I’d rather be surfing

I don’t have time to read as many blogs as I used to which is why I’ve come to appreciate (and rely on) techmeme. I started reading the “mini” (i.e. mobile) version (available at www.techmeme.com/mini) a few months back on the train ride to and from work on my Blackberry Pearl.

Overall, I like techmeme/mini, with the one caveat that the stories link back to the full web version of the sites. Here is an example… Today there is a store on techmeme about Walmart. The full techmeme version links to the nytimes.com version, which is expected. The techmeme/mini version links to the same article! My Blackberry does a solid job of formatting the page so I can still read it but it would be optimal if I was directed to the mobile.nytimes.com version of the story.

Techdirt Insight Question 1: Outline a Strategy for Developing a Mobile Application

This is a pretty tough question even for those of us who have spent years in the mobile industry. The question specifically “wants to maximize coverage while minimizing integration costs”.

This question is pretty broad, so I would first try to answer a couple of big picture questions to narrow the possibilities…

1) What exactly is meant by “mobile applications”? There are two types of mobile applications, client and web with pros and cons for each. Mobile client applications are apps written in Java (J2ME), Brew, Symbian, Windows Mobile or some other flavor of compiled code. The developer can control just about every detail of a client app but porting across devices (even on the same platform) is a nightmare and porting cross platform means one or more rewrites.

Alternatively, mobile web applications are slightly less fragmented XHTML-MP, WML and CHTML . Developers have less control over the rendering and UI of an application but portability is much easier (read: cheaper). Plus, there are some cool tools out there such as WURFL and WALL which make mobile web dev a bit less painful as compared to mobile client apps.

In terms of “maximizing coverage while minimizing integration costs”, mobile web development is the way to go. Most phones (even old ones) have decent XHTML-MP browsers (especially in the US and Europe). If you simply have to build client apps, then the question is a little more complex which leads me to my second question.

2) Who is your target market? Or more specifically, what devices and/or platforms are you targeting? Whether you are targeting all devices that are MIDP2.0/CLDC1.1 (JavaME) or Nokia Series 60 devices, the answer to this question should help you gauge the cost of porting and whether or not it makes sense to attack a specific platform. In terms of global install base, JavaME is the clear leader. (I am not sure about the numbers but I suspect MIDP2.0 has passed MIDP1.0 by now with most of these being CLDC1.0.)

3) One of the expensive issues of portability is the UI. Most apps are built with highly customized user interfaces that look and feel, well, just graphically designed. Particularly in the JavaME world, these UIs are build custom for each device and tweaked constantly to get the perfect user interaction. Porting is costly and since new devices are constantly coming online you have to constantly port your application.

Overall, my recommendation would be to developer you applications for the mobile web not mobile clients. You get more bang for your buck: development and porting is both cheaper and easier. If you must develop client apps, choose JavaME, ease up on the custom UIs for portability (costs) sake.

I had an interesting conversation with a friend of mine today.  We were discussing how to communicate with developer communities.  He said something like, “developers want to argue and they respect you more if you stand up and argue”.  It made me think about wrestling with friends as a kid.  After a friendly wrestle, both kids would end up learning some strengths and weaknesses about themselves and their peer but most importantly, both kids would garner a little respect from one another.

Part of the motivation for starting this blog is derived from the fact that there are so many interesting questions. I took this screen shot of “Latest Issues” posted in the Techdirt Insight Community.

Well, Techdirt finally did it. Their Insight Community finally made blogging something I have to do. I’ve tried a number of times to start a blog, quickly losing interest… But this should keep me motivated and should at least give me fodder (not that I really need it).

This should help my “discoverability” and “brand” as well.

So like every good programmer: Hello World.