SirSpunky's Forum Posts

  • I agree with TheWyrm that it's a problem that new feature suggestions get lost in the General forums. Good ideas get lost and people will feel less encouraged to share them. I also think this creates unnecessary duplicated posts and more work for the administrators. Just like most forums or ticket systems have bug reporting, they usually have a dedicated section for feature suggestions.

    Like I mentioned earlier, I personally think a better solution would be a new forum for it, called Feature Suggestions or similar. There would be no voting, but you could estimate popularity by number of posts, and people would not have to register on a new site to post ideas.

    I think it should be a forum for brainstorming C2:s short- and long-term future, so anything goes. Of course, it's just brainstorming, so there should be no obligation for official answers. The best official posts there would probably be estimates of how big an undertaking it would be, to help people adjust their expectations.

  • Trust me, multiplayer will definitely be put to good use eventually. Although I've yet to learn how to implement the details, I'm planning to use it in all my bigger projects now, even though they're mainly singleplayer games, which I didn't even consider before. It's an extremely powerful and game changing feature, and I really do like your serverless architecture which I think is much better suited for indie projects.

    Regarding feature requests, the only thing I miss is a dedicated forum for it. I think we would see better ideas, better discussions and less duplicated posts if we put them in a separate forum. And I think it would be easier for you to evaluate their popularity.

  • Amazing work, Roger. Works great now, thanks a lot!

  • Just to be sure, I modified the demo to use Bullet + Wrap behaviour instead of Sine, and I still get the freezes. Here's the capx:

    https://dl.dropboxusercontent.com/u/242 ... test2.capx

  • rogerwang

    Thanks for taking the time to look into this.

    Here's my log from NW v0.10.3 win32 from your last post, which gives me freezes:

    https://dl.dropboxusercontent.com/u/242 ... gpu_nw.htm

    Here's the log from my Chrome v37 installation, where I don't get freezes:

    https://dl.dropboxusercontent.com/u/242 ... chrome.htm

    I'm running Windows 8.1 64-bit.

  • —

    That's interesting. Are you running Windows 8.1 as well?

    I used the simple capx from Giganten. It uses a sine behaviour for movement, which causes freezing for me after some minutes (not in preview Node-WebKit but full export):

    https://www.dropbox.com/sh/iyia6jc30cv3 ... .capx?dl=1

    I'm still curious what difference there is between Node-WebKit previews and full exports, as there's obviously some important change that's causing the freezes to appear in the exported version only.

  • You can download the latest Node-WebKit from here:

    github.com/rogerwang/node-webkit

    Just scroll down to "Downloads" and download v0.10.3 for your platform (Windows, Mac OS etc). Unzip and run the program, and paste the following URL in the address bar:

    dl.dropboxusercontent.com/u/2425456/Nwtest/index.html

    Watch the slider for some minutes to see if a freeze occurs (give it some time, maybe a few refreshes). I tried again with the latest version for Windows. On first try I waited 2 minutes then got a 5+ second freeze. Can someone else try the same thing and post the result?

    In Chrome v37 or preview Node-WebKit from within Construct 2 I get no freezes.

    rogerwang

    I haven't managed to find Chrome 35 unfortunately. The link you posted does not seem to work any longer I'm afraid (it just redirects to the official download page for the latest Chrome version)

  • Here's a URL with the HTML export of the app:

    https://dl.dropboxusercontent.com/u/242 ... index.html

    I tried running it in stand-alone NW 0.10.2 win32. It seems to have similar freeze issues. Although it seems to work slightly better, after about 10 refreshes I got a big freeze of 10 seconds after about 1 minute. It's a bit tricky to reproduce as it requires some waiting, and seems to be randomly distributed.

    Haven't tried Chrome 35 yet.

  • Ok, this is interesting. I opened nwtest.capx from Giganten and tried it out on Windows 8.1 64-bit on an i5 2500K with Radeon HD 6850 (C2 178).

    When I preview as node-webkit from inside C2, everything works great, but when I make a real node-webkit export, the win32 program freezes occasionally for 1-2 seconds. It goes down to 0 CPU during the freeze and I cannot exit the program until after the freeze. Usually it freezes once or more just a few seconds after I start the program, but sometimes it runs fine for minutes. I haven't managed to get any freezes in the preview node-webkit so far.

    Ashley What difference is there between running a preview node-webkit and making a full node-webkit export?

    Maybe it's related to some js minification that occurs in full export but not preview?

  • Nice, that's some very nice artwork in my opinion! I love the style!

    I think the text during the tutorial was a bit small on my mobile, otherwise worked great. Good job!

  • That's a great cause. Wish we saw more charity in computer gaming. Keep it up!

  • I suddenly cannot build any more apps in XDK. I get "An error occurred while building the application" although I didn't change any settings. I'm guessing they're having temporary server issues.

  • 40 fps at 4000 stars in Chrome (on 3 year old i5 2500K, Radeon HD 6850)

  • Try Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Try Now Construct 3 users don't see these ads

  • When printing the CPU usage within the app it does seem to be lower in the compiled app compared to the Chrome mobile browser, about 50% lower. Although my fps is good at the moment, it seems to jerk/lag slightly at irregular intervals.

    I also noticed a delay when playing audio (already preloaded), which also happens in Chrome mobile browser and CocoonJS. It's about 0.3 seconds long on Samsung Galaxy S3, which is noticable and make timing-based audio difficult.

    As I'm experimenting with a music game, my biggest issue is that Audio.PlaybackTime in C2 does not take this delay into account, so I have to manually add 0.3 secs to Audio.PlaybackTime to get the actual playback time of the song. I'm mainly worried that this delay varies across devices, which might mess up the timing. Still, this is an improvement from CocoonJS where Audio.PlaybackTime did not work at all.

  • Good work, this has improved a lot since I last tried it in february! Now my main issue is performance, which is not yet on par with CocoonJS. I suspect this might be an issue with Chrome rather than XDK though, as I have the same lag when running it in the Chrome mobile browser (which I don't have in the same project in CocoonJS on my Samsung Galaxy S3).