on pull request #102: Fix #100: follow PEP440 when generating version https://git.io/JfL9s
00:08:18 <TrueBrain> that took a lot more time than I expected, but .. the result is pretty clean and nice
00:08:42 <TrueBrain> tomorrow I hope to attach a DNS entry to it, as that alllows people to test it out :D
01:41:50 *** tokai|noir has joined #openttd
01:41:50 *** ChanServ sets mode: +v tokai|noir
02:47:28 *** Wormnest_ has joined #openttd
02:49:44 <FLHerne> LordAro: Hm, I should try and get my nmlop changes finished and in before your huge reshuffle
02:49:56 <FLHerne> Otherwise rebasing them will be a nightmare :-/
02:53:58 <LordAro> FLHerne: but then *I* would have a nightmare rebase
02:54:26 <LordAro> more pressingly, we need to release 0.5.0
02:54:43 <LordAro> even more pressingly, i should go to sleep
02:54:59 <FLHerne> Yeah, but yours is automated, and not done by a mixture of insane regexes that I lost months ago and hand-editing :P
02:56:09 <FLHerne> I don't quite get your comment on the PR about hand-editing the spaces; weren't they mostly aligned sanely to begin with?
04:07:20 *** Wormnest_ has joined #openttd
06:54:00 *** Smedles has joined #openttd
07:17:10 *** Smedles has joined #openttd
07:24:07 *** Smedles has joined #openttd
07:32:02 *** Smedles has joined #openttd
07:44:22 *** Smedles has joined #openttd
07:49:26 *** Smedles has joined #openttd
07:56:06 *** Smedles has joined #openttd
08:14:20 *** Smedles has joined #openttd
08:19:36 *** andythenorth has joined #openttd
08:22:24 *** nielsm is now known as Guest23280
08:30:00 *** Flygon_ has joined #openttd
08:34:06 *** Progman has joined #openttd
08:37:44 *** Smedles has joined #openttd
08:42:21 *** Laedek_ has joined #openttd
08:46:55 *** Smedles has joined #openttd
08:48:14 <TrueBrain> morning andythenorth
08:48:25 <andythenorth> LordAro has been busy again :)
09:02:59 <TrueBrain> with a very tedious task
09:04:10 <andythenorth> it's good to see nml get some care
09:04:28 *** Smedles has joined #openttd
09:04:33 <andythenorth> it's a good tool
09:05:43 <Samu> still have the canal pathfinder bugged :(
09:07:02 <Samu> can i build a bridge over another bridge if the another bridge is 2 tiles only?
09:08:17 *** sla_ro|master has joined #openttd
09:12:26 <Samu> stupid bridge rules, only complicates the pathfinder :(
09:14:36 *** Smedles has joined #openttd
09:15:30 <andythenorth> there are a bunch of nml commit messages that don't fit rules, should I fix them in rebase?
09:15:54 <andythenorth> and a few others
09:17:01 <andythenorth> quite a lot actually, it looks messy :)
09:17:54 <Wolf01> Hmmm, should I choose games from the humble choices or just play minecraft rtx?
09:19:43 <Samu> dont waste money on games
09:20:12 <Wolf01> I already paid a year, so...
09:20:13 <TrueBrain> andythenorth: never rewrite public history :)
09:20:15 <TrueBrain> it will break ALL forks
09:20:37 <andythenorth> probably not desirable
09:20:51 <TrueBrain> in a project this size, survivable, but meh
09:21:07 <nielsm> IMO it's fine if the initial (exploratory, I guess you can say) stages of a project are messy and not up to a standard
09:21:11 <andythenorth> does that actually apply if just re-wording the message?
09:21:19 * andythenorth assumed the hash wouldn't change
09:21:29 <nielsm> yes changing the commit message also affects the commit hash
09:21:36 <andythenorth> I don't want to do it, it's just guilt, some of the commits are mine
09:21:47 <nielsm> and that affects the hash of all commits on top of that one
09:21:49 <andythenorth> I've been writing 'Docs: ' not 'Doc: ' for years and I can't stop
09:25:10 <TrueBrain> as it goes: learn from the past, live in the future
09:25:42 <andythenorth> everything is brighter in the future, and we are better people!
09:27:31 <TrueBrain> hmm .. all DNS entries on openttd.org had their TTL reset to 14400 .. that is odd, and not wanted
09:28:15 <Wolf01> TB: it's remember yesterday, think about tomorrow, but you have to live today ;)
09:28:37 <TrueBrain> Wolf01: strongly depends on your country of origin ;)
09:28:53 <andythenorth> I don't think England has any of these :P
09:29:28 <Wolf01> TB: or if you listen to hammerfall :P
09:29:40 <TrueBrain> and clearly, I do not :)
09:40:22 <andythenorth> nml changelog: I can use judgement and consolidate some things?
09:40:36 <andythenorth> there are commits to add a feature, then 2 or 3 more commits to fix that feature
09:44:35 <andythenorth> it's interesting that the convention (here and elsewhere) is to group by commit message prefix
09:44:48 <andythenorth> it doesn't really aid understanding related changes
09:44:56 <FLHerne> We already have a comprehensive list of commits, it's the commit history :P
09:45:28 <FLHerne> You mean the Add:, Fix: etc?
09:45:50 <andythenorth> depending who authored two highly related things end up separated
09:45:58 <andythenorth> I kind of wonder who the intended reader is
09:46:05 <andythenorth> it doesn't aid the casual nml user
09:46:13 <andythenorth> nor is it a reliable list of commits
09:46:14 <FLHerne> Changelog should be for users
09:47:06 <andythenorth> the correct way makes no sense to me
09:47:19 <FLHerne> Sorting them by prefix means they can ignore the whole "Codechange:" section because they don't care
09:47:36 <FLHerne> (which does raise the question of why those are in the changelog at all)
09:47:44 <FLHerne> Perhaps they shouldn't be
09:47:53 <andythenorth> lots of codechanges are in as 'change'
09:48:19 <andythenorth> I'll push a draft in a minute, just feeding the kids
09:48:38 <FLHerne> I like that Horse changelog
09:49:43 <FLHerne> Are Road Hog trucks supposed to have load sprites for metal?
09:51:59 <andythenorth> I'd have to check, my guess is some lengths do and some don't
09:52:10 <andythenorth> I'd like to work on Road Hog and do an NRT edition
09:52:24 <andythenorth> but nml...bananas...etc :P
10:03:43 <TrueBrain> I know that feeling :P
10:04:28 <andythenorth> as long as we're not the old lady swallowing a spider to catch the fly :P
10:04:38 * andythenorth just likes trains
10:07:53 <andythenorth> 70 lines or so of changes
10:23:18 *** Smedles has joined #openttd
10:28:06 <TrueBrain> downside of making a DNS mistake with 1.1.1.1 as your resolver ... a lot of people will now have your mistake cached for the next hour :P
10:28:46 <TrueBrain> its fine, they are new entries
10:30:25 * andythenorth wonders about combining retro website style with a new responsive theme
10:34:42 <andythenorth> maybe not now :P
10:48:05 <DorpsGek_III> [OpenTTD/OpenTTD] kiwitreekor updated pull request #7575: Feature: Add industry production graph https://git.io/fjc3i
11:13:28 *** mcbanhas has joined #openttd
11:14:49 <mcbanhas> https://i.imgur.com/rDHv7ye.png Sometimes you get a sense of how much of a dumpster fire the OpenTTD UI can be just by looking at the tooltips. Not even my re-writing below helps much.
11:17:41 <mcbanhas> I'm seriously beginning to consider an alternate formatting to tooltips basd on some more modern standard. Maybe using left alignment instead.
11:21:54 *** Laedek_ has joined #openttd
11:23:05 <nielsm> the entire orders UI is terrible and overloaded
11:23:14 <nielsm> and nobody has dared improve it yet
11:24:05 <nielsm> the first thing to do IMO would be to scrap the idea of shared orders, and instead change orders into "routes"
11:24:26 <nielsm> so you assign a vehicle to a route, instead of making a route for a vehicle
11:25:37 *** Xetalim has joined #openttd
11:29:04 <mcbanhas> I'm just patching text here. It would be nice to start addressing UI sometime, indeed. It's not good to keep shirking the relevance of presentation and UI.
11:30:05 <mcbanhas> My guess is that UI organization being more of subjective matter, it can spark violent discussion, hence why people avoid it.
11:31:08 <nielsm> and also people can tend to make all kinds of funky mockups in drawing programs, but it all comes down to implementation
11:31:23 <mcbanhas> On the other hand OpenTTD has been like what, 10 years in development or more? It's not good to keep sweeping this under the rug.
11:32:10 <nielsm> and started out with lots of baggage being a straight 1:1 reimplementation of Transport Tycoon Deluxe
11:33:01 <Wolf01> Like when I have to make a site from a jpg... "do you know a little of how html works or I just have to make a clickable jpg?"
11:34:54 <DorpsGek_III> [OpenTTD/OpenTTD] kiwitreekor updated pull request #8008: Feature: Double-click to rename vehicle https://git.io/Jv4id
11:35:58 <mcbanhas> Yeah, but that will happen again eventually, if we just keep adding new features without improving the UI.
11:36:31 <DorpsGek_III> [OpenTTD/OpenTTD] kiwitreekor commented on pull request #8008: Feature: Double-click to rename vehicle https://git.io/JftJs
11:37:13 <andythenorth> nobody dares touch orders
11:37:19 <andythenorth> oh you said that already :)
11:37:21 <DorpsGek_III> [OpenTTD/OpenTTD] kiwitreekor updated pull request #8008: Feature: Double-click to rename vehicle https://git.io/Jv4id
11:37:53 <Wolf01> We should remove the orders totally and let the game figure them out by itself :>
11:38:13 <andythenorth> you know which XKCD we need
11:38:18 <nielsm> had to abort in the end, couldn't make my idea work
11:38:33 <andythenorth> I tried to redesign newgrf configuration window
11:38:39 <andythenorth> and 'new game' window
11:39:06 <mcbanhas> You post that comic all the time. I bet you have it bookmarked at this point.
11:39:18 <andythenorth> it's easily found
11:40:35 <andythenorth> (in case meaning is lost in irc format), I think OpenTTD is in a good place, and we can be cheerful about development in last 2 years
11:41:08 <andythenorth> but the UX is a shitshow in places, and unfortunately orders/timetables are the worst place to start trying to fix that
11:41:26 <andythenorth> imagine walls of rainbows attached to this comment
11:41:52 <mcbanhas> Well, sometimes it can be the other way around in the sense that we can throw it all out of the window.
11:42:52 <andythenorth> we'd probably have to leave the legacy UI in place
11:43:14 <andythenorth> pathfinders all over again :)
11:44:18 <mcbanhas> I really don't like maintaining legacy features and groaners. I'd rather move forward and let natural selection do to the rest.
11:44:28 <andythenorth> we have no choice :)
11:44:38 <andythenorth> total 100% commitment to backwards compatibility
11:45:05 <nielsm> there's still people playing old and very old versions of ottd, and that's perfectly fine
11:45:18 <nielsm> if those versions are the game they want to play then by all means play that version
11:45:19 <andythenorth> generally this approach to software makes for a small cohort of happy users, but a terrible user experience
11:45:38 <andythenorth> I suspect there's a scale between ease-of-use and legacy suport
11:45:43 <andythenorth> we choose one end
11:46:10 <mcbanhas> Wait, but how does that relate to OpenTTD gameplay though? It's technically a different client that only fishes out optional graphics right?
11:46:31 <andythenorth> compared to TTD?
11:46:58 <andythenorth> I was speaking only about OpenTTD
11:47:29 <andythenorth> I have seen a myth on websites that TTD is somehow holding back OpenTTD development
11:47:49 <andythenorth> this isn't true, OpenTTD development is held back (if at all) by previous OpenTTD development
11:47:50 <mcbanhas> Because if you're talking about Legacy support in those terms it means that somehow the clients can be compatble together via multiplayer or something like that.
11:47:56 <nielsm> I will go and do other things, enjoy the daytime
11:48:28 <andythenorth> mcbanhas the support is for savegames
11:48:55 <andythenorth> specifically all savegames back to version 0 must work in current OpenTTD
11:49:04 <andythenorth> this means we can never change things like orders
11:49:16 *** Flygon_ is now known as Flygon
11:49:22 <andythenorth> change / drop features /s
11:49:46 <michi_cc> andythenorth: Of course we can change it, e.g. to routes, but we'd need to include a load converter that transfers shared orders to routes.
11:50:01 <andythenorth> and timetables, and conditional orders
11:50:39 <andythenorth> and we'd have to preserve a UI that keeps all those affordances mutable
11:51:10 <andythenorth> I'm not trying to prevent progress, I just think this is particularly hard problem :)
11:51:25 <mcbanhas> michi_cc: I smell a lot of errors coming out of such a converter.
11:51:41 <andythenorth> I looked at it before, and thought 'this is easy' we just delete all the broken crap
11:51:48 <andythenorth> but I was of course wrong :)
11:51:52 <mcbanhas> This is really where I'd rather throw TTD save game support out of the window.
11:52:06 <andythenorth> TTD is a red herring
11:52:16 <andythenorth> TTD does not have timetables, conditional orders etc
11:52:32 <andythenorth> they are mistakes in OpenTTD history, if simple order UX was the goal
11:52:43 <andythenorth> or adding more order features that actually work and are useful :)
11:53:02 <mcbanhas> So we can do routes and still keep save game compatbility?
11:53:55 <andythenorth> what do routes do? o_O
11:54:11 <mcbanhas> Not sure either, nielsm proposed that.
11:54:30 <mcbanhas> I mean, "something better that the orders we have now"
11:55:32 <mcbanhas> But what is the major impediment for throwing away all the broken crap?
11:55:56 <andythenorth> 100% savegame compatibility
11:56:01 <andythenorth> and the player base
11:56:35 <mcbanhas> What elements in save game compatbility are in direct conflict?
11:56:40 <andythenorth> also 'broken crap' is a highly subjective comment by me, I withdraw it
11:56:59 <andythenorth> just because I find timetables don't work, doesn't mean they don't work for other people
11:57:28 <mcbanhas> I think that most people do think they're crap though.
11:57:36 <mcbanhas> And that's not a big if.
11:58:06 <andythenorth> _dp_ you're right "We guarantee that every revision of OpenTTD will be able to load savegames from every older revision (excepting where the savegame is corrupt). Please report a bug if you find a save that doesn't load."
12:00:24 <mcbanhas> I gotta say this is a bit frustrating, given it puts an impediment on a lot of work that I wanted to do.
12:00:50 <andythenorth> well I might be wrong
12:00:51 <mcbanhas> And unfortunately given I'm not a programmer, I'm just stuck on having to rant about it.
12:00:54 <andythenorth> I'm not even a dev
12:01:22 <andythenorth> but OpenTTD is very easy to make progress on in some areas
12:01:39 <andythenorth> and very hostile to progress in others (the actual program, not the community)
12:02:06 <mcbanhas> I noticed. I mean people are doing radical stuff like water depth and whatnot.
12:02:08 <andythenorth> orders and timetables are almost impossible to ever change
12:02:15 <mcbanhas> But the UI is still stuck in the 90s
12:02:34 <andythenorth> there are players who have spent years building detailed recreations of specific railway systems, with realistic timetables
12:02:52 <andythenorth> and there are others using conditional orders to build self-managing networks with thousands of trains
12:03:10 <mcbanhas> Well here's the deal.
12:03:27 <andythenorth> I've proposed that so many times :)
12:03:41 <andythenorth> it just needs a fork
12:04:07 <andythenorth> currently we have one known viable fork - JGRPP
12:04:16 <andythenorth> and maybe a few more for goal servers that I don't know about
12:05:06 <andythenorth> all patchpacks that I've seen have tended to add a lot more complexity, not less
12:05:07 <mcbanhas> The problem is that by sticking to legacy we are inadvertently making the game less accessible to new players.
12:05:24 <mcbanhas> That's bad, because at some point people start losing interest.
12:05:38 <andythenorth> the most vocal part of the player base wants a lot more complexity
12:06:07 <andythenorth> the players who are overwhelmed and quit don't go on forums or reddit asking why development of complicated things is so slow
12:06:43 <andythenorth> also one person's legacy is another person's years of work on a savegame
12:06:55 <andythenorth> we're not Apple :|
12:07:52 <mcbanhas> Curious, how did you try to fix the new game screen?
12:11:59 <_dp_> it's nice to see UI talks but imo the first thing that needs fixing is that horrible bitmap font openttd comes with
12:12:50 <andythenorth> it's not very good
12:13:14 <andythenorth> the issue I was trying to solve is that starting a new game is massive yak shaving
12:13:16 <mcbanhas> _dp_: I was going to work on that after I was done with tooltip text actually
12:13:25 <andythenorth> _dp_ what's wrong with it? :D
12:13:46 <andythenorth> apart from it has curly quote apostrophe in one direction only?
12:13:49 <_dp_> to wide and too small resolution
12:13:58 <TrueBrain> content.staging.openttd.org, port 4978, now has the new BaNaNaS ingame content server running :)
12:14:28 <TrueBrain> different port, so you would have to recompile the client to use it :D
12:14:53 <mcbanhas> You'd prefer a narrower font?
12:15:15 <andythenorth> _dp_ is that the OpenGFX font?
12:15:21 <_dp_> well, it's like 150% wider than any other font
12:15:59 <mcbanhas> _dp_: which font are youtalking about exactly? Can you give examples on you'd prefer?
12:16:15 <mcbanhas> I'm saying this because I'm genuinely going to work on fonts after I'm done with tooltips
12:17:25 <mcbanhas> andythenorth: I see the problem there, you tried to put too many potatoes in one bag
12:17:50 <andythenorth> it lacks any flow
12:17:58 <mcbanhas> Personally, I do not mind NewGRF settings being separate.
12:18:28 <andythenorth> I would throw out all newgrf if I could
12:18:49 <mcbanhas> The good thing about the current new game menu is that you just have to throw in a few clicks and presto.
12:19:09 <mcbanhas> This just looks overly engineered.
12:20:08 <andythenorth> but the number of savegames I've abandoned because of forgetting to change a setting is very high
12:20:34 <andythenorth> I think it's non solvable
12:20:44 <DorpsGek_III> [OpenTTD/bananas-server] sentry-io[bot] opened issue #6: TypeError: '<' not supported between instances of 'tuple' and 'list' https://git.io/JftJM
12:20:45 <mcbanhas> Not even by printing warnings?
12:21:17 <milek7> current newgrf settings being separate is just weird
12:21:23 <milek7> it is savegame setting, not global one
12:21:58 <mcbanhas> I did some work for OpenXCom, and what they did there in regards to mods, was simply to print a warning when you load a savegame, if you're lacking anything or have different settings than the currently loaded ones.
12:22:48 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain opened pull request #7: Fix #6: fix version-check, and with that content-listing https://git.io/JftJS
12:22:53 <TrueBrain> easiest fix yet ^^ :D
12:23:08 <andythenorth> mcbanhas but how do you validate options?
12:23:35 <andythenorth> starting a game with missing GS for example
12:23:42 <andythenorth> or with 10 AI players configured
12:26:36 <andythenorth> that a is so ugly
12:26:44 <andythenorth> in the pixel font
12:27:16 <Samu> 310 ticks, 1 more tick than yesterday
12:27:25 <Samu> honestly, I expected worse
12:28:55 <milek7> TrueBrain: there's nothing there to download?
12:29:12 <TrueBrain> see the last issue and PR
12:31:59 <mcbanhas> _dp_: what do you exactly mean with this?
12:32:23 <mcbanhas> You prefer the TrueType dejavu over a pixel font?
12:32:48 <_dp_> pixel is pretty much unusable for me
12:32:57 <_dp_> either hard to read or too big
12:34:48 <mcbanhas> We talked about this before, the possibility of including TrueType fonts in the package along with the pixel originals.
12:35:22 <mcbanhas> The problem with TT fonts is that they do not really match the aesthetic of the windows and menus unless they are somehow stylized
12:35:54 <mcbanhas> So we would likely need a custom TT font for that, which *might* be possible.
12:36:42 <_dp_> imo that's a non-issue compared to having it almost unplayable
12:36:45 <mcbanhas> I'm going to work on the pixel fonts first, because I actually kind of like them, but I'm gonna try to figure out a TT solution as well.
12:37:24 <milek7> pixel fonts are fine, but not on hi-dpi displays
12:37:41 <_dp_> also idk, tt fonts look perfectly fine with aesthetic imo
12:38:21 <mcbanhas> milek7: That always seems to be the problem right? TT fonts on the other hand, make the UI scale in weird ways.
12:39:25 *** Smedles has joined #openttd
12:39:52 <Samu> it's using a bitset or bitmask, whatever it's called
12:41:04 <mcbanhas> You can also make a pixel-styled TrueType font. I'm not sure how well would that work though. It would have to be tested.
12:41:38 <mcbanhas> If it were to work well, we could actually drop bitmap font support altogether and just have pixel-styled TTs
12:42:18 <Samu> 1 = axis x, 2 = axis y, 4 = aqueduct ramp (the platform in the air), 8 = aqueduct head (the thing at the start or end of the aqueduct); 4 | 8 = it's not an aqueduct, it's a lock, which occupies both
12:44:09 <Samu> then i check for a conflict in a single tile with another function
12:44:11 <Samu> /* Check if both directions can go together on a single tile. */
12:44:11 <Samu> return !(existing_direction & new_direction);
12:44:53 <mcbanhas> andythenorth: OXC is a bit more easy, I guess, since it does not have such a thing as GS or AI modules. But if you have a savegame that is produced when mods are loaded, it will print a warning saying there are mods missing if you try to reload the save without them on.
12:45:12 <mcbanhas> Come to think of it, but OpenTTD already does that on the multiplayer lobby.
12:45:37 <mcbanhas> My question is, couldn't something similar be applicable to GS and other settings?
12:45:44 <andythenorth> I don't see how it could ever work
12:45:53 <andythenorth> I think we talk at cross purposes
12:45:54 <_dp_> just use some regular tt font, ui already has hi-dpi icons, tt fonts look perfectly fine along with those
12:46:10 <andythenorth> OpenTTD has no way to know that I didn't *want* 10 AIs in new game
12:46:19 <mcbanhas> _dp_: I personally think they look very bad.
12:46:24 <andythenorth> but then I have to abandon the game because the AIs have wrecked the map
12:46:47 <mcbanhas> The only game is low rest pixel art, you put a hi-res font on the backdrop and it doesn't match.
12:46:48 *** frosch123 has joined #openttd
12:46:56 <milek7> mcbanhas: how it's related to ottd? it enables/downloads missing content automatically
12:48:01 <mcbanhas> aaaah you mean in a new game.
12:48:17 <mcbanhas> I thought you meant in a savegame
12:50:04 <mcbanhas> andythenorth: if you want to UI design your way around that there's an easier solution than what you did in your early mockup: step-by-step new game menus, similar to games like Civ or Homm
12:50:43 <mcbanhas> So like, first you select your map settings, press next, then select your opponents, press next, and so on and so forth.
12:51:32 <milek7> certainly worse than tabbed approach
12:51:54 <mcbanhas> I don't think so, personally.
12:52:31 <mcbanhas> Functionally you can still forget things on the tabbed menu..
12:52:57 <_dp_> well, anyway, openttd looks terrible by default on modern resolutions imo
12:53:11 <milek7> ..that's not game problem, though
12:53:12 <_dp_> you literally have to go and edit cfg to make it playable
12:53:34 <DorpsGek_III> [OpenTTD/bananas-server] frosch123 approved pull request #7: Fix #6: fix version-check, and with that content-listing https://git.io/JftUO
12:54:47 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain merged pull request #7: Fix #6: fix version-check, and with that content-listing https://git.io/JftJS
12:54:49 <DorpsGek_III> [OpenTTD/bananas-server] TrueBrain closed issue #6: TypeError: '<' not supported between instances of 'tuple' and 'list' https://git.io/JftJM
12:55:25 <TrueBrain> I do not; I would have to check it out
12:55:28 <TrueBrain> haven't had the time
12:55:41 <frosch123> ok, i was hoping you would just know it :)
12:55:50 <frosch123> no worries, more important stuff at hand
12:56:31 <TrueBrain> like how to call a lambda based on an CloudWatch event ..
12:59:05 <TrueBrain> at least listing works .. downloading .. not so much :D
12:59:05 <frosch123> andythenorth: who do you consider the target audience for the nml changelog?
12:59:20 <TrueBrain> Access Denied, awwwhh
13:00:36 <frosch123> andythenorth: ah, nevermind, the changelog for the previous versions is as weird :)
13:01:35 <frosch123> maybe that changelog should be deleted
13:01:51 <TrueBrain> so two bugs: 1, the http redirect doesn't work, 2, downloads via the server don't work :P Fixing 2, 1 is more a puzzle ..
13:02:00 <frosch123> or maybe i am the only one who things a changelog is very different from a vcs history
13:02:37 <TrueBrain> okay, downloading works too now at least :D
13:04:00 <TrueBrain> owh, ofc http download doesn't work .. I am fetching that from the production server .. that is not going to work :D
13:05:55 <TrueBrain> but okay, full-chain-testing should now be possible. But I need to finish some automation first :)
13:12:06 <TrueBrain> milek7: it works for you too now?
13:13:04 <TrueBrain> that requires some more changes in the client
13:13:14 <TrueBrain> but it should fallback to the TCP connection to download
13:13:24 <TrueBrain> so from a user-perspective, it should work
13:13:34 <TrueBrain> for the http, you have to change "binaries.openttd.org" to "binaries.staging.openttd.org"
13:13:56 <milek7> dbg: [net] [tcp/http] unhandled status reply 400 Bad Request
13:15:14 <andythenorth> frosch123 very valid question
13:15:17 <Wolf01> There are at least INT_MAX ways to deal with redirects and I didn't find one which is nice, works, doesn't need to change 45 things.
13:15:48 <andythenorth> surely changelog is for end user? History is for VCS
13:16:21 <frosch123> that's what i thought. but then 3/4 of that changelog should be deleted
13:16:33 <andythenorth> mcbanhas wizards perform terribly for anyone but the newest users
13:16:36 <andythenorth> they just get in the way
13:16:49 <andythenorth> frosch123 I pushed a consolidated version, glad you agree :)
13:16:55 <andythenorth> still some tidying to do
13:18:32 <andythenorth> I wonder if previously we were quite concerned about maintainer audience?
13:18:55 * andythenorth is always scared to break stuff in OpenTTD, so many tentacles
13:20:39 <frosch123> andythenorth: maintainers of distributions are also a valid audience
13:21:01 <andythenorth> so I left in ~all the codechange stuff
13:21:04 <frosch123> but neither of the two audiences is interested in codechanges that neither affect the bundle nor the functionality
13:21:21 <frosch123> .gitignore is not even part of the bundle
13:21:44 <andythenorth> notify irc - local concern?
13:21:58 <andythenorth> release workflow - remove, or single line?
13:22:33 <LordAro> anything that's purely build system related is not relevant
13:24:36 *** sla_ro|master has joined #openttd
13:24:42 <andythenorth> "enable VT100 sequences on windows" ??
13:26:53 <LordAro> that's just coloured text output, right?
13:27:20 <LordAro> so it should say that :p
13:27:50 <andythenorth> there's a few others like that
13:28:02 <andythenorth> the commit message is masking the actual benefit
13:28:15 <andythenorth> I've trimmed some
13:28:24 <LordAro> do you now understand the sort of task putting the OTTD changelog together is? :p
13:28:32 <andythenorth> I write changelogs in other places :P
13:28:39 <andythenorth> but usually I understand the audience
13:28:59 <andythenorth> THANKS FOR DOING THE OTTD CHANGELOG LordAro :D
13:29:33 * andythenorth adopts the weird capitalisation of first word after :
13:29:39 <andythenorth> it's not a new sentence
13:29:59 <andythenorth> but rest of changelog does it so eh
13:36:25 <andythenorth> "To add insult to injury, this isn’t even a dialog… it’s a wizard"
13:45:47 <frosch123> [13:29] *andythenorth adopts the weird capitalisation of first word after : <- that just means that the previous changelogs were written by germans :)
13:47:29 <andythenorth> I'll go with it :)
13:59:40 <TrueBrain> milek7: if you receive a 400, it most likely means that your request did end up on the production system; nevertheless, on staging you are being redirect to an https, which won't work for the client .. so I have some things to fix anyway
14:00:55 <milek7> static const char * const NETWORK_CONTENT_MIRROR_HOST = "binaries.staging.openttd.org";
14:07:33 <mcbanhas> _dp_: can you help me test something?
14:08:41 <mcbanhas> If you could send me screenshots would be great too
14:09:05 <mcbanhas> I have a small screen so difficult for me to test stuff in higher res
14:09:07 <milek7> TrueBrain: <head><title>400 The plain HTTP request was sent to HTTPS port</title></head>
14:09:14 <_dp_> ok, tho will take a bit to remember how to setup fonts xD
14:09:34 <TrueBrain> milek7: how is that a 400? Lol
14:09:57 <TrueBrain> owh, HTTP/1.0 I guess
14:10:11 <TrueBrain> does OpenTTD implement HTTP/1.0? Wouldn't surprise me
14:10:59 <mcbanhas> Sure, take your time, and thanks in advance :)
14:11:26 <andythenorth> that's ready for approve or amend ^
14:11:35 <andythenorth> I'm gone as far as I can with it :)
14:15:58 <LordAro> andythenorth: it specifically doesn't block release, not really
14:16:06 <LordAro> it just makes non-releases slightly harder to work with
14:16:17 <LordAro> (though not at all noticably)
14:16:33 <TrueBrain> okay ... the NLB itself is now fully operational; if an EC2 is being scaled up/down, it reconfigures the DNS, and if eht ECS is being scaled up/down, it reconfigures nginx. W00p :D IPv4 + IPv6 NLB, without manual actions (like it was on DigitalOcean :P)
14:16:37 <TrueBrain> now to fix the bugs we have found :D
14:18:57 <TrueBrain> milek7: give it another spin if you like :)
14:19:34 <TrueBrain> even the CDN URL seems to be correct \o/
14:21:17 <TrueBrain> frosch123: historican, what was the first release with content service?
14:22:36 <TrueBrain> HA! Even 0.7 works :D
14:22:44 <TrueBrain> these things make me happy :P
14:23:35 <frosch123> you compiled 0.7 with staging server?
14:23:35 <TrueBrain> so if 0.7 works, pretty sure all other versions work too :P
14:23:45 <TrueBrain> no; I redirected some things on my local machine
14:23:55 <TrueBrain> I installed an nginx to redirect everything to the right place :P
14:24:01 <TrueBrain> and put my hosts to 127.0.0.1 :D
14:24:16 <TrueBrain> wanted to be sure an unmodified binary worked :P (and I am incredibly lazy)
14:24:24 <frosch123> i was surprised you got it so fast, and did not need lord's compile-old-stuff patches
14:24:36 <TrueBrain> I am just that good :P
14:25:01 <frosch123> anyway, i thought downlaod via http was added later
14:25:07 <frosch123> is that already in 0.7? :o
14:25:26 <TrueBrain> but as far as I know, it was also never changed
14:25:31 <TrueBrain> so if 1.10 works .. pretty sure the others work too :D
14:25:44 <TrueBrain> 0.7 uses an older protocol version, so that was the one I was most worried about
14:25:48 <LordAro> TrueBrain: i'm pretty sure no one would consider you "incredibly lazy" after the effort you've been putting in
14:26:04 <TrueBrain> pretty happy with it
14:26:19 <frosch123> lazy programmers are the best :) they code exactly to specs, with no extras :)
14:26:23 <TrueBrain> might be good to add: the whole OpenTTD infrastructure uses DNS RoundRobin (with some geo-location, but that is not important)
14:26:26 <TrueBrain> the TTL is 60 seconds
14:26:38 <LordAro> speaking of old compiles, i noticed i have clang 10.0 now
14:26:38 <TrueBrain> so if some infrastructure restarts or what-ever, it can take up to 60 seconds before it works again
14:26:57 <TrueBrain> just something to keep in mind when dealing with user-reports
14:27:11 <mcbanhas> One question, is it me or OpenTTD does not have anti-aliasing for fonts?
14:27:24 <TrueBrain> (the LoadBalancer should be fairly stable, but you never know ;) )
14:27:41 <milek7> you can enable aa for vector fonts
14:27:43 <TrueBrain> and everything is auto-healing; a big step up from Digital Ocean :P
14:27:52 <TrueBrain> so .. bugs .. lets fix bugs ..
14:28:05 <_dp_> mcbanhas, imo looks better than default but worse then dejavu
14:28:33 <_dp_> idk how it's called properly but imo openttd should use font made for reading, not a fancy one
14:29:21 <milek7> window buttons looks oversized
14:30:06 <_dp_> milek7, if you're talking about my screenshots keep in mind I have 4k 27" (160 dpi)
14:30:56 <TrueBrain> SpComb: it takes 300 CDK lines, 120 + 150 Python Lambda lines of code to write your own NLB :P
14:31:46 <milek7> I'm wondering whether to buy 4k 24"
14:31:48 <milek7> not sure if this good idea :D
14:32:20 <TrueBrain> I take 2k 27" over 4k 24" any day
14:32:30 <_dp_> I'm having the most trouble with dual monitor mixed dpi setup on linux
14:32:50 <_dp_> and some programs that use gui libs that can't scale
14:33:14 <DorpsGek_III> [OpenTTD/nml] LordAro requested changes for pull request #104: Doc: draft changelog for 0.5.0 release https://git.io/Jftkw
14:35:36 <mcbanhas> _dp_: This particular font is the official font Apple used in its OS for many years, and it's somewhat similar to the original TTD font. I was wondering how it would look on a higher resolution and if we should create something similar in TrueType.
14:36:51 <mcbanhas> I should still contact Zephiris sometime to ask him in which TrueType he based the current bitmap font on.
14:37:12 <glx> mcbanhas: IIRC tahoma is close to the original font
14:40:15 <andythenorth> not sure what GH will do if I accept the suggestions
14:40:20 <andythenorth> we can probably squash that later
14:40:33 *** WormnestAndroid has quit IRC
14:40:46 *** WormnestAndroid has joined #openttd
14:42:00 <LordAro> andythenorth: makes a lot of noise here :p
14:42:13 <andythenorth> we should get a dev channel
14:42:16 <TrueBrain> it also seems that 1 push gives 2 lines, which is weird
14:42:48 <andythenorth> these are good changes, I'll do them after I feed the kids
14:44:02 <FLHerne> TrueBrain: They're also all the identical link
14:44:06 <TrueBrain> wow .. DorpsGek received 100+ events in the last 10 minutes :D
14:44:15 <TrueBrain> FLHerne: that is normal, that is what the shortener is doing
14:44:30 <TrueBrain> same long URL results in the same short URL
14:44:35 <FLHerne> Possibly DorpsGek could suppress exactly-duplicate lines
14:44:50 <TrueBrain> nah, andythenorth was really pushing so many updates
14:44:52 <TrueBrain> so that is more than fine
14:45:00 <glx> andythenorth: dev channel exists but rarely useful :)
14:45:11 <andythenorth> I forgot the irony emoji :)
14:46:31 <FLHerne> TrueBrain: He was, but all in one go, hence the comment/update URLs being identical
14:46:49 <TrueBrain> I don't think he did it in one go tbh :P
14:46:53 <FLHerne> We don't really need to be told to go and look at the same link six times :P
14:46:59 <TrueBrain> these were individual push events
14:47:10 <glx> you can commit suggestions in one go with github interface
14:47:12 <TrueBrain> just because something happened once in ... 2 years, doesn't mean we have to act on it ;)
14:48:49 <LordAro> it's definitely happened more than once
14:49:27 <TrueBrain> but with the check_run events GitHub hooks emits, it is impossible to see what GitHub was doing exactly; which is funny to me :)
14:49:38 <TrueBrain> every CI that picks up the phone, is reported to DorpsGek_III :P
14:50:17 <andythenorth> so after nml and bananas, can we stop doing big things for a bit? :)
14:51:18 <andythenorth> funny, I am reading about cmake now :)
14:51:32 <TrueBrain> now that scares the shit out of me
14:51:54 <andythenorth> unless I badly misunderstand, cmake doesn't need ogfx, nml, newgrf docs, bananas or website updated
14:52:40 <glx> it just replaces our custom makefile generator
14:54:14 <andythenorth> hmm, sometime we might want a newgrf translator again
15:07:00 <TrueBrain> frosch123: I keep hitting issues with your api_call. When the API returns a 401, things don't really recover. redirect() returns a flask response, but this is not returned from api_call (mostly because the caller is not expecting this). Not sure what I can do to fix that
15:07:20 <TrueBrain> a function returning different output depending on a flag are always hard to deal with for me :(
15:16:28 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain opened issue #39: Allow undoing metadata overwrite on version level https://git.io/JftIi
15:19:29 <frosch123> TrueBrain: ah, yes, that's one of the differences between bottle
15:19:36 <frosch123> (eints) and flask (bananans)
15:19:36 *** mcbanhas_ has joined #openttd
15:19:43 <frosch123> in bottle, redirect throws
15:19:52 <TrueBrain> I am a bit surprised flask does not, tbh
15:19:52 <frosch123> so, i always mess that up :)
15:20:07 <TrueBrain> the world becomes a lot simpler if returns throw in these kind of applications :D
15:20:29 <mcbanhas_> I want to see how this one performs in smaller sizes
15:20:47 <mcbanhas_> I don't think it will be fitting for larger ones though
15:22:16 <mcbanhas_> Chicago will not do unfortunately, it is too streamlined
15:28:04 <TrueBrain> frosch123: it is correct that you didn't make the errors pretty on the edit pages?
15:28:43 <frosch123> yes, i just dump the json
15:29:01 <frosch123> i would need to know more about the structure
15:29:28 <TrueBrain> I hope the structure is pretty self-explaining :D
15:29:29 <frosch123> it looked complicated when i added them
15:29:32 <milek7> _dp_: eww, comparing fonts on jpg screenshots ;D
15:29:59 <TrueBrain> it is a dict, with the key of the field that failed validation, and the value is (most of the time) a list of errors
15:30:11 <TrueBrain> the only exception are where the values are lists
15:30:19 <TrueBrain> there it shows an extra layer of which element of the list failed
15:30:27 <TrueBrain> {'url': ['Longer than maximum length 95.']}
15:31:34 <milek7> this is still recompressed jpg ;p
15:31:40 <milek7> imgur started to skimp on storage
15:31:46 <_dp_> size is a bit different but overall looks same as dejavu to me)
15:31:48 <LordAro> imgur always uses jpg, the extension is irrelevant
15:32:15 <LordAro> imgur has always used a massive amount of compression
15:32:54 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain opened issue #20: No link to upload new package from the package overview https://git.io/JftLv
15:33:56 <mcbanhas_> _dp_: can you send me a pic of the main menu?
15:36:41 <FLHerne> mcbanhas_: What do you mean by 'too streamlined'?
15:37:55 <FLHerne> IMO the Chicago screenshot looks, at first glance, very much like a vector-ised version of the classic pixel font
15:38:26 <FLHerne> With the very rectangular letter shapes, and unusual weights
15:38:37 <mcbanhas_> There are plenty of differences though
15:38:38 <FLHerne> None of the others are even remotely similar
15:38:44 <mcbanhas_> the capitals are much narrower
15:39:09 <FLHerne> True, but I'm not convinced that's really important
15:39:16 <mcbanhas_> The edges also feel a bit too sharp
15:39:23 <FLHerne> It's always going to look different in detail, or there's no point :P
15:40:33 <FLHerne> Chicago fits in with the angular pixel-based UI and preserves the general feel of how it looked before
15:41:07 <FLHerne> Short of drawing a custom vector font based on the OpenGFX one, I'd be surprised if you can improve on it much
15:41:20 <mcbanhas_> Yeah that's what I thought, but I'm not completely satisfied with it yet
15:41:59 <Samu> yay 224 ticks, another breakthrough
15:42:39 <DorpsGek> Samu: 0.722580645161
15:42:57 <mcbanhas_> _dp_: your last screenshot is weird though, you sure you got the right font? It seems much bolder than the one in the link I sent.
15:43:26 <mcbanhas_> Does OpenTTD sets TrueType fonts in bold by default or something like that?
15:47:15 <andythenorth> loads of spam incoming
15:47:19 <andythenorth> I'd leave the channel TBH :P
15:47:40 <frosch123> i'll go for walk then :)
15:47:46 <_dp_> mcbanhas_, you linked the bold font, no?
15:47:52 <_dp_> I don't even see a regular version
15:48:01 <glx> andythenorth: replying to code comments ?
15:48:36 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on pull request #104: Doc: draft changelog for 0.5.0 release https://git.io/JftLc
15:49:00 <glx> andythenorth: you can reply to all doing a review ;)
15:49:03 <andythenorth> I get highlighted by each one, which means my irc icon bounces :P
15:49:06 <mcbanhas_> CMU is on the right?
15:49:07 <_dp_> openttd does shadow though
15:49:20 <_dp_> wonder if does it even for black font
15:51:10 <TrueBrain> frosch123: do dependencies work in the frontend? When I type something, I get no error and nothing happens? (just checking :D)
15:52:49 <_dp_> guess not: draw_shadow = fc->GetDrawGlyphShadow() && (colour & TC_NO_SHADE) == 0 && colour != TC_BLACK
15:52:58 <frosch123> TrueBrain: yes, you are supposed to copy&paste the contentid from the dep site
15:53:07 <TrueBrain> but no error when you did it wrong?
15:53:17 <frosch123> iirc there is an error, but only once
15:53:24 <FLHerne> TrueBrain: Apropos of nothing, kudos on the Github move and all the tooling you've set up
15:53:30 <TrueBrain> haven't seen it once, and I tried adding a bunch of shit :P
15:53:36 <frosch123> it will clear the invalid entries on first submit
15:53:44 <frosch123> well, who knows, maybe it got broken
15:53:45 <TrueBrain> frosch123: I am editing
15:53:55 <FLHerne> The whole OpenTTD-and-related development process is so much nicer now
15:53:55 <TrueBrain> k, will make an issue; not that important atm :D
15:54:13 <TrueBrain> that was the goal; happy it is delivering :D
15:54:47 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain opened issue #21: During editing, wrong dependencies don't result in an error https://git.io/JftL0
15:56:10 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on pull request #104: Doc: draft changelog for 0.5.0 release https://git.io/JftLu
15:56:54 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain commented on issue #35: /package/<content-type>/<unique-id>/<upload-date> gives 500 for unknown unique-id/upload-date https://git.io/JfIdw
15:58:21 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain commented on issue #35: /package/<content-type>/<unique-id>/<upload-date> gives 500 for unknown unique-id/upload-date https://git.io/JfIdw
15:58:54 <andythenorth> 'Added Visual Studio Editor Script' is syntax highlighting support?
15:59:43 <mcbanhas_> CMU on the right I suppose? Something seems off, the Y shape and other letters dont seem to match
15:59:58 <_dp_> yeah, should be on the right
16:00:36 <mcbanhas_> that's weird, bc i looks more like the one on the left
16:01:15 <andythenorth> LordAro wrt to removing all codechanges, are any relevant to maintainers? Otherwise I remove...
16:01:21 <andythenorth> nearly done here
16:02:10 <_dp_> mcbanhas_, yaeh, looks wrong, I'll check
16:03:03 <mcbanhas_> Yeah I think you're loading Arial or liberation serif on the right
16:03:05 <TrueBrain> there we go, first bunch of fixes :)
16:03:48 <TrueBrain> and the mandatory: run fucking black before push ..
16:05:35 <TrueBrain> completely murdered the regression-set :D Haha :D
16:06:08 * andythenorth hopes for nml release today :D
16:06:12 <andythenorth> was there a deadline?
16:06:30 <TrueBrain> personal sanity is also a deadline
16:07:51 <andythenorth> I have many amusing other things to do :)
16:07:55 <andythenorth> dunno about the rest of you
16:08:32 <andythenorth> including cleaning the house :P
16:09:08 <Samu> 283 ticks, it's less than 310, but much more than 224
16:12:03 <_dp_> dbg: [freetype] Font "/usr/share/fonts/truetype/liberation/LiberationSerif-Regular.ttf" misses no glyphs
16:12:10 <_dp_> what's that supposed to mean?
16:12:53 <TrueBrain> you should always miss glyphs :P
16:15:52 <LordAro> fontdetection.cpp:627
16:15:55 <LordAro> just a debug message :p
16:16:32 <TrueBrain> frosch123: hihi, you can add a dependency to yourself :D That is funny :)
16:17:34 <glx> can lead to interesting failure
16:18:26 <_dp_> but for 'CMU Sans Serif Bold' it does Unable to use 'CMU Sans Serif Bold' for medium font, FreeType reported error 0x1, using sprite font instead
16:19:38 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain opened pull request #22: Fix #21: invalid dependencies weren't shown to the user https://git.io/JftLj
16:19:48 <TrueBrain> frosch123: I think I found the typo why I didn't see any error ^^ :D
16:23:03 <TrueBrain> right; so main task left for me is to make sure the API and Server reload the GitHub content when there is a change on GitHub. Shouldn't be too hard :)
16:24:42 <Samu> 310 ticks 2.56 MB low frame time, 283 ticks 21.00 MB low frame time, 224 ticks 3.21 MB very high frame time
16:25:01 <Samu> 3 ways to do the same thing, I have to decide which one to keep
16:26:09 <_dp_> ok, idk what's wrong with cmu font but openttd doesn't want to use it at all
16:26:57 <_dp_> and somehow reports no apparent error
16:29:17 <Samu> 224 ticks is so desirable
16:37:06 <Eddi|zuHause> there's a possibility openttd tests for certain glyphs being available, and silently rejects the configured font
16:37:32 <Eddi|zuHause> i've never looked at how that actually works
16:46:19 <_dp_> interesting, I can use cmu font for small or large but not for medium
16:46:28 <_dp_> mono fonts seems to somehow fuck it up
16:48:13 <TrueBrain> thar, fixed the 10-or-so people who had invalid URLs in on their content :)
16:48:23 <TrueBrain> really surprised only 10-or-so had them wrong :P
16:48:30 <Samu> with PR #8091, it can go as low as 184 ticks!
16:48:56 <Samu> something that some days ago took 403 ticks, can go as low as 184 ticks
16:55:30 <_dp_> ok it fails because of missing glyphs but CheckForMissingGlyphs definitely misses error report
16:57:27 <Eddi|zuHause> _dp_ so when can we expect your PR? :p
16:58:52 <_dp_> I still don't quite understand tf is going on with font loadding xD
17:22:27 <_dp_> I give up, this shit is too broken :p
17:24:21 *** Wormnest_ has joined #openttd
17:25:06 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 approved pull request #22: Fix #21: invalid dependencies weren't shown to the user https://git.io/JftqE
17:25:16 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 merged pull request #22: Fix #21: invalid dependencies weren't shown to the user https://git.io/JftLj
17:25:17 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 closed issue #21: During editing, wrong dependencies don't result in an error https://git.io/JftL0
17:25:46 <_dp_> whole glyph check and fallback logic makes no sense, should be rewritten from scratch imo
17:26:33 <Eddi|zuHause> _dp_ so when can we expect your PR? :p
17:26:38 <frosch123> _dp_: did you already encounter the deadlock crash?
17:27:00 <_dp_> frosch123, lol what, it can do that too?
17:27:19 <frosch123> if you start ottd with no baseset, it will auto-pick a font to display the download window. if loading that font fails, it will fall-back to the sprite font
17:28:23 <_dp_> didn't encounter that but I'm not surprised
17:29:00 <_dp_> I just found out that it seemingly can miss glyphs in one font but reset everything but that font
17:32:41 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
17:34:54 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 opened issue #23: Map enums from API to nice looking user-readable values. https://git.io/Jftqd
17:38:34 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 opened issue #24: Handle multiple requests in parallel https://git.io/JftqN
17:53:09 *** mcbanhas has joined #openttd
17:53:30 <mcbanhas> _dp_, I believe the font has a problem
17:53:50 <mcbanhas> It aint working on my system for some reason, even if I try to load it on Libreoffice
17:54:09 <mcbanhas> So it's likely not an issue with OpenTTD
18:18:21 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain closed issue #20: GitHub commit message can have duplicated entries https://git.io/JfTE3
18:18:31 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain closed issue #39: Allow undoing metadata overwrite on version level https://git.io/JftIi
18:18:33 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain closed issue #35: /package/<content-type>/<unique-id>/<upload-date> gives 500 for unknown unique-id/upload-date https://git.io/JfIdw
18:18:49 <DorpsGek_III> [OpenTTD/bananas-api] TrueBrain closed issue #34: /package/<content-type>/<unique-id> gives 200 for unknown unique-id https://git.io/JfIdz
18:18:53 <frosch123> two clicks, so much spam :)
18:18:57 <LordAro> TrueBrain: are you trying to get DorpsGek_III kicked for spam? :p
18:19:03 *** berndj-blackout has quit IRC
18:19:05 <TrueBrain> it was just an efficient PR :)
18:20:24 <TrueBrain> we seem to be weeding out most of the annoying bugs :D
18:21:31 <TrueBrain> some automation is still missing, OpenTTD login-flow, and CSS of course, but .. otherwise .. it seems .. this is at least equal in quality to the current BaNaNaS :P
18:38:47 <andythenorth> LordAro done I think, I squashed all the GH commits and dropped 'draft' from commit message
18:39:05 <TrueBrain> you are such a spammert
18:40:09 <TrueBrain> andythenorth: that one change in the middle of features is triggering me
18:40:30 <andythenorth> this whole approach is weird tbh
18:40:36 <andythenorth> but who am I to argue
18:40:44 <TrueBrain> the one writing it? :P
18:40:53 <andythenorth> it should be grouped into concerns of the author
18:41:33 <TrueBrain> you can spend hours and hours on the changelog; not sure it is worth your time :)
18:42:01 <andythenorth> I will keep going until someone approves :P
18:42:13 <frosch123> + <tr class="{{ loop.cycle('odd', 'even') }}"> <- magic :o
18:42:17 <DorpsGek_III> [OpenTTD/OpenTTD] DerDakon commented on pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/JftOf
18:42:30 <andythenorth> frosch123 that's not magic, that's good magic
18:44:22 <TrueBrain> so, hmm .. OpenTTD logins .. hmm ..
18:45:23 <TrueBrain> I wanted to piggyback on the existing Django, but .. I hope we can disable that in a few weeks or so ..
18:45:37 <TrueBrain> so ... export username + sha from LDAP and store it somewhere safe?
18:45:59 <dwfreed> or just auth against ldap?
18:46:08 <andythenorth> hmm, seems I can approve my own PR in nml
18:46:28 *** Progman has joined #openttd
18:46:28 <TrueBrain> it is not possible to reach the LDAP from outside the old infrastructure; and I don't want to depend on the old infra :)
18:47:11 <TrueBrain> that is what I say: export username + sha and store somewhere safe :)
18:47:31 <TrueBrain> (sha = hashed password, to be clear)
18:47:57 <dwfreed> I mean literally just run an ldap server in the new infra
18:48:24 <TrueBrain> LDAP is a bitch to configure; and if it is only for allowing deprecated account access, I don't see the point
18:48:32 <TrueBrain> I want to move forward, basically ;)
18:49:06 <TrueBrain> I want that people can go <somewhere>, proof they are owner of OpenTTD account, and "link" it to their GitHub
18:49:17 <TrueBrain> most likely just changing the author of all BaNaNaS entries to their GitHub account
18:49:24 <TrueBrain> after that, it can forget about the OpenTTD account ever existing
18:49:30 <dwfreed> then yeah, just shove it in a table in the database
18:49:38 <TrueBrain> there is no database atm
18:49:54 <dwfreed> make a sqlite one, then :)
18:50:21 <TrueBrain> there is no persistent storage attached to any pod :)
18:50:22 <LordAro> TrueBrain: thoughts on what to do with the wiki?
18:50:41 <TrueBrain> LordAro: not 100% sure, but my suggestion would be to make authentication via GitHub there too
18:50:51 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 opened pull request #25: Change: Apply the style from the website. https://git.io/JftOB
18:50:53 <TrueBrain> good thing is, that we don't really have to migrate accounts for wiki, as .. it is a wiki :P
18:50:57 <dwfreed> so add some? or use the managed DB, though that's probably significantly more expensive
18:51:08 <frosch123> andythenorth: another chance to approve your own stuff :)
18:51:32 <TrueBrain> dwfreed: I am happy you consider it all this easy, but this requires a bit more thought first :) Just attaching stateful disk to a pod is not easy, but also a bitch to maintain
18:51:47 <TrueBrain> I rather find a solution that is elegant, maintaince free, and easy to understand :)
18:52:32 <TrueBrain> not overly happy we are serving static files via flask, but .. what can you do :D This does need a solution .. just not now :P
18:53:15 <DorpsGek_III> [OpenTTD/bananas-frontend-web] TrueBrain approved pull request #25: Change: Apply the style from the website. https://git.io/JftOg
18:53:46 <TrueBrain> I guess it is not okay to email everyone a "code" :D
18:53:56 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 merged pull request #25: Change: Apply the style from the website. https://git.io/JftOB
18:54:29 <TrueBrain> so eints will go via GitHub, needs no migration. wiki via GitHub, no migration. Only BaNaNaS is the annoying one ..
18:54:41 <TrueBrain> nothing else needs authentication I think
18:54:58 <frosch123> TrueBrain: we can avoid storing the pw. if we have the email, people could enter their email somehwere, then then they get an email with a take-over link
18:54:59 <LordAro> wiki *could* have migration
18:55:05 <LordAro> if you tried hard enough
18:55:18 <TrueBrain> LordAro: you can do a lot of things. But there is no point, is there?
18:55:31 <TrueBrain> frosch123: I rather store hashed passwords than plain-text emailaddresses :)
18:56:16 *** snail_UES_ has joined #openttd
18:56:19 <TrueBrain> LordAro: is an honest question btw; there is no point, is there?
18:56:26 <frosch123> people forget their pw all the time. emails allow manual verification
18:56:31 <TrueBrain> we have no pages restricted to a certain user or something, right?
18:56:44 <TrueBrain> frosch123: pretty sure we have more people that no longer have a valid email on file
18:56:47 <LordAro> TrueBrain: i guess not, other than losing user pages & history
18:56:56 <TrueBrain> you don't lose history at all
18:57:05 <LordAro> properly linked history*
18:57:12 <TrueBrain> nothing changes in the history
18:57:15 <TrueBrain> not sure why you would think that
18:57:20 <TrueBrain> it is not that we remove the usernames that are in there
18:57:36 <LordAro> i mean that github users do not get their changes linked with their old accounts
18:57:56 <TrueBrain> ah, yes, they do not. But .. given the amount of changes on our wiki, pretty sure nobody gives a flying rat about that :P
18:58:39 <TrueBrain> frosch123: as you currently cannot change your email nor your password, I have noticed more people have an invalid email than those that forgot their password :P
18:58:40 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 opened issue #26: Instrument edit/upload forms with labels https://git.io/JftOS
18:59:18 <TrueBrain> but okay, I guess the cleanest solution for both wiki and BaNaNaS would be to make a small pod that acts as an OAuth provider, based on the old credentials
19:00:01 <frosch123> we can now have private repos, right?
19:01:27 <TrueBrain> for BaNaNaS we can always fix it for users if they ask, I mean .. we can commit in the repo
19:01:37 <TrueBrain> nice job guys; seriously, nice job :)
19:01:50 <frosch123> i'll do some fixes to the header now
19:01:54 <andythenorth> you didn't look in the html or css :D but thanks TrueBrain
19:01:56 <frosch123> like actually make the highlight work and stuff
19:02:00 <TrueBrain> being logged in breaks :P
19:02:07 <TrueBrain> andythenorth: I skimmed it
19:02:14 <andythenorth> the header needs reworked
19:03:09 <TrueBrain> It is better than what it is now :P
19:03:41 <andythenorth> frosch123 I'd split manager, logout, home, and username onto a row first
19:03:45 <andythenorth> and see how that helps
19:04:04 <andythenorth> then I'd ask questions about libraries and end users
19:04:28 <frosch123> andythenorth: no idea what you mean, but we have a restriction here: we want to make the index pages cacheable, so they should not look different when logged in
19:04:53 <andythenorth> edge side includes :P
19:04:58 <andythenorth> yes ok, that's reasonable
19:05:24 <_dp_> can someone rewrite openttd in rust pls? c++ too broken %)
19:05:26 <andythenorth> but they won't always show username?
19:05:38 <TrueBrain> maybe give Manager a whole different header
19:05:46 <TrueBrain> so it is also more clear you are in the manager part
19:05:52 <andythenorth> I'd consider moving it right up above everything else
19:06:08 <andythenorth> then you really could do something like ESI trivially and varnish the rest
19:06:17 <TrueBrain> anyway, I got out of the UI/UX industry, kinda want to stay out :P
19:06:36 <andythenorth> I seem to sadly be in the security industry somehow TrueBrain :(
19:06:43 <FLHerne> Or is it just my browser that's silly?
19:06:49 <TrueBrain> it is a nice industry andythenorth :) Lot of stupid people :P
19:06:50 <andythenorth> FLHerne that's lolz
19:07:00 <frosch123> FLHerne: no, for me "manager" is in a new row :p
19:07:04 <FLHerne> (I did clear the cache)
19:07:34 <mcbanhas> Guys is it possible to make a tooltip window not show up at all? There's a couple of redundant tooltips I wanted to remove altogether, but if I just clear the string in the language file, it still shows a small empty tooltip window.
19:07:36 <frosch123> (yes, nav header is broken)
19:09:05 <frosch123> he, why did gh attribute the commit to me. i spend so much work looking up --author
19:09:57 <TrueBrain> that is the mistake :D
19:10:12 <TrueBrain> squashing means GitHub is recommiting the commit, and it does that as your user
19:10:30 <frosch123> huh, i always squash to add the PR to the commit message
19:10:38 <TrueBrain> it is a good thing to do
19:10:43 <TrueBrain> just .. not if you want another author :P
19:10:55 <TrueBrain> on the plus side, your commit is now "Verified"
19:11:20 <andythenorth> frosch123 where should I base branches from / push to now?
19:11:22 <frosch123> whatever that means :)
19:11:40 <frosch123> andythenorth: what do you want to do?
19:11:49 <andythenorth> help sort out the nav
19:11:57 <andythenorth> can't do it right now, much chores
19:12:10 <andythenorth> but I might get to it later, if no-one else does
19:12:10 <frosch123> ok, enjoy some free hours, until i fix some stuff there :)
19:12:30 <andythenorth> if $someone could put form labels around all the form inputs, with ids
19:12:34 <andythenorth> that would be very good
19:13:07 <FLHerne> What browser support does the CSS care about?
19:13:39 <andythenorth> there's no written policy
19:13:55 <andythenorth> I would think IE 11, and the evergreen versions of the others
19:14:03 <LordAro> "anything newer than IE11"
19:14:12 <LordAro> iirc IE11 doesn't work well with the website
19:15:09 <LordAro> frosch123: TrueBrain: "verified" means that your commit was GPG signed, which website commits can do automatically
19:15:28 <LordAro> PGP? GPG? i can never remember which is which
19:16:13 <TrueBrain> and what does it mean, I mean .. GitHub did it for you .. on GitHub .. so .. what does signing mean there? That it was done by GitHub?
19:16:18 <andythenorth> apart from "Firefox can't size inputs sensibly" I haven't written browser specific CSS for years
19:16:22 <TrueBrain> what value does it add? :P
19:16:28 <andythenorth> but I usually start from a very different place :P
19:16:28 <Eddi|zuHause> LordAro: i never bothered learning whether there is even a difference between these two
19:20:28 <mcbanhas> Anyone could help me with my question above?
19:22:25 *** sheepo99 has joined #openttd
19:23:22 <andythenorth> are we supporting non-desktop browsers? o_O
19:23:35 <_dp_> mcbanhas, my guess is that removing tooltips is only possible by patching the code but idk rly
19:25:28 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
19:25:43 <_dp_> mcbanhas, btw, are you sure those tooltips are redundant in all languages?
19:26:31 <mcbanhas> It's the tooltips for "Start/stop replacing engine" buttons.
19:26:49 <mcbanhas> The tooltip says the same thing by different words.
19:29:00 <_dp_> probably there is also something to be said about ui consistency
19:29:19 <_dp_> if one button has tooltip an other doesn't it may seem to be broken
19:46:06 <FLHerne> andythenorth: Hrm, I think "Industries: support 16 cargos in / 16 cargos out" in the changelog isn't enough info
19:46:35 <FLHerne> That includes completely removing (not even deprecating) waiting_cargo_1/2
19:46:59 <FLHerne> And a bunch of other industry stuff I don't understand :P
19:47:52 <FLHerne> Anyone compiling an industry-related grf with nml 0.5 will find it breaks in a number of ways, and "support 16 cargos in" doesn't really cover that
19:48:31 <FLHerne> or tell them what needs to be done
19:49:09 <glx> 0.5 is incompatible with 0.4.5
19:49:41 <glx> and it's documented on the wiki IIRC
19:50:35 <FLHerne> Ok, then the changelog should say that, and where to find the documentation
19:58:24 *** ChanServ sets mode: +v tokai
19:58:34 <andythenorth> FLHerne if this was my project I'd write a section about 16 cargo industry support
19:58:42 <andythenorth> and I'd attach houses to the end of that as a line
19:58:49 <andythenorth> and I'd write a section about NRT
19:58:54 <andythenorth> and a section about railtypes changes
19:59:03 <andythenorth> and then lump everything else together
20:01:04 <FLHerne> That sounds like a sane way to do it
20:18:35 <andythenorth> FLHerne I was going to say can you comment on the PR, but then you did
20:18:49 <andythenorth> can you rearrange it into sections though? o_O
20:18:56 <FLHerne> I could make up another comment :P
20:19:10 <andythenorth> I write newgrf changelogs so that I don't have to also write release notes
20:19:16 <FLHerne> This is the first release on GitHub, isn't it?
20:19:47 <FLHerne> Might as well start a better tradition
20:47:01 <frosch123> meh, i have an import recursion between helpers and session :/
20:47:55 <TrueBrain> means you should split it over more files :)
20:49:17 <TrueBrain> okay, this is tricky .. when the API commits to GitHub, GitHub does a callback to reload the index (as it has changed)
20:49:22 <TrueBrain> not sure that is wanted
20:50:04 <TrueBrain> guess the GitHub Action could check if Librarian has made the commit, and not give the callback
20:50:06 <TrueBrain> hmm, that would work
20:51:36 <TrueBrain> lol; because of a recent change I made, it also forgets in that case all pending changes :D
20:51:40 <TrueBrain> that sounds sub-optimal
20:52:44 <frosch123> it also sounds lucky that you found it :)
20:52:48 <TrueBrain> chances are very very slim, ofc, but not zero :P
20:53:01 <TrueBrain> okay, I need to rework this a bit
20:53:08 <TrueBrain> that sounds like a job for tomorrow :P
20:53:16 <TrueBrain> still not sure how to do the OpenTTD login stuff
20:53:39 <frosch123> can we handle it similar to the translator applciation?
20:56:20 <frosch123> hmm, i guess i misremembered how that worked :)
20:56:46 <TrueBrain> too bad; would love to hear ideas, as mine are running a bit thin :D
20:57:08 <frosch123> anyway, main question is: rely on pw, email or offer both?
20:58:02 <frosch123> store either or both in a private repo
20:58:16 <TrueBrain> what I would like: you login under OpenTTD account, you are asked to make a GitHub account, we move accounts over, and we remove the OpenTTD account
20:59:35 <TrueBrain> I guess my main question is: how long are we going to support old accounts?
20:59:39 <frosch123> add an api call to add a certain <authmethod2>-id to all packages with some <authmethod1>-id
20:59:56 <TrueBrain> migration of the old old OpenTTD system started in .. 2009? And I know at least 1 account migrated in 2017 :P
21:00:01 <frosch123> it's actually similar to adding more authors
21:00:13 <frosch123> just that they are assigned to the same display-name
21:00:24 <TrueBrain> the API can already handle it, yes
21:00:42 <TrueBrain> not the rewrite, but yeah, that is easy
21:01:22 <TrueBrain> ideal, we don't add OpenTTD code to the API. Keep it clean, etc
21:01:33 <TrueBrain> so when we migrate someone, we commit in the BaNaNaS repo, the API sees the new commit, reloads
21:02:55 <TrueBrain> guess it does require us knowing a bit what we want with the wiki
21:02:58 <TrueBrain> there are several user pages
21:03:08 <frosch123> i wanted to use the api for "add-user"
21:03:34 <TrueBrain> what would "add-user" do?
21:03:44 <TrueBrain> or do you mean "add author", to add another author to your package?
21:05:13 <frosch123> hmm, i guess 1) write some oauth-provider with the old logins 2) allow login to the frontend/api with openttd account 3) add a method to add github user xyz to all my packages (or one package at a time may be enough)
21:05:38 <TrueBrain> my issue with that flow is that it is kinda optional to move away from the OpenTTD account
21:05:43 <TrueBrain> I want to promote that a bit .. stronger :)
21:05:51 <TrueBrain> (as you cannot change password or email on OpenTTD)
21:07:00 <frosch123> ok, so another github app, where you login with github, and then enter your old credentials
21:07:20 <TrueBrain> yeah, a "link your OpenTTD account" app
21:07:58 <TrueBrain> but it would be weird if someone has to do this now for BaNaNaS, and later for wiki
21:08:09 <frosch123> can we send verification emails? or would the sender appear invalid and get fitlered out somewhere?
21:08:38 <frosch123> TrueBrain: i checked the oauth-plugin for wiki 2 (?) years ago
21:08:49 <TrueBrain> most emails we have on file are no longer valid, so I am not a fan of adding verification emails here .. that would give me a lot of work :P
21:08:50 <frosch123> usernames are just stored as display-name
21:09:00 <frosch123> there are no user-ids or something
21:09:12 <TrueBrain> so how do we avoid people "claiming" someone else's username?
21:09:31 <TrueBrain> say your name is now A, and there is no GitHub account A yet .. someone else can claim your name
21:09:37 <frosch123> most people also have forum accounts
21:09:37 <TrueBrain> (and take over your user-page)
21:10:04 <frosch123> for the wiki? i would just not bother
21:10:22 <frosch123> i am not going to migrate frosch to frosch123
21:10:22 <TrueBrain> that is what I am on the fence about .. those darn user pages
21:10:44 <TrueBrain> so how bad is it if another person can take control over your wiki "identity", basically
21:10:51 <TrueBrain> in terms of edits, don't care
21:10:56 <TrueBrain> but those few userpages, can be a bit nasty
21:11:09 <frosch123> it would take a long time for people to notice if you made the wiki read-only
21:11:52 <Samu> question, why is that AddList is so slow
21:12:07 <Samu> i mean, i'm adding a list to an empty list
21:12:16 <Samu> can't that be speeded up?
21:12:21 <frosch123> so unless the wiki magically exposes the email of the old owner, i would not bother
21:12:40 <frosch123> and if it does, we should just blank all personal data from the wiki db
21:12:42 <TrueBrain> over 1800 users edited the wiki over the last 20 years :P
21:13:01 <TrueBrain> no, the wiki is linked to the LDAP; it only has the display-name in its database
21:13:03 <TrueBrain> that is not an issue
21:13:11 <frosch123> but only 5 users in the last 5 years :p
21:14:55 <TrueBrain> I am tempted to see if you can lock user pages
21:15:02 <TrueBrain> simply don't allow any mutations to them
21:15:12 <TrueBrain> as then I really do not care what name the system gave you
21:15:38 <TrueBrain> which is fine; anyone can edit that page, so what-ever
21:16:09 <frosch123> the same holds for the user pages?
21:16:17 <TrueBrain> weren't they locked to the user?
21:16:44 <TrueBrain> (I can't tell either way, as I am admin :P)
21:16:56 <TrueBrain> okay, in that case I don't care
21:17:03 <TrueBrain> wiki we just switch OAuth to GitHub, and done
21:17:17 <TrueBrain> that makes BaNaNaS a bit easier; also fewer authors
21:18:15 <TrueBrain> right .. small application which can do OAuth; you go there, it asks you to login to OpenTTD. After that, it asks you to login to GitHub. Migrates your stuff from one user to the other, and sends you to the logout or something. After that, login via GitHub, and lets go
21:18:50 <frosch123> i would reverse the login order. first github (since it may involve sign-up), then ottd
21:19:16 <TrueBrain> not sure it matters, tbh
21:19:21 <TrueBrain> I can see pros and cons for both
21:19:33 <TrueBrain> hmm, maybe not an oauth app, maybe just a small webapp
21:20:15 <frosch123> you mean, people shall just enter their github id, without auth?
21:20:42 <TrueBrain> the frontend is also a webapp; you don't see anyone entering a github id there either :D
21:20:54 <TrueBrain> I was first thinking linking him in the authentication mechanism we have now
21:21:03 <TrueBrain> but we can also just make a link: migrate your OpenTTD account?
21:23:14 <frosch123> don't even think of updating that page :)
21:23:17 <TrueBrain> back in the day where NewGRF was just to manipulate existing things :P
21:24:21 <TrueBrain> I completely forgot I wrote that page, so no, I aint touching that :P
21:24:24 <andythenorth> truebrain has a birthdate :o
21:24:31 <andythenorth> 5 years after me
21:24:32 <TrueBrain> no, a creation date
21:25:06 <frosch123> andythenorth: most dev pages have
21:25:11 <frosch123> noone validated them though
21:26:42 <TrueBrain> okay, so: you login, you get a page: Login via GitHub / Login (Development only) / Migrate your OpenTTD account. You click that last button, it logs you in to OpenTTD and GitHub (in what-ever order). Via the API it adds the GitHub accounts with the OpenTTD session, and removes the OpenTTD accounts with the GitHub session to the API. After that, you continue with the GitHub session
21:26:56 <TrueBrain> next time you try to migrate, it tells you you have no packages under that username, and as such, cannot migrate
21:27:39 <frosch123> or we add a "migrate old ottd account" to the manager page
21:27:48 <frosch123> so you login via gh to the regular bananas frontend
21:28:18 <TrueBrain> the issue I have with these kind of things, that we add stuff to the frontend for migration. I have seen that, that code is never going to be removed. So I was more thinking doing this outside the API / frontend, so those are kept clean of this
21:28:55 <frosch123> that means we need a special migration app registered to gh
21:29:03 <frosch123> and users have to approve both migration and bananas
21:29:26 <TrueBrain> it just hooks into the OAuth flow that is already there, basically :)
21:32:16 <TrueBrain> if we do only username/password for now, it makes it a bit easier. We can see how much people complain about that, and add email later. Emailing is just a bit of a pita, basically.
21:32:23 <TrueBrain> just to scope this a bit :P
21:32:39 <TrueBrain> okay, I will have to draw this out, but this might work tbh
21:33:06 *** gelignite has joined #openttd
21:33:10 <TrueBrain> it is light-weight, it is easy, and it can be removed at any time
21:42:24 <TrueBrain> I hope it makes sense ..
21:42:27 <TrueBrain> it is getting late :P
21:42:40 <Samu> I got a proble. When I do "local list = other_list;" and then add an item to list, why does the other_list also get the item added?
21:42:57 <Samu> feels like they're glued together under 2 different names
21:43:53 <Samu> how do I create a copy of other_list without resorting to "list.AddList(other_list);"?
21:44:01 <Samu> because this method is too slow
21:44:46 <nielsm> yes when you juad do local list = other_list; you're making a reference to the same object
21:47:38 <frosch123> TrueBrain: there are 446 unique authors on bananas, shouldn't we just do "send forum pm with your github username"?
21:48:07 <Samu> there is no faster method?
21:48:11 <TrueBrain> owh, as text on the site
21:48:28 <TrueBrain> no, where someone would read how to do that
21:48:29 <TrueBrain> that was my question
21:48:44 <frosch123> ah yes, just a message on the frontpage
21:49:04 <TrueBrain> 2 assumptions there: 1) that people have a tt-forums account (but they can make one, ofc), 2) that the username matches their OpenTTD account .. how do we validate the validity/
21:49:19 <frosch123> to get access to your old content, message xyz and prove your identity (easy if you have a tt-forums account)
21:49:56 <frosch123> TrueBrain: main assumption: most authors have a tt-forums account, and their content links to a tt-forum thread with them as thread starter
21:50:05 <TrueBrain> it feels like a lot ofhazzle for the user
21:50:15 <frosch123> and no, people are not supposed to create a tt-forums account, if they have none yet
21:50:49 <frosch123> TrueBrain: hazzle for who? the author sending stuff, or us filling the info manually into some script?
21:51:07 <TrueBrain> authors need to do some work, which is of course not user-friendly
21:51:23 <TrueBrain> but our speed of picking up anything community related like this .. is not the best track-record, I would say :D
21:51:59 <TrueBrain> nevertheless, I am okay with it, if we can make it a bit more flexible
21:52:20 <frosch123> it's only 205, if you do not count people with a single upload
21:52:51 <TrueBrain> how about .. 1) your content has an URL to a tt-forums thread; send us a PM with that thread; 2) send as an email from the one you used during registering; 3) ???
21:53:11 <TrueBrain> 2) ofc with your username (NOT YOUR PASSWORD)
21:53:36 <frosch123> it's only 131 if you exclude scenario and heightmaps, which are never updated anyway
21:54:00 <TrueBrain> again, I am not against it. We just have to be aware that it is not really user-friendly. If we accept that, fine by me
21:54:09 <frosch123> TrueBrain: whatever the text is, i think we can handle this manually without an app
21:54:49 <frosch123> an entire app that is at best used 131 times :)
21:54:52 <TrueBrain> hmm .. they should make an issue on the github repo btw, otherwise it will be hard to get their id
21:55:18 <frosch123> pretty sure that id is public
21:57:08 <frosch123> TrueBrain: yep, curl gives me the id for a username without auth
21:57:36 <TrueBrain> I picked some random other dude, I guess :P
21:58:04 <frosch123> otherwise i would not have the 123 :)
21:58:09 <TrueBrain> okay, so we need to figure out the options we are giving them
21:58:57 <TrueBrain> 3000 out of the 5000 have a tt-forums url
21:59:01 <TrueBrain> so that should catch most
21:59:35 <TrueBrain> so let them PM me and you, I would say
21:59:36 <frosch123> 1) register to github, if you haven't already. 2a) contact xyz via tt-forums pm and a link to your github user page. 2b) or send an email to info@ with your original email, and a link to your github user page
22:00:03 <TrueBrain> with 2a) a link to a tt-forums post of any of your content
22:00:09 <TrueBrain> saves us doing any work :P
22:00:55 <TrueBrain> do we have any other means to offer
22:01:16 <TrueBrain> contact us on IRC if neither works out for you, and we will figure it out
22:02:53 <frosch123> 3) sorry for the inconvenience
22:03:00 <TrueBrain> 2b) is easy to validate for us. 2a) we have to be a bit careful, and check if the tt-forums link matches the author. But that should be fine
22:03:21 <TrueBrain> okay; so all we need is a small script that fixes the authors, basically
22:04:07 <Samu> nielsm, I need a faster method than "list.AddList(other_list);", do you know any?
22:04:19 <TrueBrain> I will take care of that tomorrow
22:04:43 <TrueBrain> frosch123: like I wrote, I also want to add a page when you press Manager, that tells you you are about to authenticate with GitHub. Currently it feels too much magic, and it is a bit scary. You agree?
22:04:44 <Samu> faster, in terms of less cpu usage
22:04:48 <nielsm> Samu well why do you need to duplicate the list? and does it need to be a full list or really something else?
22:05:16 <TrueBrain> I will add that tomorrow too
22:05:18 <Samu> I am keeping track of the list of tiles a path uses
22:05:33 <TrueBrain> after that, fixing reload in the API, making the server-reload work
22:05:41 <TrueBrain> add some notes about the 15 minute delay between edit and seeing it ingame
22:05:48 <TrueBrain> and that would empty my list for now :D
22:06:11 <TrueBrain> for now, good night :)
22:06:12 <frosch123> .tar/.zip/... upload is still open :)
22:06:21 <TrueBrain> "optional" to me :P
22:06:41 <TrueBrain> fine fine ... but that can be done after the deadline of the 30th
22:07:06 <frosch123> sure, we just delay people getting their accounts merged, claiming we are busy :)
22:07:17 <TrueBrain> if possible, I would like to aim to go live monday / tuesday. But we should not rush it in the sense that we feel we have to, but delaying any more might also not be useful
22:07:46 <TrueBrain> pretty sure we can spend weeks more making this perfect and shiny etc :P
22:07:56 <nielsm> Samu that doesn't explain why you need to duplicate the list
22:08:03 <Samu> in the AyStar.Path, I keep a list up to this point for each partial-path
22:08:07 <nielsm> is it that you need to backtrack/"undo" steps?
22:08:31 <Samu> i have a path with tiles 1 2 and 3, each of the step has a list of tiles up to that point
22:08:39 <TrueBrain> it does, how ever, require more people testing, tbh :D
22:08:50 <TrueBrain> owh well, we are also good with testing-in-production
22:08:58 <TrueBrain> I just need to make sure the binaries are being backed-up too :P
22:09:11 <TrueBrain> (the pods have no right to delete anything, so that should be fine tbh :P)
22:09:44 <Samu> I'm so bad at explaining things
22:10:02 <TrueBrain> anyway, I am going to watch a movie, so this all tomorrow :) Tomorrow is going to be another good day! Nice progress and work frosch123 :D (and andythenorth ofc :D)
22:10:39 <TrueBrain> from idea to production in ... 2.5 months?
22:10:41 <Samu> when I want to find a collision (a tile that already exists in a path), I found out that caching the list of tiles every step would speed up the pathfinder
22:11:28 <Samu> so each step has a copy of the old list + the current tile that im working on
22:11:39 <Samu> i can't manage to make a copy of the old list
22:12:11 <Samu> well, I can, via AddList, but that method eats cpu like nothing
22:12:22 <Samu> however, it results in much less ticks needed
22:12:22 <LordAro> Samu: there's usually a copy or clone method for these things
22:12:31 <LordAro> squirrel docs will tell you
22:13:37 <Samu> gonna check what I can do with that
22:13:50 <michi_cc> I'm quite sure the Squirrel clone will not work as it doesn't know anything about the internals of our ScriptList class.
22:14:28 <nielsm> yeah but then why does it need to be a ScriptList at all? if it's just a list of tileindexes?
22:14:32 <glx> local list = clone other_list; should work, but maybe it needs us to implement __cloned()
22:14:54 <nielsm> I'm quite sure using a native squirrel array is cheaper
22:17:44 <Samu> the "normal" method, which is the one currently implemented in AyStar.6, resulted in 310 ticks
22:17:57 <Samu> the native squirrel array 283 ticks
22:18:22 <Samu> but heavy cpu, like 10 x more than the others
22:18:30 <nielsm> but now you're saying the AIList is slow to duplicate
22:18:40 <nielsm> so maybe the array is faster after all
22:19:04 <glx> he said cpu intensive, not slow ;)
22:19:47 <Samu> and all I wanted to do was to clone a list
22:20:16 <Samu> this._path_tiles = AIList();
22:20:16 <Samu> if (this._prev != null) this._path_tiles.AddList(this._prev._path_tiles);
22:20:52 <Samu> wanted to do something like: if (this._prev != null) this._path_tiles. = this._prev._path_tiles;
22:21:21 <Samu> if (this._prev != null) this._path_tiles = this._prev._path_tiles;
22:21:51 <Samu> the list is empty to begin with
22:22:05 <glx> do you really need to copy the list ?
22:22:23 <Samu> yes, caches in every step
22:33:18 <Samu> 223 ticks, I gained 1 more tick :o
22:40:52 <Samu> this->modifications++ seems weird, u sure that's correct?
22:42:36 <michi_cc> Yes, it is simply a marker used in Valuate(). The actual value is not important, just that it changes when the list changes.
22:43:42 <Samu> better rebuild solution, visual studio has some weird problems from time to time
22:50:22 <Samu> eh.. not much of a gain, I hoped for better
22:51:18 <michi_cc> It should not gain you any ticks at all, just consume a bit less CPU.
22:53:18 *** tokai|noir has joined #openttd
22:53:18 *** ChanServ sets mode: +v tokai|noir
22:55:10 <Samu> be back later, testing fast forward
22:56:11 <michi_cc> Your tick count is a proxy for number of opcodes used, and native calls count the same (with a few exceptions), not matter if they use more or less real CPU time.
22:58:38 <DorpsGek_III> [OpenTTD/bananas-api] frosch123 opened issue #42: API endpoint to get URL for OAuth access review https://git.io/JftWr
22:59:32 <frosch123> must have been a boring movie :)
22:59:53 <TrueBrain> Nah. Just want to keep the flow going :)
23:05:15 <DorpsGek_III> [OpenTTD/bananas-frontend-web] frosch123 closed issue #20: No link to upload new package from the package overview https://git.io/JftLv
23:12:39 <frosch123> andythenorth: i updated my static branch, you can fork it again :)
23:14:13 <Samu> i'm back, well, in fast forward, it is faster
23:15:33 <Samu> started in 1935 jan, and now it's jun 1938 vs aug 1937, there's some improvement
23:16:35 <Samu> so instead of 10x slower, it's about 6x slower now
23:17:06 <Samu> gonna head to bed, take care
23:34:41 <DorpsGek_III> [OpenTTD/OpenTTD] michicc opened pull request #8097: Codechange: [Script] Improve copying a list into another empty list. https://git.io/Jftl6
23:38:05 <frosch123> is there no LEQ and GEQ?
23:38:21 *** snail_UES_ has joined #openttd
23:38:54 <FLHerne> There are, but I just realized it really shouldn't use them :P
23:41:02 <andythenorth> wrt nml, some weekend work stuff has come up
23:41:16 <andythenorth> it's unlikely I can finish release notes + release
23:41:24 <andythenorth> we are sooo close though :)
23:41:59 <mcbanhas> Guys I'm adjusting the following description text for the wagon removal tool:
23:42:27 <mcbanhas> ORIGINAL: "Make autoreplace keep the length of a train the same by removing wagons (starting at the front), if replacing the engine would make the train longer"
23:42:54 <mcbanhas> EDIT: "Automatically removes one or more wagons at the front of train to maintain length, if replacing with a multiple unit locomotive."
23:43:19 <nielsm> why are you introducing multiple units there?
23:43:37 <mcbanhas> Is it not the issue though?
23:43:57 <nielsm> no, locos can have different lengths while still being a single length
23:44:23 <nielsm> *while being a single unit
23:44:26 <mcbanhas> let me adjust that then.
23:44:30 <andythenorth> is there a separate tooltip for wagons part of auto-replace?
23:45:00 <mcbanhas> "Automatically removes one or more wagons at the front of train to maintain length, if replacing with a longer locomotive."
23:45:35 <nielsm> "Prevent trains from becoming longer due to replacement. This is done by removing wagons from the front of the train, until it's as long as or shorter than it was before the locomotive replacement."
23:45:42 <nielsm> is maybe not good either
23:45:48 <andythenorth> the current tooltip is inaccurate
23:46:03 <andythenorth> it's not just engines, wagons also change length
23:47:00 <andythenorth> "Automatically removes one or more wagons at the front of train to maintain length, if replacement vehicles are longer."
23:47:06 <andythenorth> not quite right but factual
23:47:28 <andythenorth> longer than what?
23:47:38 <nielsm> actually I wonder, does wagon removal do any rounding of train length?
23:47:46 <andythenorth> it's remarkably magical
23:47:50 <andythenorth> or so simple I misunderstand it
23:47:52 <nielsm> e.g. a train changing from length 4.2 to 4.5 would never be a problem in practice
23:48:08 <andythenorth> it works *really* well whenever I use it, including accidentally supporting brake vans
23:48:40 <nielsm> while going from 4.8 to 5.1 would be a problem
23:49:10 <andythenorth> the rounding of train lengths is best not looked at :P
23:49:13 <FLHerne> mcbanhas: "Remove wagons (starting at the front) if the train would become longer otherwise."
23:49:20 <nielsm> and 4.4 to 4.7 usually not a problem, unless you're relying on diagonal signal distances
23:49:23 <andythenorth> one of my kids noticed it doesn't round as a human would expect
23:49:40 <andythenorth> due to integer maths probably, unless it's....just a bug
23:49:49 <FLHerne> andythenorth: I suspect brakevans are *why* it starts at the front?
23:50:01 <andythenorth> it's pretty smart
23:50:10 <andythenorth> it's one of my favourite openttd features
23:50:32 <mcbanhas> FLHerne, thanks, but I'm going with andythenorth's version. It doesn't require brackets and is less confusing.
23:50:34 <FLHerne> andythenorth: The weird rounding is just because of train-lengths being in 16ths
23:51:00 <FLHerne> You can have commas instead
23:51:32 <FLHerne> "Remove wagons, starting from the front, if the train would become longer otherwise."
23:51:39 <FLHerne> I find andy's hard to read, tbh
23:54:18 <FLHerne> Also, "removes wagons *at* the front" is confusing
23:54:24 <FLHerne> Wagons are never at the front :P
23:55:03 <mcbanhas> Well that really depends on you ahah
23:56:35 <FLHerne> "Remove wagons if the train would otherwise be made longer. Wagons closest to the front are removed first."
23:57:03 <mcbanhas> Not bad, but I still don't like "otherwise"
23:57:08 <FLHerne> *would become longer otherwise
23:57:53 <FLHerne> Just "if replacement vehicles are longer" isn't technically accurate
23:58:05 <mcbanhas> what about "Remove wagons from the front"?
23:58:10 <FLHerne> There might be some that are longer than the originals, and some shorter
23:58:28 <FLHerne> Like I said before, I find that really confusing
23:58:39 <FLHerne> Wagons can never be at the front
23:59:30 <mcbanhas> what if you have a train in the depot that happens to have a wagon at the front?
23:59:53 <FLHerne> Then it isn't a train, so far as OTTD is concerned
continue to next day ⏵
