Package submission process

[Jeff Spaleta]

On 10/17/05, Greg DeKoenigsberg <gdk at redhat.com> wrote:
> How important do you think these metrics are?  I know that Elliot is
> working on a whole bunch of metrics -- it might make sense to start
> collecting all of these metrics ideas in a central place.

how important.. that's difficult to say. The casual spot checking on
the number of blocking FE-NEW and FE-REVIEW I've done doesn't seem to
indicate there is a backlog problem happening. So its not like there
is a burning fire that needs to be watched closely. I will say however
that without time history metrics, it will be difficult to do a good
job of optimizing the workflow any further. And if a problem with the
workflow does develop, time history metrics like I'm talking about are
going to be the best way to catch it early and try to address it.

Additionally we've no way to actually communicate that things are
operating nominally to the..impatient potential contributor.  How
important is it to give new contributors a rough estimate on how long
the process for a first package normally takes? People walking into
the process..expecting it to take 3 hours to get a first package
submitted will be a continual frustration unless they have some
upfront guidance on what the average submission process is like. The
best way I can think of is to generate the time history metrics.
Things like "last month the average amount of time a package sat in
FE_NEW state was... and average time in FE_REVIEW was...."

-jef