future content of OpenJFX
johnvalrose at gmail.com
Wed Feb 7 20:23:19 UTC 2018
OK, after Wolfgang’s comments, I will unleash my rant again in the “appropriate” thread as I feel that the lack of JavaFX adoption is very much due to the nature of the toolkit itself:
Well, not only do I think that a docking framework is *that* complex, I see it as an essential part of any serious graphics toolkit.
In general, I don’t understand all the limitations that people keep imposing on JavaFX as if we have to “settle” for what is basically a 2nd class citizen compared with toolkits in other languages.
Why can’t we just make JavaFX so compelling that developers from other languages are enticed by the feature set itself? There is no reason (other than a lack of effort) why JavaFX cannot be on par with Qt or Xamarin in terms of features, performance, tooling and adoption.
Am I the only one that believes the world’s most popular and amazing programming language, platform and the JVM deserves a first class graphics toolkit?
I understand the constraints imposed on Oracle staff but isn’t there anyone else out there in the community who shares my vision for JavaFX?
It seems the general attitude is that JavaFX just needs to be a “better Swing”.
Forget about this nonsense of “thinking outside the box”.
There is NO BOX!
JavaFX can and should be the best that we as a community can make it.
And then we just keep making it better...
If we don’t adopt such a vision and attitude then what is the point of doing anything at all? Just leave JavaFX to rot into oblivion and relegate Java to a server-side language only.
I’m sure Jyloo and others would be much more successful and get real ROI on their investment in JavaFX if JavaFX itself was “better” and more competitive.
We don’t want developers in general to ask “Hmm, so I need a graphics toolkit. I mostly know Java pretty well so I guess that means JavaFX will just have to do even though I’d use Qt if I knew C++ because it’s sooo much better”.
What we want is “Hmm, so I need a graphics toolkit. Let’s see what’s out there. Hey, this JavaFX thing looks awesome! It has all the features I need, performs really well, runs on most platforms and can be programmed by any JVM language! And there’s all those amazing IDEs and tools and Java developers are really easy to find. Yes, I think I’ll choose JavaFX for this new multi-million dollar system!”
> On 8 Feb 2018, at 00:07, Wolfgang Zitzelsberger <openjfx at jyloo.com> wrote:
> As a vendor of third party controls it's finally time for a feedback. We create look and feels and controls mainly for Swing/JavaFX desktop business applications. For us the most important things are:
> * Bugfixing - and we've already reported a lot over the past years
> * Rock solid base controls
> * An extendable API
> * Behavior API
> Because of the API, in Java 9 for some methods the visibility changed from protected to package local - in conjunction with JPMS it's pretty hard to find proper workarounds. Some of our FX controls are pretty sophisticated (e.g. our Table control) but all the open bugs makes further work pretty expensive simply because it slows development extremely down. Even if we noticed some higher activity in bug fixing, many open bugs are moved from one release to the next without getting fixed. Another big issue is the bug fixing latency - when we post a bug today it will be possibly fixed in Java 11 or later. For us this means a new product feature can't be tested and released before Java 11 - time goes by and in the meantime other technologies win.
> Johan, JavaFX is much more popular on Google Trends now than Swing because some of you guys praised JavaFX as the "successor of Swing" and often followed the "Swing is dead" mantra. If you are looking for a competitor you should take a look at web technologies or other non-Java products. IMHO Swing is *not* a competitor it's simply the older brother. Anyway, this doesn't matter in this discussion. For us creating complex JavaFX controls is more complicated than in Swing - not because of JavaFX itself, it's mainly because the JavaFX API is currently too restricted (way too many final and package-only visible methods).
> We've already invested thousands of bucks in JavaFX development without any ROI. Not a problem at all but at some point this is quite frustrating. And we are far away from seeing a huge interest in JavaFX business products. BTW: Any number of downloads can be misleading if you know nothing about the user base - business, education, hobby, bot...
> Just my 2 cents.
> Cheers Wolfgang,
> CEO of Jyloo Software
More information about the openjfx-dev