Ejecta and construct 2

Ashley

I have found this and it looks not bad - Maybe it is also interesting for you:

Please take a little Look also to the Video on the Site:

Ejecta

KMag

Very interesting!

Hi,

Yes, i also think it is amazing and hopefully Ashley can check this out for us!

KMag

the performance look faster :D it's nice !

I can't really see what this offers over directCanvas or CocoonJS. These platforms both support a lot more than just canvas and audio, and both have done a fair amount of Construct 2-specific optimisation and adjustments, as well as having services behind them that help make building and monetisation easier. So while it looks like a cool library, I'm not sure it's right for Construct 2.

Ashley

hmm...i have tested out Cocoonjs. For me it won't work. Allways i have Problem's with it on my Android Devices. When i build an APK File - in most Cases i can't use it because an Error Message Pops up that says "Parsing-Error" for example. Also i had some System-Reboots when i start it. I'm sure i have done no Error - i followed all the Tutorials Step by Step.

Now, im searching for a stable way to bring the Games "Stable and Fast" to my Devices.

I'm switched now to IOS-Devices...

Hopefully you can find the Time to Clarify if this is a better Solution for Developing to IOS-Devices...

Ejecta looks good for me and the Makers wrote "Many HTML5 Games run out of the box, or with minimal modifications ? with better performance, better sound support, Game Center integration and more."

KMag

It seems Stencyl guys solved the problem in Stencyl 3.0 by rewriting the entire engine in Haxe. With it they get Windows, Mac, Linux, Flash, iOS, Android, HTML5, and more. All native. Not sure about performance , yet i've seen mentions of it performing very well in iOS. In Windows it generates C++, so native code. This is 'confidential' info btw. Only know it because i ocasionally lurk at their forum.

This is Ejecta , it appears, is just a simplified CocoonJS like tailored to be used with ImpactJS engine.

kiyoshi - if the performance is in fact there, that actually sounds like a really good idea.

I'm hopeful that construct 2's modular export system would mean that the editor itself would require no modification to implement something like that, it could all be in the exporter.

Ashley - I know you've got your hands full now with html5, but once you've got the time/resources to work on another exporter, if haxe's performance is actually that good, it sounds like a great idea - you get tons of platforms at once with actual native performance and without the limitations/issues html5 has, that often leave you dependent on others to improve/fix. Maybe for construct 3 or something?

Yeah. I think that's a pretty reasonable idea. If Stencyl guys decided to use it, it must be viable. When they announced that they were overhauling the engine and creating native exporters for all platforms i doubted. It seemed an almost impossible feat. With Haxe it's indeed possible. More specifically, HaxeNME. I believe that's what they used. The idea is great since you can still export to HTML5 and keep the html5 vision goodness, but export nativelly to all platforms too. So it's completelly future proof.

Arima

Good Post - i agree with that. I know, Ashley work hard Day by Day, but anyway i think he should hold his eyes open and test out things like haxe etc...

KMag

KMag - have you reported the errors you got to Ludei or us with .capxs to reproduce? It would be good if we can fix those issues so other people don't run in to them. I don't see any reason those problems can't be fixed, and I would guess it would involve a lot less effort to fix those problems than shifting to a whole new engine which is a vast amount of effort.

kiyoshi, Arima - it sounds like an interesting strategy, but the big problem with using a new language is third party plugin compatibility. We already have a large library of third party stuff, and it's all JS designed to be run in a browser. I really want to avoid compatibility headaches where every plugin works on just a few certain platforms, since it can make it prohibitively difficult to find a reliable set of features that will work on a broad set of platforms.

wow... that HaxeNME looks pretty awesome,

I browse their game showcase at:

http://www.haxenme.org/showcase/

and i have that Rune Raiders game.. it's a fun game and plays pretty smooth even on my old ipod touch.

if Construct 2 can somehow incorporate HaxeNME, it will be the king of 2D game making tools <img src="smileys/smiley36.gif" border="0" align="middle">

and we no longer have to depend on any other external tools to export to other platform...

Right now we're at the mercy of CocoonJS, AppMobi, Awesomium, Phonegap when we try to export to other platform, even for the browser version, only Chrome / Firefox that somewhat play nicely with HTML5 game.

but with HaxeNME, aside from HTML5, we can even export to Flash if we deem it necessary, pretty sweet....

It's probably gonna take a lot of work though, so i'm not holding my breath here.

EDIT:

looks like Ashley posted right when i'm replying,

I would sacrifice of not using any 3rd party plugin for the platform of choice that HaxeNME offer <img src="smileys/smiley2.gif" border="0" align="middle">

(most 3rd part plugin doesn't work with CocoonJS / AppMobi anyway)

I can't speak for others, though I think many would agree, but it would be well, WELL worth it to me to get native performance and increased options with less issues if it meant not using any of the previous third party js plugins. Especially on mobile!

Admittedly, I try to avoid using third party plugins anyway generally, but 'vanilla' c2 has plenty of functionality regardless that most games don't really need them. I think it would be a lot better for the long term viability of c2 if plugin development shifted to haxe - then plugins would hit everywhere rather than just html5. I think in the long term it might actually fragment things less than if potential exporters, third party or otherwise, were made for other platforms one by one.

I understand that would be a difficult choice to make, but I think it's a vastly superior path!

To have Construct 2 without the limitations of HTML5... it boggles the mind.

Let it be known that a move would have my full support, even if it took a year with no updates, and no third party compatibility.

This really excites me. I know it would be a tough decision, but vanilla HTML5 just seems to be having so many problems.

sqiddster

Yes, i agree with you! When Ashley need some Money support from us, from me, he can have it!

KMag