On 04.04.2012 20:46, Daniel Shahaf wrote: > But one _could_ write a filesystem that implements compression and/or > deduplication aimed at .svn/pristine/'s workflow. Should effort be > spent there rather than on implementing a smarter svn_wc__pristine_*()?
Now that's an interesting idea. I propose that Ashod immediately drop his pristine work and instead implement a proof-of-concept, portable userspace compressed filesystem. I further suggest that the best test case for the reference implementation might be pristine storage for Subversion, and in order to lop a few man-years off the proof-of-concept stage, he might use SQLite for storing filesystem metadata. Oh wait ... :) -- Brane
