Dear community,
This may seem like a strange moment to introduce myself, but I think the
clarification warrants it.
Last decade, I started using a cross-platform CMS. At that point I
hadn't yet done any programming for a production software project. The
first bug which blocked me was specific to Microsoft Windows, and I hit
quite a few Windows-specific bugs. This got me to start contributing to
that big project and got me thinking about switching to UNIX.
Eventually, I realized I couldn't make a release on Windows. As a couple
of my friends used Mandrake (Mandriva) at the time, I switched to it.
Several months later, my mentor in that CMS project convinced me to
switch to Knoppix. A few weeks later, Knoppix's nature got me to switch
to Debian. At the time, Debian stable was really old, so I chose Debian
testing. When I installed Debian, I tried using GNOME for the second
time. I was hoping to switch to GNOME because it was the "pure" desktop
environment, Gnu-er than KDE. As happened in my first attempt, I failed
because I couldn't use GNOME with my keyboard layout. One of the things
which helped me settle for KDE was that at the time (2004), it was
perceived as the more popular option.
I got very involved with Debian and became for a while its top ticket
triager. I triaged KDE reports in particular and forwarded several
upstream. I have been contributing since that time, but was never very
active directly in KDE.
Eventually, I became more busy and had much less time for volunteer
work. A few years ago, this got me to finally switch to Debian stable,
to waste less time on regressions. Unfortunately, Debian stable was
still unstable. Early in 2019, it got to a point where I had to restart
KDE about weekly, even if I haven't used Debian for remunerated work
since 2012.
So I figured that well, if nobody uses Debian KDE, I would better try
something more popular. And I installed Kubuntu 19.04 when it came out.
Unfortunately, that was a disappointment too. 2 major bugs I had been
experimenting on Debian 9 persisted on a fresh Kubuntu 19.04: all notes
keep disappearing from KNotes, and KOrganizer still loses reminders.
Additionally, there is a big window manager bug and several Plasma
crashes. All of that on a 6-year old solid desktop which I carefully
shopped for quality and GNU/Linux compatibility. And I didn't even try
using KDE software for package management, browsing or mail. One of my
worst disappointments was to realize that my migration got me to lose
the KDE product which had been my favorite for more than a decade,
Amarok, which was removed from Ubuntu starting with 19.04.
Perhaps I didn't get the best version, but all of that got me thinking
once again about what I should do.
But the worst issue was a serious KOrganizer regression, which made it
basically unusable. A few months later, I got a few weeks of vacation
and attacked that one. To my surprise, I found that almost none of the
underlying issues had been reported. I filed several tickets,
culminating in a "summary"/end-user ticket against Kubuntu:
https://bugs.launchpad.net/ubuntu/+source/korganizer/+bug/1837156
I noted a number of KDE tickets in need of love during that reporting
spree and used the occasion to triage a few. This eventually got me to
report a bug against bugs.kde.org. That (plus more triaging) in turn got
me to report more bugs.kde.org issues.
I eventually identified 2 meta-issues affecting bugs.kde.org and asked
for help dealing with them in a thread I opened in November:
https://mail.kde.org/pipermail/kde-community/2019q4/005637.html
I reported 2 major issues in that thread:
1. Much difficulty getting ticket "severity" (importance) adjusted
2. Recurrent ticket mishandling from Nate Graham
There were enough replies to that thread, but answers can easily be
summarized to Harald Sitter's following assessment:
"Nate is well. He's a treasure to this community. I have read through
the reports and couldn't find him in the wrong."
This made me firmly decide to give up on recommending measures to deal
with Nate.
There was no useful answer about point 1. Probably as a result of the
thread, there was some activity in
https://bugs.kde.org/show_bug.cgi?id=272388
One trivial fix was made, but the issue remains and the ticket was
marked as resolved. I consider the net outcome as negative.
This long path has now lead me to ask this list a question which I
myself find vague, because I want it open: Am I missing a much better
way to use KDE?
I know that Kubuntu is a bit of a SCC. But if I look at all "universal"
GNU/Linux distributions, the only (apparently?) serious option I seem to
find which is focused on KDE is openSUSE.
I say serious because I want something minimally popular. I don't mind
dealing with and reporting a few issues sporadically, but I'm looking
for a product first, not a project.
And does openSUSE even have a 10% market share on the (free software)
desktop?
In fact, this brings me to an equally important question: do we have a
fair estimation of KDE's current market share on GNU/Linux? As I wrote,
I settled for KDE in part because it was dominant, but when I look at
distribution defaults nowadays, I'm under the impression that KDE has
plunged a lot. Much of that must be to blame on higher fragmentation in
the desktop environment landscape, but I now wonder whether KDE is still
more or even as popular as GNOME?
The mere extensiveness of the yearly report
https://ev.kde.org/reports/ev-2018/ shows this community still has some
momentum. At the same time, while it is very good at showing that KDE
has IRL meetings (and justifying the high share of expenses dedicated to
these), I failed to identify metrics showing increasing (or even stable)
momentum. Besides for one element: the yearly income is reportedly
"record". Unfortunately, this record is not quantified, and it's not
even clear in which year this record was set. There is not a single link
pointing to data for previous years. I did my best to report these
issues, but this hasn't helped so far.
We used to have a weekly "commit digest", with statistics on tickets.
Now, even our yearly report doesn't say a word on tickets. Were less
tickets reported than previously? Were less solved than in previous years?
Don't get me wrong - I remember where KDE on Mandrake was 15 years ago
and the progress since is enormous. KDE is much more usable, but its
architecture also evolved a lot, software on which it depends evolved
tremendously too and dependencies are more flexible.
I'm sure I can use an even way better KDE operating system in 15 years,
but I'm still worried if the pace is slowing down.
And I'm not just worried about where we'll be in 15 years, I'm also
worried about how we'll get there. A large part of the issues I faced
this year weren't there 15 years ago. If for every step forward we have
to take 1 step left, 1 step back, 1 step right and an extra step
forward, some can enjoy the physical training, but if the erratic dance
resulting is mocked by others, the journey is just a lot more exhausting
and frustrating.
I recently bought a new laptop, which unsurprisingly came with an
infamous proprietary OS. I must say that while I evaluated 1 KDE product
on Windows, I still haven't installed KDE. If there are encouraging news
about KDE, options I missed, or if someone has convincing indications
that the project is progressing at good speed, with less deviation from
the straight line to our goal, now would be a good time to tell me to
prevent me from going back to where I once came from...
Can we expect that quality control will become more respected,
rewarding, efficient and ultimately attractive, so that more users will
share the work? Or will we ensure KDE products are marketed in a way
which better reflects their maturity?
Thank you
--
Philippe Cloutier
http://www.philippecloutier.com
