I am a Java developer, designer, public speaker and a JAS 39 Gripen instructor fighter pilot. A somewhat unusual combination I guess, but I like challenges. I have created miglayout.com, migcalendar.com and Wing, a flight planning system used in several Air Forces. I find end user usability to be the most important part of a system and have therefore specialized in creating such applications. Mikael has posted 10 posts at DZone. View Full User Profile

Why the Java Store has Gotten a Luke Warm Reception

  • submit to reddit

Jonathan Schwartz recently blogged about it and it didn't take, not too much any way. Why?

I am very excited about the Java App Store, really really excited actually. But I understand why many aren't. Sun hasn't got a very good track record when it comes to client related products, especially when usability and aesthetics are involved. Maybe that's about to change, but it hasn't changed yet, so people are hesitant, reluctant even, to believe the hype.

There seem to be a general consensus in the blogosphere; Sun's marketing team is nowhere near reality. True or not, it’s still seems to be the consensus. Maybe the reason isn’t so much bad public relations as it is failure to communicate. There simply is too much that have gone wrong the last couple of years for people to give them any slack.

We can talk about Java versioning confusion, default Java icons in the Windows gutter and JavaFX premature announcements, but it wouldn’t lead to much good. Nor would it lead to anything useful since of course there’s two sides of everything and if you’ve done it, you stick too it.

I think the main reason that developers (we are talking the early adopters here, we haven’t crossed the chasm yet) are hesitant is that since Sun has such a bad track record when it comes to spinning consumer related products the sensible course of action is to wait and see. Just the opposite of the iPhone App Store, for which Apple had much mojo to collect beforehand.

My personal feeling is that much of Sun’s bad rep is because of bad timing. Many projects have been announced too early and/or rushed out the door (JavaFX, Jini, Swing). Sun has been so eager to announce the next cool thing that the marketing-ish guys have gotten the take on the tech guys and basically tried to collect points before there were any to points given. The reason why, only Sun can answer, but my guess is that tech guys are too high in the ranks and the marketing guys can’t speak tech. Failure to comunucate, again.

Take JavaFX. It was announced almost two years ago and it’s still not usable. Sure, you can use it for Java2D-like demos, but there’re no components. The language itself was changed after demos were posted and books were written. A lot of things were missing from the 1.0 release. The upcoming 1.5 release should've been 1.0, and it should have been announced six month ago.

So I hope that Sun has learned its lesson(s) and will actually announce something that is actually ready this time. They can’t announce another half made thing with a big v0.1 sticker on it and having the marketing guys climbing allover each other trying to spin it as a done deal for developers. A simple demography/psychology of their first tier audience gives that they are dealing with highly intelligent and impatient people that in hours will test and criticize any flaws in the product, not giving it any chance unless it is near perfect. Not because the ealy adopters need that, but because there's no credit left on the client account (anyone that has read the excellent "Crossing the Chasm" knows that early adopters are actually quite lenient when it comes to usability and bugs, as long as they can fiddle with the product).

Only a flawless product with as good as Apple, or preferably better, usability will do it for Sun this time. There simply are no more leeway for Sun, they have to deliver, or the last credibility would be gone, at least when it comes to consumer related products.

I which I were more positive about the outcome of this one, I really do. Especially since it would help our next product a lot (secret as of yet). But, recent comments from Sun developers, that they are basically hurrying this one out the doors in time for JavaOne, makes me not hope for a new beginning. Doing development like that is bad at best. Conference driven development, or any time constraint driven development for that matter, is as bad for the long term as quarterly optimized financial reports. It might work for a couple of quarters but will fail miserably in a few years.

So why am I writing this? To try to reach the Sun decision makers and make them understand that good things can’t be rushed, neither can they be prematurely debited for goodwill unless there is a lot of goodwill to collect (there aren't). The only thing that will impress in the long run is a bug free, ready, and really really user friendly product. Apple has been doing this for a couple of years now and strangely enough no one has even been able to copy it.

Sun, listen to the people, but don’t put them all in the driver’s seat. Harness their knowledge but don’t design by committee. It’s a thin line, but it is possible to walk it. You start by holding your head high and looking firmly ahead, without fear of falling and with the mind set to do great things, not leaping ahead.

Mikael Grev

p.s. I think Java needs a better layout manager in the JDK. :)

Published at DZone with permission of its author, Mikael Grev.

(Note: Opinions expressed in this article and its replies are the opinions of their respective authors and not those of DZone, Inc.)


Thierry Milard replied on Sun, 2009/05/24 - 6:22pm

I agree 100% with this post.

For Swing : If only on there 1.1 version that just could have change the Layout () and make a better thing than manual programm Listener for any gui event. .... and of course make theSwing gui closer to Web things.

I still do not get why Sun in 2004 did not produce a "Swing 2.0" with many big improvements.

They just lost their momentum here on the web community. With javaFx October 2008 Release, Sun trie to catchning up ... but with a version 0.8 : a miracle if it had worked out well.

And I won't be too mean by .... still noticing that the java Player on a standard 2 year windows PC takes a bulky 7 or 13 seconds to be ready to display an standard Web-applet.

Sun say it's better now with update 13. My god it is better for sure. Othewyse java would be dead by now !

May I recall to any Sun java developper that the "horrible" Flash Player takes less than 2 seconds to load. How can we java developper can conpete with a 7-13 seconds start up jre. Ooops I forgot Sun marketers say ... it will be better with java7....

Sun looks sometimes like an old Russian Communist : 5 year planification system.... 

Yes it does work in a closed door dictature, not in a free web world. By-by I'm a Flash developer.

Finnally I will finish with 2 big wich for 2009 
- Sun Marketer please find a startup job and free sun from your coming mistakes and the bulky sun payroll.

- Sun java core developers , can't you for god sake produce QUICKLY a Windows jre launcher that is cloth to Flash launcher ? Shit are Adobe the only ones to have "fine tuning" developers ? Come on make an effort for the sake of the future of the java platform !


Otengi Miloskov replied on Sun, 2009/05/24 - 9:47pm

Yeah, Sun stop with the huge marketing, sometimes marketting does not work. Sun be pragmatic and creative dont copy, We dont want a C#, we want a real Java with own great things and improvments. We need Swing2 with modern layouts and optional GUI declarative as xul. We want innovation with Open Source or Propetary but no more bullshit.


PS. Kill JavaFX its just a piss dream, get to work on something more pragmatic and that just works as "Swing2".

Christophe Hanon replied on Mon, 2009/05/25 - 1:45am

Ufortunately,I have to agree. Sun has a long track record of late half delivering higly expected promised technologies. I think their marketing decided years ago not to listen to customers and developpers and prefers announcing products they would like instead of those developped by their engineers.

 I really like Java but I profoundly regret that this platform is unable to deliver boxed/publicly available polished products which is basically what the so called Java stored shoul be aimed at. So this will be the most powerful empty store on the Internet. 

Casper Bang replied on Mon, 2009/05/25 - 1:51am

Perhaps there's a major missing link in communication between Sun engineers and marketing folks. The comments above this one are spot on, Sun always sucked on marketing and that last finishing touch. I'd like to be proved wrong by say JavaFX but it looks like the majority of people take the same defensive stand as I, which could also explain the Java store reactions. I don't really recall having seen real innovation from Sun in a really long time, just an occational bump to try to match certain competitive producs (Visual Studio, C#, Flash etc...).

Osvaldo Doederlein replied on Mon, 2009/05/25 - 8:59am

As a JavaFX enthusiast, I must admit to also agree with Mike. And the most irritating problem is that client Java problems are not all a matter of resources or technical chanllenges. I can understand a technical challenge like the difficulty of creating a runtime that offers fast loading time but with enormous APIs and portable code. I can understand that Sun is tight on resources and they can't fix all bugs as fast as I want. But I cannot understand boneheaded issues like the Java system tray icon active by default - that crap is intrusive to end users, it's only useful to developers, Flash doesn't do that, Silverlight doesn't do that, Java shouldn't do that either. Fixing this enormous user experience bug doesn't even require JRE changes, it's a simple installer patch so it won't enable the "Place Java icon in system tray" setting by default. It's almost like Sun doesn't want to succeed.

Well, having said that, Sun's behavior has certainly improved a lot in recent times. We may complain of a rushed JavaFX 1.0 release, but this is very minor compared to past failures. And it's even defensible on the grounds of time-to-market and competitors being just as immature in their initial releases. Remember Silverlight 1.0? No component package either, only JS supported for programming... basically a "developer release" just like JavaFX 1.0. It was basically a flop with the real world, I never saw a Silverlight 1.0 app outside microsoft.com (well except cases 'bought' by MS like NBCOlympics.com); adoption only started to happen with v2.0. The difference of course is that Microsoft, different from Sun, has a track record of success with desktop products, and if an initial release is weak they have the resources to keep pushing it fast and hard until it succeeds. Plus they released v1.0 1,5 years before JavaFX 1.0, so timing is a real problem for JavaFX, in the end I'm forced to agree with Sun's decision to release last October because there was no other choice. I for one, didn't take the time to even look superficially at JavaFX before v1.0. Not everybody has time for pre-1.0 betas. The JavaFX community and uptake, however small, are much bigger today that they would be if Sun decided to wait and only release v1.0 now. Because it went FCS, I became an early adopter and promoter, having written three full-length articles on JavaFX for the Brazilian Java Magazine. Because it went FCS, I reported a few bugs to the JavaFX JIRA, so the upcoming Marina release may be slightly better than it would otherwise, with a much smaller group of people outside Sun testing it.

Richard Osbaldeston replied on Mon, 2009/05/25 - 9:38am

I think it may be a little early in the evolution of JavaFX, but I don't think Sun have much choice at this stage, time is no longer on their side. Personally It's nice to know in advance where they're going with JavaFX for once. At least by building this now its a positive gesture and vote of confidence in JavaFX. It has the added advantge that any buyer of Sun that dosnet necessarly share the same JFX vision (can you think of any?) will now be faced with the unnenviable task of shutting this site down and facing a possible public blacklash. It's far easier to kill a half-finished API that nobodies seriously using for much more than fancy demos - time for JavaFX to step up and be seen as a polished product.

..even if it isnt quite ready. I think the issues come from the immaturity we developers can't really ignore. The lack of available handsets with JavaFX preinstalled.. the heterogeneous nature of the market their entering offers serious headaches for the developer. Unlike say the iPhone or Android phones theres no standard handset, what size is the screen, how much memory, cpu power, touch screen or keypad?, accelerometer?, camera? what features and interfaces are available and how they can be accessed in a graceful degradating manner both cleanly and simply from JavaFX seems unanswered (except maybe not - use J2ME for hardware features? only for now?). Testing your apps on all capable handsets is probally just not possible for the small time appstore/vector developer, how is that really going to work?

So will Vector somehow address some of these 'not all handsets are equal' issues or not much different from JFXStudio with a payment gateway?


GeekyCoder coder replied on Mon, 2009/05/25 - 10:54am

I can hope that Sun succeed this time with Project Vector as I feel that Sun still have much to improve on execution in software system. Though I like to say that JavaFX improves on Sun's ability to execute, it leaves me wanting as there is a lack of update and concrete roadmap on JavaFX and JWebPane. Microsoft and Adobe engineers look more willing to share and blog about their technologies which is very engaging and encouraging. It is very surprising that there is not much public discussion of JavaFX 1.5 and JWebPane at this point. It is still very closed-door development. Hope too that Oracle's acquisition give Sun more resources and support to get thing done.

Peter Karussell replied on Mon, 2009/05/25 - 2:57pm in response to: Casper Bang

> Perhaps there's a major missing link in communication between Sun engineers and marketing folks


Thats the point I think! Great tools like Java, NetBeans, OpenOffice ...!

But bad or even lazy marketing?

(One example are the latest version numbers of Java: 1.6.0 update 13 ... has to be translated for us to get memorizable: Java6 update13, uhm ... and it gets even more confusing if a JavaFX 1.0 or 1.5 is bundled ;-))

Ivan Lazarte replied on Mon, 2009/05/25 - 10:37pm

I think this is what everyone is expecting for the app store (which isn't too far from my JavaFX experience)

1. Click search
2. Choose App
3. Download App? Click Yes/No
4. App Store Certificate not Verified, Continue Anyway Yes/No
5. This App Wants Secure Access! Continue Yes/No
6. Accept License? Yes/No/Cancel
7. Confirm App Purchase/ Visit App Store/Abort/Continue/Retry
8. Download Commenced -
9. Grey Screen of Death!
(user quits in disgust and never loads it again)

Thierry Milard replied on Tue, 2009/05/26 - 2:39am

Hello java javaFx develpers.

One final thought about future superbe idea .... trying to copy the Apple Store.

Check this article on tech Crunch. It says ..... well it says it is finally not such an enormous sucess..




PS: One more reason to step by step improve the user experience of java (hide and java tray icon, still long startup, ect... ). Before doing bad Apple copycat. Sun marketers, just listen to us java dev and java users.


Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.