On Saturday, July 11, 2020 2:36:04 PM MST Neal Gompa wrote:
There's a difference between half-baked and a roadmap of
incremental
improvement. This Change is an example of the latter. If anything,
your fanciful and speculative conjecture has done only harm for your
credibility. You continue to attack and lambast every change proposed
for the past few cycles,
On the contrary, I've only been against Changes that I've considered to be
harmful to Fedora. There have been a surprisingly high number of those this
release cycle, but that doesn't mean that that I'm against all Changes.
and you provide no useful alternatives.
The alternative to this would be.. not enabling EarlyOOM, which will kill our
users' processes needlessly.
Additionally, you continue to insult our intelligence by assuming we
haven't done our homework and attack the Workstation working group for
doing a lot of legwork to try to solve very real problems that users
see.
I'm sorry if it came off that way, but that wasn't the intention.
With respect to EarlyOOM, I *wanted* to introduce this to all
desktop
variants at the same time last cycle, but the amount of research and
effort we spent to figure out what we should even *do* meant that
there was no time to do any coordination work last cycle to bring this
everywhere. The evidence from user reports and reviews has been clear:
Fedora 32 Workstation Edition had a phenomenal user experience
improvement in terms of desktop responsiveness over its predecessors
as well as many of its counterparts that released at the same time by
our fellow Linux distribution communities. Much of the work we're
doing for Fedora 33 and beyond is a continuation of that effort.
The real question is whether or not EarlyOOM has anything to do with this
"phenomenal user experience", which seems unlikely. Generally, users'
processes getting killed isn't a good thing.
My personal mission is to bring all the improvements we've done
to the
Linux desktop experience in Fedora Workstation to at least Fedora KDE,
if not all Fedora desktop variants. And where it makes sense, I will
attempt to bring these improvements to *all* Fedora variants.
That sounds like an excellent idea, but I'm not convinced that killing users'
processes, while there's still tons of memory free available, is actually an
improvement.
--
John M. Harris, Jr.