(CC vs) C2 performance (community?) benchmark #1.

Discussion and feedback on Construct 2

Post » Sun Feb 09, 2014 10:28 pm

@liaeb - is that with WebGL support? Nice to see the results are pretty close.
Scirra Founder
B
403
S
238
G
89
Posts: 24,653
Reputation: 196,143

Post » Mon Feb 10, 2014 10:12 am

Sorry, Yes that's using WebGL.
B
9
S
3
G
1
Posts: 112
Reputation: 1,827

Post » Wed Feb 12, 2014 7:17 am

Pretty surprising, actually. Now (when and if we get more results) i'll have to cook up something to compare the physics engine and other such features...
B
10
S
1
Posts: 163
Reputation: 1,436

Post » Tue Feb 18, 2014 6:31 pm

bump. anyone? The amount of results posted is ridiculous - you guys really should participate to see what the performance of C2 is on differenct setups / how it compares to CC (optional).
B
10
S
1
Posts: 163
Reputation: 1,436

Post » Tue Feb 18, 2014 8:10 pm

In relation to the cc vs bit
Perhaps a lot dont want to install an older construct version.
Me neither, the tool is absolete and would require me to learn some basics from an older tool which I will not be using in any near future.

In relation to the c2 tests
Looking at the details fo the benchmark test, I see no comparision to how I would run and use things. In fact, I would probably make my programs opposite of your test whahaha

Seriously, when creating events I tend to check the effects of each piece trying to create every event to work as efficient as possible.

Knowing the max limitations of a flawed programming approach is useless. Not calling your test flawed, but, if someone would implement something like that in his or her game .. that would simply be bad.

Once a project starts to get more shape and getting bigger, then its about time to test that programs effects on different hardware.

You could perhaps, make or take an existing game, which works and runs decent, and use that for tests.
You could do it with different engines and game types, now that would be usefull information.

Who dares wins
B
57
S
17
G
21
Posts: 1,878
Reputation: 19,592

Post » Wed Feb 19, 2014 4:48 am

@lennaert You know, i'm not really sure i get your point.

1.) I'm not making anyone download / learn CC, they just have to download the pre-made binaries and run them, then compare the results in the spreadsheet.

2.) Also, it's older, but nowhere near obsolete. Sure, it has bugs, but it is still ahead of C2 because it's built to work specifically in Windows with DirectX 9.0 and as such it has all sorts of useful features (good luck with file manipulation within the mess that is C2 (in this regard, at least) - cross platform compatibility has its drawbacks), and also has much better hardware support (none of that WebGL stuff, just the old pixel shader 2.0 requirement).

3.) It's a benchmark and it's made for a very specific purpose - testing the rendering performance of sprites and the process of transforming them in various ways. It's made using the most basic features in CC/C2, no behaviours or plugins are used to make sure that just the developers' work is being compared. I'm not sure how else you would make a sprite rotate or rescale, mainly because there's only one way for each of those actions, so i don't really have the slightest idea what you are going on about there.

4.) Flawed programming approach? Oh, you mean Java with its terrible performance issues (burn)? I don't see how CC is flawed in any way, apart from the fact that there are unresolved bugs in the editor (C2 has plenty of those too) and that C2 is being updated and CC isn't (not that it couldn't, it's open source) - that doesn't make it flawed in my view, at least.

5.) You didn't read the post, did you? I said rendering test. I'm not porting an entire game, especially if the reception is this bad.


Either way, i guess i'll just let the post die, seeing as there is no interest in it - just the next time someone claims that C2 is faster than CC (unless it's proven by a trusted source, e.g. the developers) or vice versa, just link them here...
B
10
S
1
Posts: 163
Reputation: 1,436

Post » Wed Feb 19, 2014 9:01 am

1. One the binaries would give me an indication of CC performance. I have no use for that. Thus comparing to c2 means nothing to me.

2. It is obsolete. It is no longer being development and there is a newer version that has taken its place, which is being further developped with far more features.

3. objects created/manipulated every few ticks in a continious stream, thats not a game mechanic, thats more like a stress test.

4. I meant exactly what I said, like the mechanic used in point 3. Its not the program (cc or c2) thats flawed, the created mechanic for the test is flawed in relation to comparring it to something usefull like parts of an actual game engine.

5. yes I did read it, and rendering is fine, but i just called the aqcuired information through these tests useless. Thats why I suggested doing similair tests with programs that actually mean something to the majority of developers here, providing relevant information. That is why people are not interrested. At least, I think.
And because your not porting an entire game, you will not see the performancce difference in rendering from one game. which is something I am betting a lot more would participate in.






lennaert2014-02-19 09:03:29
Who dares wins
B
57
S
17
G
21
Posts: 1,878
Reputation: 19,592

Post » Wed Feb 19, 2014 10:21 pm

@lennaert
Oh, so what you are saying is that the drawing performance and the capacity to handle large amounts of sprites are not important factors when choosing an engine to develop a game on? And that stress tests are irrelevant (e.g. the guys who made Furmark wasted their time)?
Pfft, people these days...
B
10
S
1
Posts: 163
Reputation: 1,436

Post » Thu Feb 20, 2014 1:31 am

really nice test man!! i think ashley should update the performance post
C2 is almost catching CC with the native performance

but for the OP, if you don't mind my humble opinion and about your friend concern on CC being a viable option: use C2(unless you really love CC); we all know that CC have a good native performance but choosing a engine only because of the FPS/object count is not a wise move (specially in this case since C2 is almost there); by the time his project hit the medium size, C2 probably will be more reliable and with more possibilities, is growing faster each day, already have tilemap object which give a boost in performance and development speed and by the way...
multiplayer is coming!!
B
48
S
5
G
6
Posts: 210
Reputation: 6,297

Post » Thu Feb 20, 2014 6:01 am

I think it's interesting, and wish more people would try it out (I'll try it when I have time). Reason being:

CC may not be developed anymore, but it's speed was pretty amazing compared to other 'visual programming' game builders of the same era. If C2 can whomp it, or even get close to it (with the handicap of C2's javascript engine vs CC's native export) then that's a real feather in C2's cap.

Obviously, it wouldn't be a good idea to start a major project in CC at this point in time. I don't think that's the point of this post
B
44
S
10
G
10
Posts: 1,106
Reputation: 9,202

PreviousNext

Return to Construct 2 General

Who is online

Users browsing this forum: No registered users and 1 guest