Re: [Gnu-arch-users] GCC v. Arch address@hidden: Regressions on mainline]

[Tom Lord]

    > From: Andrew Suffield <address@hidden>

    > On Wed, Jun 23, 2004 at 10:15:30AM -0700, Tom Lord wrote:
    >> I'd be worried that committers will be frustrated by the lag time
    >> between when the commit and when what they've committed appears in
    >> mainline.  For example, it would mean I couldn't say to you, on the
    >> phone from far away, "Oh, I have something for that that's been
    > tested.  Lemme go ahead and check it in and then call you back in 15
    >> mintues after you check it out."  Instead it would be "Oh I have
    >> something for that .... I'll check it in and call you back tomorrow
    >> after archive-side testing completes."

    > Oh, do think about it; you tell them you committed it to your own
    > branch :P

When that's ok, it's ok, but often it won't be.

If 10 rbcollins' were hacking GCC, sure -- that's definately what
would happen.  Lots of cherry-picking and lots of users who, like
rbcollins, are very comfortable with that.

Absent a bunch of users in that headspace, though, I think that just
bouncing everything off a shared star-merge hub is the route that will
least often lead user confusion.

That said, there's the whole other topic of being able grow and shrink
star-trees temporarily.   I.e, go from this:


                main
               / |  \
              A  B   C

(but now A and B want to work together without mucking up main) to:

                main
               /    \
             A+B     C
             / \
             A  B

and then back again.   Nothing hard about that but makign it very
trivial (so that users don't have run `archive-setup', for example) 
might be desirable.


    > [Otherwise yes, running tests between every pair of revisions is
    > stupid and wrong and useless - can we stop talking about it now
    > please? :P]

Yes.

-t