Confirmed.

I don't know if it should work in a different way but I would enclose the lines with CDATA comments

<fix>
		<item><![CDATA[
			Oprava položky u data 15-12
		]]></item>
</fix>

Thank you ReLater,

when I used CDATA, it is OK.

Yes, but I couldn't find a reason why it doesn't work without CDATA. Tried a lot to enforce UTF-8 while parsing... checked encoding of your file... added <?xml version="1.0" encoding="UTF-8"?> ...

Maybe encoding of the webserver itself is still not utf-8 (just to be sure)?

If all is set up for utf-8 it should not need the CDATA in my opinion, but I might be wrong.

I've tested it also with my custom site (and a custom changelogurl on it) that runs with utf-8 for sure.

Maybe the httpRequest (I don't know how it's called correctly) gets the wrong encoding ? The wrapped lines happen very early in the PHP parser logic.

I've found a very old comment yesterday https://www.php.net/manual/en/function.xml-parser-create.php#100099

The original parser sometimes wrapped the text before the first diacritics appearance.

@ReLater The comment after that one reads also interesting, encoding ... even if 15 years old it could still be relevant.

I didn't try with UTF-8 plus BOM because that wouldn't be practicable for devs. Perhaps @klucon wants to try? Just for the sake of clarity.

Hello,
now I have converted all the files to UTF-8 with BOM and the same problem.

Here is a UTF-8 ZIP file with BOM:

https://build.klucon.cz/mod_kluconholiday_with_bom.zip

If I use CDATA, everything is alright

Maybe encoding of the webserver itself is still not utf-8 (just to be sure)?

I am testing on the XAMPP (compiled Jun 5th 2019) and Windows 10
Apache/2.4.41 (Win64) OpenSSL/1.1.1c PHP/7.3.12

Yes, but I couldn't find a reason why it doesn't work without CDATA. Tried a lot to enforce UTF-8 while parsing... checked encoding of your file... added <?xml version="1.0" encoding="UTF-8"?>

when I added <? xml version = "1.0" encoding = "UTF-8"?>, no change took effect

@klucon
Thank you for testing!

tested with and without bom - no change :(