Re: Moving patches from head to 3_1-branch

[Phil Edwards <phil at jaj dot com>]

On Fri, May 17, 2002 at 01:51:45PM +0200, Paolo Carlini wrote:
> I have just started moving my stuff (beginning with 6648), but, then I 
> noticed something in the Changelog which made me think me quite a bit:
> 
> 2002-05-14  Release Manager
> 
>     * GCC 3.1 Released.
> 
> If we start moving patches predating 3.1 to the branch keeping the 
> original commit date, wouldn't we end up giving the completely wrong 
> impression that those fixes are present in 3.1.0 (vs 3.1.1) too??

On the trunk the changelog reads from newest to oldest, but (my opinion only)
this need not be true on the release branch.  The text of the date tells
us when the patch was originally conceived, and the order of the entries
tells us the relative order in which they were applied.  On the trunk
those two concepts are synonymous, but on the branch, they might not be.

So if I see

    2002-05-10  Some Person

        * Another fix.

    2002-05-14  Release Manager

        * GCC 3.1 Released

    2002-05-10  Some Person

        * A fix.

then I know that "a fix" was included in 3.1, and that "another fix" was
applied to trunk at that time, but not moved over to the branch until
after 3.1 was out.  (We don't get to see the actual date of the move,
but that can be pulled out of "cvs log".)


This is just me thinking aloud, trying to explain this to myself.  :-)
It's confused me for a while too, but I think it finally came to me today
at lunch.


Phil

-- 
If ye love wealth greater than liberty, the tranquility of servitude greater
than the animating contest for freedom, go home and leave us in peace.  We seek
not your counsel, nor your arms.  Crouch down and lick the hand that feeds you;
and may posterity forget that ye were our countrymen.            - Samuel Adams