https://bugs.kde.org/show_bug.cgi?id=343246
Xavier Guerrin <***@kindwolf.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |***@kindwolf.org
--- Comment #34 from Xavier Guerrin <***@kindwolf.org> ---
I intended to comment on #341143 (which has the advantage of being reopened)
but, following loop.rw's suggestions, here seems to be a better place.
Many things have already been said in the reports listed by loop.rw and I
concur with most of them. Basically, I think the key word here is "regression"
(or "drop"). The word itself does not appear in so many comments, but this is
clearly what we have here: people have been able to set a different wallpaper
on each virtual desktop for years (at least 11 years as far as I can say,
possibly more).
Like many options, not everybody use it for the same reasons: some like the
resulting aesthetics in grid/cube view, some like it to quickly switch to a
wallpaper that fits their mood, some use it to recognize on which desktop they
are and feel completely lost without it, etc: it is hard to put everybody in
simple, nifty user cases/scenarios.
One thing is sure here: it is now considered a basic feature for a desktop
environment. Some desktop environments chose not to implement it, sometimes to
remain lightweight, sometimes to promote an "everybody-should-live-this-way"
approach, sometimes because of lazyness or ineptitude. On the other hand, KDE
has always been praised for its lot of features and options, i.e. for its
ability to let end users configure and organize their desktop environment the
way they like it, not the way some UX guru decided things should be.
From that perspective, KDE activities are neither the culprit nor the solution.
Some people love KDE activities, and it is good as it means they have not been
implemented for nothing. On the other hand, some people do not recognize them
as a proper replacement to per-desktop wallpapers/widgets, and their voices
ought to be heard.
A few other things to consider:
On the technical feasibility of the feature:
I have read the regression is "due to architectural changes" and that because
of this, KDE "can't really get back". Let's be completely honest on the matter:
the feature itself has been possible for years, so there is no way KDE can make
the world believe it is not possible anymore.
Of course, we can understand the re-implementation would be particularly
painful because of the new underlying software architecture. But, again, coming
from the team who chose to start two major rewrites in the last decade, this
sounds phony. Over the years, the KDE community has been able to implement many
programs with a tremendous amount of features and options (my personal favorite
being KIO/slaves) and suddenly, what has been possible for years is not
possible anymore? That simply won't do.
On the anger one can feel when reading some comments:
Well, I assume KDE developers are used to angry users depicting a particular
bug as the new apocalypse that will strike the whole world of Free Software
from the universe. However, it seems to me it is very easy to relate to the
anger of those users: most of them discovered the issue right after discovering
that the jump from 4 to 5 would have them reconfiguring Plasma entirely (a
particularly user-unfriendly policy). Some discovered it after having painfully
solved the switch from X.org + kdm to wayland + sddm. Most discovered it was
the worst regression but not the only one. Most discovered this while hoping
the KDE team would have learned something from the rather painful KDE3 -> KDE 4
migrations: software are expected to evolve, gaining features and stability as
time goes on. Having to reconfigure something feels particularly wrong in 2015,
but not being able to reconfigure it the way it was before the upgrade is one
of the most frustrating things for end users so, naturally, anger ensues.
This being said, and as #341143 has been reopened, I feel like this topic needs
technical clarification in order to progress: would it be possible to detail
the exact, technical constraints that make it so hard to bring this feature
back? What makes Plasma so strict regarding wallpapers and widgets? Could the
feature be delegated to an external application/process/plugin? Many thanks in
advance.
--
You are receiving this mail because:
You are watching all bug changes.