Editor's Daily Blog

Virtual Insanity

Is Java ME's development cost zero or infinity?

There's an interesting and rather contentious difference of opinions on Danny Coward's blog this morning. In Numbers that Count, he looks at how many mobile phones are out there (2.7 million phones are sold every day, according to research cited by the SDN article Java ME Technology: Everything a Developer Needs for the Mobile Market), which he contrasts with a much smaller number:

Zero is the cost of entry to develop an application for Java ME, which is deployed today on many of that staggering number of devices (for a complete list, see here). No complicated agreements to get the Java ME SDK, the development of the platform is out in the open, so everyone sees it unfold at the same time. Better the free visual development environment for Java ME has been pretty great for some time now.

This assertion, while compelling, drew a pretty sharp rebuke from reader Sam Halliday:

Your "zero" does not count the cost of paying ridiculous amounts of money to the Unified Testing Initiative if you want to obtain a certificate to access most of the useful APIs. Plus manual purchase of certificates for the older handsets that don't support UTI. More complaining http://javablog.co.uk/2007/08/09/how-midlet-signing-is-killing-j2me/

So what is the cost of developing an ME application for the real world? Zero, because you don't have to pay for tools? Infinity, because you have to pay to get tested and signed for a potentially unlimited number of devices and carriers (some of which don't accept third party apps at all)? Something in between?

The latest JavaOne Community Corner Podcast is j1-2k8-mtH08: Underworld - the Java EE 5 Backend For Wonderland by Adam Bien. Wonderland is an interesting 3D collaboration application. It uses the darkstar server as backend. Project http://underworld.dev.java.net goal is porting the Wonderland's communication and persistence layer to Glassfish v2 (later v3) to leverage its non-functional capabilities like monitoring, management, deployment and scaleability. In this shorttalk, especially the architecture and design, as well as, challenges, hacks, and workarounds will be discussed.

In Java Today, Patrick Curran has published an article on Java, Standards, and Free Software in Europe, with a particular focus on European participation in prominent new JSRs. "In the May issue of JDJ, I wrote about Java and free software in Brazil. This month, after some recent visits to Europe (to Antwerp for JavaPolis late last year, to London for the QCon conference in March, and to Paris for a JCP Executive Committee meeting in May), it seems logical to follow up with an article about Java in Europe."

Alexis Moussine-Pouchkine has seen a common problem pop up over and over again in the forums: I'm moving from the (NetBeans) GlassFish development server to a production server and my application won't run! Help! "NetBeans auto-magically creates all the resources required in the GlassFish runtime (JNDI resources, connexion pools, and other configuration), so directly deploying an application (.war, .ear artifacts) in a newly-installed GlassFish instance will most likely fail because the resources the application replies on are not present. To fix this you have several options..."

Over on Javalobby, Chui Tey has posted some interesting Questions for Sun over JavaFX. He wants to know will using the scenegraph API conceivably consume less resources than a comparable app developed using DHTML, why interpreted JavaFX script is slow, when interpreted javascript is fast enough for client-side work, and whether Sun be shipping a JavaFX Script interpreter.

The latest java.net Poll asks "How have job prospects in your field changed in the last year?" Cast your vote on the front page, then visit the results page for current tallies and discussion.

In today's Forums, jezzbel hopes to create a LWUIT TextArea. "I need to create a bit advanced text editor where I can highlight, underline, make bold words. How could I achive that in the LWUIT framework? I need to have a menu where one can choose some options for text changing and in one text area one can have a text differently formated.I've noticed that I can add any commands to the text area and can't to anything with the text."

sundararajana discusses a possible BD-J markup in Re: Show file alternative syntax. "Actually, GRIN XML syntax support is being developed. We expect to check-in code changes for this in near future. To set proper expectations, let me add this - XML schema being considered is *not* a subset of some standard format like HTML, SMIL etc. It'd would be more or less XML encoding of GRIN text format with some changes. Once we get there, we may use XSLT to convert various subsets of popular media XML formats to GRIN XML syntax."

In today's Weblogs, Varun Nischal wonders about Implementing Dynamic Binding while Coding? "An Interesting discussion took place during NetCAT 6.5 Second/Third Week...Topic was whether in NetBeans, the user can navigate to the implemented method (of an interface), analogous to the concept of dynamic binding?"

If all you need is just one more RSS feed, Jan Haderka writes, "Magnolia brings you new module that allows to combine multiple feeds into one and provides you with simple and manageable way to use content of the feeds to enrich your website."

Manfred Riem blogs about Joining FishCAT "I am one of the people selected for FishCAT, so let me know what some of the bugs are you think are irritating."

Current and upcoming Java Events :

• August 18-22 - J2EE Training Philippines
• August 22-24 - Central Florida Software Symposium 2008
• August 25-28 - Dev Week India 2008
• August 25-28 - Java Power Tools Bootcamp in Wellington
• September 12-14 - New England Software Symposium 2008: Fall Edition
• September 15-19 - Java Training Philippines
• September 19-21 - Pacific Northwest Software Symposium 2008: Fall Edition
• September 23-26 - Java Power Tools Bootcamp in Melbourne
• September 29-October 1 - The Ajax Experience
• September 29-October 2 - Java Power Tools Bootcamp in Sydney
• September 29-October 3 - JavaEE Training Philippines

Registered users can submit event listings for the java.net Events Page using our events submission form. All submissions go through an editorial review before being posted to the site.

Archives and Subscriptions: This blog is delivered weekdays as the Java Today RSS feed. Also, once this page is no longer featured as the front page of java.net it will be archived along with other past issues in the java.net Archive.

Comments

• To answer your question of Zero or Infinite costs on JavaME, the answer is Infinite, unless you have something that the operators feel can bring them in money, then they might lower the cost or even be slightly helpful. :)

  Having worked on JavaME for over 9 years (yup, back when kvm, and Spotlets where just starting in 1999) and working with Ericsson, Handmark, Disney, etc.. I have seen how the more money you can throw at a problem, the less money you need. It's totally stupid.

  The independent developers have been forsaken because we are seen as the enemy, and trying to corrupt the mobile operators environment. While I am sure there are those BlackHats that are jerks, there are more of us honest developers just trying to make cool things for everyone to use and make life easier and maybe fun.

  Of course there are so-called "developer" that just have no business in mobile, or for that matter developing code period. These clowns make awful code that brings down the network, or devices, rendering the device useless. As a protection mechanism operators have to treat us all like idiots, instead of just the idiots. So if they make the process complex, and costly that will deter the idiots. Nope, by definition they are too silly to realize that these challenges are to deter them, and those of us that know what we are doing are too annoyed to bother dealing with the operators silliness.

  On the other hand folks like Apple, and Rim get it, and provide everything needed for a one time fee of $99. Wow, that is perfect, that is reasonable, and I have am totally on board with these folks. I can only hope Google's Android follows this path, and not the Certs out the wazoo path that JavaME has fallen into. I would hope that JavaME could get out of this pit of doom, but with the JavaCertified program that seems to only promote this lunacy.

  These opinions are mine, and do not represent the views or opinions of anyone that I have, do or will work for. :)

  -Shawn

  Posted by: sfitzjava on August 22, 2008 at 07:01 PM

• And the main cost factor in most any development effort that's not taken into account here is manpower (and the hardware those people need, from desks, telephones, coffee machines, and toilets to computers and phones to test the product on).
  Those are very real cost items even for a one man shop. If your product doesn't recover the effort you put into it you're not going to be able to pay the bills after all. If you spend 2 months writing that thing you need to recover at least the money you'd have earned doing a paid job or you're running a loss.
  

  Posted by: jwenting on August 25, 2008 at 10:09 AM

This work is licensed under a Creative Commons License.