[cvsnt] Question about CVSNT and Unicode files

Bogdan Farcasanu bfarcasanu at yahoo.com
Wed Feb 20 08:15:22 GMT 2008


I am using:

WinCvs 2.0.2.4 (Build 4), on Vista 64, all updates installed
and
CVSNT 2.0.51d, on Windows 2003 Server, 32 bit, all updates installed.

Flags on the file ar -ku. What happens is that even if the file is indeed Unicode, a conversion to UTF-8 is done at commit time, and this is where the weird characters arfe from.

How do I set a specific code page for CVSNT?

Thanks,

Bogdan.


----- Original Message ----
From: Arthur Barrett <arthur.barrett at march-hare.com>
To: Bogdan Farcasanu <bfarcasanu at yahoo.com>; cvsnt at cvsnt.org
Sent: Tuesday, February 19, 2008 10:27:35 PM
Subject: RE: [cvsnt] Question about CVSNT and Unicode files

Bogdan Farcasanu,

Part 2...

> The problem is 
> that When CVS commits the changes the file it makes a few 
> changes (like adding  at the beginning of the file, and 
> replacing some special Eastern Europen characters). 

What is the codepage of the server and client? What versions of CVSNT are you running on the server and the client?  What OS are you running on the server and the client?  What is the output of cvsdiag on the server (or if the server is linux/unix what is the /etc/cvsnt/Pserver file settings)?  If you do 'cvs ver' from a sandbox what is the answer?  If you enable the server to 'allow clients to trace server' then do a trace commit on the client 'cvs -ttt ci -m "" filename.ext' what does the trace say the server and local codepages are?

Regards,


Arthur


      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ 


More information about the cvsnt mailing list