I think alot of people find the concept of workspaces a bit challenging. I certainly did. Mostly due to misconceptions. I kept thinking of it as something akin to a RDBMS schema; a user's section of the db with all of their related data structures. However, workspaces are nothing like this.
To me, the best way to understand them is to think of the JCR repo as if you were writing a version control system. In CVS, for example, the repo has a virtual set of the files. Each user can check out a state of those files into their own "workspace", where they can make changes to those files and even later merge those files back into the whole. I hope this helps a bit. Also, the specification is quite informative and worth reading closely. On Sat, Feb 13, 2010 at 11:25 AM, Rakesh Vidyadharan <[email protected]> wrote: > I was playing around with workspaces and ran into unexpected behaviours. It > seems that regardless of the workspace my session is bound to, I see the same > set of nodes (using same credentials). I created the workspace with just > name (without option to clone an existing workspace), then got a session to > the newly created workspace expecting no nodes (other than system generated > ones), but was surprised to see the exact same nodes as I usually see with > the default workspace. I then switched to the security workspace and saw the > same data. > > I tried putting some debug information on each node to ensure that the node's > session was indeed bound to the workspace I wanted, and that part looks fine > (at this point I believe it is not my code issue). > > Any explanations on this would be most welcome... > > Thanks > Rakesh
