It would be nice to synchronize the content automatically with a subversion repository. This could be done e.g. with a RepositoryListener.
Yes, this would be very valuable! If this became an alternate repo for lenya (like jsr 170), we could also utilize the versions, properties that svn maintains
(In reply to comment #1) > If this became an alternate repo for lenya > (like jsr 170), we could also utilize the versions, properties that svn maintains That would of course be great, but I think we could start with a simple synchronization, i.e. commit everything that changes (including the Lenya rcml files). That would allow to backup and restore the exact state of a website.
ATM every change adds and removes backup files. This makes it a PITA to sync the content with SVN. Maybe we could add an option to store all backups in a single file, or at least use non-changing file names (backup-1 .. backup-n)?
(In reply to comment #3) > ATM every change adds and removes backup files. This makes it a PITA to sync the > content with SVN. Maybe we could add an option to store all backups in a single > file, or at least use non-changing file names (backup-1 .. backup-n)? i don't think this makes much sense. even if you had persistent unique file names, *all* old versions would pile up in svn, cluttering your repository in no time. it is never advisable to pile two different mechanisms on top of one another that both attempt to do the same yet are agnostic of each other (think of tunneling tcp/ip over tcp/ip as some VPNs do - two layers of transport control inevitably mess things up). we should either make svn our main repository and get rid of backup files altogether, or discourage using svn on our existing repo. what you could do if you need to get things done is instruct svn to only pull the current versions and ignore the backups. that would enable you to get a reasonably complete revision history in svn, provided you pull more often than users update, or you are content with snapshots at fixed intervals regardless of missing the odd update.
(In reply to comment #4) > (In reply to comment #3) [...] > we should either make svn our main repository and get rid of backup files > altogether, Implementing that would be overkill for my usage scenario (small company website). > or discourage using svn on our existing repo. But SVN is so convenient ... > what you could do if you need to get things done is instruct svn to only pull > the current versions and ignore the backups. that would enable you to get a > reasonably complete revision history in svn, provided you pull more often than > users update, or you are content with snapshots at fixed intervals regardless of > missing the odd update. Good idea, I'll give this a try.
BTW, this issue will certainly become important when we start to manage the Lenya website with Lenya.