Re: merge-tracking and svn 1.5

On 12/13/06, Ben Collins-Sussman <sussman@red-bean.com> wrote:

I am glad you sent this email. I was talking to Paul Burba yesterday and
getting ready to send something similar.

> So, here are the questions I'm throwing out to the list:
>
> 1. Should svn 1.5 block on the completion of merge-tracking?

I agree with the sentiment that there is plenty in trunk to warrant a 1.5.
However, I also mentally defined 1.5 as being "merge tracking". If it is
going to take 3-4 months to finish, then 1.5 should go without it. If it
just needs refinement, then it should block 1.5 to get the refinement done.

I also agree with Dan Berlin's statements. If we define 1.5 as "merge
tracking" it is going to be hard to control expectations, and possibly the
subsequent disappointment.

In an uneducated and uninformed way, I am concerned about the introduction
of SQL Lite as a new repository requirement and what unintended side
effects/problems that will bring.

> 2. How do we test merge-tracking effectively?

I think it has to be stabilized in trunk as best it can be and then loaded
on svn.collab.net prior to any 1.5 RC. Perhaps even prior to the 1.5branch.

Mark

PS - The features I really want from 1.5 are the ones that make things like
checkout. update more tolerant of existing files and the enhancement to
allow something to be moved multiple times without a commit. Both of these
will really help IDE integrations.
Received on Wed Dec 13 22:48:20 2006