I'm looking for a limited number of testers for a new project. Testers should have Fusion 2.5 and ideally some experience with networking extensions for Fusion such as Moo/Lacewing.
Type: Posts; User: James; Keyword(s):
I'm looking for a limited number of testers for a new project. Testers should have Fusion 2.5 and ideally some experience with networking extensions for Fusion such as Moo/Lacewing.
Should be a good one.
For what it's worth, I use F2.5 exclusively in Parallels.
That Curl object is only for Android. A cross-runtime Curl object would be great.
We had a WebGL renderer once... we can probably publish the source if anyone wants to try to get it to work.
Does this example help?
Click to launch the object.
Which extensions are you using?
I didn't discontinue it, I just dropped it from liblacewing to turn it into a separate library. I'm now maintaining the relay protocol separately here, but I haven't yet updated the MMF2 extensions...
The one you used for Windows should work just fine! Only thing to port would be the actions/conditions/expressions to the iOS SDK.
It's worth trying the profiler in DDMS to see if it can point out something specific as your bottleneck. I've been meaning to write a short guide for this.
Both of those extensions are Android compatible.
It's not present on http://forums.ouya.tv:
Proof of concept: http://clickteam.com/james/temp/StringCompressor.zip
A property allows you to specify which characters to exclude from the...
So pack the string and use byte 255 as a continuation marker, which can be followed by 0 (255), 1 (=), 2 (NULL) or 3 (newline). Then your packed string grows by only one byte for each occurrence of...
We're just waiting for some other fixes to be tested before we release the next Android beta - IAP and the crash fixes will be in there.
I don't think running Mac OS X in VMware is supported by Apple.
Which extensions are you using in the level loading code?
If you extract the RuntimeAndroid.zip in Data/Runtime/Android, you have the runtime. You can then add new extensions as classes to the Extensions package, and then to distribute them just zip up the...
Thanks, fixed in the next beta.
Yes, the parameters of the expressions are broken. This will be fixed in the next beta, sorry about that.
Should be fixed in the next beta.
Is there any reason you need to do this?
Just an update on IAP - in their quest to make it as difficult as possible to develop for the OUYA, the OUYA guys expect developers to complete and sign a tax declaration form before they'll allow...
There's a way this can be implemented, but it's quite complex. I'll look at it for the beta after next.
You don't have to load the APK onto the phone manually if you use Build & Run (for which you'll need to install the ADB driver and enable debugging mode), but you'll still need the phone plugged in.
We are working on IAP as a matter of urgency, and should have something released shortly. Sorry that we left you waiting for this.
I know how useful things like sub-conditions would be, but a lot of changes like this that seem pretty simple would actually require major changes in the editor and all of the runtimes, under the...
Or for a simpler solution if you know HTTP - Lacewing Webserver on the server-side, and the Get object on the client-side.
The Active Picture object is supported - this is the (Background) Picture object, which currently isn't.
Just a tip - a power of two sized motif will be a lot faster than a non-POT one in the ES1 renderer. For a non-POT motif, the runtime has to draw each tile separately.
It might be too small to display the content? The grey square is just the Android background color for a list.
The resolution of your application changes depending on which scaling mode you set in the application properties. The buttons etc. are then resized proportionally to how your application was resized...
It's not, however, possible to build a NES game with MMF2, whether you want to load it onto the cartridge or not.
Yes, you can read the state of both sticks and all of the buttons with the OUYA object.
Game is looking cool!
I've uploaded the latest version I can find here.
I'm sorry for the lack of updates - my priorities had to shift a bit to accommodate the Android runtime, but now that's released I'm planning...
Window controls such as buttons should be resized automatically to keep the same proportions relative to your frame across different devices. Can you send me an MFA?
There should be expressions for left stick X, left stick Y, right stick X, right stick Y.
I thought it was Nifflas I wrote the extension for, but turns out it was actually Pkeod (of Subsoap). This was before I worked for Clickteam, and I don't have a copy of the extension, the source...
I've re-homed my old extensions here. Please don't use the aquadasoft.com links, as that domain won't be around forever.
Not that I don't respect your opinion, but nobody here said anything about cloning the game or creating a "shameless ripoff", as you suggest. The question was about a moving background effect, and...
You know, using the same kind of moving background effect doesn't make the game remotely similar.
Sorry, this isn't currently possible. I can add it in a future update though.
Lots of 8x8 objects are going to take more memory and longer to draw than, for example, a single 256x256 object.
Except from the file selector. The "Load Filelist" and "Load directory list" actions are now implemented for both List and Combo.
I believe shader support is planned for the iOS runtime, so I'll be sure to port that over to Android when it's implemented.
The class name should be com.yourcompany.yourapplication.Main. In your case, that'd be com.brevo.comeonrio.Main.
We do have the source code, but the Moo extensions are not maintained by anyone.
The Lacewing extensions have always been backwards compatible with old MFAs.
Is it working OK? I don't think this was tested much in the beta.
I'm planning a big update for the Lacewing extensions soon.
Bootcamp is just a dual boot, it's not an emulator. There's no difference between Windows installed via Bootcamp and Windows installed on a machine that isn't a Mac.
Bootcamp with which version of Windows?
If it's Windows XP, you don't have to run anything as an administrator (because there's no UAC).
Same goes for Android, for the record (although with Android you can choose between OpenGL ES 1 and ES 2 in the application properties).
Can't you add your string objects to a qualifier and spread a value to assign a runtime ID?
INI storage on Android is virtualized, so you can just make up any filename you like (it will never write a physical file, as your data will be stored as SharedPreferences).
On Android, the...
I've been busy with Android stuff so haven't had any time to update this extension. I'll get round to it before too long.
You should now be able to register your Android exporter and receive the forum badge.
You can also try using WINE, which is free (and CrossOver is basically just a WINE distribution, albeit much easier to configure).
There's also VirtualBox as a free (much more limited) alternative...
It is indeed!
There will be an update shortly after release to fix the outstanding issues - but we really want to get this out there so that everyone can start bringing their games to Android.
Which object are you trying to load binary data with?
It will be in the first update.
INI saving on Android is virtualized using SharedPreferences, so you won't get a physical file on the device. But it should still save and load the data correctly.
IAP isn't supported yet, sorry.
Yes, you should include the path (better still, just use the file selector).
I confirm there's a bug, but I haven't tried to debug it yet.
Sorry Koji, this will have to wait until the first update after release (because we're currently preparing the first release build).
Another build sent your way...
Scottige, I sent you a PM.
One of your extensions using the EDIF SDK isn't installed properly. Might help to see which files you have in Data/Runtime with the extension ".json".
Is this the latest version of the extensions? If so, they need a simple change to work with the latest version of the runtime (removal of com.clickteam.runtime prefix from imports).
With DDMS (in the Android SDK tools), there's an option to "Start method profiling". Can you try this when your game is running - let it profile for a few seconds and then stop it. Then you should...
All of the scaling is done by the GPU with simple OpenGL settings, so I don't think there's much we can do in this space.
You'll need the appropriate driver to use Build & Run.
Well, you wouldn't be able to identify which controller was used if you used "Read button state" (although it would work).
Thanks for tracking this down. I'll look into it ASAP
3) Yes, but I still have to work out the best way to implement this.
Which exporter are you using?
The final build is RC6. There won't be another RC now, so anything reported will have to be fixed in the first public release.
There's an expression in the Android object to retrieve the data storage directory, which should be on the internal storage. I'll probably make that the default.
I'll change this for the first release, thanks.
Yep, looks like trying to load a non-existent file will cause a crash, which I'll fix for the first release build.
In your example, you're currently trying to save/load without an absolute path....
Hmm, I just realised you'll also have to manually extract any extensions you used. So I'm not sure this is practical after all.
We provided this feature more for extension developers really:
1. Extract RuntimeAndroid.zip from Data/Runtime/Android
2. Hold shift when selecting the option to build, and continue holding...
You'll need the WRITE_EXTERNAL_STORAGE permission, but I think that's about it. I did test array saving/loading before posting this build, but I'll double check your example shortly.
It looks like I helpfully got the flag the wrong way round. Enabling the "Play samples when not in focus" option will cause the sounds to be paused in RC6.
Sorry about this - will be fixed in the...
I'll be honest with you - I really can't understand this bug report, and the resizing modes seem to work fine on my devices and for other games. Maybe if you can isolate a specific problem I'll be...
Yes, please make a new thread for each report.
TCP is stream oriented, so you might receive the data one byte at a time, lumped together into one big packet or anything in-between. That's the purpose of the length being in the relay header.
Yes, it seems to be fixed now.
Isn't this what "Adjust window size" is supposed to do? Maybe I misunderstood the doc I was given on this.
Can you try with this version of the object? (for Data/Runtime/Android):
If you can send me the MFA of this frame - I'll see if this happens on any of my devices. Otherwise I'm not sure this will be fixed before the first public release.
Some ink effects are supported (monochrome only with OpenGL ES 2.0). Shader support will arrive after it's implemented in the iOS runtime.
There is a bug in RC5 in which ink effects can't be set...
I confirm this is fixed in the final build. Thanks again!
Thanks for the report, and sorry for the late reply.
Android is basically completely ignoring the mode I'm using to set the control size, so I'm going to use something else for this in the final...
Does C:\Android\android-sdk\tools\ant\build.xml exist?
Thanks, will be fixed for release (or in RC6 if there is one).
Sorry - I took the code for this action from the Java runtime, and it turned out to be just an empty routine. Fixed in RC6.
Can you send me your MFA?
There's also a profiler in the Android SDK tools (launchable via DDMS).
I just can't get this to happen on my devices. I might send you over a special build with some logging to try see what's going on.
I still haven't been able to reproduce your stuck touch issue, though. Is that still happening for you?
I just copied that expression from the Java runtime - it looks like it does return 0 if the string you're searching for is longer than the string you're searching in, which I've now fixed. Could that...
Fixed in RC5 (maybe today).
Ah, now I understand. I've fixed this - maybe we'll release a new build today, I'm not sure yet.
The INI object is currently just a simple key value store in Android and doesn't parse INI files. I'll integrate an INI parser in the final version.
That file doesn't exist?
Which resizing mode is this?
I've fixed the touches over frames thing, but I can't reproduce the touches getting stuck on either of my devices. I'll let you know if I find anything in the source that might be causing this.
Hopefully the OUYA should arrive later this week, so I can take a look this weekend.
Hi, sorry for the delay. There are some issues with the Date & Time object common to several runtimes (including Java and XNA), and we'll fix them all together. Unfortunately this might not be...
Doesn't seem to be Android-specific.
RGBAt isn't supported in Android, as the image isn't kept in memory once it's uploaded as a texture. I can probably do something like reloading the image from the CCN when this expression is used,...
The Mouse movement isn't supported on Android. You might want to look at the Drag & Drop movement instead, which also supports multitouch.
I've copied some changes from the iOS runtime to Android, and it seems to have fixed this. Very weird bug.
I think I know a better way to implement the show/hide keyboard actions, but it'll have to wait until the final build now.
Changing the text of a string is an expensive operation, because the string has to be rendered and uploaded as a texture each time. Strings that change infrequently shouldn't be a problem for...
Did you make sure it was completely uninstalled before installing it to the secondary HDD?
You can send it to me, but I'm not sure if there's anything I can do about this.
I think a frame with Start of Frame > Next frame should work for this.
Yes, asynchronous loading will be supported in RC4 (via an action in the Android object).
You can replace substrings with String Parser, which is supposed in iOS.
Sorry, but I don't think that will happen before release.
Yes, this is implemented in RC4.
String Tokenizer and String Parser 2 are supported at least in Flash, Android and iOS (ported by Clickteam).
Sorry, github.com/clickteam-sdk is now github.com/ClickteamLLC. Last rename, I promise!
Sorry for the wait - this is fixed in RC4.
Please could you include logs with crash reports?
Looks like the Camera object is saving a bad image file. This should not cause the Active Picture object to crash, though (in the next build, it will fail silently instead).
Hmm, you have to set wrap mode off for the image to be scaled. I can't work out what wrap mode does, but I'll resolve this for the next build.
It should be in the Contents/Resources directory of your .app (if you right-click it and select "Show package contents").
Could you post your anaconda.log?
If you do end up using CrossOver, it'd be great if you could let me know how well it works. I'm still using VirtualBox to run MMF2, but running a virtual machine is a pretty big waste of resources...
New build is up. Thanks for reporting this.
Did you enable the INTERNET permission?
I forgot to disable the debugging mode before releasing RC3, which will have a serious effect on performance. I will release a new build as soon as possible, but it might be sensible to revert to RC2...
Ah crap, it looks like I left debugging information in there. You should revert to RC2, I'll release a new build as soon as possible.
Hmm, this is a problem with the ES1 renderer. You might want to try the ES2 renderer as a temporary workaround.
Well, I added an optimisation to make sure off-screen sprites aren't drawn unnecessarily, but there was a issue with the code that calculates what's on screen.
If you build a blank application and look at the APK, you'll find the size of the runtime. Image compression is the same as iOS, so I doubt it's a problem with that.
I think this is a known issue in RC2 with sprites close to the edge of the screen. If so, it's fixed in the next build.
If you're using one of the scaling modes, any control objects should be resized automatically.
Are you sure this condition doesn't work at all? It's very commonly used, and nobody else has reported it as not working.
Looks like the hotspot was being ignored. Fixed in RC3.
Sorry for the late reply - this is fixed in RC3.
This is a known regression - fixed in RC3.
The black screen is due to the Android 2.x bug with > 1 MB resources, yes. We're going to hack around this in RC3.
OK, the crash was related to the rotation of collision masks. It's fixed in RC3.
This sounds sensible - I'll add it to the default manifest. Thanks!
I've reproduced this - looking into it.
OK, I downloaded your MFA and enabled the touch controls group. Do I need to do something else to enable the asteroids and saucers?
Really nice game, by the way.
I'm looking into this today - thanks.
What are the exact actions you're using?
Scaling and rotation only work when wrap mode is off (this is a limitation with the way the renderer currently draws patterns - might be changed in a later build).
I think it's just android.intent.action.DIAL. You don't need to add a category.
Now you can just hit "Build & Run" from MMF and it should start up on your device.
Sorry I missed this - fixed in RC3.
A very strange assumption to make. Have you benchmarked this?
Do you have an example?
Can you send me the MFA? No need to isolate anything, but some steps to reproduce would be useful.
Are you sure you didn't want "Stretch to fill" instead of "Fit inside and adjust window size"?
I think Android only accepts the full intent name, and not the shortened version you can use in Java. So where you have Intent.ACTION_VIEW, it has to be android.intent.action.VIEW.
Looks like the only way around this is to disable ZIP compression for the CCN file,...
Fernando - which Android version is the device you got that log from?
I think using ACTION_SENDTO instead of ACTION_SEND should filter out the applications that aren't e-mail.
Things like the "Repeat while mouse..." should still work correctly with multi-touch, though (if those conditions don't work, it's a bug).
Is it a multi-line edit box?
I can't reproduce this on my device, but I just fixed a similar issue (disappearing sprites), so hopefully this should be fixed in RC2 as well.
I can't reproduce this. What kind of counter is it?
Fixed in RC2.
Thanks, fixed in RC2.
Sorry, this is out of my control - I'll ask Yves about it.
Looks like you need to install the SDK for API level 14.
Fixed in RC2.
Yes, the Get object is supported.
Does it work if you unselect the "Play from disk" option for the sample?
It looks like the menus you describe have been deprecated by "action bars" in Android 3.x: http://developer.android.com/guide/topics/ui/menus.html
Sorry, extensions aren't supported in the demo version.
If you send me your MFA, I can try to profile it to see what's causing the slowdown.
Check the "VIBRATE" permission in the application properties.
1) 3) This is a limitation of how sub-applications are rendered on Android - because there's only one GL surface, sub-applications are rendered as sprites and can't appear over window controls.
Does this also crash in the Java runtime? Can you isolate the issue and send me an MFA?
Interesting - I've seen this crash log in a few different "corrupted" MFAs, so you might just have found the cause.
Any chance of a crash log?