On Oct 13, 2014 7:42 AM, "David Mason" <dma...@ryerson.ca> wrote: > > On 13 October 2014 04:54, Tony Papadimitriou <to...@acm.org> wrote: > > The claim that once you shun a 0-length file you will not be able to > > commit another 0-length file again is not entirely true. If you first > > delete the existing 0-files, and re-enable the shunned SHA1, then > > you can add 0-length files again. > > Yes, when I said "never" I should have added "unless you unshun that > id." I wasn't necessarily advocating a change in behaviour, just > pointing out what a big stick this was. Perhaps pointing out this > edge-case in the documentation or the shunning web page would be > sufficient (or a pop-up warning if the user enters the 0-length UUID > on the shunning page).
I would think documentation is the best course. One could make almost the same argument about any really short file, or the (extremely unlikely) case of a hash collision (though at that point there are far worse problems to deal with). Rather than trying to catch all possible potentially bad uses of the shun stick, just improve the quality of the documentation.
_______________________________________________ fossil-users mailing list fossil-users@lists.fossil-scm.org http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
