Dear Volker,
I will try to address some points from my own context which is that I've
been using Qt for our open source cross platform desktop application since
early Qt4. I've stuck with Qt because of the close to write once (mostly) just
compile and run everywhere kind of thing. I code all day, try to run my
company, hire people, manage projects, manage contracts so frankly I don't have
the time to _learn_ the deep underpinnings of Qt to fix a bug. If it really
only took an hour or so to fix I might actually consider contributing. And at
this point since Qt 5.15.2 will not have any more releases there is absolutely
*no* reason to contribute to Qt.
I think some of the hurdles to putting in bug patches is the following:
* Overly difficult to setup all the needed accounts, git pulls, PRs, Gerrit
review processes.
* Finding time in my own schedule to match up with time from a Qt engineers
schedule.
* My bugs probably may not align with TQtC priorities
* Writing a unit test is always a blast. Again, time waiting for a code review.
* Ensuring it works cross platform when most may not have access to all of the
needed testing systems.
I agree with the others that the LTS releases need to go out to open source. I
understand that The Qt Company is a business and needs to earn revenue to pay
its own engineers. On the flip side though TQtC yanking the licensing rug out
from underneath my company puts _my_ company at risk. I don't have a solid
answer for how the licensing should be done, but clearly the current practice
_isn't_ well received by the Qt community. I don't think the community is
asking for much to have the 5.15.8 released to OSS.
Respectfully
Michael Jackson
Owner, BlueQuartz Software, LLC
On 2/24/22, 10:39 AM, "Interest on behalf of Volker Hilsheimer"
<[email protected] on behalf of [email protected]> wrote:
Hi All,
Thanks for keeping it civilised.
Yes, Qt Quick Controls - and largely the entire Qt Quick framework - were
originally designed for mobile and embedded platforms, and indeed, that shows
when using them for the desktop.
I’m happy that at least in The Qt Company we are now in a position that
allows us to put more focus on the desktop, and that we are are able to do more
than maintenance and catching up with what’s happening on the underlying
platforms. That includes the journey of making Qt Quick Controls a great
toolkit for the desktop as well. In Qt 6 so far we have had first
implementations of the native styles - yes, those require more work; we have
made a number of improvements to item views, including a TreeView now in Qt
6.3; a first set of standard dialogs is in Qt 6.2 and more are coming in 6.3.
We have worked on some architectural issues that are problematic on the
desktop, such as keyboard navigation and focus handling, and there is a fair
amount of more work needed there as well.
I’m not going to claim that all things will be wonderful any moment now;
there’s plenty of work that needs to be done. But things do get better, both
with Qt Quick Controls, and with Qt Widgets as well.
What keeps confusing me personally is how few people in the community seem
to find it interesting to contribute to either of our UI frameworks in Qt. If I
take one of the QtWidgets issues that came up in this thread: "QTBUG-6864 is 12
years old, has 47 votes”. I sat down on Tuesday evening to check what it would
take to implement hiding of rows in a QFormLayout; after a few hours I had a
working implementation, which is right now on its way into the dev branch. The
hardest part, as it so often is, was writing a unit test.
Now, I understand that not everybody finds it fun to do that kind of thing
on a Tuesday evening. But given the apparently high interest in this feature,
that nobody seems to have tried to give it a shot in 12 years is really
puzzling me. When Nokia acquired Trolltech, it didn’t take a crystal ball to
see that the focus won’t be the desktop. And one answer to this was to move Qt
under Open Governance so that anyone could contribute to Qt and make sure that
it stays awesome also for domains that Nokia won’t care much about.
Evidently, the people commenting in this thread care deeply enough about Qt
on the desktop to participate in the discussion. And I suppose most of us on
this list are software engineers, many perhaps for more reasons than to put
food on the table. My question to you is: how can we make it easier, or more
fun, or more motivating to contribute to Qt, and to help with making things
better?
Volker
_______________________________________________
Interest mailing list
[email protected]
https://lists.qt-project.org/listinfo/interest
