Some recent changes

[Paul W. Frields]

On Tue, Oct 20, 2009 at 11:53:28AM -0400, Matthias Clasen wrote:
> On Tue, 2009-10-20 at 08:48 -0700, John Poelstra wrote:
> > Matthias Clasen said the following on 10/20/2009 08:33 AM Pacific Time:
> > > On Tue, 2009-10-20 at 08:06 -0700, John Poelstra wrote:
> > > 
> > >> I think removing the hide desktop button is a really bad change too.  I 
> > >> also don't understand why these changes continue to pile on when we are 
> > >> way past past alpha and feature freeze.  These do not seem like "bug 
> > >> fixes."  Are more changes planned too?
> > >>
> > >> Is there a compelling reason (e.g. user interaction study, etc.) why 
> > >> this change is being made and why it has to be made now?
> > > 
> > > This change has to be made now because we want F12 to be as polished as
> > > it can be. Dropping dead and refraining from any further changes at beta
> > > time might be comforting to rel-eng and qa, but it is not conductive to
> > > getting the best release we can. 
> > > 
> > > I strongly deny that we have to provide a 'user interaction study' for
> > > every change we make.
> > > 
> > > 
> > 
> > Please quote the section above where I said you "have to provide a user 
> > interaction study." I still don't see the compelling reason for this 
> > change beyond that you thought it should happen.
> 
> Right, you didn't actually say that. I just read it that way. Sorry.
> 
> 
> > Is this an upstream decision or maybe you "are upstream"?  I'm still 
> > trying to understand how all this works :)
> 
> There is no upstream in the default panel configuration - this has
> always been our own. Look for redhat-panel-default-setup.entries in our
> gnome-panel package. So yes, this is just our effort to come up with the
> best initial panel configuration for the desktop spin.

In a previous email, I recall one member of the Desktop team writing
that help from other people was needed and appreciated -- not in the
sense of a design by committee, but in the sense of finding problems
and testing changes.  The current example seems like a very
ineffective way of getting even that kind of help.

I appreciate the effort to further improve the Desktop spin, but the
timing of these changes really undercuts those earlier points.  By
making these changes post-Beta, we're eliminating a good bit of that
potential testing.  Mairin's post was enlightening for me, because she
showed how the "Show Desktop" button was documented to cause problems.
It would have helped to be able to tell people ahead of time in our
many Beta announcements, "Here are some changes to defaults that some
people will experience, and this is why they were made, in service of
making a better user experience."

I do appreciate the fact that Matthias went to the trouble of making
some entries in the release notes to bring them up to date.  I believe
the Docs team should be able to get those changes into the final copy
for F12 GA.

And I completely understand that the code changes overall in the
Desktop spin take a huge amount of time in the cycle.  Many people,
including me, do as much as possible to champion that work, and show
how usable it makes the free software desktop today.  And there isn't
a huge amount of time left for polishing things at the last minute.
That's even more true when that process also requires courteous (and
downright necessary) heads-up for the very people who are doing
testing, documentation, and triage.

So how can we make this better in the future?  Some ideas...

* Keep a list of these small target issues somewhere, in BZ, the wiki,
  or whatever is the preferred, easy location where interested people
  can read it

* We get together a few weeks before final freeze to see what is being
  contemplated for picking off the list

* Someone from Fedora Marketing team (and that might even be me) makes
  a post about the possible changes, as a heads-up to the community,
  and writes up a wiki page showing experts how to reproduce old
  behaviors if they care; we link this in the notes.

-- 
Paul W. Frields                                http://paul.frields.org/
  gpg fingerprint: 3DA6 A0AC 6D58 FEC4 0233  5906 ACDB C937 BD11 3717
  http://redhat.com/   -  -  -  -   http://pfrields.fedorapeople.org/
  irc.freenode.net: stickster @ #fedora-docs, #fedora-devel, #fredlug