To be honest, I'm starting to get a bit disillusioned with html5 myself. I wasn't happy with the idea of html5 only initially, but was eventually convinced that it was a good idea since the whole industry was moving to support it, but after years of waiting I'm still frustrated and disappointed by several things.
- Html5 has no manual memory management. You cannot dump anything from ram, you have to wait for the browser to do it for you, and you have no idea when it will do so, and garbage collection can cause stutters. Layout by layout loading only handles what textures are sent to the video card, so does not help with this. This problem makes large mobile games pointless to bother with because since they don't free up memory they just crash. I wanted to put loot pursuit on mobile, but it's just not an option, and I only found that out after months of work, which was quite upsetting. I've been using construct for about 7 years and consider myself pretty knowledgeable about it, so it's quite likely less experienced users won't realize this limitation without first putting in wasted work too. C2 loading everything at the start also causes long loading times, but hopefully something can be done about that while still using html5.
- Javascript's poor performance. On desktop it's not as much of an issue, but on mobile it's obvious, and is made worse by the fact that iOS doesn't allow compiling JavaScript in apps, making things like cocoonjs like 2/5ths of the speed of safari in my tests, and safari was already behind native. It's also disappointing that on all platforms, C2 will always be steps behind native exporters in speed (this segment is again about the speed of the code, not the rendering - C2's rendering is written in webgl, which as I understand, IS basically running natively and in one of my older tests was faster than CC's rendering). That said, on desktop I very rarely encounter speed problems on an amd 4400+, which is old enough that it can't even decode 1080p video smoothly, so that says good things about the speed there, but we're still always behind what we could have, and what the competition has.
- Reliance on third parties for export, which despite all their efforts, which are appreciated, have proven to be still not ready for prime time, years later. Relying on third parties means scirra cannot control the quality of their own product, which is disappointing as scirra has a terrific editor.
This also means we're beholden to their whims, like chrome, and therefore node webkit, disabling hardware acceleration on all XP and vista systems. I mean, seriously, wtf.
The truth of the matter is C2's competition - almost all the competition - has native exporters, and what we're experiencing is why. Clickteam has it. Game maker has it. Unity has it. The list goes on and on.
Ashley - if you read this, when you respond to the idea of native exporters (I could be wrong about this, I apologize if so but it's the impression I get), it seems like you think we're talking about you making your own complete browser engine, because you talk about how you couldn't compete with google or such's team of engineers. We're not. What we want is just native games, no web tech that games don't need. Yes it would make a lot of third party plugins obsolete, but I just don't care as I don't use them that often. i really think html5 is holding C2 back.
I don't believe that it's impossible to create native exporters that work better than chrome. There's thousands, possibly hundreds of thousands of games out there that do exactly that. Chrome is trying to support everything a web browser can do. Games need only a fraction of that. We don't need CSS or whatever web tech. Trying to support them all is not an achievable task for a small team, but supporting all of the functions needed just for games has been achieved by many, many small teams out there, including you guys with the native exporter in construct classic.
Yes, it might take a while to code. I understand that C2 is solidly on the html5 train, and it's not good to try to get off a train while it's in motion. However, if someone could be hired to work on it concurrently, then work on the current to do list would be able to continue at the same time and no delay in updates to c2 would occur.
However, even if you can't find someone to help, then even working by yourself I think when most or all of the stuff on the to do list is done, then a switch to native should happen. I don't know if 3d could be considered at that point as well, but I also think it should be incorporated, and that might be a good place to do so, or to at least lay the framework for it to be implemented later on.
C2 has the best editor out there that I've tried, hands down. It shouldn't be held back by its exporters. Exporting to native would make it a powerhouse would dominate the competition. As it is, and this is really painful to say considering how much I love this program - I'm finding it harder and harder to recommend c2 to people unless they fall into very specific categories and don't mind the caveats.
Well, nobody is making you use C2, perhaps you should try some other software.
Here we go with the weekly native thread disguised as something else.
That's really not the right position to take. Scirra's event editor isn't available anywhere else, and without it I wouldn't be making games at all as I am terrible at traditional code. Also, it just drives away people who would otherwise become paying customers.
Scirra's got a great product, but it is not without its flaws, and we shouldn't discourage discussion of them, as discussing them constructively can result in a better product. If a lot of people are complaining about something, then it's likely a real problem.