2009/2/5 David Berry <[log in to unmask]>:
> Scenario... I create an AST branch within which I start to make
> experimental changes. At the same time, people keep on sending me AST
> bug reports (as they do) that need fixing on master. So I am swapping
> fairly frequently between the master and experimental branches.
>
> Am I right in thinking that each time I swap branch, I should do a
> complete rebuild of AST from scratch, since all the built files will
> be pertinent to the other branch? For instance, let's say I add a new
> class to AST, so the Makefile.am on the experimental branch contains
> source files that don't exist on master. So if I've been working on
> experimental, and have corresponding versions of Makefile.in and
> Makefile, and then swap to master, I will need to run configure again.
>
> This could become a pain if swapping branch becomes frequent. Is
> there any way I can use different directories to hold checkouts of
> different branches?
Would a solution be to have an entire seperate repository for the
experimental work? Sounds a bit like taking a sledge-hammer to crack a
nut...
David
|