Re: Symmetric Merge -- Algorithm
From: Julian Foad <julianfoad_at_btopenworld.com>
Date: Tue, 17 Apr 2012 20:01:43 +0100 (BST)
Paul Burba wrote:
> Julian Foad wrote:
In brief, my plan is to develop a way forward through discussion with the community. Thank you for engaging and helping with this.
> I worry about proceeding
Yup, me too.
> And what about other merge "edge" cases, e.g. shallow merges, shallow
Yup. FWIW, the way I work, and from the position from which I'm entering, the most difficult step is describing what we want the behaviour to be, and then implementing it is (still hard but) the lesser task. So I'm concentrating on trying to create a descriptive model of what merging should do.
You tested a number of scenarios (mainly by making the test suite run my symmetric-merge initial trial implementation and analyzing what happened) and that was very useful. I found that I needed to give more thought to the functional design and that's why I didn't reply directly to those emails.
This is an archived mail posted to the Subversion Dev mailing list.