From: Junio C Hamano Date: Wed, 25 Aug 2010 18:58:24 GMT Subject: Re: kill filter-branch --remap-to-ancestor? Message-ID: <7v39u2senj.fsf@alter.siamese.dyndns.org> In-Reply-To: Csaba Henk writes: > I had some headache with the issue to which the cure is the > "--remap-to-ancestor" option of git filter-branch -- back when > "--remap-to-ancestor" was not yet > (cf. http://thread.gmane.org/gmane.comp.version-control.git/112068/focus=112838). > > This time, in a similar situation, I had some headache with finding out > that the panacea is already there and it's called "--remap-to-ancestor". > > Why not cut back on complexity and get rid of "--remap-to-ancestor" > while automatically enable the implied behavior for filters including a > path? This was already proposed by Junio and the author of the option > had no objections against this idea (see > http://thread.gmane.org/gmane.comp.version-control.git/130949/focus=132684). I am not a heavy user of filter-branch myself (I used it once during the past 6 months and it was with pathspec but the project did not have any tags, so that particular exercise does not matter). I do not offhand see a reason not to. Please make it so, if you are so inclined, or perhaps somebody else may volunteer?