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.
Hello Tony, Tony Hoyle <tony.hoyle at march-hare.com> writes: > [...] > http://www.cvsnt.org/manual/html/commit-files.html > http://www.cvsnt.org/manual/html/Variables.html (REAL_CVSROOT and > VIRTUAL_CVSROOT are missing from that description though). Aaah, maybe I should check my eyes... Thank you for the pointer! This page must be relativly new - I've never seen this when scrolling through the manual (or my manual was too old for too long time). Then REAL_CVSROOT seems to be the answer to my commitinfo problem (other post). I will try this out asap. > > * description of the environment one can expect inside a script. This > > may be platform and setting specific. > > It's so specific it's next to impossible to document. Scripts > shouldn't be written to expect anything about their environment anyway > (too many potential security holes). My main concern was about referencing scripts added to the CVSROOT directory of the used repository using checkoutlist. The CVSROOT variable seems to contain only the VIRTUAL_CVSROOT. This can not be used to call the script. > > * the kind of linefeed characters a script has to use in it's > > output. (I now know it has to be unix LF and it's defined in the > > cvsnt client/server protocol... but it's not documented somewhere > > to find it when writing a script.) > > For most scripts it's automatic - the standard output is binary and > they pick this up. I believe python doesn't.. not sure about others. Maybe this is my problem. I'm really unexperienced in python and have to sort things out. If it's only a problem in python I understand why it's not mentioned in the manual or on other ressources. Thanks for the helpful response. You still keep cvsnt ahead of svn in support and features! Best regards, O. Koltermann