The Big Question

Discussion and feedback on Construct 2

Post » Fri May 11, 2012 2:05 pm

How does appmobi's 'Direct Canvas' compare to Cocoonjs?
I am very interested in knowing. Tests, thoughts, anyone?
B
90
S
30
G
24
Posts: 3,189
Reputation: 32,400

Post » Fri May 11, 2012 2:16 pm

They should both be about equally fast, although because neither supports both platform yet you can only compare different devices, which of course makes it unfair. The real difference is how each platform does publishing. Ludei haven't yet sorted that out though, they're still working on their build system.
Scirra Founder
B
359
S
214
G
72
Posts: 22,952
Reputation: 178,580

Post » Fri May 11, 2012 2:27 pm

If that's the case, I would be interested in seeing some performance comparisons - i.e. the 'how many sprites on screen' test or something like that, perhaps for a range of devices, just to see how much we are 'allowed' to do before encountering FPS drops.
B
90
S
30
G
24
Posts: 3,189
Reputation: 32,400

Post » Fri May 11, 2012 2:30 pm

Good idea, I'll do some performance tests at some point. I doubt they'll show up a CocoonJS vs. directCanvas performance difference, but they should still give a good rough idea of sprite counts you can get away with and mobile vs. desktop performance.
Scirra Founder
B
359
S
214
G
72
Posts: 22,952
Reputation: 178,580

Post » Fri May 11, 2012 2:38 pm

Exactly, that would be perfect. You should show mobile browser vs directcanvas/cocoonjs vs desktop performance.
B
90
S
30
G
24
Posts: 3,189
Reputation: 32,400

Post » Mon May 14, 2012 8:24 pm

@Ashley, something VERY important if possible would also be to compare performance directly with native apps.
I understand javascript will be slower by default, but how much is the question.
B
90
S
30
G
24
Posts: 3,189
Reputation: 32,400

Post » Mon May 14, 2012 9:13 pm

@sqiddster - as our recent blog post said games are 90% rendering, so once you start using hardware accelerated rendering most of the important differences disappear. Of course our squares-test shows higher performance on native, but that's mainly because the test actually measures the CPU time to issue a draw call.
Scirra Founder
B
359
S
214
G
72
Posts: 22,952
Reputation: 178,580

Post » Mon May 14, 2012 10:35 pm

Are you saying, then, that games exported with these softwares are proven to have performance almost identical to that of native games?
Because that would be amazing.
B
90
S
30
G
24
Posts: 3,189
Reputation: 32,400

Post » Tue May 15, 2012 2:23 pm

Well... the best test would be to write Space Blaster in native and compare the framerates. However that's a lot of work!! I think it's best to describe performance as "native-like", because it is using hardware accelerated rendering which is the main (90%) point. However the logic will still be 5-10x slower as usual, but for games that don't have much logic it will be more or less native equal.
Scirra Founder
B
359
S
214
G
72
Posts: 22,952
Reputation: 178,580

Post » Tue May 15, 2012 10:35 pm

Here's my test.
Displaying a simple background image,60 fps (vs. 30 fps Old HTML5 Web Export)
But whenever there is a single sprite moving around it goes down to 10~15fps.
Idk exactly what is this directCanvas,but it's not native.
Yet,wait for a year and I bet HTML5 Games will be playing like native over mobile devices.
B
5
S
1
G
2
Posts: 44
Reputation: 1,470

Next

Return to Construct 2 General

Who is online

Users browsing this forum: No registered users and 11 guests