gles.js - a lightweight WebGL renderer for Android

Discussion and feedback on Construct 2

Post » Mon Mar 30, 2015 12:26 pm

@Ashley Well I wouldn't mind the customer needing to install proper DirectX files, even now we are finding that customers need to install DirectX updates or the latest versions of their graphics drivers in order to avoid the game having the strangest glitches/bad reviews, so it's not really that much of a difference in our case.

There probably are nasty bugs hiding somewhere in each layer that you are working on top of for sure, but the ability to fix it or do a work-around at one of the lower levels rather than relying on a slow giant corporation to contact another slow giant corporation seems better for consistent stability (as a driver or DirectX/OpenGL issue will still exist in our games but be further out of reach as it may require a driver update, and also modifications to Node Webkit, or even Chrome itself). I know well that it's tough enough getting anything to run properly across a variety of machines, so it definitely is nice that the runtime is able to slowly expand its compatibility and stability every day, but it also feels like it's taking a lot longer for fixes to trickle down the chain than even making an in-house webkit style runtime would take to fix some of the most pressing bugs right now.

Yeah, it is pretty cool that Google is putting some help into fixing their own product, but that's also a situation where you are just as helpless as your customers, as the problem is further up the chain. At least with the AMD bug you were able to fix it and then release it as soon as you had the solution. Google could tell us the secrets to all of our Node Webkit issues but if we aren't compiling the source code ourselves we can't fix them.

And that's great, I really love C2 and all the hard work you've done since your Tigerworks days. I do see the future of HTML5 and WebGL, but my customers aren't buying the future they just want to play my 2D game on a PC that is more than fit to play Half Life 2, and any bugs that I created/I am able to fix myself in my capx, or that you can fix in a future update, are A-Okay :)
"Construct 4 lets YOU make advanced games! (but not play them)" Construct Classic - Examples Kit
B
95
S
33
G
14
Posts: 2,120
Reputation: 16,493

Post » Mon Mar 30, 2015 12:43 pm

@Ashley

I'm frustrated too, and I can see this comes at a bad time after all the nonsense with CocoonJS


C2 apps works smooth on iOS thanks to @Ludei
C2 apps works much better than Crosswalk on Android thanks to @Ludei

so maybe it's good time to make peace agreement with company that helps so much to C2 devs?

you should understand that people are using CocoonJS not because they are Ludei's fanboys,
but maybe they have no alternative? [ just to mention @Egyptoon case ]
B
18
S
6
G
1
Posts: 783
Reputation: 4,167

Post » Mon Mar 30, 2015 12:52 pm

Jayjay wrote:At least with the AMD bug you were able to fix it and then release it as soon as you had the solution.

I think you are missing the point. For 9 months we were completely hosed on this problem, and I did not believe we would ever fix it because AMD were impossible to reach and absolutely no help at all. We fixed it by the good fortune of having someone randomly pass by with a way to work around the driver bug, hardly a process that can be relied upon. That is nowhere near us being responsible for the technology and being able to resolve issues independently, even when working with native tech. Google by comparison are amazingly reachable and it's possible to communicate directly with engineers working on the relevant components, and updates go out every 6 weeks (which is a double edged sword, but compare that to driver updates that effectively never happen for anything not currently on sale).

@szymek - we do support Ludei's Webviews. Last I heard, Canvas+ still had memory management problems, which was the #1 complaint, and still has a long list of features not supported (Web Audio, WebRTC, XML parsing...) that lots of users want. Even if you can get by without them, a lot of other users need that, and only the webviews can provide it.
Scirra Founder
B
367
S
217
G
82
Posts: 23,263
Reputation: 183,841

Post » Mon Mar 30, 2015 1:13 pm

@Ashley hmm, my point was that the same bug being someone else's problem (while still technically being our problem as paying customers and therefore also a problem for you) doesn't make it any better. Google release updates that try to be one-update-fixes-all, and so when they introduce bugs for us it's not really their problem and it takes more effort to get the whole community to report the bug so that they realize it's important.

Sure, they have connections and contacts that you don't, but they also have way more projects and people to support in their day job. Perhaps hiring an expert in low-level driver/OpenGL issues who can experiment and debug and fix while you can focus on the next features would make native extremely viable, and everyone would benefit even if the cost of C2/C3 was increased for it.
"Construct 4 lets YOU make advanced games! (but not play them)" Construct Classic - Examples Kit
B
95
S
33
G
14
Posts: 2,120
Reputation: 16,493

Post » Mon Mar 30, 2015 6:47 pm

Egyptoon wrote:@digitalsoapbox
Can you please mention 1 single professional game made in C2? just one :)
do you know why ? Because of what i already said in my reply.
Check games made in Unity and compare :) it'll be fun to do.
and I'm still saying C2 is Awesome (to play with) in your spare time and have fun with it, not for real serious game production.
C2 is just about 1 step close to be perfect, and this step requires REBUILDING it from scratch and include its OWN native exporter, without this it'll stay USELESS.


I suppose that depends on your definition of professional. How many "professional" games can you think of that were made with Unity when it was in its second iteration? Because I can't think of any, and that's when I started using Unity - which I still use, often for my day job. I didn't see any "professional" games start popping up until v3. As for C2, The Last Penelope is as polished as any 2D game I've seen done in Unity.

Your other points are comments that can be made about any development software that's purchased early in its life cycle. None of what's said above is remotely unique to C2's development.

To be clear: you're welcome to criticize all you like, but that doesn't make you right. *shrug* to each their own.
B
57
S
33
G
13
Posts: 413
Reputation: 13,175

Post » Mon Mar 30, 2015 6:55 pm

Jayjay wrote:@digitalsoapbox as someone who has made both 2D and 3D games from scratch I can confirm that it actually is better in some cases than being completely held hostage by a third-third party (as it's not just Construct 2 and driver vendors we are waiting on, it's also Chrome, Node-Webkit, and the HTML5 standard in general).

The main reason we don't re-invent the wheel is because we want to actually finish the games we make instead of producing tonnes of unfinished abandonware before we finally settle on choice of programming language, memory usage/allocation, design pattern(s), and other low-level things. Prototyping is a great time to switch engines around, and being half-way finished or further is not.

Middleware that works roughly the same way every time can be accounted for with work-arounds, while random engine-breaking updates can stall or kill a project entirely.

The reason why I personally want native desktop is because I know that Scirra can and has done an AMAZING native DirectX 9 runtime before with Construct Classic: http://www.scirra.com/construct-classic

Literally the C2 editor exporting to CC would be all I'd ever need ( as CC had a buggy editor but pretty solid runtime...except on Vista :P ).

Point is though, we want to use Construct 2, as a game dev tool it is literally the best we've ever seen, especially for 2D games. However, when you want to release commercial titles and pay say $500 for a copy of a "professional business edition of a game development tool" for each member of your indie studio to do so, you really want said game engine to actually perform the tasks it promises, especially after you've just raised all your funds in a Kickstarter and now have to release a product on the budget you initially planned (and other, less specific, cases).

CC is based on entirely different, and outdated, technology. I'm sure there's multiple good reasons for it not being directly importable into C2, but from what I've seen of CC porting stuff as a more manual process, while potentially frustrating, seems entirely doable. It's also the nature of changing engines - it's not like Unreal or Unity have had painless upgrade or porting paths either.

Unity is where a lot of serious C2 developers seem to be going, and each one I've seen hasn't blamed the editor in Construct 2, just the lack of control that Scirra has over runtime/export.

Also as a side note I've been following Sombrero for some time and it's looking really good, reminds me of the Friendly Strike series I played in my Clickteam TGF/MMF days :)


Except using Unity, you're still held hostage. It's just a different company's tech. There isn't even enough hours in the day for me to list all the bugs I've run into in Unity over the past 8 or so years, including some that have persisted across multiple major releases. Don't even get me started on their GUI tools. I've seen few bugs with C2 that can't be solved by rolling back to an earlier release of C2 or NW, not depending on beta software releases - which you shouldn't count on anyway as being production-ready.

If you want to do 3D, use Unity; if you want fast cross-platform 2D that isn't especially hard to get running at 60fps on almost every system I've tried it on (FireTV is a bit slower), I still think C2 is the way to go. And speaking of Unity: see if they're still offering the $75/mo deal, I've been paying for Unity that way since they announced that plan and it's a very reasonable price considering what Unity is capable of overall, including outside of game-specific development.

With MS announcing Universal App support for XBox One including WebGL/HTML5/Javascript-based apps, there is now little in the way to prevent C2 games from ending up on major consoles. In fact, I'm meeting with a MS rep next week to discuss doing that very thing with Sombrero (thanks for the kind words on that btw). Any info I find out about getting C2-based stuff on XBox One I will, of course, share on these forums.
B
57
S
33
G
13
Posts: 413
Reputation: 13,175

Post » Mon Mar 30, 2015 7:56 pm

What would really make C2 exporting better would be some competition for desktop (and potentially, console) distribution. Right now it boils down to: NW.JS. And that's it.

Firefox may be terrible now, but project silk looks to be a dramatic improvement that should bring the fox on par with chromium (when chromium is behaving that is). What about Mozilla making their own version of NW.JS?

Ditto for Microsoft: what about a 'redist runtime' version of Spartan that offers support across Windows Desktop, Mobile, and XB One?

As for mobile, it would be great to hear from more people using CJS Webview+ to distribute on iOS8, because so far all I've heard is that it is phenomenal performance-wise, but the only person I know of who is using it for sure is silverforce.

I'm becoming more and more convinced that Android is a poor platform to begin with, given it's poor commercial potential, poor performance relative to the raw system perf (even for native apps), high input latency, massively fractured OS/Hardware, dismal state of chromium/crosswalk for android, etc...
Don't lose your work. Backup your game with Dropbox.
B
43
S
10
G
10
Posts: 1,106
Reputation: 9,147

Post » Mon Mar 30, 2015 8:27 pm

I gotta say that I can totally understand @ashley for not "simply rewriting the entire game engine" and sticking to HTML5 and JS.

I mean jesus christ people :shock: Just re-writing an entire engine in a different language is a MASSIVE piece of work! Even without considering that JavaScript with its prototypical style is quite different from most other programming languages.. Even more importantly having everything work with Chrome / FF / IE is already a struggle; Imagining to have to write the plugins for Windows, IOS, Linux, Android, IPhone, other smartphone OS'es, keeping track of all their quirks + additional differences in between versions and flavours - and doing all that as a 2 man team??? WOW! What a nightmare!

Mobile performance is a bitch in JS and the best we can do is optimize and reduce draw calls, keep the JS idle and wait for V8 improvements, ECMAScript 10? (LOL) and moores law to take care of the performance gap.

You can call me a fanboy, I really don't care, because I'm not.. I can see the limitations. But this is a 2 man team building a HTML5 game engine that ports well... to HTML5. And you bitch about that :lol: ? I guess I am gonna yell at the guy running the chinese restaurant in the city next time I eat there because he does not serve pizza :roll:

It's all good to complain and vent. But when you bitch at a product that does what it was designed to do because it does not do what you want even tho you know what it is designed for.. well... I think it is time to pack your things and fuck off :)
B
40
S
11
G
4
Posts: 665
Reputation: 8,413

Post » Mon Mar 30, 2015 8:40 pm

@Schoening yeah i agree i think some of these people(not all) don't realize what going into a native exporter. Look at unity they have 100s of people working on their exporters and they still suffer on alot of devices because android like @Tiam said has just become bloated and a lot of manufactures have terrible drivers or have skins that make gaming impossible(looking at you Samsung)
So you have to write things in your exporter just for these devices. Android has a terrible fragmentation issue. Your not even guaranteed OpenGl they all have different versions sometimes they even have a custom version. Making a Native Android exporter would be a massive under taking that even Unity or Unreal cant really get right(they export to a much smaller subset of devices).
Last edited by volkiller730 on Mon Mar 30, 2015 9:31 pm, edited 1 time in total.
B
42
S
17
G
2
Posts: 850
Reputation: 6,209

Post » Mon Mar 30, 2015 8:53 pm

@Ashley: I know you're probably sick of me pointing to this, but I would really love for you to read these two articles and then tell me if this is not a technology worth considering:

http://notes.underscorediscovery.com/haxe-from-1000ft/

http://notes.underscorediscovery.com/haxe-entry-point/
B
31
S
4
Posts: 57
Reputation: 1,520

PreviousNext

Return to Construct 2 General

Who is online

Users browsing this forum: AltairVitor and 1 guest