non "standard" characters are not handled anymore by SC -- URGENT

Recently all accented letters are shown incorrectly in all grids and all forms.
E.g. an “?” stored in the db correctly is shown in SC as “è”

Also the Euro sign “?” is now shown as “?”.
This is not a character stored in the DB.
It’s generated by SC, as per the SC settings for a currency field.
E.g. what was before shown as “? 5.400,00” is now shown as “? 5.400,00”,

I think this is bug with one of the latest SC update, as all grids and forms were working correctly and the DB itself and all DB connections defined in SC were not modified.
The issue is in both the development and production environments.

This is the error SC returns if I try to update e text field in a form by inserting just the letter ‘?’:

Incorrect string value: ‘\xE8’ for column ‘name’ at row 1

The view SQL link shows that the SQL generated by SC is:

UPDATE train_and_cert_person SET id_brand = 0, id_contact = 5977, id_area = 3, id_type = 2, id_delivery_type = 4, id_status = 1, name = ‘?’, ident_data = ‘’, status_notes = ‘’, cost = 0.00, start_date = null, last_renew_date = null, end_date = null WHERE id_train_and_cert_person = 19

I pasted the above update sql statement in HeidiSQL and run it successfully.
Then I refresh the same record in SC and the “?” (stored in the db correctly) is shown in SC as “è”

SOLVED

the project properties for locales\project languages were set set to Western (ISO-8589-1) while the correct value is Unicode (UTF-8).

I suspect the change was performed by an SC update, as I set those values to Unicode many months ago and I’ve never changed them since.