On Mon, Feb 12, 2001 at 01:27:46PM -0800, Mr . Bad wrote:
> >>>>> "TC" == Tavin Cole <tavin at mailandnews.com> writes:
> 
>     TC> Oh, the pain, the agony, the sheer abomination of having to
>     TC> type "cvs co -r 0.4 Freenet".
> 
> Well, I went through a relative pain in the ass helping you and Oskar
> do it. Remember, Mr. Sassafras? So maybe it IS easier to have the
> fewest number of people possible working on the branches.

Which never would have happened if we had just branched 0.3 off the trunk
and then let Oskar commit everything.

Anyway, the point is no matter how we organize the branches, people can
learn to type "-r xxx" or they shouldn't be messing with the code anyway.


>     TC> If trunk == experimental there will be a lot less merging to
>     TC> do, since you presumably make the 0.x branches when the
>     TC> codebase is sufficiently stabilized, then the trunk
>     TC> metamorphoses again.
> 
> How, exactly, is merging from A to B easier than merging from B to A?
> If they both have the same starting point, CVS has to go through all
> changes since the starting point, regardless. 

I'm not talking about what CVS has to do, I'm talking about what humans
have to do.  If trunk != experimental we will always have to deal with
merging experimental back into trunk, but if we just keep branching
0.x off the trunk, the *bulk* of the only merging we'll ever have to do
is so stable can catch up with the 0.x's (which we have to do either way
anyway).  That, and bugfixes/feature enhancements that are relevant to
the future codebase would have to be merged from the current 0.x to trunk.

So I think sheer lines of code is the key..  as generally the
voluminous lines of code changes in the future codebase obsolete the
associated older code (i.e., very little merging from old codebase into
future codebase).


>     TC> I'm just glad it's you who has to merge experimental back into
>     TC> trunk ;^>
> 
> Hey, I got no problems with it. It seems like there won't be an
> incredible amount of overlap, according to what I've seen in
> experimental. So, I can deal.

I do appreciate your efforts.  Thanks.


-- 

/* tavin cole
 * composer of e-mail messages
 */

_______________________________________________
Devl mailing list
Devl at freenetproject.org
http://www.uprizer.com/mailman/listinfo/devl

Reply via email to