[PATCH 1 of 5 v3] rebase: move base calculation from rebasenode() to defineparents()

Mads Kiilerich mads at kiilerich.com
Fri Dec 5 09:52:43 CST 2014


On 12/05/2014 02:42 AM, Pierre-Yves David wrote:
>
>
> On 12/04/2014 04:51 PM, Mads Kiilerich wrote:
>> On 12/05/2014 01:09 AM, Pierre-Yves David wrote:
>>> On 12/03/2014 07:12 PM, Mads Kiilerich wrote:
>>>> # HG changeset patch
>>>> # User Mads Kiilerich <madski at unity3d.com>
>>>> # Date 1416364731 -3600
>>>> #      Wed Nov 19 03:38:51 2014 +0100
>>>> # Node ID a20a470ffce95d5266b8e332c0e01c72604fc528
>>>> # Parent  c237499a7fba65c88a2da721a22b66df4f39cf4e
>>>> rebase: move base calculation from rebasenode() to defineparents()
>>>>
>>>> We want to collect all calculation in one place.
>>>
>>> We want to? Why? Can you be a bit more verbose about the goal you are
>>> aiming to ?
>>
>> It is just a refactoring preparing for the next patches that change the
>> whole calculation.
>
> Hints:
> - You are not charged at description length. You can add more details 
> there.
> - I still do not get what is you actual plan beside rewriting this 
> whole logic.

The rewrite simplifies the code - I think that has value in itself.

It also makes it easy to handle special cases separately and in special 
ways - for example 
http://selenic.com/pipermail/mercurial-devel/2014-November/064134.html

> (I assume your end call it is related to issue4076 but I've no idea 
> what you plan to do here. I would naively expect some bid merge 
> related solution from you)

http://selenic.com/pipermail/mercurial-devel/2014-November/064134.html 
was one attempt at solving my problem ... which probably is the same as 
4076. That attempt was however wrong. We can discuss it in details once 
we have something we can point at and discuss separately.

/Mads


More information about the Mercurial-devel mailing list