Is JavaFX not to late?

[Stephane Clinckart]

The question is simple...

Why should JavaFX be used instead of Flash or Silverlight?
More efforts for less results :-(

May be I'm wrong ?
--> I know not enough about all these technologies... I play to less with them...
But... in average with technologies... when the efforts is to big for a small result at the end... the tech isn't choose by the communauty of developers.
Certainly today... with this world crisis where developpers has to build qwick and durty code to be rentable.
May be Oracle will change that by putting some money in these technology? I'm not sure... but who knows?

Try to convince me if I'm wrong.

Kind regards,

Stephane

[Mr Obi Orjiekwe]

I totally agree on the JavaFX. If Flash is so readily available and browser ubiquitous why bother with new tech? I've seen a few demos and have been in impressed but what is the incentive to learn (apart from learning something new)

quick and dirty code to be rentable? no compadre. My recent experience is the opposite. recruiters and the general community are looking for quality more than ever - clean architecture, clean code, documentation at the right level and audience.

[Stephane Clinckart]

Mr Obi Orjiekwe wrote:I totally agree on the JavaFX. If Flash is so readily available and browser ubiquitous why bother with new tech? I've seen a few demos and have been in impressed but what is the incentive to learn (apart from learning something new)

quick and dirty code to be rentable? no compadre. My recent experience is the opposite. recruiters and the general community are looking for quality more than ever - clean architecture, clean code, documentation at the right level and audience.

I agree... they want more... but pay less.
At a moment there is no possibility to do it... or you die at work.
--> You get payed for 40h/week... but need 60h to finish "correctly" the job.
This is the "tendance"... at less here arround me.

Some people do the job well (take job at home, stay longer, ...) other people provide "durty" code...
@ the end... nobody win... People who take home at work or stay longer after a while get divorced, be in depression ,....
and change opionon about "theyr job". At the end they provide "durty" work ...
Later, they get fired... So what s the benefit at the end???

But... please, convince me that JavaFX is a good product... Why should I invest time in it more than in another product?


Regards,

Stephane

[James Ward]

In my opinion JavaFX is indeed late.

Flex based applications are quite mature these days, and Flex(Flash) front-end with J2EE back-end is now becoming popular too.

Also - it is very easy for a Java Programmer to pickup Flex as the Scripting language used in Flex (ActionScript) is very much java like (Flex IDE is basically an Eclipse IDE flavor!).

But, you may want to consider the following too:
1. Silverlight too is a bit late.
2. There may be space for more than one player(Flex,Sliverlight,JavaFX) here - so JavaFX may have a future - But - there does not seem to be any noise/news/push from Sun/Oracle in this direction these days - dont know why. And when was the last time you saw a real-world application in JavaFX (it is yet to reach that tipping point).

[Mark Spritzler]

JavaFX is actually pretty cool, we are using it for one of our production tools for uploading photos to our website and it works great.

To add to James post about J2EE, You can also hook up Spring Framework to your Flex app, which is also very neat. Of course, since JavaFX is Java, you can easily hook up Spring to it too.

Also note, I have not had time myself in my life to code either Flex or JavaFX, so I don't quite have an opinion on one being better than the other, just a couple of pieces of information that I have heard.

Mark

[Stephen Chin]

There is always room for good new technologies, and a little competition in the RIA space can only make things better.

With the caveat that I know enough Flex to get myself in trouble, and zero Silverlight, here are some of the advantages I see that JavaFX brings to the table:

Seamless Mobile Development - write your code once and run it on desktop, mobile, and (soon) TV

Java Integration - Use your favorite Java libraries from JavaFX, wrap JavaFX in Swing and vice-versa... everything compiles down to Java bytecodes

JavaFX Script - This is probably the most controversial, but I see the JavaFX Script language as a huge productivity booster. It combines the best of a statically typed language like Java with declarative markup like XML, and adds in powerful features like binding and animation. Once you try writing code in JavaFX Script it is hard to go back!

This is just scratching the surface of what JavaFX has to offer.

[Tom Eugelink]

The statement "is a bit late" is correct. However, Flex / Flash suffer from their legacy, mostly based on the fact that flash initially is an animation framework. You can bolt only so much on top, before the foundation starts to cause problems. You can see this from the changes that have been made to the scripting language, as it is being matured to support full fledge applications.

In Java the GUI design also suffered from legacy; Swing works fine, but it is considered complex and difficult-to-do-right. So Sun also had a problematic foundation in its framework for build GUI's. And they decided to break with it and do a complete rework; JavaFX is a completely new technique, targetting a new level of GUI (hardware rendered 2D and 3D), but based on the experience of Java2D and Swing.

Silverlight is also such an attempt, but the difference with JavaFX is the JVM. The JVM is still faster and wider spread (supported OS wise) than DotNet's VM. And you can use existing Java, so JavaFX's software library base is enormous.

Therefore I feel one has to congratulate Sun for being so couragous and dare to break with existing Java GUI development and try to push it to a new level. THe concepts in JavaFX to me are mind boggling; you throw in a few node and hoppa! hardware rendered graphics. You can do games and forms with the same platform. Naturally they must to be able to mix old (Swing) and new, but that is the case.

Now, is it all perfect? No. JavaFX also suffers from the JVM; it's big and even chopped into pieces it takes way longer for the JVM to come up than the flash engine. So ATM there is no replacing Flash when it comes to small decorative animations on websites. But if you want to go fancy full size applications, I think JavaFX has a real chance here!

But Sun does need to keep pushing now:
- mature that technology ASAP
- make sure the foundations are well thought through from all perspectives
- good support for the most popular EDI: Eclipse

[Dean Iverson]

Well said (written), Tom. Those are my thoughts exactly. Well, except that I prefer NetBeans to Eclipse.

(but I'm still 100% behind the concept of better IDE support for both Eclipse and NetBeans)

[Andriy Tsykholyas]

Stephane Clinckart wrote:The question is simple...

Why should JavaFX be used instead of Flash or Silverlight?

Silverlight is not Linux friendly. Yes, there is Moonlight on Linux, but it's not as good as Silverlight.

Stephane Clinckart wrote:
May be I'm wrong ?
--> I know not enough about all these technologies... I play to less with them...
But... in average with technologies... when the efforts is to big for a small result at the end... the tech isn't choose by the communauty of developers.
Certainly today... with this world crisis where developpers has to build qwick and durty code to be rentable.

As far as I know only 20% of software development expenses go for initial build. The remaining 80% go for maintenance. So "quick and dirty code" is not "rentable" for companies.

[Gregg Bolinger]

Andriy Tsykholyas wrote:As far as I know only 20% of software development expenses go for initial build. The remaining 80% go for maintenance. So "quick and dirty code" is not "rentable" for companies.

When posting numbers like this it is a good idea to have a reference to back it up. Otherwise, they are meaningless. Especially when prefixed with "As far as I know". To me, this 20/80 makes no sense. I just spent 4 weeks building a project for a client and I'd say at this point, 90% was development and we've had about 10% maintenance. Of course, we write good software also. ;)

[Andriy Tsykholyas]

Gregg Bolinger wrote:

Andriy Tsykholyas wrote:As far as I know only 20% of software development expenses go for initial build. The remaining 80% go for maintenance. So "quick and dirty code" is not "rentable" for companies.

When posting numbers like this it is a good idea to have a reference to back it up. Otherwise, they are meaningless. Especially when prefixed with "As far as I know". To me, this 20/80 makes no sense. I just spent 4 weeks building a project for a client and I'd say at this point, 90% was development and we've had about 10% maintenance. Of course, we write good software also. ;)

You have spent 4 weeks and you have already maintenance expenses. If you project would run in production for a while, then the real maintenance expenses would reveal
And about reference. I don't remember exactly where I've read those numbers. But it impressed me much. So, I still remember the numbers, while forgot their source
Nevertheless a quick googling ended up with those 3 links at the top:
1) https://www.thedacs.com/databases/url/key/5881
2) http://books.google.com.ua/books?id=JA0pV1KVG4cC&pg=PA248&lpg=PA248&dq=softrawe+maintenance+costs+80%25&source=bl&ots=SOac62KsK6&sig=oSLcH8qYcqFWLWh4l7gLa0HuCC0&hl=en&ei=SSrGSuS1KYXe-QbYtdE8&sa=X&oi=book_result&ct=result&resnum=3#v=onepage&q=&f=false
3) http://fox.wikis.com/wc.dll?Wiki~SoftwareMaintenance~SoftwareEng

P.S. As far as I know JavaFX is statically typed. This is a good thing. Especially in the maintenance phase.

[Gregg Bolinger]

I've no doubt maintenance will rise. I just wanted some links, and you've provided some. Thanks! See how your numbers don't seem so made up now?

[Philippe Lhoste]

Stephen Chin summarized well the advantages. I will add to "JavaFX script" that the latter is compiled to Java, and so uses a powerful JVM with good JIT. A number of comparisons I saw shows that JavaFX is faster than ActionScript 3, and even than Silverlight, despite a perfectible scenegraph (figures were even better with latest releases...).

On the minus points often cited: big runtime that you have to install (Flash is often shipped with browsers, although to use the latest features, you have to upgrade, but install is small and fast); relatively slow startup time (and no good start screen yet); lot of security warnings for some "simple" operations like accessing a foreign Web API or clipboard.

To address (at least partially) these points: with the big runtime, you have out of the box lot of Java libraries, while each Flash application have to embed some code to do the same thing, resulting in bigger applets; startup time is improved, latest Java releases worked on this (so JavaFX tech was useful to Java programmers too); and I prefer to be warned of potential issues rather than running a dangerous applet (reading clipboard, sending content to some random site...).