On Sun, Sep 12, 2010 at 04:21:03PM -0700, Daniel Trebbien wrote: > I removed the normalization counting code without much thought in my > hastened efforts to produce a version of `svnsync` that I could use to > mirror the GNU Nano repository. Currently, I am thinking that Stefan > Sperling's idea of a `svn_subst_translate_string2` function is the way > to go.
FWIW, I agree that knowing how many properties were tweaked by svnsync isn't that interesting. It could really be debugging message. However, telling users that the source repository has problems is a good thing, but we should provide more information than "X of your props have bad line-endings and/or the wrong encoding". Would you be willing to work on a follow-up patch that improves svnsync's reporting about problems in the source repository? E.g. it could print per-property information as it processes revisions. Something like: Copied properties for revision 42. Normalized line-endings to LF in: svn:ignore (/trunk, /trunk/src) svn:externals (/trunk/ext) svn:log Re-encoded to UTF-8 from ISO-8859-1: svn:log svn:author Having a similar reporting feature in "svnadmin verify", and ways to fix these problems with "svndumpfilter" or maybe even "svnadmin recover" wouldn't hurt either. Stefan
