Build 295.10 - Release Version

Quote from marbenx

Was posting about this on the discord, but there might be some kind of bug with collision events/qualifiers/flags

Please login to see this attachment.

Both objects in the frame have "Physics - Static" movement. Flag 0 is OFF for the green object and ON for the purple object - but the collision event seems to affect both even with a condition that checks the flag. (I can work around this by running a foreach loop on the object and testing in the loop instead)

If you change it to 'overlapping a backdrop' instead of 'collides with the background' it works.

Quote from aenever

If you change it to 'overlapping a backdrop' instead of 'collides with the background' it works.

Correct - but collisions are much more performant than overlaps (especially in my intended case where I'll have possibly hundreds of these objects on the screen at once)

I've noticed the same thing appears to happen with the 'animation has finished playing' condition as well - scoping on the immediate if type conditions seems not play nice with other conditions that might scope the same object/qualifier

Hello, would it be possible to add the "Display properties" option to the objects when you work in the Global event editor?

I often open the properties of the objects when I work in the Event Editor so I can change or add values to the objects.

But I can't do that in the Global Event Editor and I have to manually find the object in the Objects folder or on the Frame.

Can you add an option that exports the events as pseudo code that we can use to feed-in to ChatGPT ? This might help (me) with troubleshooting ,

Quote from Yves

When you copy events to the clipboard it also copies them as text, does this help? (hmm I think you need the 2.5+ DLC for this)

Wow, thanks, I have never thought of that. It works like a charm. It simply does not export the number of the event , but everything else is perfect and ChatGPT understands it. Thanks !

Quote from Yves

Hmm, a bit risky if the object you want to modify exists in several frames and is different in these frames, you would need to make sure the base frame is set to the one you want to modify...

So it's not possible because of the possibility of several objects with the same name but with different values and animations?

Quote from Yves

It's "possible" but the change would affect a single object only, in the base frame, (unless it's a global object with editor synchronization options). This could be confusing from a user point of view.

Ok, thank you. It can't be helped.
It would be useful but I get it that it's problematic.

Quote from Ach

Can you add an option that exports the events as pseudo code that we can use to feed-in to ChatGPT ? This might help (me) with troubleshooting ,

Hey!

I am wondering, how you go about this? I would love to have assistance troubleshooting ometimes.

For example, say you copied the events from CTF2.5, and pasted them as text. What do you ask ChatGPT? Does it know what CTF2.5 is?

Quote from punchbird

Hey!I am wondering, how you go about this? I would love to have assistance troubleshooting ometimes.

For example, say you copied the events from CTF2.5, and pasted them as text. What do you ask ChatGPT? Does it know what CTF2.5 is?

Hi, for example - Google Play reports crashes that I still can't troubleshoot (4-5 months later). I asked ChatGPT what might the root cause; it game me 3-4 options. Yesterday I uploaded my whole game (I use only Global Events so this is easy) in it and asked to implement the first solution . It pointed me to the events that it thinks are problematic and gave me a fix - it suggested adding a new global variable that checks for a state (in my case). For example - this is an event that is suggest (with the fix) :

* User clicks with left button on button_AddShot
+ frame > Menu_frames
+ Requested_shots = 0
+ Only one action when event loops
+ AdMob: Is Banner Visible?
Special : Set Requested_shots to 1
AdMob : Destroy Ad Banner
please_wait-Global : Set direction to ...........
please_wait-Global : Set position at (585,660)

So far it looks pretty reasonable. This week I plan to implement its suggestions and check whether it has solved my problem or not.

There is one cool feature missing that I think could easily be added. Look at (X,Y)

Please login to see this attachment.

Quote from Ach

Hi, for example - Google Play reports crashes that I still can't troubleshoot (4-5 months later). I asked ChatGPT what might the root cause; it game me 3-4 options. Yesterday I uploaded my whole game (I use only Global Events so this is easy) in it and asked to implement the first solution . It pointed me to the events that it thinks are problematic and gave me a fix - it suggested adding a new global variable that checks for a state (in my case). For example - this is an event that is suggest (with the fix) :

* User clicks with left button on button_AddShot
+ frame > Menu_frames
+ Requested_shots = 0
+ Only one action when event loops
+ AdMob: Is Banner Visible?
Special : Set Requested_shots to 1
AdMob : Destroy Ad Banner
please_wait-Global : Set direction to ...........
please_wait-Global : Set position at (585,660)

So far it looks pretty reasonable. This week I plan to implement its suggestions and check whether it has solved my problem or not.

Display More

That's actually pretty incredible. I am going to purposefully create some issues, and see if it can solve them!!!

Quote from punchbird

Hey!

I am wondering, how you go about this? I would love to have assistance troubleshooting ometimes.

For example, say you copied the events from CTF2.5, and pasted them as text. What do you ask ChatGPT? Does it know what CTF2.5 is?

ChatGPT *thinks* it knows what Fusion is, but in my experience it often has no clue. It has a tendency to make up nonsense, like really specific instructions that relate to features that simply don't exist at all in Fusion. I find ChatGPT helpful with coming up with general equations that I use in my code, but when it comes to stuff specific to Fusion, I find it to be of little help at all.

on that subject.... imagine having AI built into fusion X.x ; where you could simply add an object, go to it's AI tab, and tell it what you would like. eg; i want this object to move about in this zone (x,y-x,y) or i want this object to be a platform character for the player, he can double jump, slide, has inertia... etc and it just popped the code out in a group ( or the objects events )

at the rate AI is moving, i dont think this is unrealistic.

Yeah I think you'll have more success with ChatGPT or other AIs if you ask generalized questions about algorithms, than for technical questions or syntax. Most of the simpler questions of how functions work and platform movements and capabilities of Fusion can best be answered by just pulling up the old school help tab and searching them up

Is there any chance that the below feature/amendment could be made for the next version?

Quote from Volnaiskra

I requested this once before, but I'm not sure if it was noticed. This would (hopefully) be a relatively small tweak to the UI, but it would be a massive quality of life improvement for me (and presumably others).

Please login to see this picture.

Would it be possible to introduce an option that allows selected layers in the Frame Editor to be invisible? Currently, clicking on the eye button of the currently selected layer doesn't do anything. I often find myself wanting to quickly toggle the current layer's visibility - either to see what's behind it, or as a way of seeing what's on the layer itself (by seeing what disappears when I hide it). But the process is needlessly convoluted for something that should be really quick and simple:

1. You want to hide the visibility of layer (A), but you realise you can't because it's the current layer
2. So you click on a different layer (B) so that A is deselected.
3. You hide the visibility of A
4. You realise that by selecting B, you've just inadvertently made it visible, because it was invisible before, but Fusion automatically shows any layer you select.
5. You want to make B invisible again like it was supposed to be, but you can't now, because now it's selected, so now hiding it is forbidden.
6. You start looking around for some other layer you can temporarily click on just so that you can re-hide B. You need to be careful to find one that's currently visible because you don't want to make the same mistake again.
7. You find another layer that's not currently invisible (C) and click on it
8. Now you can finally re-hide B.

The above annoyance would be eliminated if Fusion would just let any layer be visible or not, independent of whether or not it's selected. This is not unorthodox behaviour: lots of programs work this way (eg. Adobe programs like Photoshop, Illustrator, After Effects). Even the Fusion event editors work this way, allowing you to deactivate events regardless of whether they're selected, or select deactivated events without forcibly activating them.

So my request is for something like a "allow selected layers to be hidden" option in the preferences. This would have two effects:

1. the eye button would successfully toggle the visibility of a currently selected layer
2. selecting a new layer would no longer forcibly show that layer if it was previously hidden.

Display More

I have been asking for a video playback object that works on all platforms for years. But there is no answer at all...

Can add the ability to output anm files from a runtime exe?
It would make it so much easier to create user mods if we could import a custom image into the active object while it is running with a frame load action and package it as anm.