We're currently using tortoisecvs (1.8.31) to commit/access our cvs repositories.
We've recently wanted to support Japanese text in comments, so that viewvc can be used for browsing. However, it doesn't appear that you can set the encoding of comments being sent by Tortoisecvs are properly converted/sent as UTF8.
Example of Tortoisecvs output:
In D:\Workspace\dummy\src: "C:\Program Files\TortoiseCVS\cvs.exe" -q commit -m コミットテスト日本語 dummy.py
CVSROOT=:pserver:[email protected]:/cvsroot/dummy
/cvsroot/dummy/src/dummy.py,v <-- dummy.py
new revision: 1.10; previous revision: 1.9
/opt/viewvc-1.0.5/lib/cvsdb.py:198: Warning: Incorrect string value: '\x83R\x83~\x83b...' for column 'description' at row 1
cursor.execute(sql, sql_args)
I've confirmed that the cvs repository/viewvc supports japanese by committing through Eclipse, where you can define the encoding of comments being sent to the repository. But, most of our development isn't done in eclipse.
Is there a way to get tortoisecvs to assign the proper encoding to comments? Or, if it's commandline related issue, is there a way to change windows to default to utf-8? (Would that fix it?) (Tortoisecvs seems like it's just calling a command line tool, which I'm assuming then just applies the default system's encoding to the text)
If not, what windows CVS clients (if any) support proper handling of encoding in comments?
Update Removed SVN, tested and SVN correctly handles encoding and transmits comments in UTF-8. Thanks, Stefan for clarifying this.
Found this request in the tortosiecvs project on sourceforge:
It's been open since 2006. I guess it's not a big issue for a lot of people.