Build 295.10 - Release Version

Quote from Yves

Works fine...

That is unfortunate because it's 100% reproducible on my system... Is there anything else I could share to help isolate this?

EDIT just in case: I'm running Windows with 150% display scaling.

Quote from Fadzior

I made an example of how zoom could work in fusion, unfortunately it only works on dx11 because the scaling is different in dx9. Unfortunately, resizing the window is only possible when it is minimized

Please login to see this attachment.

Is there a chance of a built-in zoom function in the future?

Oh wow that actually works pretty well

Quote from Yves

I've tried on several machines, 150% and 200% scaling, no problem.

Do you use the stand-alone version of Fusion? If so, maybe try to re-install the latest update patch?

I installed a fresh copy from Steam just to be sure and it still happened... It looks like this when I run dpi.mfa from the editor.

Quote from PkR

Sure, here it is. I use an AMD GPU if that helps.

Tested on my NVIDIA machine and didn't have any issues.

Additional info: My monitor resolution is 2560x1440. The issue doesn't happen if I set the desktop resolution to 1920x1080, but it returns when I set it back to 2560x1440.

Quote from PkR

Additional info: My monitor resolution is 2560x1440. The issue doesn't happen if I set the desktop resolution to 1920x1080, but it returns when I set it back to 2560x1440.

I use a Ultra Wide monitor (3440*1440, 100%) and i don't see any issue, i tested windowed and full screen. I also tested on my Surface Laptop Studio (2400*1600, 150%) and no issue.

Seems to be your installation or system that have a problem.

When you build an EXE, does it have the same issue on your system?

Quote

When you build an EXE, does it have the same issue on your system?

It only happens when I run it from the editor. I don't know what kind of system issue would cause this only in Fusion editor and nowhere else, and only with that combination of resolutions...

Update: I made a version that breaks at the 1920x1080 resolution, and a user on Discord was able to reproduce it. Here it is.

It seems that for the glitch to happen, the application window must be exactly half the resolution of the screen, e.g. 960x540 if you have a 1080p screen, or 1280x720 for 1440p in my case.

Quote from PkR

Update: I made a version that breaks at the 1920x1080 resolution, and a user on Discord was able to reproduce it. Here it is.

It seems that for the glitch to happen, the application window must be exactly half the resolution of the screen, e.g. 960x540 if you have a 1080p screen, or 1280x720 for 1440p in my case.

Can confirm halving the resolution of my monitor reproduces the issue on my end using a 3440x1440 ultrawide.

Quote from PkR

Update: I made a version that breaks at the 1920x1080 resolution, and a user on Discord was able to reproduce it. Here it is.

It seems that for the glitch to happen, the application window must be exactly half the resolution of the screen, e.g. 960x540 if you have a 1080p screen, or 1280x720 for 1440p in my case.

With this mfa dpi_1080 the problem does appear on my PC

-------------------------------------------------------------------------------------------------------------

With dpi_1080.mfa the error appears and in the compiled version too
To make the error disappear in the exe version, you have to minimize it from the taskbar and restore it again
But when starting the application the bug appears

The fact that minimizing and restoring makes it look fine makes me think the fix might be easy, who knows? Maybe some internal variable isn't updating correctly?

Quote from Yves

Thank you for the example! I've zero idea why this happens but I was able to fix it in the build 296.

Is there a timeframe for when the update will be available? I'm wondering whether to wait for update 296 or continue updating applications in the current version.

Hello, I found a bug where when placing a value name or string in a loop, it corrupts the loop index.
Please login to see this attachment.

actually it may not be an error but you need to define the loop name and then it will work

Please login to see this attachment.


Please login to see this attachment.


maybe it's a bug though, because it should work in a loop too

Request for Expanding the iOS Exporter Limitations

I am preparing the iOS version of a game that has already been released on Android, but I am facing a critical issue.

On iOS, I have discovered that if the number of events per frame exceeds 8192, events without immediate conditions start getting ignored sequentially from the earliest ones.

The provided URL contains a sample MFA file where you can check a counter that does not work properly on iOS.

Please login to see this link.

My app’s main frame contains over 25,000 lines of events, making it impossible for the game to function correctly.

The game is already complete, and at this stage, splitting it into multiple frames would be extremely difficult.

I would greatly appreciate your support in resolving this issue.

Thank you.

Quote from SONNE

My app’s main frame contains over 25,000 lines of events

25000

Do you really need 25,000 events to do what you're doing?
I think you should start optimizing that.

Can you show a link to your game (Android Store)? I'd like to see it.

I'm up to 46377 events in one project's frame, its one of the reasons we had to deal with the max event cap on the .exe compiler and Yves was able to solve it there, hope the iOS one can be solved (and that I can keep mind under 2^16)

Quote from TonyThunder

Can you show a link to your game (Android Store)? I'd like to see it.

Thank you for showing interest in my game. Here it is:
Please login to see this link.

Although the design is quite simple and not very polished, it's an RPG with enough content to take thousands of hours to fully complete.
To minimize loading times during frame switching, I intentionally kept the frame structure minimal, which resulted in a large number of lines.
I focus on compact event structuring and efficient event group management to optimize performance and prevent unnecessary processes from running.

Quote from Pixelthief

I'm up to 46377 events in one project's frame, its one of the reasons we had to deal with the max event cap on the .exe compiler and Yves was able to solve it there, hope the iOS one can be solved (and that I can keep mind under 2^16)

That sounds like an incredible scale! I can't even imagine what a project that could potentially reach 2^16 might look like, but I'm sure it's something amazing.
Wishing you great success!

Quote from Yves

Sorry for this issue... In the iOS project, search for the assemblePrograms routine (in Events / CEventProgram.m), and search for this line at the beginning of the routine:

I appreciate your quick response! I'll follow your instructions, apply the fix, and test it out.
Thanks a lot!

Quote from Yves

Sorry for this issue... In the iOS project, search for the assemblePrograms routine (in Events / CEventProgram.m), and search for this line at the beginning of the routine:

WORD evtAlways, evtAlwaysPos;

and then replace WORD by DWORD:

DWORD evtAlways, evtAlwaysPos;

This fix will be in the build 296.

I followed your instructions and changed WORD to DWORD. The original issue was successfully resolved!

However, after making this change, the game now crashes at the location shown in the attached image. This issue did not occur before changing to DWORD.

When I asked an AI for advice, it suggested that the crash might be caused by changes in the memory layout of the structure due to the DWORD modification. However, as I am not an engineer, I do not fully understand the details.

Is there a way to avoid this issue? If manually fixing it is difficult, I will wait for update 296, hoping that it will resolve this problem.

I truly appreciate your hard work and support.
Best regards,

Please login to see this attachment.