Community technical support mailing list was retired 2010 and replaced with a professional technical support team. For assistance please contact: Pre-sales Technical support via email to sales@march-hare.com.
On 15/2/2006 21:02, Tony Hoyle wrote: > Flávio Etrusco wrote: > > This makes sense. Seems like a routine is expecting an utf-8 string, >> detects a character code above 127, decides it'll use the next two >> bytes for decoding the character but the bytes don't actually encode >> an utf-8 char, so it converts to '?'... >> > > That's entirely possible. The code is a backport from 2.6 which > operates in unicode always. > > Try switching on the UTF8 server in the control panel. > > Tony > Tony, Switching the server to Unicode (I didn´t see this option before) solved the problem with the database. Maybe a simple test into the Init method of the MSSQLRecordSet (cvsapi/mssql/mssqlRecordSet.cpp, 188) avoiding the conversion in the case of a non-Unicode server should be enough to allow the usage of the audit without char code problems. To change my server to Unicode I need to checout my projects again because of the error below. cvs server: Updating . cvs update: Translation from server codepage 'UTF-8' to client codepage 'CP1252' lost characters cvs server: cannot open directory /PhSGQ/Tecnologia/Gerncia de Desenvolvimento/Implementao/SISPACC/POS/src: No such file or directory cvs server: skipping directory After the new checkout almost everything worked fine. I had to commit my modules2 file again, because the modules defined there were not recognized. cvs -w -d :sspi;username=clovis;hostname=192.168.1.1:/PhSGQ checkout -P -d SISPAC_ -- SISPAC (in directory X:\Ok\b121705) cvs server: cannot find module `SISPAC' - ignored