This post originated from an RSS feed registered with Java Buzz
by Fred Grott.
Original Post: Code, Design, And Innovation
Feed Title: ShareMe Technologies LLC-The Mobile Future
Feed URL: http://www.jroller.com/shareme/feed/entries/rss
Feed Description: A Weblog about Java programming and digital convergence on mobile devices in such areas as P2P and collaborative technology.
The PhoneGap demo of reading GPS via javascript and a demo of my own Xspot doing the same thing will not be up until Monday or Tuesday and since I now have Google's attention I might as well cover some important topics for them to view on Monday. First, why a Loopt clone?
Originally, I was interviewed for a development position with Loopt back in the first year of their founding when they were called Radiate and the founder was still conducting the interviews himself. At that time the founder of Loopt wanted an html and ajax interface in mobile which just was not possible at that time implementation wise the way he wanted to implement it.
Ever since than I have had a better idea of how to implement a Loopt clone as a more html ajax way of implementation. the geenral idea was to sue something like Zk or Mojax to target J2ME GPS handsets. But as you know that server side implementation often weighs down mobile start ups in costs and what you have to give up revenue sharing wise to get someone to invest in that server sdie infrastructure.
Several months back I got somewhat fed up with all the people and entities proposing to put the developer at 100% risk without even putting their own money in and thus while another person/entiy was attempting the revenue sharing abuse game wanting a Loopt clone I started native development of a Loopt clone. That of course the financial crises happened which served to cut off in immediate seed capital in the $300,000 ranges.
Thus what to do? So I went back to my notes from all those GillMor gang talks aobut social applications, clouds, an etc as I had been participating as an ustream chat participant in those talks since about April of this year. Thus, the new implementation was to use clouds like FriendFeed, Brightkite, Google and etc and tie it together with Gnip and thus no longer need to spend time developing a server infrastructure.
Why web based as in embedded Webkit via webview classes? The second higest expense in something as large as Loopt behind server infrastructure is that porting expenses in porting from J2me to Symbian to Windows Mobile and etc. I knew that I could not totally eliminate the native side however I knew that as I was reading about embedding webkit via webview classes that maybe someone may have started doing it on iPhone with a native library to read GPS and make those objects accessible via javascript.
That is when I found the PhoneGap project and noticed that they had started an Android code base. Since that would cover both the top and 2nd mobile platforms as far as sales growth and etc in 2009 it suddenly became imperative that I change the implementation from full native to this new web based approach. plus, the 3rd top platform namely Symbian S60 would enable me to port PhoenGap to Symbian at some point and thus give back to the PhoneGap project.
What about business a business model? Apple, Microsoft, Yahoo, Google, and etc have been in most cases making both indirect and direct bets that the business model for mobile is ad based revenue supporting mobile application based services. I believe some combination of sales through iTunes, Android Market, and etc coupled with ads may be the right business model mix in that having those revenue streams pay the operating costs of the sevice that the Xspot application provides. I would like to further develop the business model to a more full ad revenue stream set-up but that will depend upon the makeup or mix of liquidities of the seed capital that I acquire and how long I can hold down costs to conserve that seed capital.
Why the videos of your development progress? In mobile application development the impact of the user interface in terms of users wanting the application, code size, development time, and etc approaches 85%. Thus it is highly important that I get the right user interface. Now I could adopt an approach similar to Motorola, T-Mobile and etc in that completely weight the process down with al these ossifed layers of thsose who do not get it as far as mobile development and design and have a user interface designed by committe. While that may work for IDE development it does not work for mobile consumer application development.
While development of consumer mobile applications is closer to game development it cannot be a closed shop whereas feedback is cut off from those who may have the experience to guide through indirect feedback. Thus, who better to judge if I have the user interface right than the skilled mobile developers of the Android community. And it is relatively painless as far as the amount of work required to analyze the feedback as I only have to see which videos score the highest in hits totals.
But what about scaling? you are right in your implied assumption as the Xspot application faces the same exact scaling problems that Twitter faces at a higher impact level as its a mobile application that has a publisher and subscriber model like Twitter. Part of that scaling probem is off-loaded by using Gnip to interact with different APi such as Brightkite, FriendFeed, and etc rather than face large latency times in dealing with each API on a direct and individual basis. The other part of the scaling problem involves balancing the ajax and javascript loads across the components of webkit.
For example, iPhone like page and item transitions can be best enabled with 95% of it being CSS animation and 5% of it being javascript and thus freeing up the webkit javascript engine to handle other more important tasks sas far as scaling and latency. By avoiding the time taxing processing and modifying the DOM to enable page transitions via javascript I am able to tap that unused power to reduce the scaling issues.
Innovation wise its probably the first time an Android application has been set up to have server infrastructure replaced by implementing the use and integration of the social clouds approach and be html and ajax based at the same time. The exact future of mobile application that Google dreamed about that was one of the core focuses when both Googl ena dOHA developed the Android platform. This is going to get real interesting.
Previous Topic
