Build 292.26 - Release

Why can't I always update these extensionsPlease login to see this attachment.

When use Sub-Application more than one, then object size and position change.
Also size of frame for sub application change object size and position change.
See video
Please login to see this link.

The Screen Zoom object is supported the Android runtime. Set scale action worked fine, but, Shake, Linear, Smooth, Elastic, Hot spot, Destination are not work.:(

Please login to see this attachment.

Quote from wuxuechen

Why can't I always update these extensionsPlease login to see this attachment.

IIRC CEM can only update e-install extensions.

Feature Request / Bug : Incorrect Path

If there's a wrong path somewhere on any frame, then none of the frames will run. Can we please have an option to ignore an incorrect path and still run the frame?

To overcome this - I copied the particular frame I need to a blank .mfa - but it keeps asking me for the right path for all the objects in the original .mfa even though not a single one is in the frame that I'm copying and pasting.

Feature Request / Global Values

Any chance that we could have the option to copy / paste the names of global values between different .mfa's - see above why I need to.

I have two major issues in this update
1. ControlX slows the game to about 1/3 speed. I'm using controlX to emulate key presses when using an XBox controller, but when this is active it makes the game run at 1/3 speed.
2. Physics seem to work differently based on framerate now. The game I'm working on runs at 90fps, and now some (but not all) of the movements are 1.5x as fast, only going back to normal if I lower the framerate to 60 (which I can't do)

What can I do about these? I was not aware there was going to be an update so I didn't make a backup. Would I be able to revert to an old version?

You can revert to prior versions from the Product Owner's Lounge.

Yes. Download the previous version and run it.

Marv

Guys, he said that his Fusion was updated without him realising it. Unless he downloaded and manually installed the EXE while sleepwalking, that means he's a Steam user (as indeed his icon confirms) ;). So I don't think the Product Owner's Lounge is applicable to his version.

[MENTION=31107]theknewgreg[/MENTION]: I don't think there's a way to revert to an older version on Steam - at least not that I know of. But is there a chance that somewhere along the line you checked/unchecked the "machine independent speed" setting in your application settings? The problems you're describing sound like the sorts of problems that that might cause. Also, while I don't use the Physics system, I do use ControlX, and I haven't noticed any slowdowns.

Please login to see this picture.

Also, you can try verifying the integrity of the files via Steam, just in case the problem was caused by a faulty update.

Please login to see this picture.

Sorry if neither of these suggestions helped.

Quote from ASD

The Screen Zoom object is supported the Android runtime. Set scale action worked fine, but, Shake, Linear, Smooth, Elastic, Hot spot, Destination are not work.:(

Please login to see this attachment.

this was mentioned in previous post only angle and zoom are ported for now.

Hmm I like how there's the Collision Mask thing that returns if a position is inside an obstacle
So I was wondering if something like this would be possible for other kinds of objects, it could be separated into two parts, one that's general where you can specify which object by the name (accepting qualifiers as well) to check if in the exact position it's part of the collision mask of the object, and a second part where the collision mask condition comes from inside the object (mostly actives), where it would tell if a pixel is part of the collision mask but relative to the objects position (just like the thing to get a pixel color but for getting the value from the collision mask of the object)

I think this would be an great addition, since there are a lot of people that do intricate collision systems that don't really use backdrops

Edit: Another thing that could be interesting would be being able to check if the object was in another position, if it would be overlapping a obstacle or another object

Found a Bug:

If i have 2 windows in tile mode, one in event list editor and one in frame editor. Then click on deactivate group event, fusion crashes on the window pop-up.

Havent been able to recreate it, unsure if its this specific MFA or if its something im missing. Attached the MFA in question, go to event 18, put fusion in Tile mode (window-> Tiled) And then have the event list editor and frame editor open at the same time, double click on the Actions in event 18 and it should crash.

"Event editor: object list corruption if the frame contains more than 2000 unique active objects. This limit has been increased to 30,000."

Thanks for this one.

Major (for me anyway) bug:
If you have an Active DirectShow set to "Use Media Foundation With DirectX 11" you can not get the duration of the clip.

Someone mentioned a possible bug where double clicking and event (number or checkmark) sometimes makes some objects do not appear on expression editor (this sounds similar to that bug with qualifier of multiple object types inside of behaviors)

Please login to see this attachment.

I'm having a lot of issues with this version and activating / deactivating groups - when a frame is restarted multiple times. It seems that the on group activation is not being run once, but being run multiple times - I've replaced it in some cases with "run this event once" and it fixed the problem.

For example, a pause where I activate / deactivate a group works 100% the first time, but when the frame is restarted and I activate it - then it does crazy stuff - that I didn't even program in - for example, there's an object that on group activation it's meant to create once - now when I deactivate the group - it's creating multiple copies of that object at the point that I deactivate the group i.e. it runs the group activation event continuously until it no longer can.

In trying to make a simple example I did notice this:

Line 23 is not being executed when I would expect it to - I would expect that if line 22 adds 1 to global value A and it now equals 4 - the frame would restart - the start of frame events would run and when it gets to line 23 it will reset global value A = 0. But that is not what's happening - you can see by the pattern of objects that are created.
However, if I move line 23 into group 3 - then it resets correctly. But where it is - Fusion is just skipping over the chain of commands when global value A = 4.

I don't know if this example will help or is tied into the above, I just noticed it as I was trying to make a simple example to illustrate the first 2 issues.

Quote from fwtep

Major (for me anyway) bug:
If you have an Active DirectShow set to "Use Media Foundation With DirectX 11" you can not get the duration of the clip.

Never mind. It's working. It's my brain that's not.

Quote from Janette5

I'm having a lot of issues with this version and activating / deactivating groups - when a frame is restarted multiple times. It seems that the on group activation is not being run once, but being run multiple times - I've replaced it in some cases with "run this event once" and it fixed the problem.

For example, a pause where I activate / deactivate a group works 100% the first time, but when the frame is restarted and I activate it - then it does crazy stuff - that I didn't even program in - for example, there's an object that on group activation it's meant to create once - now when I deactivate the group - it's creating multiple copies of that object at the point that I deactivate the group i.e. it runs the group activation event continuously until it no longer can.

In trying to make a simple example I did notice this:

Line 23 is not being executed when I would expect it to - I would expect that if line 22 adds 1 to global value A and it now equals 4 - the frame would restart - the start of frame events would run and when it gets to line 23 it will reset global value A = 0. But that is not what's happening - you can see by the pattern of objects that are created.
However, if I move line 23 into group 3 - then it resets correctly. But where it is - Fusion is just skipping over the chain of commands when global value A = 4.

I don't know if this example will help or is tied into the above, I just noticed it as I was trying to make a simple example to illustrate the first 2 issues.

I don't know exactly how those "Storyboard Control" actions work, but I do know that they don't activate immediately. So, for instance, if you do this, the frame will never end (if you reverse the order of the actions then, conversely, the frame will never restart):

Please login to see this picture.

That tells us 2 things. Firstly, the later storyboard control action ("restart current frame") is overriding the earlier storyboard action ("end the application"). Secondly, it shows that even after we've told it to end the application, Fusion keeps reading and executing the events.

So, in your example, you increase Global Value A in #22 and tell it to Jump to Frame. Fusion will Jump to Frame, but first it will continue to read and execute the rest of the events....so it gets to #23, sees that Global A is 4 (since you just made it 4 in the previous event), and sets it back to 0. Then, when the frame eventually restarts, the global value is already at 0 (and not 4, as you would have expected), and event #3 never gets executed.

By outputting to the debugger at various points, we can confirm that this is what is indeed happening. Global Value A is only ever at 4 for the briefest moment:

Please login to see this picture.

Thank you [MENTION=15682]Volnaiskra[/MENTION] - I kept thinking I was missing something obvious but couldn't put my finger on it. It's strange that I hadn't picked up on this before.

The other issue was my mistake too - one of my start of frame events lost its Start of Frame command and so was continuously running on a restart when a value triggered true. Sorry.

I don't know if I already asked for this, but could be possible to add a expression to get the name of the current loop?
This could be used for scoping an such