On 03/08/2011 11:48 PM, Arthur Barrett wrote: >> The main trouble is that CVSNT is a closed solution. > > CVSNT is free, open source software, we publish the source code. CVSNT > uses open protocols to communicate between client and server. I cannot > find anywhere a definition of 'closed solution' that fits, but if > Michael can supply a reference to one I'll happily read it. Definition > of free software: > http://www.gnu.org/philosophy/free-sw.html
I am aware of this, and purposely did not claim that CVSNT is not "free software" or not "open source software". AFAIK CVSNT abides by the letter of those definitions. I am referring to the following types of "closedness", all AFAIK: * The vendor only makes available the source code of the most recent version of CVSNT to paid customers. * CVSNT is not developed in the open, with a publicly available source code repository or public mailing lists. In no way do I dispute the right of Arthur's company to develop and distribute CVSNT in the manner that they have chosen. But it is important for potential users to understand that the closed nature of CVSNT development implies certain disadvantages for CVSNT users, in particular the lack of a vigorous community comparable to those organized around CVS or other "truly open" software projects and a resulting poverty of community-contributed tools. > CVSNT uses the standard RCS file format to store the version history. > The standard RCS file format can be read by many many tools. The RCS file format is analogous to XML, in that it is a very extensible generic format. CVSNT indeed uses the RCS format, but (1) its "schema" includes components that are alien to standard RCS or CVS and (2) it stores some data (e.g., the contents of some binary file revisions) outside of the RCS-format files. Michael -- Michael Haggerty [email protected] http://softwareswirl.blogspot.com/
