Qt3>4 was a very very very large change which impacted a lot of applications requiring source changes before they compile again. Qt4->5 had significantly less breaking changes, and Qt5->6 only had a small handful of changes.
KDE just recently got KDE6 out. Qt Company serves a lot of embedded customers who need stability.
I think the 6 series is here to stay for a while. At least until there’s a very good reason to break forward compatibility which I haven’t heard of anything even near the radar.
Yes, the reason for doing major version bumps like that is to clean out deprecated features/APIs and in order to make major changes that would be difficult to do in a maintainable and backwards compatible way. Things like changing rendering backends, switching build tooling, fixing architectural mistakes etc. If tech debt hasn't slowed down the development of Qt6 yet and all features on the roadmap are still considered reasonably implementable then there's no real point in doing a Qt7 yet.
The point I was making was the trolls spent the past decade doing a great job of that tech debt cleanup (as well as making the framework use new C++1x features.
Or maybe Qt company will troll them and they'll be forced to release Plasma 7 and completely drop support for X11 a couple years earlier than planned lol
I wish we could do something about the long tail of niche closed-source commercial applications that seem permanently stuck on Qt 4, with nothing properly motivating them to upgrade.
This went from mildly annoying to outright infuriating the moment I switched to a HiDPI display on my Linux system. Heck, at this point I'd even take a specially patched/shimmed variant of Qt 4 as an alternative.
I probably have in the past. The problem is that HiDPI compatibility mode on Windows works "well enough" that bigger customers who actually matter to them aren't complaining. Also, when a company has been resting on their laurels on Qt 4 for a decade, asking them to upgrade to Qt 5 or 6 just to please an extreme minority of users is a bit of a tall ask.
Right now the biggest culprits that I use are everything from Segger (J-Link software, Ozone, SystemView, etc), and the software for the Beagle Total Phase protocol analyzers.
None of these are what I'd call "typical end-user applications", but they are things I regularly use for embedded development work.
Beyond Compare also took a while to update, but they had a better excuse and have now been on Qt5 since v5 of their app released last summer.
Zeenss@reddit
Next qt 7 or qt 6.10?
throwaway6560192@reddit
It will be many years before another major version bump for Qt, as far as I know.
the91fwy@reddit
I’ve heard nothing about Qt7.
Qt3>4 was a very very very large change which impacted a lot of applications requiring source changes before they compile again. Qt4->5 had significantly less breaking changes, and Qt5->6 only had a small handful of changes.
KDE just recently got KDE6 out. Qt Company serves a lot of embedded customers who need stability.
I think the 6 series is here to stay for a while. At least until there’s a very good reason to break forward compatibility which I haven’t heard of anything even near the radar.
Salander27@reddit
Yes, the reason for doing major version bumps like that is to clean out deprecated features/APIs and in order to make major changes that would be difficult to do in a maintainable and backwards compatible way. Things like changing rendering backends, switching build tooling, fixing architectural mistakes etc. If tech debt hasn't slowed down the development of Qt6 yet and all features on the roadmap are still considered reasonably implementable then there's no real point in doing a Qt7 yet.
the91fwy@reddit
The point I was making was the trolls spent the past decade doing a great job of that tech debt cleanup (as well as making the framework use new C++1x features.
ilep@reddit
Exactly. Qt 6 was one with notable changes. For example, newer compiler standard and changing build system to be based on cmake instead of qmake.
It would need to be something of similar magnitude before Qt 7 is justified (notable difference in compatibility).
Qt versioning is different from Linux kernel, which is just time-based model these days (major numbers are just numbers without significance).
equeim@reddit
Version numbers are not decimals.
__konrad@reddit
If you increase minor version it will just back to 6.1 ;)
vyashole@reddit
What?
__konrad@reddit
6.1 and 6.10 decimal numbers are equal
vyashole@reddit
But version numbers are not decimal. Am I in a whoosh moment?
webmdotpng@reddit
Maybe 6.10, because KDE follows Qt versions, and will take some time for Plasma 7.
p0358@reddit
Or maybe Qt company will troll them and they'll be forced to release Plasma 7 and completely drop support for X11 a couple years earlier than planned lol
ACNAIsNotChristian@reddit
I mean...y'know...
nevadita@reddit
Yeah, they do know how to
jlpcsl@reddit (OP)
The QtReleasing wiki page says 6.10 https://wiki.qt.io/QtReleasing
dkonigs@reddit
I wish we could do something about the long tail of niche closed-source commercial applications that seem permanently stuck on Qt 4, with nothing properly motivating them to upgrade.
This went from mildly annoying to outright infuriating the moment I switched to a HiDPI display on my Linux system. Heck, at this point I'd even take a specially patched/shimmed variant of Qt 4 as an alternative.
jcelerier@reddit
If you pay for these apps why not call their support to report the issues you have with them ?
dkonigs@reddit
I probably have in the past. The problem is that HiDPI compatibility mode on Windows works "well enough" that bigger customers who actually matter to them aren't complaining. Also, when a company has been resting on their laurels on Qt 4 for a decade, asking them to upgrade to Qt 5 or 6 just to please an extreme minority of users is a bit of a tall ask.
jlpcsl@reddit (OP)
What are the examples? I don't use any tht does not use at lest Qt5 or have some free/libre opensource replacement that is Qt6/5
dkonigs@reddit
Right now the biggest culprits that I use are everything from Segger (J-Link software, Ozone, SystemView, etc), and the software for the Beagle Total Phase protocol analyzers.
None of these are what I'd call "typical end-user applications", but they are things I regularly use for embedded development work.
Beyond Compare also took a while to update, but they had a better excuse and have now been on Qt5 since v5 of their app released last summer.
GreyXor@reddit
Still not dark fusion native theme ?
Monsieur_Moneybags@reddit
I much prefer Qt to GTK+. I wish there were a Qt version of Emacs.
imbeer7734@reddit
I havent tried to write GTK applications, what is the problem with it?
PM_ME_YOUR_REPO@reddit
nice
rohmish@reddit
qt nice
Altruistic_Ad3374@reddit
nice
Beautiful_Crab6670@reddit
Nice
zlice0@reddit
nice
alerikaisattera@reddit
Nice