Announcement

Collapse
No announcement yet.

The Qt Company Launches Qt Marketplace For Free + Paid Qt Extensions / Add-Ons

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #11
    The KDE developers should probably fork Qt at this point...
    It would be better than doing a complete rewrite every 6-7 years...

    Comment


    • #12
      Originally posted by frank007 View Post
      Linux needs a completely free gui library.
      Free and not shitty. Not some not invented here OOP / reflection framework built on top of legacy C (GTK, EFL). Modern GUI techniques require decent support for many advanced OOP features: intersection types, proper unified object hierarchy with bottom and top types, proper variance rules for higher kinded parametric polymorphism, reified generics, lifetime management not tied to static scopes, recursive product/sum types, etc. Note that most of these feature require advanced typing, not void* type punning. Orthogonal requirements can be identified when integrating with modern GPU processing pipelines.

      Comment


      • #13
        Originally posted by frank007 View Post
        Linux needs a completely free gui library.
        One that you can use for whatever you want, can modify however you like and can freely distribute both verbatim and modified versions of?

        We've already got a few of those, and one of them is called "Qt".

        Comment


        • #14
          Originally posted by ddriver View Post
          They just keep on pushing Qt to be everything except what it needs to be - a stable, bug free and capable software development framework.

          It saddens me to think of the developer effort wasted by their terribly misguided management.

          So many wheels reinvented over and over again, so much catering to short lived fads, so much bloat and code duplication, and still - missing important functionality to make it a complete solution, introducing new bugs faster than solving the existing ones.



          Any evidence of starving families there?
          As someone who runs KDE on a daily basis, I have to say its far ahead of GTK+ in that regard. I don't think I've ever had an issue that can be attributed to QT.

          Comment


          • #15
            Originally posted by archsway View Post

            One that you can use for whatever you want, can modify however you like and can freely distribute both verbatim and modified versions of?

            We've already got a few of those, and one of them is called "Qt".
            Not quite. QT basically forces you to open source your software unless you pay for their commercial version (which costs even more than Microsoft's solutions!). While open source is a good idea, not every company that makes their bread and butter from software can do this. Quite frankly, if I wanted to deliver a commercial software package built on QT today, I couldn't do it. The startup costs are quite simply too steep.

            Comment


            • #16
              Originally posted by betam4x View Post
              Not quite. GPL basically forces you to open source your software unless you buy the Free Software Foundation (which costs even more than Microsoft's solutions!). While Free Software is a good idea, not every company that makes their bread and butter from software can do this. Quite frankly, if I wanted to deliver a commercial software package built from GNU today, I couldn't do it. The startup costs are quite simply too steep.
              FTFY.

              Comment


              • #17
                "QT basically forces you to open source your software unless you pay for their commercial version"

                Ehh! No they do not. Sine Qt is LGPL, that is no problproblem.

                ​​​​​

                Comment


                • #18
                  Originally posted by carewolf View Post
                  Because with almost all of the framework being available as LGPLv3 there is little money to be made where they are. Their options are to push to GPL or make more products. Anyway the marketplace seems orthogonal to that. It is mostly just making existing products more visible, including lots of products from third parties in the Qt ecosystem.
                  You can use Qt under GPL as well, there are many modules that are only available under GPL or commercial. Most of what they deem as moneymakers is licensed this way.

                  They make little money because their commercial licenses are ridiculously overpriced. And Qt is not a vanity booster product, which is what people are typically willing to throw needless money at for some bling. Big companies tend to use the platform native SDKs, even if not portable, so Qt catering to big players is perhaps not the best strategy, because those have money to throw at native development. The portability aspect is far more enticing to smaller companies or independent developers, on which Qt could make a lot of money if they could swallow their greed and offer full (not crippled) commercial licensing that smaller development outfits actually can afford. A % of revenue would be the best solution IMO, with no ridiculous upfront cost, and you pay proportionally to what you make - that seems reasonable and fair. Instead they loudly voice the complete opposing notion - they claim you cannot even start a product under LGPL and switch to commercial when the product makes enough money to justify that licensing...

                  I don't think they need to make more products, they need to make Qt into something worth adopting, people more creative than them will make both the products and the encompassing ecosystem. Their architects clearly don't have vision, thus the framework is wandering about, not quite sure what it wants or needs to be. Way too many issues, way too much feature duplication, way too many still gaping holes in the functionality stack. Years of development wasted by misguidance - for example the now pretty much completely rewritten for Qt 6 QML, after they needlessly put heavy emphasis on JS in the blind belief this will somehow gain them millions of web developers as new users. Their "new" products are always a little too little and a little too late, clearly the company cannot neither innovate nor compete at those things, I think they should really focus on their core product, and only branch away after it is sufficiently refined.

                  Qt is pretty much without competition in its own category, but being relatively the "best" solution is not that much of a big deal when it is in a context of a "one of the few available and the least sucky and incomplete" victory. There is still a lot to be desired, and they are not addressing it in the slightest. The lack of direct competition also gives them a false "bigger than the world" sense, which seems to make it redundant for them to improve on the core product.

                  Instead they are squirming to pocket a little more money every which way, not thinking big (even if acting bigger than they are), not thinking long term, having no big picture to paint, wasting dev effort firing blindly into random directions, hoping to hit something profitable, which neglects what they should focus on and degrades is further.
                  Last edited by ddriver; 12-03-2019, 02:53 AM.

                  Comment


                  • #19
                    Originally posted by mzs.112000 View Post
                    The KDE developers should probably fork Qt at this point...
                    It would be better than doing a complete rewrite every 6-7 years...
                    Why would anyone do a rewite? KDE 5 was a major change because KDE chose to significant change their own structure, not due to anything with Qt.

                    Comment


                    • #20
                      Originally posted by caligula View Post

                      Free and not shitty. Not some not invented here OOP / reflection framework built on top of legacy C (GTK, EFL). Modern GUI techniques require decent support for many advanced OOP features: intersection types, proper unified object hierarchy with bottom and top types, proper variance rules for higher kinded parametric polymorphism, reified generics, lifetime management not tied to static scopes, recursive product/sum types, etc. Note that most of these feature require advanced typing, not void* type punning. Orthogonal requirements can be identified when integrating with modern GPU processing pipelines.
                      Yes, not everybody can do a good gui libs. But can be done.

                      Comment

                      Working...
                      X