Build 295.10 - Release Version

Quote from Yves

Yes that's what I do of course, and I can't reproduce the problem you mention, the mouse cursor is not visible.

Please look at this video.

Please login to see this media element.

Hi! I've tried it on half a dozen devices and Android versions (v10-13). Two devices I have here at the moment are a Galaxy S23 Ultra (Android 13), and a BLU View 2 B130DL (Android 10).

Thanks!

A minor feature I think would be logical and helpful and probably not break anyone's existing code, would be if multiple "break" statements could exit out of multiple layers of child events. So 1x break exits one layer, 2x break exits 2x layers, etc. Right now a break can only exit out of one control structure, no matter how many you put, so there's no real way to control logic flow in nested child events other than to use flags to track them or disable a code group containing them.

Quote from Fernando

It seems fine here. Tested it with Android 10 (Lenovo tablet) and Android 14 S23 Ultra. Note that you set min 9 and API 11 with build 34. Why don't you try with min 5.1, target 14, and build with API 34? Let me know.

Just to be sure, you may download the APO object for android here. The same mentioned in a previous post and available also in the discord Clickteam channel: Please login to see this link.

Thanks! Changing the min and target builds did it! The kcpica.zip file that you linked to here seems to be more recent than the one on Discord, but it was changing of the min/target that did the trick, because after it worked I went back and set them back to what I had, just to see if it was that or the kcpica.zip. Thanks for your help with this!!!

Quote from Pixelthief

A minor feature I think would be logical and helpful and probably not break anyone's existing code, would be if multiple "break" statements could exit out of multiple layers of child events. So 1x break exits one layer, 2x break exits 2x layers, etc. Right now a break can only exit out of one control structure, no matter how many you put, so there's no real way to control logic flow in nested child events other than to use flags to track them or disable a code group containing them.

This would be useful. Or a break function with an input parameter for how many parent events to break out of.

Quote from SirEatAlot

This would be useful. Or a break function with an input parameter for how many parent events to break out of.

DarkScript features this already

At the moment I'm working on normalising object selection/access to Fusion's internal variables inside the DarkEdif extension SDK. Right now, every platform does object selection a bit different.

But I've now got all sorts of variants of foreach and fastloop working on Windows, Android, iOS and Mac.
In one action, you can:

• stop a fastloop after it runs that iteration's events (Fusion's Stop Loop action),
• get it to skip just the reset of the events for that loop iteration (a non-Fusion continue;),
• or stop it instantly (a non-Fusion break;).

With the foreach, you can choose the behaviour you want. If you run a foreach on Active with qualifier Good, should only "On Each Active" trigger, or "On Each Group.Good" trigger too, or both, etc.

And you can abort function-inside-function calls until it hits a matching function name.

I'm a bit confused, I just opened Fusion 2.5+ and got an update, but it's still showing as 295.9. Checking for updates reports that my version is up to date. (not Steam version).Please login to see this attachment.

Quote from Phi

DarkScript features this already

At the moment I'm working on normalising object selection/access to Fusion's internal variables inside the DarkEdif extension SDK. Right now, every platform does object selection a bit different.

But I've now got all sorts of variants of foreach and fastloop working on Windows, Android, iOS and Mac.
In one action, you can:

• stop a fastloop after it runs that iteration's events (Fusion's Stop Loop action),
• get it to skip just the reset of the events for that loop iteration (a non-Fusion continue;),
• or stop it instantly (a non-Fusion break;).

With the foreach, you can choose the behaviour you want. If you run a foreach on Active with qualifier Good, should only "On Each Active" trigger, or "On Each Group.Good" trigger too, or both, etc.

And you can abort function-inside-function calls until it hits a matching function name.

Sounds great! Did the object selection crash bug get fixed?

Dear Clickteam,

Congratulations on the new forum!

I found a few issues but I guess you're already working on them. First is when I post my post I see next to the New word on its title that there's a Disabled mentioning in green. Not sure what it means. And at the bottom of the post it says: This post will remain inaccessible for others until approved by a moderator. Don't remember that it was like this in the old forum.

Besides that, many of my product badges are not showing under my profile picture to the left. I guess it's something that you're working on, but just wanted to mention it as I see that some of the missing product badges are showing for other members.

As for the issues with the product that I would like to bring up:

I less work on Clickteam nowadays, but sometimes use it for building mockups and for loading old project for fine tuning them, and found this issue that you may want to fix, as many would probably appreciate it:

In the Frame Editor, if you set the zoom level to 100% and copy and paste objects (which is commonly used when building the levels layout), the pasted object appears near the mouse pointer correctly. However, if you're working on small resolution projects and zoom to a bigger zoom level, such as 200%, the pasted object appears in a diagonal offset from the pointer to the bottom right. And it's getting even worse if you zoom to bigger zoom levels - the more you zoom-in the bigger the offset the object get pasted. And in a very big zoom level the object is even pasted off from the visual screen and you have scroll around to see it.

I also have a suggestion to make the Paint Mode better: it would be very nice if the selected object that we pick from the objects library could be snapped to the grid in preview when just hovering with it on the grid, before actually placing the object on-to the grid with the left mouse button - it's something that was exists back in Klik & Play and The Games Factory 1, which from some reason was removed in Fusion. And due to the offset issue I described above, drawing tiles with the Paint Mode (especially when you zoom in to work on small resolution games) feels like a joke, as whatever you paste appears in a ridiculous offset, and you always have to exit from the Paint Mode to manually select and drag the offset tiles where you want them to be. So if a snap to grid preview could be re-added to Fusion that'll make things much easier. I guess that others would appreciate having it back.

Thanks

Quote from SirEatAlot

Sounds great! Did the object selection crash bug get fixed?

Yea, that one was to do with how actions looped through objects, and only turned up when you created 2+ objects in a fastloop (not a DS loop) that ran 3+ times. I eventually ported it to Android runtime, where I could see what it was doing, and it was isolated and fixed thusly.

Run application from current frame doesn't work if we use the Run option -> command line option. It probably overwrites or deletes the /J{Frame number} command. However, it works if we add a '/' before our own command, but I wouldn't want CF to require the use of '/' for my own commands to work correctly. By the way, could you please add a combo box with the option to choose the frame to run during the editing of global events? Switching between global events and the tested frame when there are many frames is cumbersome. I suggest that the combo box automatically sets to the frame from which we entered the global event editor or the one we were last in. Similarly, when we are in global events and click on the event editor or event list editor, it should navigate to the selected frame.

I already posted this on the Discord but I'll post it here as well just in case:

On Android, if a shader file uses LF line separators instead of CRLF, it will crash the application at load time. The getVertexBlock and getFragmentBlock methods in CEffectBank try to use the index of "\r\n@End" in the substring method, which causes the crash with LF files since it doesn't exist, and the if guard only checks if "@End" exists. The "\r\n" should probably be removed.

Quote from Yves

I see what you mean, but this desn't happen on my machines...

I can duplicate this phenomenon.
That happen on build R295.10 and previous build is no problem.

Suggestion: A way to find deactivated events