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 Mon, 23 Feb 2004 19:44:37 +0100, Manuel Martin <martin at ferber-software.de> wrote: >Hello, > >> This was actually changed due to a bug report as RCS 5.7 >> behaves in the >> 2.0.30 way (Bug#188). I'd expect ViewCVS to actually cope with it >> better now as it's closer to 'proper' RCS output. > >just an update: it does not very well, as it has been tweaked to cope >with the "cvsnt-way", according to my test with Bo's installer (at least >the included version of viewcvs's standalone.py) and 2.0.30. Now, >error-messages are generated. > >I'll look into this later on this week, as for now, I probably better >use the old cvs.exe. > >bye, mm Sourceforge is too busy now to allow any checks, but I did an update to HEAD and found that two files have been updated from when I built the installer: lib/ezt.py lib/viewcvs.py I tried to diff but I am getting an error message all the time. Anyway the viewcvs.py in the installer is max 2 revisions back from 1.232 so if you can diff these (1.232 and 1.230) then you can see what has changed in the meantime and find out if a solution has been included or not. In the installer there is an included cvs.exe (2.0.11), it gets used by ViewCvs if the detected version of cvs.exe is older than 2.0.4, otherwise I point viewcvs to use the same as is used by the CVSNT service. HTH /Bo (Bo Berglund, developer in Sweden)