Noorul Islam K M wrote: > Index: subversion/svn/copy-cmd.c > =================================================================== > --- subversion/svn/copy-cmd.c (revision 1042948) > +++ subversion/svn/copy-cmd.c (working copy) > @@ -79,7 +79,7 @@ > SVN_ERR(svn_cl__eat_peg_revisions(&targets, targets, pool)); > > /* Figure out which type of trace editor to use. > - If the src_paths are not homogeneous, setup_copy will return an error. > */ > + If the src_paths are not homogeneous, try_copy will return an error. */ > src_path = APR_ARRAY_IDX(targets, 0, const char *); > srcs_are_urls = svn_path_is_url(src_path); > dst_path = APR_ARRAY_IDX(targets, targets->nelts - 1, const char *);
Thanks, Noorul. I couldn't understand this comment very well until I dug in to the history using "svn blame" and "svn log". For the record, r862493 and r862501 are relevant. "Trace editor" seems to have been a predecessor of "notification", so I changed that term. The relevance of the comment about homogeneous paths is that a check for homogeneity was removed because it is not needed, so I clarified that. I have updated the comment in r1043380. I note that our current trend is to check for homogeneity of targets (in other words, being either all URLs or all WC paths) in this layer as well as in the client library, so we might want to change this. - Julian
