Re: svn blame miscategorizing a file as binary

Mark Benedetto King wrote:
> If there are truly binary revisions, the blame information is likely
> to be extremely unreliable.

My case is a counter example. A normal XML file was once "xml/svg"
then became "text/xml". The blame information was perfectly reliable.

You assume SVN's way of determining "binaryness" is better than it
really is...

> I think the current approach is the safest. However, we may need
> more knobs:

Safest maybe, but not the most useful (IMHO)...

I think using the "binaryness" of the revision you're viewing makes
the most sense, rather than the "binaryness" of the pessimistic union
of all prior revisions.

If I've marked revision NNN as being not binary, why should "svn blame"
try to second guess that decision?

-Archie

__________________________________________________________________________
Archie Cobbs * CTO, Awarix * http://www.awarix.com

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Oct 11 18:30:54 2004