> Is that really important to see xref autogenerated content in real time? In > our opinion, the answer is no.
Yes, the point regarding links which Kevin makes is the really important one. > If we were satisfied by using version control systems like RCS or subversion > (i.e. check out, work on a copy, check in, solve merge problems) to manage > our documentation, we wouldn't have developed XMLmind Document Repository. Underlying any mature VCS are data and storage models and algorithms that have taken many years to develop. The check in/check out workflow is only incidental. Many other tools can be built on top of them. By equating RCS and Subversion you really make my point for me. Subversion was created due to longstanding disatisfaction with RCS-based CVS. Other VCS's like mercurial, bazaar, darcs take different approaches. These all represent people who were not satisfied with the CVS era, and have developed alternatives. I just thought that rather than embarking on this journey yourselves as well, it would be good to find common ground with existing work and build on it, gaining the benefit of that work's development community. For example, here's a project that uses git to version Python data structures: http://www.newartisans.com/blog_files/git.versioned.data.store.php Or look at darcs's theory of patches: http://darcs.net/manual/node8.html -- jean . .. .... //\\\oo///\\
