What kind of charset declaration do you have? (You can see it if you look at the source code of one of your pages after export.)

I often write in German myself, and I have no problem writing ä, ö, ü (I don't have them on my keyboard, so I have to type first the ¨and then the letter). I don't have the 'sz' either, so there I need the ß

At the top of the German sites I make I have this:

They are looking at the source code of the internal preview. That appears to me to be the only problem.

Yeah, I guess you're right.

Dear Inger, my HTML head looks exactly as you suggested.
And it worked well, also using the ¨+ letter keys.
But now my new Win10 is a German installation while before I had a
Taiwan/English Win7. Already I downloaded the English language pack and
changed the Win display language but to no avail.
Martin (oakhouse.de)

Have you tried the 'Search for undocumented characters', which can be found under Tools, then about half-way down the list of options?
You may write your German letters as normal, and then, when you are done, run that search and choose Replace, or Replace all. then your ö's and ä's etc. will be replaced by HTML entities.

Dear all, I focused my search on this character coding and found the straight forward solution.
Because my wife’s Taiwan installations made some adjustments in the ‘Administrative language Settings’ and the ‘Language for non Unicode programs’ had been set to Chinese (Taiwan) with UTF-8 set ON. I changed that back to English and UTF-8 to OFF and good old CoffeeCup feels like home finally!
Thanks for the encouraging discussion, Martin
By the way, will there be a Unicode version of CoffeeCup? I guess, back in 2016 Liehcheng had the same problem.

"The ccId's should only appear in the temp files used for previewing, and
not in saved files."

Not for me: ccID IS added to the saved/published html, and remains in the code after deselecting "Clicking tags in code selects elements in Preview" and "Clicking element in Preview selects its code". Deselecting those options only prevents new ccIDs from appearing.

It's a very irritating feature, as I don't want to publish anything that unneccessarily inflates every element on the page by 11 bytes. As each ccID is unique, once they are generated the only way to remove them before saving is one-at-a-time for every single element on the page.

For me, the cost of this feature far outweighs the benefits. Or am I missing something?

Removing the old ccID references could almost certainly be done in bulk by using a program like Text Crawler with Regular Expressions.

Frank