I stumbled on this cultural model of the software development space. It represents well the influences and constraints. In the mobile and ubiquitous world, this is one set of constraints developers must face (Development with Constraints). Another, lower level of constraints would be the technical, economical and physical limitations (Design with Constraints).

Development with constraints

I got kicked out of the restaurant car in a Swiss train because I was reading my NetNewsWire with my 12″ PowerBook. I was shown a the written “restaurant” policy booklet containing rule 11: “notebooks are NOT allowed”. Reading newspapers, books, and yelling on a mobile phone are allowed. Even smoking cigars is a more than accepted activity. I did not go into explaining that I was actually reading the news with my small electronic device and having an IM conversation with GPRS. It is interesting to notice that a notebook is still perceived as a working tool and not a communication device. Until december 2005 (when smoking will fortunately be banned from all trains) smokers are more welcomed than geeks in swiss train’s restaurants.

Google has an open-source program mainly to support young geeks and tie into open source. In that program, Googles is functioning as a university as it creates a new kind of distributed lab. Chris DiBona, open-source program manager, has a great quote on the importance for software engineers to properly apply the tools of the networked economy (message boards, phone, VoIP, email, instant messaging, video conferencing, and blogging):

“Developers should get used to the idea of globally distributed work groups all over the place, all over the time zones. It’s the future of software development.”

Via Google and the networked economy on ballpark.

Miguel Angel Cordobes Aranda of the Auna (now Ono) Innovation Center gave a presentation at UPF on the results of the first WiMax trial in Spain. He knew half of the acronyms and only understood a few of the technical numbers and telco charts. However Miguel provided an overview of the current and future of WiMax from an operator’s point of view. In a few punchy lines:

• WiMax cannot compete with xDSL, HFC in terms of capacity
• WiMax will net be able to compte with mobile networks (GSM) in terms of reach
• WiMax brings more problems than solutions for fixed services (e.g. user has to be close to the window)
• The added value of WiMax is to offer mobility… and the current standard does not support it (no handover).
• WiMax is good for nomadic mobility but not complete mobility. Current technology insures connectivity up to 50-60 km/h. So no usage in trains!.
• WiMax is focused on the midium-size metropolitan area! The countryside, rural ares and big agglomerations are not of a big interest (financially).
• Apparently, particulars won’t be able to proclaim themselves providers by installing their own broadband base stations.

It is very interesting that operators view WiMax as a technology for mobility even though it clearly lack of coverage (it probably won’t ever reach GSM’s overall coverage) and lack of connectivity at average speed. All these technological and financial constraints strengths my current interest in the moving sands that mobile and ubiquitous applications and users must deal with. All wireless offers come from their technical limitations and financial constraints. Currently, no wireless technology eats the other, but rather collocate to and stack on each other with extremely poor interoperability. Each context (controlled indoor environment to uncontrolled outdoor environments) calls for its sets of technologies. Ubiquity, to its strict sense, is still far from reach.

Jim Waldo gave a talk in the NYJavaSIG May meeting about Jini being the base for Service-Oriented Architecture (SOA, buzzword alert!).

While Jini does enable all of the advantages promised of a service-oriented architecture, it also enables some others that are somewhat more surprising, such as much higher levels of reliability and the ability to change, upgrade, and extend the system without there being any service interruption.

JXME 2.1 (Tantà) has been released.

The purpose of JXTA-J2ME is to provide a JXTA compatible functionalities on constrained devices using the Connected Limited Device Configuration (CLDC) and the Mobile Information Device Profile 2.0 (MIDP). The range of devices include the smart phones to PDAs. Using JXTA for J2ME, any MIDP device will be able to participate in P2P activities with other devices within the JXTA network.

In Introduction to Mobile Game Development, Nokia defines 6 limitations to developping mobile applications:

• Small screen size: developpers must still optimize their applications for different phones
• Limited color and sound support
• Limited application size
• High latency. Latency is the aount of tie it takes between the moment a machine makes a request and the moment it receives a response. While the carriers are working to expand the amount of bandwidth available to mobile phones, they have not made latency a priority.
• Interruptibility is crucial. the application must be able to pause and recover without crashing, causing play problems
• Evolving technologies. Need to support hanset specific APIs

Sunrise Switzerland annonced a flate rate GRPS offer (9 CHF, for 4 months with monthly data transfer limit of 100MB/month). It pushed me to finally setup my PowerBook to get Internet access via Bluetooth using my Nokia and GPRS. Ross Barkman’s GPRS Info Page provides all the infos with a link to his Apple GPRS scripts. Good feeling of cheap ubiquity…

Arvind Gupta delivers again and again, with a straight forward Symbian code that autostarts a MIDlet using a Recognizer and the Push Registry.

2 articles explaining how to automatically start a MIDlet by using the Push Registry

- How can a MIDlet be launched automatically?
- The MIDP 2.0 Push Registry

Thanks Arvind Gupta for the tip!