On Mon, Nov 28, 2011 at 06:05:35PM -0700, Colin Campbell wrote:
On 11-11-28 11:39 AM, Phil Holmes wrote:
>This presumes that the Bug squad can and do run Git, which is a
More specifically, I believe it requires a Linux build environment,
to be truly effective.
If you want to test that the patch actually does what it claims to
do, then yes. If you just want to check if the patch exists in
master, then of course it's not necessary.
OTOH, with the lilybuntu VM and lily-git.tcl, doing
"Update source", drop to a shell, cd ~/lilypond-git/build and
running make, should be within reach for most of the sort of folk
who want to be bug squadders.
I honestly believe that less than 50% of bug squad volunteers are
actively working after 4 weeks. Am I incorrect as a matter of
Assuming that I'm not factually incorrect here, we should not make
it harder for bug squad volunteers to get started. Once we hit
80% retention of volunteers after 4 weeks, I'm open to increasing
the difficulty. I would hope that this goal is easy -- as long as
we're up front about what's required, maintain an encouraging
environment where questions are welcome (within the bug squad, and
I'm not claiming to be interested in taking part in that), and
ruthlessly update the CG section on Issues -- but it *does*
require skilled developers mentoring new volunteers.
Short term: sure, let's make "verify a patch" only a task for the
bug meister. The bug meister can either check patches in master
via git, or else just mark any Patch issue as verified without