On 5 Jun 2006, at 4:06 am, Joe Landman wrote: > Just a simple postgresql saving of compressed deltas with a simple > front end. SVN wasn't popular at the time, and cvs didn't look > like it could handle it. Even svn might blow lots of time in diff > calculation. Oooh yes. I've seen what happens when even small things like bacterial genomes are kept in CVS repositories. It's not pretty. Putting nt in it? *shudder* I'm not sure I'd trust diff and patch to do the right thing anyway, especially with repetitive or highly similar sequences. Does diff use enough context to be reliable for DNA sequence? I doubt it. Diff is essentially a sequence alignment algorithm anyway, and we all know all about those. :-) Tim