Construct 3 - many questions (native exporterts)

Discussion and feedback on Construct 2

Post » Sun Oct 04, 2015 5:36 pm

shinkan wrote:@TheRealDannyyy
I'll be able to do that probably within a month or so. I have NDA over my head so I can't show anything from that game until it is finished and released to public :)

Let's change it up a bit, can anyone else in the C2 community upload a smooth running android game for me to test please?
It doesn't have to be uploaded to Google Play, you can also PM me the link to download it.

(Once again I'm just curious about that, I dont want to support any of the groups here.)
ImageImageImageImage
B
56
S
20
G
76
Posts: 635
Reputation: 43,359

Post » Sun Oct 04, 2015 7:14 pm

I think that the only way to show the world that a game engine is relevant on a platform (android) is to build a game on android with it that runs at a normal speed - on statistically the most common device (samsung galaxy?).

In that case, scirra could make game jam games to test out its own engine and actually publish them as free demos on google store. Then when selling the engine, include the source code of one of those said games to show people how it was made to run at a normal frame rate.

Does scirra have a google store account? Are they also a publisher?

In the case of GODOT engine, the guys who made it actually made it to make games with it themselves-commercial games, published a number of games to different devices and then open sourced the engine - giving it to the entire world for free.

Their marketting model is different, but their engine has actually proven that it is relevant for commercial games.

In that case, then scirra could get in touch with some of the developers of succesful games on steam made with construct and ask them to share the source file of a level/demo. As for android games, I havent seen any commercial android games made with construct. In that case I have to agree with a lot of people here that the framerate is terrible when run in xdk container. But that was a year ago when i tested it and technologies do evolve. I still believe that html5 will become more usable
B
38
S
14
G
4
Posts: 426
Reputation: 5,686

Post » Sun Oct 04, 2015 7:48 pm

There is already an android game showcased as "made with C2" example. (Click HERE)
I don't want to be offensive with the following list but:
  • It has constant framedrops. (Mostly when objects are created into the layout)
  • Quiet long loading times between layouts. (Compared to a 3D game made with other engines)
  • Huge filesize (Because of the 3rd party wrapping)

It is quiet hidden from the other "successful" games featured on the main page
but I guess that is just how the webmaster choose it to be. (OR maybe on purpose?)
I would like to hear a statement about that game from @Ashley or the developer himself if possible.
Maybe some words by @Ed Farias himself answering the following question for example:
What do you think about the current methods of exporting, did you experience any issues?

P.s. Before people start blaming my mobile device, I got a LG phone with 500mb ram free to use, quad core processor 1,5ghz and a quite good graphics card. (Nvida Tera)
ImageImageImageImage
B
56
S
20
G
76
Posts: 635
Reputation: 43,359

Post » Mon Oct 05, 2015 10:21 am

@paradine - please, link me to every example of CPU-bottlenecked games you can find. I want to investigate them but am so far unaware of most of them. I assure you I do not intend to ignore them, but with hundreds of posts made to the forum every day it's very easy to miss posts.

I am also skeptical of polls for features in general. We did feature polls that showed the vast majority of users out of hundreds of voters said they wanted multiplayer. I was concerned that even with C2 doing much of the heavily lifting and hard work, networking is still a difficult subject to work with in general, so it would probably still be hard to use. Even when I specifically described how it would likely be difficult to use, still people voted for it in huge numbers. Then when we released it, very few people actually used it. As far as I can tell it's because... it's a difficult subject to work with in general!

To me this proves large numbers of users will vote for features based purely on hype. They imagine that everything will be perfect and brilliant and there will be no shortcomings or tradeoffs and it will be super effective and easy to use. Of course reality rarely ever lives up to that.

So I don't really care if a million people vote for native exporters. I regularly see that people blame HTML5 for performance issues when really it's a hardware bottleneck. I can see that people think "just make a native exporter, it's not that hard right?", whereas I can see it's a colossal project with severe ongoing impact on the product development. So even if everyone voted for it, I am not persuaded they have a proper understanding of what they are voting for, as happened with the multiplayer polls.
Scirra Founder
B
387
S
230
G
87
Posts: 24,249
Reputation: 192,240

Post » Mon Oct 05, 2015 1:24 pm

Other 2D game making softwares (with native exporters) can make nice performance mobile games easily. They have a lot of such mobile game examples:

Fusion:
http://indiegames.clickteam.com/category/itunes

GameMaker:
http://www.yoyogames.com/showcase

Stencyl:
http://www.stencyl.com/game/showcaseMobile/

GamesSalad:
https://gamesalad.com/creator/featured-games

Construct 2 was released in 2011. There are a lot of users. But how many mobile games were released.? Any nice performance mobile game was released?
B
7
S
1
G
1
Posts: 34
Reputation: 804

Post » Mon Oct 05, 2015 2:21 pm

I have an idea that might be easier to make instead of a full native exporter.
@Ashley since you tend to look through all the CPU-bottlenecked games:
...please, link me to every example of CPU-bottlenecked games you can find. I want to investigate them...

What about this, a build-in event debugger that checks for those, performance breaking events?
So I mean a system that checks for mistakes like this for example:
Image

Tells the user that the events on line XX (in this case 4 & 5), can be optimized.
And if the user wants those events to be optimized, the event debugger will change it to this for example:
Image


With this kind of system we could keep running on the HTML5 track and also make games with well performance.
What do you guys think about that, @Ashley what do you think about this kind of system?
ImageImageImageImage
B
56
S
20
G
76
Posts: 635
Reputation: 43,359

Post » Mon Oct 05, 2015 2:27 pm

@Ashley
intel XDK and crosswalk 11 - it runs 40-45 fps on my mobile device.

The newest Intel XDK and the newest Crosswalk - it runs 45-46 fps on my mobile device.

Native APK - it runs 57-58 fps on my mobile device.
B
44
S
18
G
33
Posts: 115
Reputation: 19,941

Post » Mon Oct 05, 2015 2:30 pm

@TheRealDanny, I must confess that I don't see how your example shows a more efficient alternative. I don't know how those events would be rewritten in JavaScript, but the inclusion of a function that way would surely result in a few more lines of exported code where the function is declared and referenced. More lines of code equals more for the CPU to do....
I only occasionally visit - I'm learning C# for Unity, but c2 is still a respectable game engine imo....
B
73
S
19
G
66
Posts: 2,198
Reputation: 42,188

Post » Mon Oct 05, 2015 2:45 pm

BoyHK wrote:Other 2D game making softwares (with native exporters) can make nice performance mobile games easily. They have a lot of such mobile game examples:

Fusion:
http://indiegames.clickteam.com/category/itunes

GameMaker:
http://www.yoyogames.com/showcase

Stencyl:
http://www.stencyl.com/game/showcaseMobile/

GamesSalad:
https://gamesalad.com/creator/featured-games

Construct 2 was released in 2011. There are a lot of users. But how many mobile games were released.? Any nice performance mobile game was released?

non
even airscape don't have good performance ! it's just so laggy !
i made this game in 1 week and i spend more than 2 weeks to optimize it :D
and still...well see it yourself... :roll:
generals-war-hd_t162049
Last edited by mahdi71 on Mon Oct 05, 2015 3:01 pm, edited 2 times in total.
B
16
S
6
Posts: 243
Reputation: 1,755

Post » Mon Oct 05, 2015 2:57 pm

Colludium wrote:@TheRealDanny, I must confess that I don't see how your example shows a more efficient alternative. I don't know how those events would be rewritten in JavaScript, but the inclusion of a function that way would surely result in a few more lines of exported code where the function is declared and referenced. More lines of code equals more for the CPU to do....

This is the recommended way of eventing by a professional C2 user, however that was just an example. (Link to the post is HERE)
To be honest @Colludium, I don't think that Ashley will have the time to look through thousands of games and optimize them the way they should be.
This debugger system would do the work for him, that was my main idea behind that.
ImageImageImageImage
B
56
S
20
G
76
Posts: 635
Reputation: 43,359

PreviousNext

Return to Construct 2 General

Who is online

Users browsing this forum: Darth Crusher, Litoghvgvgv and 10 guests