--- Log opened Fri Sep 21 00:00:23 2018 --- Day changed Fri Sep 21 2018 20180921 00:00:23< celticminstrel> Ripping out GUI2 is one thing; I could get behind that. Porting everything to a brand-new engine? That's basically writing a new game from the ground up. 20180921 00:00:35<+wesdiscordbot> Yes. 20180921 00:00:39<+wesdiscordbot> But I feel it's worth it. 20180921 00:01:06< celticminstrel> I can't say with confidence that it definitely wouldn't be worth it. 20180921 00:01:33<+wesdiscordbot> It saves us the trouble of having to deal with HiDPI and the accelerated rendering and the backend OGL/Vulkan business, for example. 20180921 00:02:09<+wesdiscordbot> and it gives us the opportunity to ditch so many of the things that just don't work very well, like WML. 20180921 00:02:21< celticminstrel> FTR, even if we did choose to go to a new engine, I'd still want to keep WML. 20180921 00:02:44<+wesdiscordbot> That's a very bad idea. 20180921 00:02:44< celticminstrel> Deprecate ActionWML, maybe, but definitely keep WML as a data language. 20180921 00:02:56<+wesdiscordbot> eh 20180921 00:02:59< celticminstrel> It's not even bad as a data language. It's no worse than XML or JSON or YML really. 20180921 00:03:00<+wesdiscordbot> possibly 20180921 00:03:11<+wesdiscordbot> but actionwml would definitely go byebye 20180921 00:03:15< celticminstrel> And keeping it means keeping backwards compatibility with current Wesnoth. 20180921 00:03:29<+wesdiscordbot> well, the point would be to ditch backwards compatibility 20180921 00:04:03<+wesdiscordbot> such an endeavor would be our one shot to do things properly. if we're trying to maintain backwards compatibility, we're hampering ourselves. 20180921 00:04:26< celticminstrel> Yeah, I can't get behind any effort that ditches backwards compatibility. Sorry. 20180921 00:04:58<+wesdiscordbot> ActionWML must go. 20180921 00:05:23<+wesdiscordbot> and I'm in favor of ditching WML entirely 20180921 00:05:32< celticminstrel> It cannot just go, it must be deprecated first. If you want it to be completely gone in this hypothetical rewrite, you're gonna need a separate 1.16 prior to the rewrite. 20180921 00:05:41< celticminstrel> Also FTR, if we went to UE I'd be totally willing to write code that reads GUI2-like WML (probably a modified form of it mind you) and uses it to build up a UMG widget. 20180921 00:05:41<+wesdiscordbot> No. 20180921 00:06:08<+wesdiscordbot> There's no reason 1.14 cannot just be the last 1.x release series. 20180921 00:06:20<+wesdiscordbot> people want WML? they stick with 1.14 20180921 00:06:28< celticminstrel> Sure, but there's also no reason a 2.x release series can ditch backwards compatibility. 20180921 00:06:30<+wesdiscordbot> people want new thing? they use 2.x 20180921 00:06:45<+wesdiscordbot> You're missing the point here... 😐 20180921 00:07:00< celticminstrel> I think you're the one missing the point here, but whatever. 20180921 00:08:20<+wesdiscordbot> the point is to take this opportunity to design something that works well. WML does NOT work well as a scripting language. Other things like IPFs and individual image paths are also endemic of the current design. If we try to work within the current framework, we won't be able to come up with something good. 20180921 00:08:41< celticminstrel> IMO IPFs are a very cool feature that we should keep. 20180921 00:09:03<+wesdiscordbot> Perhaps 20180921 00:09:05< celticminstrel> Their implementation may be suboptimal ATM, but the concept is nice. 20180921 00:09:13<+wesdiscordbot> Maybe. 20180921 00:09:34-!- stikonas [~gentoo@wesnoth/translator/stikonas] has quit [Remote host closed the connection] 20180921 00:09:34<+wesdiscordbot> but WML and its surrounding paradigms must go. 20180921 00:10:07<+wesdiscordbot> it's not worth the effort at all if we try to make some sort of Frankenstein release. 20180921 00:10:09< celticminstrel> That's the attitude that automatically makes me go "nope". 20180921 00:10:50< celticminstrel> If you're gonna go to something new that's a completely new engine with no compatibilty, I'd rather you hand over the reins of this project to someone else and consider your thing a separate project. 20180921 00:10:58<+wesdiscordbot> Why? 20180921 00:11:16< celticminstrel> Because people can continue to work on and improve Wesnoth 1.14 / 1.15. 20180921 00:11:22<+wesdiscordbot> We would still keep 1.14 available, at least for a time. 20180921 00:11:36<+wesdiscordbot> 2.x would not immediately replace it. 20180921 00:11:38< celticminstrel> But if you go off to rewrite it in a new engine, there's a high chance that it'll never even get off the ground. 20180921 00:12:01<+wesdiscordbot> we would gradually phase support for it out, but 1.14 can remain on legacy support for a few years, even. 20180921 00:12:25< celticminstrel> And you're not going to get users switching to this new engine if you have zero compatibility. 20180921 00:12:35<+wesdiscordbot> Says who? 20180921 00:12:48< celticminstrel> You might get brand-new users who never played the old engine, but I'd wager the majority of existing Wesnoth users would stick to 1.14. 20180921 00:12:49< celticminstrel> Why? 20180921 00:12:55< celticminstrel> Because they can't play their campaigns. 20180921 00:13:01<+wesdiscordbot> If we design things well and make creating UMC 10x easier than in 1.x, I bet you many people will switch over. 20180921 00:13:14< aeth> not just campaigns, very popular scenarios 20180921 00:13:20< celticminstrel> Because they can't just tweak a few things to get their scenarios to work. 20180921 00:13:25<+wesdiscordbot> Yes 20180921 00:13:31< celticminstrel> Why do you think Windows never took off until version 3? 20180921 00:13:31< aeth> Most MP scenarios aren't just maps that could be easily ported, they do fancy things in WML and/or Lua 20180921 00:13:38<+wesdiscordbot> That's why 1.14 would have a sunset period. 20180921 00:13:51< celticminstrel> It's because version 3 could play DOS games in a window. It had compatibility with the older OS. 20180921 00:14:02< celticminstrel> Breaking compatibility means losing users. 20180921 00:14:10<+wesdiscordbot> It also means gaining new users. 20180921 00:14:16< aeth> no, it doesn't 20180921 00:14:16< celticminstrel> If you're lucky, it could. 20180921 00:14:26< celticminstrel> But only if you're lucky. What are the chances you'll be lucky? 1%? 20180921 00:14:28< aeth> Games make most of their publicity and get most of their users around launch. 20180921 00:14:39< aeth> You'd get more players from "Wesnoth 2" than from "Wesnoth 2.0" 20180921 00:14:51< aeth> (a proper sequel rather than a new version of Wesnoth) 20180921 00:14:58< celticminstrel> Netscape 5 made the mistake you're advocating for. That's why there was never a Netscape 5 released. 20180921 00:15:11<+wesdiscordbot> We cannot continue as we have. 20180921 00:15:17<+wesdiscordbot> That at least is evident 20180921 00:15:21< celticminstrel> Maybe that was the wrong number, I can't remember. 20180921 00:15:25< celticminstrel> Yeah, we can't continue as we have. 20180921 00:15:45< celticminstrel> But reverting accelerated_rendering would be a far more effective way to kickstart the master branch than what you've proposed. 20180921 00:15:45<+wesdiscordbot> It is inevitable that more and more, we will need to do things that we just cannot do easily or within a reasonable amount of time 20180921 00:15:47<+wesdiscordbot> high dpi support 20180921 00:15:52<+wesdiscordbot> accelerated rendering 20180921 00:15:56< celticminstrel> I can totally understand why you'd not want to do that. 20180921 00:15:59< aeth> Imo: Start deprecating the ugliest parts of WML, and turn WML into a purely configuration language. 20180921 00:16:02< celticminstrel> You put a lot of work into it, after all. 20180921 00:16:14< celticminstrel> That work would go to waste, at least to a degree. 20180921 00:16:19< aeth> Don't break compatibility with everything all at once, get some of the really ugly parts replaced. 20180921 00:16:25<+wesdiscordbot> sigh 20180921 00:16:29< celticminstrel> But it also goes to waste if master just sits there stagnating. 20180921 00:16:57<+wesdiscordbot> and what, we just never get accelerated rendering and stick with the game as-in for eternity? 20180921 00:17:02< celticminstrel> No. 20180921 00:17:15< celticminstrel> We just start over on implementing accelerated rendering. 20180921 00:17:21<+wesdiscordbot> sigh 20180921 00:17:27< celticminstrel> On a feature branch, the way you were doing it before. 20180921 00:17:29<+wesdiscordbot> You assume that we can do that ourselves 20180921 00:17:40< celticminstrel> It might even be possible to build off your original branch from before you merged it. 20180921 00:17:47<+wesdiscordbot> When the damn goalposts keep moving, and the skills required get more and more complex. 20180921 00:17:51<+wesdiscordbot> and no, no it is not 20180921 00:18:03< celticminstrel> I think we probably can. We have Jyrki who knows a lot about this stuff, and I have a little experience with it too. 20180921 00:18:03<+wesdiscordbot> When we started, we needed OGL skills 20180921 00:18:14<+wesdiscordbot> Now we need Vulkan and/or Metal skills 20180921 00:18:24< aeth> Get Wesnoth to the point where its content could, eventually, possibly be ported or have large chunks rewritten. This doesn't mean making things more complex, it actually means simplifying. Right now, scripting is done in both WML and Lua and cannot be done solely in either. Huge complex mess. 20180921 00:18:24< celticminstrel> I'm pretty sure we can still manage with OGL. 20180921 00:18:28<+wesdiscordbot> OGL is going the way of the dodo. macOS has deprecated it. 20180921 00:18:34< celticminstrel> Deprecated != removed 20180921 00:18:46<+wesdiscordbot> the industry will move away from it. 20180921 00:18:49<+wesdiscordbot> and say we get to ogl 20180921 00:18:57-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 00:18:59<+wesdiscordbot> then in 5, 10 years we need to go to metal/vulkan 20180921 00:19:01<+wesdiscordbot> what then 20180921 00:19:11< celticminstrel> OGL over metal/vulkan, duh 20180921 00:19:19<+wesdiscordbot> and who will code that! 20180921 00:19:25< celticminstrel> Both already exist. 20180921 00:19:29<+wesdiscordbot> People already did yeah 20180921 00:19:32< celticminstrel> Though one is sadly propretary. 20180921 00:19:34<+wesdiscordbot> if we use another engine we can let those engine devs manage that 20180921 00:19:35< aeth> "The industry" isn't going to go to Metal. The largest general purpose engines may have a Metal backend, but Apple is absolutely entirely irrelevant for desktop gaming. It's all Windows. Apple has no power to force major changes on desktop gaming. 20180921 00:19:38<+wesdiscordbot> we don't have to worry about it 20180921 00:19:44<+wesdiscordbot> It's not about the engine 20180921 00:19:50<+wesdiscordbot> Necessarily 20180921 00:20:05< celticminstrel> Also yeah, OpenGL isn't deprecated on Windows, right? So worst-case is temporarily dropping support for Macs. 20180921 00:20:06< aeth> Apple deprecating OpenGL isn't going to kill OpenGL. Microsoft deprecated OpenGL like 15+ years ago and it's still used on Windows, the only platform that matters for PC gaming. 20180921 00:20:07<+wesdiscordbot> I think at this point I don't need to explain to you how libraries work 20180921 00:20:20<+wesdiscordbot> aeth: Well you're somewhat incorrect there. 20180921 00:20:22<+wesdiscordbot> It comes down to whether we're willing to reinvent the wheel when there are easily accessible tools right there. 20180921 00:20:23< aeth> celticminstrel: technically the last OpenGL on Windows is like 1.1 or something really ancient and the rest is a big hack due to third party drivers 20180921 00:20:27<+wesdiscordbot> Microsoft deprecates stuff and allows it to continue existing. 20180921 00:20:28<+wesdiscordbot> Apple does not. 20180921 00:20:37<+wesdiscordbot> When Apple deprecates something it's a death sentence for it. 20180921 00:20:38< celticminstrel> That's true. 20180921 00:20:58<+wesdiscordbot> Windows is full of deprecated APIs that were introduced in NT 3.1 and Windows 95. 20180921 00:21:09<+wesdiscordbot> Apple removes deprecated APIs every couple of releases or so. 20180921 00:21:19<+wesdiscordbot> IIRC iOS is where a lot of Wesnoth Inc's income comes from too, unless that's changed. 20180921 00:21:24< aeth> shadowm, my point is no one in the industry cares about macOS. It's 3%. They'll support it if portability is easy and their architecture is portable (e.g. if they use OpenGL instead of DirectX), but they're not going to do something like write two renderers just for 3% 20180921 00:21:35-!- gfgtdf [~Daniel@x4e34250c.dyn.telefonica.de] has quit [Ping timeout: 252 seconds] 20180921 00:21:36<+wesdiscordbot> @Pentarctagon 100% 20180921 00:21:43<+wesdiscordbot> aeth what are you even talking about. 20180921 00:21:48<+wesdiscordbot> People in the industry do care about macOS. 20180921 00:21:59<+wesdiscordbot> so maybe of outsized importance to us then, if nothing else. 20180921 00:22:07<+wesdiscordbot> It's just that a solution for the OpenGL deprecation problem exists even if it's awkward and not the most performant. 20180921 00:22:10< aeth> shadowm, generally support for macOS or Linux comes from two things: (1) Feral porting AAA games or (2) games built on portable engines like Unity. 20180921 00:22:30<+wesdiscordbot> celmin: In the name of "backwards compatibility", you are willing to let us and future project members kill themselves trying to make sense of the mess of our Brobdingnagian codebase, ancient-ass tech, all while trying to reinvent the wheel to do what more skilled peple have done ten times over already and better than we ever could. We cannot do that. 20180921 00:22:44<+wesdiscordbot> Anyway Wesnoth will never reach the level where the performance of a translation API needs to be a real concern. 20180921 00:22:53<+wesdiscordbot> It's a turn-based strategy game, not a first-person shooter. 20180921 00:23:10<+wesdiscordbot> We can at least work on a prototype as dave and jetrel suggested 20180921 00:23:23<+wesdiscordbot> that feels like a feasible plan. 20180921 00:23:56< celticminstrel> Oh sure, it's feasible. It's also a good suggestion in a way because, when it inevitably fails, you won't have wasted too much time. 20180921 00:24:06< aeth> Vultraz, you don't need to port to an engine. But what Wesnoth does need to do at the very least is to port to existing libraries as much as possible. Wesnoth is a very big case of NIH for so many things. 20180921 00:24:09<+wesdiscordbot> but when you look at the challenges we're facing now, they are all things switching to a new engine will means we can avoid all that. 20180921 00:24:31<+wesdiscordbot> we can avoid* 20180921 00:25:02< aeth> Have you investigated renderers that are not part of an engine? 20180921 00:25:20< aeth> Just about every component in making a game engine exists as a standalone C or C++ engine, not just as an all-in-one thing. 20180921 00:25:41< aeth> s/C or C++ engine/C or C++ library/ 20180921 00:25:42<+wesdiscordbot> no, but it's not just the renderer. the entirety of the wesnoth engine is a bloated MESS. Especially it's damn UI toolkit. 20180921 00:25:52< aeth> There are dozens of UI libraries. 20180921 00:26:01<+wesdiscordbot> Would piece-mealing it really be much less effort though? ie: removing ActionWML will break almost all add-ons anyway. 20180921 00:26:05<+wesdiscordbot> aeth you are about as unhelpful as you could be. 20180921 00:26:38<+wesdiscordbot> nor would you be doing the work. 20180921 00:30:05< celticminstrel> I'm just going to come out and say this: Vultraz, I think you should step down as project lead. 20180921 00:30:27< celticminstrel> Or whatever you call your position. 20180921 00:30:38< aeth> Game engine development is literally my specialty, Vultraz. 20180921 00:31:08<+wesdiscordbot> celticminstrel: Do us all a favour and back up statements like that with actual reasoning before posting them. 20180921 00:31:16< celticminstrel> Sure. 20180921 00:31:28<+wesdiscordbot> The emotive discourse going on here on both sides isn't advancing the discussion in a constructive direction. 20180921 00:31:55< celticminstrel> 1. Vultraz made the ill-advised decision to merge accelerated_rendering which is (at least partly) what got us here in the first place. 20180921 00:32:01<+wesdiscordbot> This is why this kind of thing is better discussed on forums or mailing lists where people can't go and post whatever comes to mind one second before pressing Enter. 20180921 00:32:15< celticminstrel> 2. From my perspective, Vultraz seems to demonstrate a lack of understanding of what the project needs. 20180921 00:32:37< celticminstrel> For example: We want to keep existing users. We need backwards compatibility, at least on a temporary basis. 20180921 00:33:09<+wesdiscordbot> there is that thread linked earlier, in fact 20180921 00:33:39< celticminstrel> 3. Vultraz repeatedly recommends taking the worst possible action - a rewrite. Rewrites rarely pan out. In many cases they've been cancelled (see the Netscape example). 20180921 00:33:49<+wesdiscordbot> Oof 20180921 00:34:07<+wesdiscordbot> People who have nothing to add to the discussion please refrain from posting. 20180921 00:35:02<+wesdiscordbot> celticminstrel: Can you remind me for how long you've been a developer? 20180921 00:35:08< celticminstrel> Uhh. 20180921 00:35:12< celticminstrel> Two or three years, I think? 20180921 00:35:17<+wesdiscordbot> In terms of version numbers. 20180921 00:36:19<+wesdiscordbot> People who've been around for a short while will be quick to jump to conclusions and decide that there's nothing wrong with the status quo and that Wesnoth is doing just fine. 20180921 00:36:39<+wesdiscordbot> And from mine, celmin, you demonstrate a lack of willingness to make the decisions that need to be done and possess a stubborn desire to stick t the way things have been done for no other reason than they have been so. I just told you 1.14 would be kept around for a time in order to not leave every single UMC author out in the cold, and you say I don't care about the existing users. You say we "need" backwards 20180921 00:36:39<+wesdiscordbot> compatibility and will lose the existing modder base if we drop it, yet ignore the fact that we make our prominent modders like shadowm and inferno8 jump through the ugliest of hoops to make the simplest fancy thing happen. That is not a sustainable environment, nor is it fair to the people doing the actual content creation. 20180921 00:37:10<+wesdiscordbot> People who've been around for as long as I am or longer are divided in two camps: those who see that the project has several inherent issues, and those that also believe that there's nothing wrong with the status quo because any kind of change involves an expenditure of time or energy. 20180921 00:37:11< aeth> Pentarctagon is absolutely right about iOS being important for Wesnoth, though. This greatly restricts the available options. 20180921 00:37:32< celticminstrel> I think the first released version with any of my code was 1.13.2. 20180921 00:38:05<+wesdiscordbot> I've seen Wesnoth's development pace steadily deteriorate due to its exponential unmaintainability. 20180921 00:38:30<+wesdiscordbot> I had to explicitly call for people to join the project because of it and you're one of the people who answered that call. 20180921 00:38:41< celticminstrel> I'm not saying we shouldn't do anything, but making people jump through hoops is unrelated to the question of backwards compatibility. 20180921 00:38:47<+wesdiscordbot> The situation since then hasn't improved, it has worsened. 20180921 00:38:53< celticminstrel> So why are you using it as an argument to drop backwards compatibility? 20180921 00:39:11< celticminstrel> Yes of course we need to improve things so people don't have to jump through hoops to do fancy things. 20180921 00:39:35<+wesdiscordbot> Because we cannot remove the hoops without dropping backwards compatibility 20180921 00:39:40< celticminstrel> And we probably have to drop compatibility with some things to get there, which is fine. But dropping everything is not fine. It'll drive people away. 20180921 00:39:44<+wesdiscordbot> Because it's not just syntax changes 20180921 00:39:48<+wesdiscordbot> it's paradigm changes 20180921 00:40:02<+wesdiscordbot> It will drive some away, yes 20180921 00:40:08<+wesdiscordbot> but I calculate that is an acceptable risk 20180921 00:40:17< celticminstrel> I think you're vastly underestimating how many it will drive away. 20180921 00:40:41<+wesdiscordbot> I think you're vastly overestimating how many and how it will drive them away. 20180921 00:40:54<+wesdiscordbot> Did Wesnoth 1.13.x drive people developing content for 1.12 away? 20180921 00:40:55-!- gfgtdf [~Daniel@x4e36ec8d.dyn.telefonica.de] has joined #wesnoth-dev 20180921 00:41:19<+wesdiscordbot> No (but it has driven at least me away after the 1.14.0 release because of the sheer amount of crap the devs broke). 20180921 00:41:20< celticminstrel> People from 1.12 don't need to rewrite their entire campaign from scratch to get it to work on 1.13.x. 20180921 00:41:36< celticminstrel> Well, most people don't, at least. 20180921 00:41:47<+wesdiscordbot> I mean, what do you think is better. That wesnoth languishes in its current state while people leave and barely anyone new comes, or we revitalize the project with a better modding experience and drive a few away while attracting more talent. 20180921 00:41:51< aeth> A game like Wesnoth depends on its mods for longevity. There's only so much you can do with standard maps or standard campaigns. You need a migration path at the very least. 20180921 00:41:56< celticminstrel> And when I say "from scratch" I mean from scratch. 20180921 00:42:09<+wesdiscordbot> A lot of people don't even bother porting their add-ons as it is. 20180921 00:42:28<+wesdiscordbot> That's not necessarily a bad thing depending on what exactly isn't being ported. 20180921 00:43:05<+wesdiscordbot> I mean, honestly, it's annoying when the developers remember that user made content exists to make a point for an argument when the rest of the time they make changes without even asking us about their implications. 20180921 00:43:54-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy] 20180921 00:45:00<+wesdiscordbot> I do have something to add besides oof. As I end user for quite a long time, as long as wesnoth remains the same turn based game with mods/add-on’s, I’ll stick around. I’m willing to be that most end users feel the same way 20180921 00:45:15<+wesdiscordbot> You also assume rewriting a campaign under the new system would be just as much work as under the old. 20180921 00:45:20<+wesdiscordbot> Players don't really care as long as they can keep playing the game. 20180921 00:45:44<+wesdiscordbot> In the multiplayer case that means playing with bloated eras and weird scenarios. 20180921 00:45:45< celticminstrel> Uhh, Vultraz, it might not be the exact same amount of work, but it would be a pretty similar amount generally speaking. 20180921 00:45:46<+wesdiscordbot> I mean, look at that paste shadowm pasted about 2 days ago. 20180921 00:45:54<+wesdiscordbot> how long do you think it took her to figure that out 20180921 00:46:04< celticminstrel> Even if it's less work, it's still a lot of work. 20180921 00:46:09<+wesdiscordbot> hours, minimum, days, realistically. 20180921 00:46:11<+wesdiscordbot> If the development team cares so much about user-made content I'm sure you'll be willing to help people port popular stuff. 20180921 00:46:15< celticminstrel> ... 20180921 00:46:16< celticminstrel> No way. 20180921 00:46:33<+wesdiscordbot> (Wait which paste) 20180921 00:46:35< celticminstrel> For a long campaign like HTTT, I doubt you'd get it done in less than a month. 20180921 00:46:47<+wesdiscordbot> @shadowm AtS animation + fog hack 20180921 00:46:53<+wesdiscordbot> Oh yeah that. 20180921 00:46:56< celticminstrel> For a short one like AToTB, maybe your estimate could be accurate. Obviously that depends on the actual engine. 20180921 00:47:18<+wesdiscordbot> It took me several hours. 20180921 00:47:19< celticminstrel> It's not about how easy it is to use the engine, it's about the fact that you literally have to recreate all your content in the new engine. 20180921 00:47:41< celticminstrel> So it'll be about the same amount of time as creating a new campaign from scratch. 20180921 00:47:42<+wesdiscordbot> dave and jetrel are recommending against Anura, so our choices are something like Godot or maybe UE. I'm told we can't use Unity even though it's very good because of end licensing. 20180921 00:47:43<+wesdiscordbot> It would've taken me longer if I didn't come armed with the knowledge that it was the only way to know it (because I'm an engine developer too and know Wesnoth's limitation. 20180921 00:47:48<+wesdiscordbot> *to do it 20180921 00:47:57< celticminstrel> Which might be less than it would have been in the old engine, but is still quite a significant amount of time. 20180921 00:48:04< aeth> celticminstrel: HtTT would be pretty easy to port even though it's long. Most of the data is simple stuff that would be ported with the base engine. Maps aren't hard to port. I wrote a little script to parse Wesnoth map files and it wasn't much work. 20180921 00:48:34< celticminstrel> Well, if you can automate porting, then... there's literally no reason not to have backwards compatibility. 20180921 00:48:36<+wesdiscordbot> Maps are trivial to port, aeth. 20180921 00:48:44<+wesdiscordbot> Everyone in this channel knows that maps are trivial to port. 20180921 00:48:47< celticminstrel> Heh. 20180921 00:48:47<+wesdiscordbot> Does UE have end licensing problems where we can't have an open source license on our game? 20180921 00:48:57<+wesdiscordbot> What isn't trivial to port is executable code (i.e. event WML and Lua). 20180921 00:48:58< celticminstrel> I'm pretty sure it doesn't but I'd have to double-check. 20180921 00:49:22<+wesdiscordbot> would you recommend UE? 20180921 00:49:30< celticminstrel> IIRC there's some allowance for using it for free games, not quite sure on open source. 20180921 00:50:03< celticminstrel> I dunno whether I'd specifically recommend UE. I'm familiar with it, so I'd be slightly more likely to help if it was UE. 20180921 00:50:30< celticminstrel> I was thinking about using UE over Unity for my own games but only because it's C++ rather than C# which is probably a dumb reason all things considered. 20180921 00:50:31< aeth> UE would probably have a license conflict with GPL, but probably not with a permissive license. 20180921 00:50:55< celticminstrel> Well, a license conflict with GPL could be problematic, depending how much code you want to lift from the old engine. 20180921 00:51:22<+wesdiscordbot> That is true 20180921 00:51:34<+wesdiscordbot> Godot is completely open source and your end product can be too 20180921 00:51:58< aeth> UE is more of a 3D engine, anyway. 20180921 00:52:11< celticminstrel> True, but so is Godot from what I can tell. 20180921 00:52:13< aeth> Imo the options are Godot or a framework or a set of libraries. 20180921 00:52:18<+wesdiscordbot> godot has both 20180921 00:52:18< aeth> celticminstrel: Godot was 2D first. 20180921 00:53:24<+wesdiscordbot> godot already has an example of a 2d hex map app 20180921 00:54:42<+wesdiscordbot> the only thing I don't really like at first glance is GDScript, but I'd have to look at that further. I'd prefer if all our content could be created in Lua. 20180921 00:54:53< celticminstrel> You could probably just use Lua. 20180921 00:55:20< celticminstrel> Ideally with an official plugin mind you, but just doing it the way we do it in Wesnoth would work too. 20180921 00:56:07<+wesdiscordbot> If we can do it, I definitely support the idea of Lua as our scripting language 20180921 00:56:27<+wesdiscordbot> We could even see if any of the Lua API is salvageable. 20180921 00:56:38<+wesdiscordbot> But ActionWML would be out 20180921 00:57:26< aeth> https://github.com/perbone/luascript 20180921 00:57:31< celticminstrel> So if you want to do what Dave and Jetrel suggested and make a prototype, possibly with one or two others helping you, go for it. Maybe we could be lucky and it goes well, and if not, you haven't wasted that much time. If it doesn't end up going well? Just bite the bullet and start over with a new master, as described in my upcoming post. 20180921 00:59:39<+wesdiscordbot> I do want to make a prototype, yes 20180921 00:59:58<+wesdiscordbot> not immediately, because as I mentioned i have a webdev contract right now 20180921 01:00:18< celticminstrel> Keep in mind that rewriting almost certainly loses you at least two team members (Jyrki and Tad). 20180921 01:00:28< celticminstrel> And possibly me as well. 20180921 01:00:41< celticminstrel> Which I'm sure is less relevant. 20180921 01:00:57<+wesdiscordbot> Tad hasn't been doing anything for months 20180921 01:01:06<+wesdiscordbot> You are focusing on 1.14 20180921 01:01:19<+wesdiscordbot> Jyrki too, and he seems very unenthusiastic about working on master. 20180921 01:01:28< celticminstrel> Gee, I wonder why. 20180921 01:01:37< celticminstrel> *cough*accelerated_rendering*cough* 20180921 01:01:41<+wesdiscordbot> There's no reason he couldn't start the OGL work 20180921 01:01:47<+wesdiscordbot> master runs 20180921 01:01:53< celticminstrel> Pretty sure he already has started it on a branch somewhere. 20180921 01:01:53<+wesdiscordbot> there's just no in-game UI 20180921 01:02:06<+wesdiscordbot> he's made one or two minor commits 20180921 01:02:16< celticminstrel> But maybe if you rolled back a_r you'd suddenly get people more enthusiastic about working on master. 20180921 01:02:27<+wesdiscordbot> if we rolled back a_r it would just be 1.14 20180921 01:02:42< celticminstrel> Obviously you'd do more than a pure rollback. 20180921 01:02:51< irker627> wesnoth/wesnoth:master newfrenchy83 a39275331d Update message.lua AppVeyor: All builds passed 20180921 01:03:06<+wesdiscordbot> i've touched too much on master to easily separate things 20180921 01:03:10<+wesdiscordbot> removed too much 20180921 01:03:15<+wesdiscordbot> rewrote too much 20180921 01:04:02< celticminstrel> I have a feeling things are more easily separated than you think, but they're probably also less easily separated than I think. 20180921 01:04:55< celticminstrel> Anyway, just read my post or whatever. 20180921 01:05:21< celticminstrel> I tried to summarize stuff there. 20180921 01:05:23<+wesdiscordbot> "sane people" 20180921 01:05:29<+wesdiscordbot> how charitable 20180921 01:06:14< celticminstrel> But accurate. 20180921 01:06:35<+wesdiscordbot> I also really don't care what tad thinks 20180921 01:06:41 * celticminstrel shrugs. 20180921 01:07:57<+wesdiscordbot> You also assume Jyrki would continue working on the ogl branch. 20180921 01:08:02<+wesdiscordbot> I doubt he will at this point. 20180921 01:08:11< celticminstrel> Why do you say that? Why don't you ask him directly? 20180921 01:08:12<+wesdiscordbot> and I don't blame him 20180921 01:08:31<+wesdiscordbot> and say he rebases to current master before a theoretical rollback 20180921 01:08:42<+wesdiscordbot> then he has to deal with the problems master is having 20180921 01:08:45<+wesdiscordbot> just on his branch 20180921 01:09:10< celticminstrel> Maybe so. 20180921 01:09:13<+wesdiscordbot> He needs my work because it lays the ar groundwork 20180921 01:09:33< celticminstrel> Well, in any case, maybe you should ask him. 20180921 01:10:16< celticminstrel> Still... if he truly does need to build off your work, then there's nothing preventing him from doing so. The branch that used to be master could still be merged in at a later time. 20180921 01:11:32<+wesdiscordbot> I do not believe it's worth his time. 20180921 01:11:40< celticminstrel> Maybe you should let him be the judge of that. 20180921 01:11:47<+wesdiscordbot> Again - why do we keep trying to reinvent the damn wheel 20180921 01:12:21< celticminstrel> I'll admit, that's a pretty good question. 20180921 01:13:14<+wesdiscordbot> Do you really believe one person will be able to make something as good as a large team of people working for several years in a reasonable amount of time? 20180921 01:13:19<+wesdiscordbot> And what when jyrki leaves 20180921 01:13:31<+wesdiscordbot> We'll have a rendering engine no one is maintaining or understands 20180921 01:14:19< celticminstrel> True, he could turn into another mordante if you're not careful. But mordante's work was a problem because it was done in master, not in a branch as it should have been. 20180921 01:15:06< celticminstrel> If it's on a branch and the maintainer of that branch leaves? You can just ignore it. Start a new branch on the topic if the old one is incomprehensible. 20180921 01:15:36<+wesdiscordbot> .... no.... that is nor why mordante's work was a problem 20180921 01:15:46<+wesdiscordbot> you were not around when mordante was 20180921 01:15:57<+wesdiscordbot> he was basically out the door when I came in, too. 20180921 01:16:03< celticminstrel> Well, that's the impression I got from what people have said, anyway. 20180921 01:16:09<+wesdiscordbot> @shadowm is the one who can fully elaborate on the problems of mordante 20180921 01:23:31<+wesdiscordbot> GUI2 had design flaws which he didn't realize existed until it was too late and most of it was already used in production. 20180921 01:23:36-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Ping timeout: 260 seconds] 20180921 01:23:53<+wesdiscordbot> Then he left in the middle of trying to correct one of those flaws (the listbox implementation). 20180921 01:25:03<+wesdiscordbot> While he might have attributed his departure to other factors when he briefly returned years (!) later, it's pretty evident that he was burnt out after 6 years of trying to finish a task that was too large for him. 20180921 01:25:24<+wesdiscordbot> That is, it was evident before he even left/disappeared. 20180921 01:34:19< celticminstrel> Okay. 20180921 01:37:57<+wesdiscordbot> Six years developing GUI2 and only GUI2. 20180921 02:00:19-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180921 02:02:46-!- wesdiscordbot [~wesdiscor@wesnoth/bot/discord-bridge] has quit [Remote host closed the connection] 20180921 02:02:56-!- wesdiscordbot [~wesdiscor@wesnoth/bot/discord-bridge] has joined #wesnoth-dev 20180921 02:02:59-!- mode/#wesnoth-dev [+v wesdiscordbot] by ChanServ 20180921 02:04:30<+wesdiscordbot> question: why isnt wesnoth growing as it is right now? 20180921 02:07:38-!- gfgtdf [~Daniel@x4e36ec8d.dyn.telefonica.de] has quit [Read error: Connection reset by peer] 20180921 02:14:56<+wesdiscordbot> Cause it’s a niche game, with not a lot of press? 20180921 02:18:54<+wesdiscordbot> I don't feel like it's that niche? 20180921 02:20:02<+wesdiscordbot> the fantasy genre isn't small at all 20180921 02:20:06<+wesdiscordbot> and turn-based strategy isn't either 20180921 02:20:24-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Ping timeout: 252 seconds] 20180921 02:30:24-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180921 03:24:47< irker627> wesnoth/wesnoth:1.14 Nils Kneuper e48473238a updated Spanish translation AppVeyor: All builds passed 20180921 03:56:00-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] 20180921 03:58:08-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev 20180921 05:41:41< irker627> wesnoth/wesnoth:master Nils Kneuper 639c5087b9 updated Spanish translation AppVeyor: All builds passed 20180921 06:09:19<+wesdiscordbot> My focus has been primarily in merging pull requests. I strongly believe that merging PRs should have higher priority than my own work, because it's extremely annoying for a contributor when their work is just ignored. 20180921 06:09:43<+wesdiscordbot> Regarding that, it's not going well, unfortunately. Pull requests are coming in faster than I can merge them. 20180921 06:10:11<+wesdiscordbot> Leaving me absolutely no time to fix bugs (some things added to the 1.14.x milestones are still not fixed), let alone the OpenGL work. 20180921 06:10:27<+wesdiscordbot> Also, the OpenGL work does not need Vultraz's groundwork. 20180921 06:12:10<+wesdiscordbot> In fact, it might be easier to start it from the state of 1.14, instead of master. Master has the disadvantage of bugs, and starting from there means that I'd need to fix those bugs (when I can't even tell if the bug was caused by OpenGL or a_r, and in the latter case I'm not the author of the code and it may be difficult for me to understand how it was intended to work). 20180921 06:50:29<+wesdiscordbot> when is the 1.14.5 string freeze lifted? asking regarding https://github.com/wesnoth/wesnoth/issues/3554 20180921 06:50:56<+wesdiscordbot> It has already been lifted. 1.14.5 was already released. 20180921 06:51:49<+wesdiscordbot> alright, I thought I remembered the freeze lasting a bit after the release in case a critical issue was discovered, or something along those lines. 20180921 07:24:08<+wesdiscordbot> Only before it's announced. 20180921 07:26:48<+wesdiscordbot> @jyrkive Regarding the last thing you said about a_r above, was this properly and clearly communicated to vultraz before he merged his branch? If so, when and where? 20180921 07:28:52<+wesdiscordbot> @jyrkive I could help out with pull requests. Which one would you recommend me to look into as a priority? 20180921 07:38:41<+wesdiscordbot> @shadowm Not really. 20180921 07:40:16<+wesdiscordbot> It is true that OpenGL integration is easier when the rendering pipeline isn't assuming the same behavior as software rendering (e.g. that reading the framebuffer is fast). It's one reason why I wasn't opposed to merging accelerated_rendering. 20180921 07:40:41<+wesdiscordbot> And now you want it off master after it was already merged. 20180921 07:41:09<+wesdiscordbot> I wasn't aware of how badly it was broken. Sorry. 😦 20180921 07:42:41<+wesdiscordbot> I realize that throwing Vultraz's work away isn't nice at all, too. I wouldn't even say that I want the changes to be reverted: I'm merely saying that the "OpenGL needs a_r" argument is false. 20180921 07:43:03-!- Rhonda [~rhonda@wesnoth/developer/rhonda] has quit [Quit: Going down the drain....] 20180921 07:45:28<+wesdiscordbot> @loonycyborg Jostephd's pull requests have been of good quality. #3528 and newer PRs from him are the ones I recommend considering. 20180921 07:48:27< irker627> wesnoth/wesnoth:1.14 josteph b72e7b51f3 Commandline: --campaign-skip-story skips AppVeyor: All builds passed 20180921 07:57:10-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity] 20180921 07:59:19-!- Rhonda [~rhonda@wesnoth/developer/rhonda] has joined #wesnoth-dev 20180921 08:02:52< irker627> wesnoth: josteph wesnoth:master d99f5b8a7689 / src/ (menu_events.cpp play_controller.cpp play_controller.hpp): Reevaluate [show_if] conditions and delayed variable expansions before displayin https://github.com/wesnoth/wesnoth/commit/d99f5b8a7689ab939f76d47535d1ff0ce6d625cf 20180921 08:02:54< irker627> wesnoth: josteph wesnoth:master 9f58a8063561 / src/ (play_controller.cpp play_controller.hpp): Add const annotation to new function https://github.com/wesnoth/wesnoth/commit/9f58a8063561a1ce4897fa88c1982e68dece33bc 20180921 08:02:56< irker627> wesnoth: Sergey Popov wesnoth:master 63d06f8a3853 / src/ (menu_events.cpp play_controller.cpp play_controller.hpp): Merge pull request #3555 from jostephd/objectives-prestart-for-master https://github.com/wesnoth/wesnoth/commit/63d06f8a38536629acdf8c2688b931abe180dc2e 20180921 08:03:33< irker627> wesnoth: Pentarctagon wesnoth:1.14 5b7e8975a9a4 / data/campaigns/tutorial/scenarios/01_Tutorial_part_1.cfg: Revert "Tutorial S1: adjust string to keep string freeze" https://github.com/wesnoth/wesnoth/commit/5b7e8975a9a454cc0e0e95e94c623fca9ab44c98 20180921 08:03:35< irker627> wesnoth: Pentarctagon wesnoth:1.14 cc01ccae843e / data/campaigns/ (17 files in 17 dirs): Revert "Revert "Menu: re-order and re-rank campaigns (including new ranks) (fixe https://github.com/wesnoth/wesnoth/commit/cc01ccae843e9fa4b6e56301ff9f91fbc935908c 20180921 08:03:37< irker627> wesnoth: Pentarctagon wesnoth:1.14 e43d989d80b9 / data/campaigns/ (5 files in 5 dirs): Revert "Revert "Menu: campaign order tweaks"" https://github.com/wesnoth/wesnoth/commit/e43d989d80b9ca978aa1a86b502a129d570638c6 20180921 08:37:59-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev 20180921 08:58:31< irker627> wesnoth/wesnoth:1.14 josteph f04ce5f1f8 Commandline: --campaign-skip-story skips AppVeyor: All builds passed 20180921 09:08:03<+wesdiscordbot> @jyrkive should I do something about your pending review in PR #3528 ? 20180921 09:08:37<+wesdiscordbot> seems initializations you requested are in PR, probably added with force-push or something 20180921 09:09:03<+wesdiscordbot> Or I can just merge? 20180921 09:11:53<+wesdiscordbot> If the initialization is now properly done (I suspect it's indeed with history rewriting), you can just merge. 20180921 09:13:40< irker627> wesnoth: josteph wesnoth:1.14 9b5d5a6171d0 / / (9 files in 3 dirs): Commandline: Add --campaign-skip-story option https://github.com/wesnoth/wesnoth/commit/9b5d5a6171d03afbc13fc1b496a949ad7548fb41 20180921 09:13:42< irker627> wesnoth: josteph wesnoth:1.14 7951dc254a53 / src/ (play_controller.cpp play_controller.hpp playsingle_controller.cpp): Commandline: Encapsulate skip_story in play_controller https://github.com/wesnoth/wesnoth/commit/7951dc254a534927f876104c6a0cf0a7296619a5 20180921 09:13:44< irker627> wesnoth: josteph wesnoth:1.14 f770d6b76a5d / src/ (play_controller.cpp scripting/game_lua_kernel.cpp): Commandline: --campaign-skip-story skips [message]s during prestart and start ev https://github.com/wesnoth/wesnoth/commit/f770d6b76a5d67f09a5d393cde4634a16be4589a 20180921 09:13:46< irker627> wesnoth: Sergey Popov wesnoth:1.14 faf167a253db / / (12 files in 4 dirs): Merge pull request #3528 from jostephd/skip-story https://github.com/wesnoth/wesnoth/commit/faf167a253db9a1e7f8eadc1bc8c1bcb928789ce 20180921 09:40:34-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has quit [Read error: Connection reset by peer] 20180921 09:41:37-!- aeth [~Michael@wesnoth/umc-dev/developer/aethaeryn] has joined #wesnoth-dev 20180921 10:01:59< irker627> wesnoth: josteph wesnoth:master f7be872da244 / / (9 files in 3 dirs): Commandline: Add --campaign-skip-story option https://github.com/wesnoth/wesnoth/commit/f7be872da24474a47894becd7c6d5f013c495a2f 20180921 10:02:01< irker627> wesnoth: josteph wesnoth:master 1d1d26035091 / src/ (play_controller.cpp play_controller.hpp playsingle_controller.cpp): Commandline: Encapsulate skip_story in play_controller https://github.com/wesnoth/wesnoth/commit/1d1d26035091108e33ea8af57754ba520ec27d35 20180921 10:02:03< irker627> wesnoth: josteph wesnoth:master b64f4b42d659 / src/ (play_controller.cpp scripting/game_lua_kernel.cpp): Commandline: --campaign-skip-story skips [message]s during prestart and start ev https://github.com/wesnoth/wesnoth/commit/b64f4b42d659d90ec33fd0cfc139e0a058b52f04 20180921 10:47:05< Soliton> looks like we should offer jostephd developer status. 20180921 10:47:43<+wesdiscordbot> Agreed. 20180921 11:24:45<+wesdiscordbot> ya know, I don't want to seem like I'm offloading blame, but the whole reason I decided to merge the AR branch immediately was you approved of the idea 😐 20180921 11:25:42<+wesdiscordbot> IIRC, I hadn't even intended to merge it until it was more feature-complete. But you recommended I merge right after splitting 1.14, so I did so. 20180921 11:26:41<+wesdiscordbot> Ah, I think you misunderstood my words. I hadn't intended to recommend merging a half-baked branch. 20180921 11:26:50<+wesdiscordbot> I see... 20180921 11:27:00<+wesdiscordbot> That's very possible 20180921 11:28:24<+wesdiscordbot> IIRC, what I said was something along the lines "right after creating a new stable branch is the right time for risky changes", and with that I was referring to the kind of last-minute refactorings you had been making to the master branch prior to branching 1.14 off of it. 20180921 11:29:06<+wesdiscordbot> I guess I was also motivated by a desire not to have to keep rebasing. 20180921 11:29:34<+wesdiscordbot> It is a perfectly reasonable desire. Rebasing is additional work. 20180921 11:35:00<+wesdiscordbot> I think there are a few big questions now, setting aside any ideas 2.x series prototypes: 1. Do you even still want to build an OGL-based rendering engine from scratch anymore? 2. Given the time such an endeavor takes and your skill, how long do you think it would take to produce something acceptable? 3. You have already stated before you don't intend to stick around with Wesnoth forever. This relates to point 1, but 20180921 11:35:01<+wesdiscordbot> more specifically, how confident are you we could maintain your engine without you without it becoming a black box? 4. How confident are you the project could attract new people with the necessary skills to work on the engine should improvements be needed, or to perform additional work like porting it to Vulkan or Metal? 20180921 11:35:38<+wesdiscordbot> These are things I considered when writing my proposal, ftr. 20180921 11:39:01<+wesdiscordbot> And also, none of that is meant to insinuate you couldn't build it. I'm sure you could 20180921 11:40:01<+wesdiscordbot> But I have to wonder whether we really want to continue down the NIH path with something as core as the rendering engine. 20180921 11:43:33<+wesdiscordbot> 1. First of all, it's not "from scratch". We do have an existing engine, and the process would be about converting it to use OpenGL, rather than fully rewriting it. And yes, I'm still willing to perform that conversion. 2. The task is too large to be able to even give an estimate. 3. I have a counterpoint to this: a third-party engine such as Godot is, by definition, a "black box" that Wesnoth developers can't easily 20180921 11:43:34<+wesdiscordbot> maintain (at best we can send them pull requests). With your question you're essentially implying that a custom engine needs to be "maintained" but a third-party one doesn't. 4. I think attracting new developers isn't very likely. Of course, a full rewrite of Wesnoth isn't a project that's easy to pull off with very little manpower, either. 20180921 11:47:53<+wesdiscordbot> You have a point there, but my thinking was more along the lines of "it wouldn't be our problem anymore". It would be the problem of the Godot or Unreal Engine or insert engine here devs (who are more numerous and likely more skilled than we) to provide a working product. We would just use it and eliminate the maintenance burden of that component. Using something a lot of people use also means a larger pool of knowledge 20180921 11:47:53<+wesdiscordbot> on the implementations those engines use. With a custom engine, only we know how it works. 20180921 11:48:16<+wesdiscordbot> What is rendering engine maintenance, anyway? 20180921 11:53:22<+wesdiscordbot> I view it as the general falling behind of the engine in terms of features. When wesnoth was developed, you didn't have to worry about accelerated rendering, now you do. You didn't need to worry about HiDPI screens, now you do. And we and modders have increasingly wanted to do fancy graphical effects (animated terrains, or even something as simple as playing a rendered cutscene) and have instead run smack dab into engine 20180921 11:53:22<+wesdiscordbot> limitations that prevent that. We then have to try to figure out how to implement such features, which means we run up against insufficiencies in the original design, and the end result is sub-optimal, if it exists at all. Take animated water. It was a great idea! The graphics were pretty easy to make. But then the performance was shit and we had to spend however long (you were one of the people who did so, IIRC) trying to optimize 20180921 11:53:23<+wesdiscordbot> it. Which we did, somewhat, but the performance is still not great, especially on large maps! 20180921 11:57:25<+wesdiscordbot> Accelerated rendering is a valid point (although I'd say that the writing was on the wall even back in 2004; 3D games had completely stopped supporting software rendering in late 1990s). HiDPI doesn't concern the rendering engine. It's about the GUI framework instead. The problems with animated water and the like are caused by use of software rendering, and wouldn't exist with hardware rendering. Video cutscenes don't use 20180921 11:57:25<+wesdiscordbot> the rendering engine in modern games, so they're also irrelevant. 20180921 11:59:33<+wesdiscordbot> Point taken on the cutscenes. But they into the territory of "thing we need to implement ourselves but really shouldn't because any capable third-party engine should be able to allow that out of the box" 20180921 11:59:49<+wesdiscordbot> I'd say that the largest change rendering engines have had after mid-2000s is the switch to deferred shading. Deferred shading has higher overhead than forward rendering, but scales better for multiple dynamic lights (especially in areas with a lot of occlusion, such as indoor areas). And guess what? For a 2D game, it's entirely irrelevant. 20180921 12:00:40<+wesdiscordbot> Quite frankly, I don't think the rendering engine would need "maintenance" afterwards (and it still wouldn't need it today if it had been written for hardware rendering from the ground up). 20180921 12:01:54<+wesdiscordbot> Perhaps so 20180921 12:05:55<+wesdiscordbot> Honestly, I also see such an engine switch as not just a way to avoid having to do the AR transition the hard way, but also to sweep away a lot of the cruft that is just bad, like WML for scripting. 20180921 12:06:37<+wesdiscordbot> Yeah, WML is a terrible scripting language, and I see the temptation of just restarting everything from scratch. 20180921 12:08:00< Soliton> one thing to consider when starting from scratch is that your new design isn't automatically going to be perfect either. 20180921 12:08:09<+wesdiscordbot> I'm also afraid we'll end up tying ourselves in knots desperately trying to maintain backwards compatibility for things like separate files for every sprite or IPFs for each image that we'll end up with a mess. 20180921 12:08:23<+wesdiscordbot> Remember that huge discussion we had over spritesheets? 20180921 12:08:40-!- gfgtdf [~Daniel@x4e36ec8d.dyn.telefonica.de] has joined #wesnoth-dev 20180921 12:09:04< Soliton> the good thing is you can just try and do it. you do not need to convince everyone to follow you before there is anything to show. 20180921 12:09:06<+wesdiscordbot> I don't recall what the technical problem was with runtime spritesheet generation. 20180921 12:11:41<+wesdiscordbot> Soliton: I'm not saying we should be lemmings. I just think an engine port needs be looked at as a serious possibility for a path forward. Either way we go is going to be a ton of work, we can't get around that, so I'd rather we took the path with the most long-term benefit. 20180921 12:12:06<+wesdiscordbot> every time it's been floated before it's been shot down. 20180921 12:13:03<+wesdiscordbot> I think wesnoth2 was a good idea back in the day, but made the mistake of also trying to change the gameplay. 20180921 12:13:13<+wesdiscordbot> Yes. 20180921 12:13:22<+wesdiscordbot> As of this point, I'm not sure if we have enough manpower left for a rewrite. 😦 20180921 12:13:30<+wesdiscordbot> And now we're dealing with Dave wanting to sunset Anura 20180921 12:13:58< Soliton> what is the issue with anura, btw? 20180921 12:14:08< Soliton> not enough man power as well? 20180921 12:14:16<+wesdiscordbot> He started using Unity and saw just how much better Unity was to Anura. 20180921 12:14:18< Soliton> and/or better alternatives. 20180921 12:14:20<+wesdiscordbot> (He = Dave) 20180921 12:14:25< Soliton> ok, i see. 20180921 12:14:58<+wesdiscordbot> Even with all the time and effort sunk into Anura (which does, FTR, have some really nice things), it doesn't stand up to Unity in his estimation 20180921 12:16:20<+wesdiscordbot> And I see the possibility of the same thing happening here. We could sink years of time and effort into implementing ar and video cutsenes and hidpi support and a better ui toolkit and whatever else we want, and then look up and realize we need to implement some new thing and that some other engine does what we do 10x better. 20180921 12:16:36<+wesdiscordbot> A similar thing happened at Rovio. Multiple teams in the company tried Unity, discovered that it was far better than our own engine (especially regarding tooling), and at one point management decided to stop development of the engine and move its future maintenance to the Nibblers team. 20180921 12:16:50< Soliton> so i think people are looking seriously at what you're proposing and some even readily agree. i'm just saying that the next step will have to be just doing it (or a prototype or whatever you want to call it). you will not be able to convince everyone with "just words". 20180921 12:16:50-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev 20180921 12:16:54<+wesdiscordbot> (But AFAIK, none of the existing projects was rewritten.) 20180921 12:17:12<+wesdiscordbot> indeed 20180921 12:18:27< Soliton> one thing why you won't convince everyone is that people contribute to wesnoth for fun. fun they have had even if ultimately it's a waste (if you want to see it that way). 20180921 12:19:23< celticminstrel> 20180921 11:29:06<+wesdiscordbot> I guess I was also motivated by a desire not to have to keep rebasing. 20180921 12:19:24< celticminstrel> 20180921 11:29:34<+wesdiscordbot> It is a perfectly reasonable desire. Rebasing is additional work. 20180921 12:19:25< celticminstrel> IMO for major feature branches you just shouldn't bother. Just merge master in every so often, like Aginor did. It's ugly, yes, but I think it might be significantly easier. 20180921 12:20:20<+wesdiscordbot> I did address that. And like I said, past contributions don't just vanish. They helped make the 1.14 that people are currently enjoying. It's not a waste. 20180921 12:20:33<+wesdiscordbot> But working with many areas of the engine is just... not fun 😐 not fun at all 20180921 12:20:48<+wesdiscordbot> Depends on who you ask. I like it. 20180921 12:20:58< celticminstrel> \o/ 20180921 12:21:09<+wesdiscordbot> I've raged at GUI2 I dunno how many times. 20180921 12:22:03< Soliton> people contribute to the parts of wesnoth that are fun for them. 20180921 12:22:35<+wesdiscordbot> yes 20180921 12:22:41< Soliton> it's nice of you if you work with GUI2 even if you dislike it but perhaps that's not the best course of action. 20180921 12:23:09<+wesdiscordbot> someone had to 20180921 12:23:11<+wesdiscordbot> :/ 20180921 12:24:11<+wesdiscordbot> I'm not even sure how necessary the GUI1-to-GUI2 conversion has. I wonder if there would have been a way to get GUI1 to use hardware rendering as well. 20180921 12:24:16<+wesdiscordbot> *was 20180921 12:24:34<+wesdiscordbot> It wasn't the rendering that was the problem, it was the layout engine. 20180921 12:25:32< celticminstrel> GUI1 has no layout engine. 20180921 12:25:41< celticminstrel> Each widget just has coordinates. 20180921 12:25:49< celticminstrel> But that doesn't scale. 20180921 12:26:16< celticminstrel> There was the ThemeWML layout engine, which I suppose you might've been able to support for general GUI1 dialogs, but it's not very flexible either. 20180921 12:27:44<+wesdiscordbot> So, what did we do? Spent the last 10 years working on a replacement that's 10x better yet still inflexible, messy, and broken in its own way. 20180921 12:28:10< celticminstrel> It's far more flexible than GUI1 was. 20180921 12:28:24<+wesdiscordbot> Think about that. The GUI2 project has been WIP for the past 10 years. 20180921 12:28:37<+wesdiscordbot> what were you all doing 10 years ago 20180921 12:28:42< celticminstrel> It does have some problems. I don't know whether it's worth working to fix its layout engine or if we should just tear the code out and use some pre-existing layout engine. 20180921 12:29:04<+wesdiscordbot> Certainly the latter would be a part of an engine switch. 20180921 12:29:41< celticminstrel> Ideally we'd be able to reuse the existing dialog definitions as much as possible in a new GUI engine though. 20180921 12:29:47<+wesdiscordbot> You don't need 15 years working in Anura or use Unity or Unreal to know there are some few issues wrong in Anura (or everybody eventually has itches about everything). Nothing's perfect. 🀷 20180921 12:30:13<+wesdiscordbot> I dislike the dialog definitions because they're so. damn. LONG. 20180921 12:30:25< celticminstrel> IIRC someone suggested something called CEGUI as an alternative to GUI2. 20180921 12:30:32< celticminstrel> Yeah they are quite long, I agree. 20180921 12:30:38<+wesdiscordbot> I bet there are GUI frameworks out there which can be used without changing the whole game engine. I think such a framework should have been chosen back then, instead of GUI2. 20180921 12:30:40< celticminstrel> The format is overly verbose IMO. 20180921 12:31:04< celticminstrel> I have no idea anymore if CEGUI would actually be sufficient, but it's one possibility to consider. 20180921 12:31:37<+wesdiscordbot> Yes, writing a UI toolkit from scratch was a mistake. 20180921 12:32:09< celticminstrel> Heh, CEGUI's almost the same age as Wesnoth. 20180921 12:32:45<+wesdiscordbot> Indeed, GUI2 window definitions are too long. For reference, Windows Presentation Foundation (WPF) window definitions can be like this (77 lines, although for a fairly simple window): https://sourceforge.net/p/keppi/code/HEAD/tree/tags/3.71/VisualizationWindow.xaml 20180921 12:33:10<+wesdiscordbot> looks like html 20180921 12:33:35<+wesdiscordbot> It's XAML, a subset of XML. 20180921 12:33:54<+wesdiscordbot> To name one. Argentum Age would break a few times is past FFL wesnorh2 was to be a true game. πŸ˜… 20180921 12:34:18<+wesdiscordbot> s/is/if 20180921 12:34:30< celticminstrel> I can't find a good list of supported widgets for CEGUI, but if anyone else wants to poke around: http://cegui.org.uk/ 20180921 12:34:40-!- behalebabo [~behalebab@unaffiliated/behalebabo] has quit [Ping timeout: 272 seconds] 20180921 12:34:42<+wesdiscordbot> I think everyone pretty much agrees Anura is the wrong choice right now 20180921 12:35:02<+wesdiscordbot> if Dave himself can't recommend it, we won't use it. 20180921 12:35:54< celticminstrel> From my brief perusal of that site it does seem like it could work. 20180921 12:36:00<+wesdiscordbot> I wouldn't recommend it. But neither current Godot (sadly!). 20180921 12:36:28<+wesdiscordbot> how come? 20180921 12:38:50<+wesdiscordbot> Asking who? 20180921 12:39:00<+wesdiscordbot> you 20180921 12:39:27-!- behalebabo [~behalebab@unaffiliated/behalebabo] has joined #wesnoth-dev 20180921 12:40:21<+wesdiscordbot> I'm afraid Godot might be 500 people yelling at each other, instead of a team. 20180921 12:40:38<+wesdiscordbot> heh 20180921 12:41:05<+wesdiscordbot> Uh... even if it was, it wouldn't matter to us. We would be only using the Godot engine, not joining their development team. 20180921 12:41:35<+wesdiscordbot> Indeed 20180921 12:54:22-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc] 20180921 12:56:44-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds] 20180921 12:58:08-!- celticminstrel is now known as celmin|away 20180921 12:58:20< celmin|away> Interesting how the conversation peters out just at the right moment for me. :P 20180921 12:58:29-!- celmin|away [~celmin@unaffiliated/celticminstrel] has quit [Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] 20180921 12:58:45<+wesdiscordbot> but you're away πŸ˜› 20180921 13:02:05-!- irker627 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Quit: transmission timeout] 20180921 13:05:05-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180921 13:55:35<+wesdiscordbot> Hopefully if some of you make a wesnoth2 prototype in Godot then I'd knew that one team is making something with Godot. I think I saw a consistent pattern of only solo dev when I was browsing the Godot showcase. I eventually had to cease looking because it didn't feel productive. So here I am. 20180921 14:09:30-!- irker748 [~irker@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev 20180921 14:09:30< irker748> wesnoth/wesnoth:master Sergey Popov 63d06f8a38 Merge pull request #3555 from jostephd/o AppVeyor: All builds passed 20180921 14:26:56-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has joined #wesnoth-dev 20180921 14:35:55< zookeeper> i'll try to write up my very short thoughts related to the whole engine thing, but i do very much protest the notion that the water graphics were "pretty easy to make". dunno how you arrived at that conclusion. :p 20180921 14:39:36<+wesdiscordbot> They look great tho. πŸ‘ŒπŸ» 20180921 14:39:51-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: mattsc] 20180921 14:45:05<+wesdiscordbot> maybe they derived it from the general approach to things observed with you guys πŸ˜„ 20180921 14:51:15<+wesdiscordbot> shorthanded, bad-tempered and plain uninformed - leading to appropriate resultz 20180921 14:51:56<+wesdiscordbot> How nice of you 20180921 14:54:45<+wesdiscordbot> yw I just describe how I receive my one week ban recently 20180921 14:55:08<+wesdiscordbot> writing reza something in the lobby while the lake of insults ingame was unknown 20180921 14:55:17<+wesdiscordbot> well heck who cares, ban that fool for a week, put a stone on his neck 20180921 14:55:59<+wesdiscordbot> Okay we don't talk about mod issues in public and you were already told this. 20180921 15:02:27-!- mattsc [~mattsc@wesnoth/developer/mattsc] has joined #wesnoth-dev 20180921 16:06:13-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 16:12:25-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has joined #wesnoth-dev 20180921 16:24:51-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy] 20180921 16:29:23-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 16:30:09-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit] 20180921 16:31:19< irker748> wesnoth/wesnoth:1.14 Sergey Popov faf167a253 Merge pull request #3528 from jostephd/s AppVeyor: All builds passed 20180921 16:35:50-!- stikonas [~gentoo@wesnoth/translator/stikonas] has joined #wesnoth-dev 20180921 17:31:38-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 17:32:02-!- noy [~Noy@wesnoth/developer/noy] has quit [Client Quit] 20180921 18:02:30-!- crimson_penguin [~crimson_p@wesnoth/developer/crimsonpenguin] has joined #wesnoth-dev 20180921 18:06:49-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 18:57:24-!- ToBeCloud [uid51591@wikimedia/ToBeFree] has quit [Quit: Connection closed for inactivity] 20180921 19:31:40-!- irker748 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Quit: transmission timeout] 20180921 19:32:58-!- irker643 [~irker@wesnoth/umc-dev/bot/shikadibot] has joined #wesnoth-dev 20180921 19:32:58< irker643> wesnoth/wesnoth:master josteph b64f4b42d6 Commandline: --campaign-skip-story skips AppVeyor: All builds passed 20180921 19:33:47-!- gfgtdf [~Daniel@x4e36ec8d.dyn.telefonica.de] has quit [Quit: Leaving] 20180921 19:43:07-!- mattsc [~mattsc@wesnoth/developer/mattsc] has quit [Quit: So long and thanks for all the fish.] 20180921 19:50:26-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy] 20180921 20:07:00-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 20:45:09-!- boucman [~rosen@wesnoth/developer/boucman] has joined #wesnoth-dev 20180921 21:00:14-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180921 21:00:20-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180921 21:49:54-!- zookeeper [~lmsnie@wesnoth/developer/zookeeper] has quit [Ping timeout: 252 seconds] 20180921 21:56:14-!- boucman [~rosen@wesnoth/developer/boucman] has quit [Remote host closed the connection] 20180921 22:33:13-!- irker643 [~irker@wesnoth/umc-dev/bot/shikadibot] has quit [Quit: transmission timeout] 20180921 22:53:06-!- noy [~Noy@wesnoth/developer/noy] has quit [Quit: noy] 20180921 23:05:27-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has quit [Remote host closed the connection] 20180921 23:05:33-!- janebot [~Gambot@unaffiliated/gambit/bot/gambot] has joined #wesnoth-dev 20180921 23:21:06-!- noy [~Noy@wesnoth/developer/noy] has joined #wesnoth-dev 20180921 23:35:18-!- valdar [~atarocch@93.56.164.28] has quit [Ping timeout: 252 seconds] 20180921 23:39:34-!- valdar [~atarocch@93.56.164.28] has joined #wesnoth-dev 20180921 23:55:20-!- celticminstrel [~celmin@unaffiliated/celticminstrel] has joined #wesnoth-dev --- Log closed Sat Sep 22 00:00:43 2018