All of the sudden all special characters that I paste into writer or calc turns into illegible signs. ä becomes ä,
å becomes Ã¥ and so on.
My version of LibreOffice is 6.4.6.2 and OS is Manjaro KDE Plasma
I assume these characters are generated from within LibreOffice itself (menu Insert → Special character). Does the same thing happen if you copy/paste from an external source (like Character Map) for the special characters?
You must check it if the used Font has the specific special character or not.
Can you upload a small ODF tyxpe sampel file here whith your formatting properties and the inserted special characters?
What I do is I copy text from another document, for instance Thunderbird or a text editor like Kwrite. The text is often in one of the Scandinavian languages and uses special characters like Æ,Ø,Å,Ö.
As the text copied from Thunderbird may come with some formatting I will “paste unformatted text”, but it does not matter if I paste like that or if I paste normal.
So I write some text in Kwrite:
“Å få en tekst til å se normal ut, er ikke greit. Jeg skjønner ikke bæret, det ser ut som om noen har ødelagt tastaturet mitt”
Pasted into Write or Calc I get:
Ã
få en tekst til å se normal ut er ikke greit. Jeg skjønner ikke bæret, det ser ut som om noen har ødelagt tastaturet mitt.
What is the encoding in KWrite? UTF-8 or some variant of ISO-8859? When you paste as “unformatted text”, you lose any helper data about the encoding and your text is pasted as a sequence of bytes. It is then interpreted as UTF-8. If you start from some ISO-8859, you end up with an invalid UTF-8 string
So, I found what had changed. I did some changes in the regional settings in KDE Plasma some time back and that seems to have messed this up.
I changed the settings back again and Voila.
Thank you all for your suggestions!
*"So I write some text in Kwrite: “Å få en tekst til å se normal ut, er ikke greit. Jeg skjønner ikke bæret, det ser ut som om noen har ødelagt tastaturet mitt”
Pasted into Write or Calc I get: à få en tekst til å se normal ut er ikke greit. Jeg skjønner ikke bæret, det ser ut som om noen har ødelagt tastaturet mitt."*
It looks like some character coding issue. What character coding are you using in the source document? Unicode, UTF-8 -16, or some other)?
Well, it doesn’t seem to matter which source document I use. I can copy from the browser, Kwrite or Thunderbird, but the result is the same.
Where would I be able to see what coding the source documents uses?
In the Thundebird you can use the View - Character coding (or some similar menu item ) for the textual contents of the mail. (I have Hungarian locale setting in my Thunderbird: Nézet - Szövegkódolás)
When the coding of the source and the target are not match, then occurs such issues.
The coding of a textual message depends on the settings of the mail client/service and on the editor software (of the embedded text or the attachment) what the sender had used.
Thank you 
Then in Thunderbird it is set to Unicode.
The coding in Kwrite I could not find, but the coding in the browsers seems to be Unicode too.
The weirdest thing is, everything worked just fine until recently, and I do not know what (I) may have changed.
“Then in Thunderbird it is set to Unicode.”
The proposed setting enforces the View function of the Thunderbird TO USE that coding for the display. That not mean if the received text has such encoding… But when the appearing is good, then (possible) that the choosen encoding of the text and the decoding of the Thunderbird does match.