Sean Sheedy's Blog

November 2007 Archives

Android's "do not fragment" clause...

The OHA members have signed an anti-fragmentation clause. "They've basically agreed not to fragment nor do things that would result in different versions of the platform", says Rich Miner in the PC World interview. And the JCP has that too - called MSA. But as Java ME developers know very well, there are many facets to fragmentation extending beyond the platform and well into the surrounding ecosystem.

A strict "do not fragment" clause is impossible to implement without putting every aspect of the platform and ecosystem through a committee, or relinquishing creativity and possibly IP to a central authority. This is because the other face of fragmentation is called innovation and differentiation.

If many vendors are playing similar roles in the ecosystem then I don't care if there is an anti-fragmentation clause in OHA, you are going to see some fragmentation. Some of this is beneficial: you need to allow new features to appear in devices and new services to be offered. "Fragmentation management" is needed so you don't eliminate the innovative process when eliminating redundancies that appear as a result of that process.

What this clause is, however, is a huge, essential step in the right direction. It establishes a positive spirit of cooperation. It reminds competing parties why they need to be at the table. It justifies discussions on processes throughout the ecosystem, such as application certification, signing processes, distribution, rewarding the creation of new features and services, open issue tracking and sharing, collaboration on documentation and developer support, interoperability testing and "bake offs", etc.

One of my objectives in my candidacy for one of the open Java ME EC seats is to drive the EC to expand its scope to address the areas of fragmentation that go beyond the platform. Having everyone sit down and sign a "do not fragment" statement, no matter how symbolic, is perhaps a great way to set the tone for the collaboration required to do what everyone already wants: turn the symbolism in concrete practices.

On the Android announcement - a bullet list

• Android has existed for probably about 4 years, created by people who have real handset experience, so a delivery timeframe of 2H08 is not unreasonable. Google acquired Android back in August 2005 and it had already been running for 22 months by then.
• This old (March 2007) article tells us perhaps more technically than most of the news about the phone that came out in the past day.
• Given that article's date, and its lack of mention of a lot of other players, and the growing leaks of Gphone news in the press prior to the announcement, the other players in the announcement were probably brought on board shortly after this and probably have been working on this coming on 4-6 months...
• Sun was noticeably absent but Jonathan Schwartz blogged his full support yesterday and promised full support in NetBeans. This is really great news for Java developers.
• If you peruse the press releases from the companies in Open Handset Alliance you will find a little more tech info which shows that it's a Java play as far as developers are concerned and quite possibly that Linux will not be exposed to developers (or at least not to anyone but close partners.) Could be wrong about that last part, and this is what we're all dying to see in the Nov 12 release of the SDK.
• This is underscored by the presence of Esmertec and Aplix who have created the Java ME platforms for a lot of mobile phones.
• If you look closely at all the companies in the OHA list, and do a little dreaming, you can piece together a vision for a handset with some pretty intense and useful capabilities that will be yet another nail in the coffin of Frankenphones we have to live with today. (I'll provide my definition of Frankenphones later, but the short answer is that it is what you get when product managers are allowed to make major functional and design decisions for handsets.)
• The presence of Nuance as part of the voice solution suggests that this may provide a stepping stone towards giving the VOIP industry the vision it needs to get out of its saturated but under-realized state. Personal IVRs may be right around the corner. Home phones where you have one line that rings through your whole house and you can't make a call while someone else is on the phone will be as passé and unacceptable as party lines would be (though you might see stuff like Selective Ring, since you may want all handsets in the house to ring, but need to know who in the house it is for.) OK, maybe this is stretch here, but VOIP is an industry ready for a focusing as much as applications and technologies we are discussing here have focused mobile data after its early fits in starts in the late 90s and early 2000s.
• Expect to see telephony integration in the PC taken to the next step as a fallout of this.
• That they cobbled together 34 companies for the OHA announcement so secretly is amazing. The NDA must have some pretty big teeth. Even people who should have been involved in this did not know about its full extent until yesterday.
• Apple has proven that a sufficiently sized outsider is a gorilla in this space and will ravage the china shop if ignored.
• You cannot ignore Google; they are a bigger gorilla than Apple both in market cap (about 40% greater) and arguably, penetration of their respective industries, depending on what industries you consider.
• Still, there must be a pretty darn good business case for the parties involved in Google's pitch deck. I would love to see an unedited version of that.
• Not to mention that those on the OHA missed the iPhone boat, so there may be a bit of fear of not being on board for this "second chance". Operators are saying that they recognize the days of closed phones are nearing the end (this is mentioned a ways down in the article by one of those interviewed.)
• How open is OHA? At the end of the FAQ on the alliance on www.openhandsetalliance.org is an email for questions about joining OHA. Of course, I sent them an email asking what the process is for joining. (I already know the general answer, but why not get an official one?)
• Interesting that Qualcomm is on the list, obviously because of chipset implications, but also because this directly competes with their relatively closed BREW business. What does this spell for BREW?
• Also interesting, but not surprising, that AT&T is not joining, but they already have their own pet gorilla.
• Nokia is also not on the list which is also not surprising as it has its own plans for a soup-to-nuts ecosystem around its handsets as evidenced by its recent location acquisitions. At last month's CTIA conference, NAVTEQ held a full day session which emphasized the many yet-unrealized possibilities surrounding location, including taking location to the pedestrian level - which is where Google will play with Android. In the Wall Street Journal, Nokia said they already have an open system in their high end phones.
• PC Magazine reports that OHA members are not dropping other OSs for Android.
• Additionally, there are reports that signing on to the alliance does not mean that a company will create a "google phone". This makes sense, as Google said that it did not want to build a phone - it would probably not want to deal with phones any more than it wants to deal with a Google PC - but it does need an environment that meets its needs.
• So given this, and with OHA's website not really being more than a link or two deep, is OHA more than a bunch of press releases? Today, probably not. Will they eventually standardize this work, or push the standards work out to places like the JCP, W3C, OMA, IETF, 3GPP and CDG?
• I also have to comment that the video on the OHA website plays like the tours of Pixar in the DVD extras of films like Toy Story. Don't all the robots look just totally cool?
• Finally, a reference to the JCP. Opening a dialog with OHA should be a priority - if Android has meat (which it does by all indications) then working with them is going to be imperative.

Note that this is all written before seeing the Android SDK. However, there are a lot of smart people behind this, and with people like Jonathan Schwartz supporting it (who probably has insight into what is under the covers - I will leave the commentary on this for others), there will probably be some very interesting stuff for Java developers.

Sean

Voting is Good but Making Your Voice Heard is Better

After spending several years in the Java ME standards circuit for a major operator, I'm now on my own as a consultant, and running for one of the open Executive Committee seats as an Individual. I'd really like to know what burning topics Java ME developers would like to see the winning candidates bring to the EC.

Java ME has been fabulously successful with well over 2 Billion devices shipped. A lot is being done right, but it does not mean that the platform is without issues. Some are longstanding, and current EC members tell me that the desire for correcting them is universal. However, the nature of corporate life makes initiating these discussions difficult. How do you tell your boss that you need travel budget to discuss fragmentation with your biggest competitors? "You're going to discuss what with who?"

An Individual can facilitate discussions on these issues. They have the advantage of not having corporate encumbrances. They are freer to start conversations among competitors to bring those to the table who need to be there for change to happen. They can highlight other Members' public actions that demonstrate their commitment to the Community Process, without being accused of "aiding the enemy." They are freer to question private actions within the EC that undermine those public actions, without burning vendor/customer relationships.

In my position statement, I listed many potential objectives, but with only 15 people in the EC, you have to focus in order to get anything done. So I am interested in hearing what people consider to be the first issue for a new EC member to raise after they are elected to the Executive Committee.

In the JCP, a Community with over 700 members, about 2/3 are Individual Members and of the remaining 1/3, a significant number are smaller companies. However, between the EC and MSA - the two bodies that drive ME - all but one member is a large corporation. This is not to say that these companies do not represent developers. From personal experience I know they do, but developer interests are represented through a corporate lens.

What do individual developers want to accomplish when that lens is removed?

SE/EE Executive Committee Candidates

Being focused on Java ME, I found that after perusing the candidate list for the SE/EE EC on jcpelection2007.org, I still was not sure who to vote for. So I put this together to be able to compare the candidates and understand their motivations for running for an EC seat.

I'm posting this so that maybe a few people or the candidates themselves can help fill this out, or comment on who they are voting for and why.

All candidates should be commended for selecting EC participation as their way of contributing to the Java community and helping to guide Java's future.

I paraphrased what I found in their candidate statements and things they linked to, so there are bound to be errors. Please let me know if you find any and I will correct them.

For each, a couple of questions came to mind and are listed. For example, I was hoping to see more information in the candidates statements about why they were running for a seat - only two of the seven candidates provided this information.

Disclaimer: I am a candidate for a seat on the Java ME EC.

Voting ends 11/12.

Continue Reading...

This work is licensed under a Creative Commons License.