On Wed, 2020-04-15 at 17:14 -0400, Stephen Gallagher wrote:
Right now, we have two main criteria around the default
background/wallpaper for Fedora releases:
Basic Criterion: "The default desktop background must be different
from that of the last two stable releases."
Final Criterion: "The proposed final Fedora artwork must be included
and used as the background on release-blocking desktops. All Fedora
artwork visible in critical path actions on release-blocking desktops
must be consistent with the proposed final theme."
The reason for the Basic Criterion is largely to ensure that people
have a simple visual reminder that they are not using a stable
release. I don't agree with the exact phrasing there.
Proposal 1:
Basic Criterion: "The default visual experience of Fedora pre-releases
must be sufficiently differentiated from stable releases so as to
avoid easy confusion." We can then expand on that to use the current
criterion as an example of how that may be accomplished (but it need
not be the only way). [1]
The reason for the Final criterion is a fit-and-polish one. We want to
ensure that the final product is as clean as we can make it. However,
I don't think we necessarily should block the release just for the
background.
Proposal 2:
Final Criterion: "Fedora may not ship release-blocking desktops with
visibly[2] unfinished artwork."
We would then add the following to the list of Automatic Freeze Exceptions[3] :
"Changes that modify only the aesthetic of Fedora, such as the default
wallpaper or window manager themes."
This would allow us to get late fixes in for the wallpaper and similar
artwork, but not require us to slip for it.
[1] I envision a world where we could theoretically have the
"Background Logo" GNOME extension display a "pre-release" notation
or
something similar.
[2] Defining "visibly" as "an average user would consider it out of
place, such as UI elements being completely missing".
Honestly, I don't really like...any of these. I kinda get the intent
but they all feel icky, mushy and squishy. We give an automatic FE to
*anything* that can be claimed to only modify "the aesthetic of
Fedora"? That seems like a loophole big enough to drive a truck
through. Proposal 2 seems very vague and suggests that if we
accidentally put in artwork that's completely wrong or incredibly ugly,
but not "unfinished", we're stuck with it...
I honestly think the main problem we have with this stuff is not the
criteria. It's the process: the fact that the packages are kind of
complex and involve a bunch of moving parts, and the fact that it seems
like there's really only poor finalzone maintaining this stuff and he
doesn't necessarily have the bandwidth to keep it all up to date
promptly given how complicated it all is.
The criteria that currently exist were actually specifically written to
enable a particular plan for improving the packages that we developed
at one point, and which Kevin was going to work on in his copious spare
time. Entirely inexplicably, given all those reams and reams of spare
time he has, this hasn't happened yet. That's the real problem here.
Re-arranging the release criteria deckchairs isn't going to help, I
don't think. Fundamentally, the things the criteria is covering should
not be *difficult* things, we just kind of suck at doing them.
--
Adam Williamson
Fedora QA Community Monkey
IRC: adamw | Twitter: AdamW_Fedora | XMPP: adamw AT happyassassin . net
http://www.happyassassin.net