FESCo wants to ban direct stable pushes in Bodhi (urgent call for feedback)

[Mike McGrath]

On Fri, 26 Feb 2010, Garrett Holmstrom wrote:

> On 2/26/2010 11:07, Jesse Keating wrote:
> > It'll require some enhancements to how bodhi is used for people
> > consuming testing updates, and it may require a more active role on part
> > of the maintainer to seek out somebody to at least give the update a
> > smoke test.  Instead of treating updates-testing users as a black box
> > that you may or may not get any output from, we should have a more
> > direct relationship.  Maybe something in the Fedora Engineering Services
> > initiative could be to spend some time smoke testing updates-testing
> > stuff.
>
> Forgive me if I'm missing something, but isn't that QA's job?  Maybe
> what we really need is a push for more QA people.
>

This is a fair question but I don't think it's QA's job to actually test
every application we push.  Some apps can really only be tested by people
who actually know how to use the them.  QA can (and should) do automated
testing and provide automated testing platforms, etc (they're working on
that).  But, for example, if someone pushes a nagios-plugins update out, I
don't think we can expect QA to know how to use them and give the thumbs
up.  This is especially true for the volume of updates we push out.

Though, in theory, fewer updates means a higher percentage of them can be
tested which means quality goes up.

	-Mike