Problem with font in Russian

[mdvadim]

Dear Phil,
Something is wrong with the latest version of Archi. Part of the relationship names can not be read due to an incorrect font.

[Phil Beauvoir]

Please explain to me what the problem is with the font. I don't know what I am supposed to be looking for. Also, more information would be very helpful - Windows version, Java version, what you expect to see, what you actually see, etc.

[mdvadim]

Dear Phil,
Starting with version 3.0 and up to version 4.0b9 inclusive, all fonts of all signatures of relations in Russian were displayed absolutely correctly.
However, after moving to version 4.0Pre1, part of the rows with relationship types changed names of relation type to a meaningless set of characters. You see this in the screenshot, in the list of model elements, the top elements of the relationship have a readable view of the Cyrillic alphabet, and those below are already missing. All other elements of the model have preserved the correct display of the name in Cyrillic.
I also note that the introduction of a new name instead of "spoiled" is correct in Cyrillic.
Windows 7.0
Java - see screenshot

[Phil Beauvoir]

I really don't know how to diagnose this. So what changed between versions? Is it the Java version? The Eclipse version? Only happening in Cyrillic? Any help would be appreciated.

[Phil Beauvoir]

If 4.0b9 was OK, it could be the Java runtime shipped with it. b9 used 1.8.0_101, PR1 uses 1.8.0_131. You can delete the "jre" folder and replace with an older version to try.

[Phil Beauvoir]

Is it the Java runtime? Is it the translation files? I can't solve this without any feedback and help.

[mdvadim]

One of previous version of *.archimate file shows that was misterious failure: No any uncorrects for Cyrillic fonts in Archi 4.0b8 and 4.0Pre1 both. I haven't got Archi4.0b9 and can't compare its impact on files with Cyrillic.

[mdvadim]

The decoder determined that the wrong characters are text encoded in UTF-8. Manually wrong text translated from the line from the encoding UTF-8 to Windows Cyrillic. The model looks like before.

[Phil Beauvoir]

The decoder determined that the wrong characters are text encoded in UTF-8. Manually wrong text translated from the line from the encoding UTF-8 to Windsor Cyrillic. The model looks like before.

What "decoder"?

[mdvadim]

It's online service, Phil.
_https://2cyr.com/decode/?lang=ru

[mdvadim]

Recheck
I uninstalled and reinstalled Arch 4.0 Pre1. Below are screenshots before and after Russification. No changes to the file occurred. All fonts in the names of the model elements are displayed correctly.

[Phil Beauvoir]

I'm confused now - is that good? Is it working?

[mdvadim]

Sure. I'm made just recheck. It's working absolutelly correctly. It was miracle failure.
Once. For no apparent reason. Without consequences.

Fuh-h-h-h. I can relax.