Ian is the Director of Marketing for the Eclipse Foundation and live in Ottawa, Canada. Ian is a DZone MVB and is not an employee of DZone and has posted 165 posts at DZone. You can read more from them at their website. View Full User Profile

What Sun Should Do

11.26.2008
| 6665 views |
  • submit to reddit

Tim Bray has written an insightful piece describing his perscription for a healthier Sun, at the end he suggested others might want to do the same, so here is my version.

Product Strategy

Sun’s problem is not that they don’t have good products, it is they have too many products for them to sustain.  As Tim mentioned they need to focus on their strengths and that would Glassfish and MySql.  MySql has a great reputation and community, they need to protect it.  In the last two years I have seen significant momentum in Glassfish, in particular they are staking out leadership in OSGi.  Keep it going.

Like Tim suggested, Sun needs to give up on JavaFX.  Someday it might be great technology but it doesn’t matter, Sun just doesn’t have the reputation or the reach into the graphic design houses that produce the cool web sites to feature this type of technology.  Ajax, Flash and Silverlight will battle it out for the hearts and minds of RIA style applications.

I am not going to say anything about NetBeans, since I am obviously biased but let me say there are many ways to attract developers.

Sales Strategy

Sun’s biggest problems are on the sales and marketing side.  How is it they are still so overly reliant on sales from the financial services industry.   Sun got wacked in the dot.com crash due to reliance in revenue from the banks and they are getting wacked again.   They need to diversify their sales into other industries.  This means hiring different sales people that know how to do this.

Open Source Strategy

Sun needs to re-consider its open source strategy.   Claiming that all Sun software must be open source provides no opportunity for license revenue.  They need to adopt what the Matt Aslett at the 451 Group calls an ‘open-core’ strategy.   Basically get wide spread adoption of the base platform, like MySql or Glassfish, and then sell up with specialize tools.   It seemed MySql was going down this path before the acquisition; it needs to be reconsidered.

Java & JPC Strategy

Lets hope the Sun leadership team listens to at least one of Tim’s recommendations:

As for Sun’s role as Steward of Java, and in particular the Java Community Process, let it go already.

Please just do it, the JCP is busted, Sun needs to let it go; just do it already.

So what do you think Sun should do?

From http://ianskerrett.wordpress.com/

Published at DZone with permission of Ian Skerrett, author and DZone MVB.

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

Comments

Jacek Furmankiewicz replied on Wed, 2008/11/26 - 9:06am

I think it would be a great thing. Something equivalent to the Linux Foundation (funded by the bigweights like Sun, Oracle, BEA, IBM, etc.), with some benevolent dictator on top (James Gosling?) having the final say.

 

Thierry Milard replied on Wed, 2008/11/26 - 7:50pm

Quote : "Like Tim suggested, Sun needs to give up on JavaFX.  Someday it might be great technology but it doesn’t matter, Sun just doesn’t have the reputation or the reach into the graphic design houses that produce the cool web sites to feature this type of technology.  Ajax, Flash and Silverlight will battle it out for the hearts and minds of RIA style applications."

===> I go the opposite on this. I  do not think javaFx is bad. I tested it. From my pov, it is in fact good

I think it is ashame swing API was not more simpler to program and better looking (my flash Web-developper friend calls this Swing API : "An API made for eye impair engenneers" not false: difficult to program, ashame to demo).

But with javaFx (and update 1.6_11) Sun is (at last !) I think back again giving us small not pure web companies one opportunity to easier program  good looking web services. Web services/ RIA is a important need, Sun is right to go with it. I think it will be at least a minimum successfull because many java Swing developers desesperatly need this to simply 'upgrade to the Web' their application.

Just look at the many web service going out every week : there are built all in Flash. 

Now with javaFx we can make java/javaFx Applets as good looking.  At last !

 

Otengi Miloskov replied on Wed, 2008/11/26 - 8:11pm

I agree Sun should let already go Java and JCP, We need Java Foundation.

JavaFX does not bring nothing new to the table is another scripting language to learn and the API is not mature enough as Swing or Flex, It will take 10 years to JavaFX to be in the position of Swing or Flex, I think JavaFX was a bad move again. Also Swing is one of the most popular GUI API around more than WinForms, We just need a xml declarative way as xul to design Swing GUI apps that was the only thing that Sun needed to do before anyway JRE6_10 it was great step forward.

But who the hell cares about desktop GUI anymore, The Net and The Web is the future, The Web is the way to go.

Martin Brehovsky replied on Wed, 2008/11/26 - 8:28pm

Ian, I believe you misinterpreted Tims though about JavaFX. Tim says "For actual business apps, the kind that our servers spend most of their time running, the war for the desktop is over and the Web Browser won. I just totally don’t believe that any combination of Flash and Silverlight and JavaFX is going to win it back. AJAX is increasingly central and we need to make sure that our Web Suite and its tools support it well. But as for GUIs and the client side, let it go already." 

But you say: "Sun needs to give up on JavaFX.  Someday it might be great technology but it doesn’t matter, Sun just doesn’t have the reputation or the reach into the graphic design houses that produce the cool web sites to feature this type of technology.  Ajax, Flash and Silverlight will battle it out for the hearts and minds of RIA style applications." 

There is a huge difference between those two. Tim believes AJAX is the technology of the future and all proprietary client solutions including Flash and Silverlight are not perspective. OTOH you simply don't believe in JavaFX because according to you "Sun just doesn’t have the reputation or the reach into the graphic design houses that produce the cool web sites to feature this type of technology".

Still I have to note the huge difference between AJAX and JavaFX is the fact AJAX is currently usable only on the connected desktop. While this is perhaps most of the PCs connected to Internet today, it is definitely not most of the devices which might be connected to Internet. What about all those mobile phones, TVs, cable set top boxes and other devices in the future. Is AJAX scalable enough to deliver content to those devices? I don't think so and this is exactly where JavaFX (and other technologies) has potential of delivering a viable platform for next generation of apps. 

Also I still have to see an AJAX app which is able to deliver rich content which is comparable with the experience you get with the proprietary solutions. While there are nice pieces of the technology (e.g. dojo.gfx is perhaps my favorite these days), they still lack the performance, the feature set and the usability aspects of the proprietary technologies and I'm afraid it is still going to take a substantial effort to get to the levels where for example Flash was 4 years ago, especially when it come to rich media such as video.

Carl Dea replied on Thu, 2008/11/27 - 3:15am

Ian,

 Wow, I think you might have ruffled some feathers. I'm not rattled though (hehe).

My thoughts on the Software or the Language area of Sun:

I do not agree with some of the statements relating to JavaFX.

---> "Like Tim suggested, Sun needs to give up on JavaFX."

It seems there is a group of Java programmers that are cynical and are FUD makers. Which know how to complain about Java's past, and still can't get over it. Sun will and make mistakes, but JavaFX is not a mistake. Dropping JavaFX would'nt be good stewarts of Java, it would be suicidal.  JavaFX is the logical step for a mature language like Java to leverage on devices, kiosks, e-book readers, BlueRay, browsers, games, etc. The future is RIA based technologies believe it or not.  I didn't drink the Kool-Aid like you might think. But, imo Sun is doing the right thing with JavaFX because it will expose more opportunities for the company than with Java (as a language) will ever have (yeah, i said it!). However, Sun has to ... (not try to) do the following with JavaFX as it becomes mature.

  • Not have old myths suddenly appear in JavaFX. (mitigate customer/developer expectations)
  1. Swing is slow
  2. Deployment is scary
  3. Start-up speed
  4. Grey rect.
  5. Applets vs. Flash (ubiquitousness)
  6. Layout is hard
  7. EDT, SwingWorker
  • Believe in themselves that they should be where Flash is now, and take back the browser. Make Applets Not Stink (MANS). Make Applets Fun Again (MAFA)
  • Help the guys with Inkscape (http://www.inkscape.org) and make sure the next build has export *.fxd files ability. But, still play nice with Adobe (Chet works there) or partner.
  • Make sure NetBeans 6.5 pluggin for JavaFX 1.0 is available.
  • Create a simple tutorial to show us old Swing developers to see the light. Demonstrating how to create a simple form based application with JavaFX (like Swing Set). There will already be tons of snazzy demos that aren't useful (which is what that group will complain about). This tutorial should show us a workflow from SVG to JavaFX Forms to Java-code and back.
  • Market this thing RIGHT. Unify the JavaFX websites, contests, YouTube it, etc.
  • Continue to support Swing. (make it a separate jar? dunno)
  • 3D, sound, codecs, printing, etc.

A side note: Adobe is focused (photoshop, illustrator, premier, pdf, flash). Sun's Java division should do something similiar.

I'm sure JavaFX isn't a drain on the company books. I'll bet its the wierd acquisitions that companies make to make the books look good. Then when the fallout of the merger occurs it causes more harm then good. Some of the money drain was the past law suits and the economy I'm sure.

I don't agree with the Sun Web Suite either. Another weak attempt to tackle the Server development market. Sun was a little slow in the beginning when it was NetDynamics,NetScapeAppServer,iPlanet, Sun Studio. I think NetBeans is a very nice product, and Sun should make sure NetBeans can connect to IBM Web Sphere, Weblogic, JBoss, etc. This is a similiar strategy when Quatro took over Lotus123 an then MS Excel took over Quatro. Is there a JSR that makes a Create Project Once Program Anywhere(C-POPA) (Don't say Ant)? An XML Java IDE Project standard should be made so that you can export a project from Eclipse to NetBeans and back (CheckStyle rules). When you make developers' lives simple they will love you back.

 

I believe the Virtualization product has great potential but VMWare is still out on top. IMO Microsoft wins many server battles because they make an Admin's job easy. I've often heard it isn't easy to find Solaris Admins. Usability is key. Is GNOME or KDE harder to use than Windows2000?

 

My thoughts on the Hardware area of Sun:

I don't know how well Sun is doing in this area to really comment, but here goes:

What's up with all these storage companies and cloud computing? Is this a big con? Sun should sell computers and services that save companies money with top notch user-friendly tools/customer/consultant services.

Learn how IBM made a comback by using Linux on their hardware.

Learn how Apple made a comback with mac and iphone.

 

-Carl

 

 

Thierry Milard replied on Thu, 2008/11/27 - 3:48am in response to: Carl Dea

Very good reply Ian. I aggree 100% with your thoughts.

You say Microsoft in virtualisation is better because they understand more than sun that :

"Usability is key".

 

I also think this is very true for java : Just look at this;  last week I spent 2 days last week just publishing an Applet because ... well JNLP is still a thing to learn for example. Why don't Sun ease the pain of Applet deployement with a simple Web service where you uplaod the code, answer a few question click OK... and it is publish as an Applet.

 

I think Sun should be (why not) build 4 people team just with the only task to see how , in all the software and tools simple change like this one could ease the user/client with usability.

 

Ian Skerrett replied on Thu, 2008/11/27 - 5:11am in response to: Martin Brehovsky

[quote=mb118009]

Ian, I believe you misinterpreted Tims though about JavaFX.

[/quote]

 

Martin, you could possible be correct that I misinterpreted Tim's remarks.  If I have then I certainly apologize.  Regardless of Tim's comments, I stand by my statement that Sun should drop JavaFX.

This has nothing to do with an assessment of the technology.  It is based on three things: 1) Sun needs to focus, it has too many products to sustain, especially since they are reducing their workforce, and 2) Sun needs to generate revenue very quickly.  JavaFX is years away before they have a chance of generating significant revenue. 3) There are some sizeable competitors to JavaFX and Sun is late to the game, they are playing catch-up.  Sun needs to lead from its established strengths.

 

Osvaldo Doederlein replied on Thu, 2008/11/27 - 11:02am

JavaFX may not win over AJAX, but that's not its goal, and there's a boatload of Java AJAX frameworks for that. JavaFX's goal is to compete reasonably in the RIA wars - against Flash, Silverlight etc. It's possible that it becomes a third-place or runner, or a nich option (eg corporate apps) but this wouldn't make JavaFX a failure. It has at least some serious technical advantages, like much richer APIs and much better runtime performance. Flash is good for animations, simple games, and smart data visualization (stock quotes, e-books, etc.) but that's all, I never saw any "real app" written in Flash.

See for example, LogMeIn's Java viewer applet - few people use that viewer because they have superior native viewers for both MSIE and Firefox, but some people use other browsers (eg Google chrome) or need to use some internet cafe where users cannot install ActiveX/NSAPI plugins. So LogMeIn offers a Java viewer that has the same functionality. Now go and try implementing that in Flash, I bet you can't (or performance would be much worse) otherwise they wouldn't have picked Java (and that business decision was taken years before 6u10 or JavaFX). All is not lost for Java in the desktop and RIA.

And yes, my Java IDEs are perhaps the only rich-client Java GUIs that I use on a daily basis, but they do rock (both Eclipse and NetBeans) and I attribute much of their success in being superb IDEs to the Java platform... I don't see IDEs of similar quality and functionality written in native languages (I was a big fan of other IDEs in the past - Borland C++, VisualC++, Delphi - but at least the latest VisualStudio, which I still use very infrequently, is a loser compared to the Java IDEs).

Don't forget also Java ME's dominance on the mobile low-end, it pretty much owns the cellphones in the CLDC hardware range. For the high-end (expensive smartphones and PDAs), the CDC configuration didn't succeed, and that's exactly the target of Java FX Mobile so if this helps Sun to increase Java ME / CDC's share of this fatter piece of the market. Competition is hard (iPhone, Win Mobile, Android etc) but Java ME is already a success in the mobile market and when you're fighting a fight that's half-way won, you don't throw the towel, you go for KO. Even in the successful CLDC/MIDP subplatform Sun is pushing LWUIT to better support modern handsets. The strategy is all right IMO, now it's a matter of execution, let's hope Sun gets their act right. For one thing, Java FX Mobile has some important advantages over Android - not the least of them, the fact that Sun is a serious full-platform provider, they have an insane amount of experience and technology in everything from CPUs and operating systems up to networking, middleware, management etc. Dalvik is cool and all, but in the end of the day it's a purely-interpreted VM, I'll take the latest CLDC HotSpot any day. Even with Android's earlier release and hype, Sun only loses this to Google if they screw up badly or if they just can't compete with Google's bags of cash.

Carl Dea replied on Thu, 2008/11/27 - 1:14pm in response to: Osvaldo Doederlein

Osvaldo,

Yes, Flash isn't famous for "Real Apps", but people(consumers and developers) even know what Flash is. Because of all the little applications like: videos, webisodes, and small games all add up. JavaFX has the best of both worlds and I believe it will surprise people. I think most Java developers are server side developers and don't care much about the desktop world. I think Sun should pull out the big guns and promote Java better... talk to Microsoft IE..., Mozilla Firefox, Google, Apple to make sure the latest and greatest JRE is pushed (seamless). So, in the future we will have multiple RIA formats in the browser, and another war starts, so what? There will always be wars... It makes things better for the consumers anyways (choice and competition is a good thing). Leveraging legacy Java code is one of the greatest things about JavaFX (nice MVC and separation of concerns). I want UI people to focus on the front-end and seasoned Java folks to focus on Architecture - JPA, WS, etc. Can you imagine TuboTax written in JavaFX w/ Java?

Two Languages are better than one, and better than three.

It just came to me... I don't think most Java people understand the world of Adobe photoshop, Illustrator, and Flash. That world they live in fits JavaFX very nicely, which equates to two languages. Java and FX script. Probably a great compromise. We can't make each world conform, it'll never happen. It's like C and Borne Shell (like peas and carrots). I'd rather have two languages that make frontend and backend developers happy than Web Apps which has (xml,js,dhtml,css,taglib, jsp, spring API,xpath,dtd, etc.) Web Apps & AJAX are so overated, and in need of an overhaul. Web Service-ize the thing and use something like Grizzly-comet and use Java and (Inkscape save as *.svg and convert with project Nile) & JavaFX. This is the same strategy Adobe is doing with Flex, Air and *.sfw file formats. Flash in the browser is what is popular not RIA Adobe Air (except AIR runs on Flash). JNLP webstart is more popular than Air, so with JavaFX,Java, & JNLP, Java will surprise folks. Sun just has to get the bugs ironed out first. (Test, test, test, and test).

Great comments on JME... I didn't mention it at all. I'm just tired of stupid patents!

 

Many little visions help make a big vision, and I think companies mess up when they have one big vision that is somewhat unrealistic that can make a company spread way too thin. They need to be leaner and start thinking of creative business models for the change and keeping their talent.

 

Java doesn't hurt Sun.  Sun hurts Sun.

 

-Carl

Dimitris Menounos replied on Fri, 2008/11/28 - 8:55pm

I can't see Java catching up with Flash on the client anytime soon.

 Instead of competing with Adobe I wish Sun would partner with them and work together. Sun could help with VM technology and Adobe could make tight integration with JavaEE and MySQL servers, or adopt Netbeans? How about a Java to Flash bytecode compiler, like how Google did Java to Javascript?

Sun should be riding the waves not fighting against them. Waves like Linux, Flash, GWT, Android, or anything... a rising tide lifts all boats.

Osvaldo Doederlein replied on Sat, 2008/11/29 - 11:48am in response to: Dimitris Menounos

[quote=agnus]

I can't see Java catching up with Flash on the client anytime soon.

[/quote]

Neither I; Flash probably will always be the best tool for some apps. I see Java FX as a better solution for more complex apps, ones that Flash is not well-suited for. Perhaps the new Air platform would be address Flash's limitations here, I'm not up to speed with the latest stuff from Adobe but in this case, it's a new market niche that Adobe doesn't own yet so FX has a verey reasonable chance to compete.

[quote=agnus]

How about a Java to Flash bytecode compiler, like how Google did Java to Javascript?

[/quote]

The Eclipse Foundation is doing something in this avenue, the next version of their RAP (Rich Ajax Platform, a GWT lookalike) will have a Java-to-ActionScript converter, so you write Java/SWT graphics code and it generates an equivalent SWF on the fly. Pretty hot stuff if the implementation turns out good.

[quote=agnus]

Sun should be riding the waves not fighting against them. Waves like Linux, Flash, GWT, Android, or anything... a rising tide lifts all boats.

[/quote]

IMHO, in Android's case it's Google that is riding Sun's wave, since it's largely a copycat of Java / Java ME. And I think Java FX Mobile has potential to be a much stronger platform than Android. Sun is building on top of more mature parts: Savaje, JavaME/CDC, phoneMe, Linux - the single "fresh new" item is FX Script. (In contrast, Android was messing with core APIs weeks before v1.0.) Even in the parts that look similar, I put my money on Sun - any company with a few good engineers can customize and maintain a Linux-based kernel, but Sun certainly runs circles around Google in UNIX OS expertise. Not to mention Java of course (CLDC HotSpot can wipe the floor with Dalvik, which is a purely interpreted VM). I know that tech specifications are not a sufficient recipe for success, but we developers deserve to have this choice.

Comment viewing options

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