Free Goods, Pricing

One need not be a landlord to realize that offering free goods and services often leads to "tenancy risks" of one kind or another. This goes back to Oliver Williamson's basic notions of economic realism: "Opportunism" and "Bounded Rationality".

So, it is no surprise when AT&T finally pulled the plug on a free service that was actually being used by a significant, albeit small minority, to its promised limits. It stopped offering unlimited data plans for it mobile subscribers.

Number Portability Actually Works

I originally wrote this entry on September 29, 2004, and published it on blogs.sun.com.

I don't know how many of you have changed mobile operators, but I've done it quite regularly: PacBell, Verizon, AT&T Wireless, Cingular and now T-Mobile. Part of the reason is that I want to change mobile devices or try different operators while receiving the device discounts. Sometimes, I like a device or an operator and stay longer than usual. What matters to me most is that I can sign a 1-year contract.

In the past, when subscribers changed operators, they would end up losing their mobile number, partly because the Supreme Court kept admitting requests for delays in implementing the FCC-mandated number portability feature. The operators were asking for delays because, they argued, the systems had not been prepared adequately to make the process smooth.

In any case, now, the process seems to work like a clock. I put in my request last night. My new "mobile station" (to use GSM-ish language), which was originally attached to a temporary number, has been working with the transfered number since this morning. (For all I know it could have been operational right after my number transfer request last night. I didn't get time to check.) A Sun partner visiting our offices just called me to remind me of the time he was arriving. He called my old number, and I got him on my new phone.

Number portability actually works and is quick!

Should I go or should I not . . .

I originally wrote this entry on September 26, 2004, and published it on blogs.sun.com.

I was supposed to attend (as vice chair and participant) the Open Mobile Alliance - Mobile Web Services working group meeting in Orlando early next week but Hurricane Jeanne has closed the airport until Monday afternoon, and going through 15 hours of travel and airports (not including the delays) while ending up with the opportunity to attend (possibly) less than one day of a meeting is not my idea of resourcefulness. So, I think I'm going to have to cancel this trip. This will incur Sun some cost but the savings in time, energy and individual productivity outweigh the losses. So, Sunday morning when I wake up, my first move will be to put a cancellation order in motion. Missing the meeting will unfortunately mean that I will also miss what promises to be a great presentation on Fast Web Services by Paul Sandoz. (He seems to have gotten to Orlando, from France, just before the airport closure.)

Some have expressed curiosity regarding how the OMA board could have allowed the meeting to be arranged during a hurricane-prone season in Orlando? Can Mickey Mouse really produce such an irresistible pull for our European OMA colleagues? I doubt it very much.

Danger Gets Dangerous

I originally wrote this entry on September 27, 2004, and published it on blogs.sun.com.

I'd been waiting for more than a couple of years to get Danger's Sidekick. No, this is certainly not like Enkidu, the sidekick to Gilgamesh, that primordial myth-model of heroism, from Mesopotamia, a land we now know better as Iraq. Nevertheless, every hero needs a sidekick, even if it is to be a cyber-sidekick. Gilgamesh had Enkidu, and I got my Sidekick at the Palo Alto T-Mobile outlet on University Avenue this Sunday. If I'm satisfied (and I get 14 days to decide), I will then have to see about number portability. I'm told it now takes a very short time to install. I will write about this more in a couple of days.

My criteria for the purchase of a sidekick had been a color screen and tri-band radio capabilities. Sidekick II meets both. Not only that, I now have IMAP/SSL accounts I can set up. I haven't had time to set up an IMAP account yet but I had enough time to take a picture of Roberto Chinnici. He visited me (or was it a visit to my Sidekick II) just a few moments ago. More about Roberto later.

Trying a "Mobile" Weblog

I originally wrote this entry on September 27, 2004, and published it on blogs.sun.com.

On my first day of owning a Sidekick II device, I've been able to start and manage a mobile weblog in less than 3 minutes, with great ease and directly from the mobile device itself. The device design and service integration are strikingly smooth. Many efficiencies are included.

SIP capabilities, a good introduction

I wrote this entry, originally, on July 5, 2004.

If you are looking for a good introduction to Session Initiation Protocol (SIP), you can find one titled "The Future of Communication using SIP" in the Ericsson Review, No. 1, 2002.

You might want to start by reading the section titled "Capabilities supported by SIP." These are on pages 31 - 32 and give you an ideal perspective on (the technical) environments where SIP would be useful.

Mobility and Single Sign-On

I wrote this entry, originally, on July 3, 2004.

For mobile telecommunications operators, single sign-on (SSO) is not just a cool technology. It is increasingly an economic necessity.

One may argue that mobile telecommunications operators already have a form of single sign-on operating under the covers.

As the "mobile station" (the technical, GSM term for the mobile phone) moves, it needs constantly to access distinct resources (or services).

When a mobile station moves, the cellular tower giving it access to the public land mobile network (PLMN) will eventually hand off to the next cellular tower. The radio network server will use signal strength comparisons between the towers as an indication of when and where hand-off needs to occur.

However, single sign-on to network resources and services is a truly general concept in need of a more broad treatment. In fact, the Liberty Alliance, a standards organization has addressed the problem of single sign-on in its ID-FF (identity federation framework) set of specs.

So, where does the economic necessity come from?

Mobile telecommunications providers are focused in providing connectivity, access, identity and other voice and data services to their customers. From their point of view, a service is a kind of network interaction for which the subscriber is willing to pay. However, as we move to the a mobile age filled with perviously unimagined forms of voice and data communications, various "services" will need to be composed to provide the required features in a sophisticated and billable service.

A simple example may help.

Let's say you're a network operator serving 15 million people, offering them Multi-media Messaging Service (MMS) capabilities. Your subscribers are carrying mobile stations that can take pictures, which they can send to others through MMS or other means. Your subscribers also receive multi-media messages, say pictures or video clips, from friends, family or other services (e.g. sports, whether, etc.) to which they have subscribed.

What if your subscriber wants to maintain an album for all these in-coming and out-going multi-media messages? Should you be providing that service or should you be "outsourcing" it to another business that specializes in maintaining large multi-media albums and libraries? You'll probably choose the latter course because the business of providing digital albums for multi-media content is not within your core competency. The album business is a separate business. In fact, your subscribers will want to access their albums in multiple ways and will probably want to preserve a distinct identity with their album provider.

This example simply shows that the decision to host a broad SSO solution can be an economically rational decision for the mobile telecommunications operators simply because economics of sophisticated services will push them to "sub-contract" or "outsource" many of the "services" which they will have to compose into a billable service.

In the presence of sophisticated service composition, one would need a very good SSO and identity model. Hence, the advent of the Liberty Alliance.

Joobin Mobile Phone Software for Persian Speakers

Joobin has some amazing software to install on mobile phone. These are primarily for Persian speakers. (Some call the language "Farsi," although "Farsi" is the word Persian's use to refer to the language of Persian.)

The variety of software is certainly worth a good try.

Mobile Phone Concepts

Check out the new mobile phone concepts, particularly the multi-layer phone from KDDI--the Ply.

SIP capabilities, a good introduction

If you are looking for a good introduction to Session Initiation Protocol (SIP), you can find one titled "The Future of Communication using SIP" in the Ericsson Review, No. 1, 2002.

You might want to start by reading the section titled "Capabilities supported by SIP." These are on pages 31 - 32 and give you an ideal perspective on (the technical) environments where SIP would be useful.

Mobility and Single Sign-On

For mobile telecommunications operators, single sign-on (SSO) is not just a cool technology. It is increasingly an economic necessity.

One may argue that mobile telecommunications operators already have a form of single sign-on operating under the covers.

As the "mobile station" (the technical, GSM term for the mobile phone) moves, it needs constantly to access distinct resources (or services).

When a mobile station moves, the cellular tower giving it access to the public land mobile network (PLMN) will eventually hand off to the next cellular tower. The radio network server will use signal strength comparisons between the towers as an indication of when and where hand-off needs to occur.

However, single sign-on to network resources and services is a truly general concept in need of a more broad treatment. In fact, the Liberty Alliance, a standards organization has addressed the problem of single sign-on in its ID-FF (identity federation framework) set of specs.

So, where does the economic necessity come from?

Mobile telecommunications providers are focused in providing connectivity, access, identity and other voice and data services to their customers. From their point of view, a service is a kind of network interaction for which the subscriber is willing to pay. However, as we move to the a mobile age filled with perviously unimagined forms of voice and data communications, various "services" will need to be composed to provide the required features in a sophisticated and billable service.

A simple example may help.

Let's say you're a network operator serving 15 million people, offering them Multi-media Messaging Service (MMS) capabilities. Your subscribers are carrying mobile stations that can take pictures, which they can send to others through MMS or other means. Your subscribers also receive multi-media messages, say pictures or video clips, from friends, family or other services (e.g. sports, whether, etc.) to which they have subscribed.

What if your subscriber wants to maintain an album for all these in-coming and out-going multi-media messages? Should you be providing that service or should you be "outsourcing" it to another business that specializes in maintaining large multi-media albums and libraries? You'll probably choose the latter course because the business of providing digital albums for multi-media content is not within your core competency. The album business is a separate business. In fact, your subscribers will want to access their albums in multiple ways and will probably want to preserve a distinct identity with their album provider.

This example simply shows that the decision to host a broad SSO solution can be an economically rational decision for the mobile telecommunications operators simply because economics of sophisticated services will push them to "sub-contract" or "outsource" many of the "services" which they will have to compose into a billable service.

In the presence of sophisticated service composition, one would need a very good SSO and identity model. Hence, the advent of the Liberty Alliance.

Multi-core performance on Solaris

Jignesh Shaw presents @ PGCon on multi core performance of Postgresql on Solaris.

What Postgresql can learn from MySQL

Susanne Ebrecht presents @ PG Con 2008...

PGCon 2008

PGCon 2008

Mobile Device Sales Decline in the U.S.

Not every child seems to need a mobile phone when the economy is not doing well.

Sara Silver of The Wall Street Journal reports:

U.S. purchases of new cellphones declined in the first quarter for the first time in several years, signaling that worries about an economic slowdown are hurting the handset market, according to two new studies.

The drop was concentrated among poorer customers using prepaid plans and among households earning $75,000 to $99,000 a year. The upper end of the cellphone market -- phones featuring full keyboards for email and text messaging, and extra features for music downloads and video viewing -- continued to see growth.

The overall decline in handset sales also comes amid finds that growth in wireless-service subscriptions is slowing as the market reaches saturation, with 83% of the U.S. population owning a cellphone now.

...

This trend benefits makers of feature-rich handsets such as Samsung Electronics Co., LG Electronics Co., BlackBerry-maker Research in Motion Ltd. and iPhone maker Apple Inc. Motorola has struggled to add high-end phones and has lost market share as a result.

At the end of the first quarter, Motorola held 25% of the U.S. cellphone market, compared to 34% at the end of the first quarter of 2006, according to Strategy Analytics. During the same period, Korean competitors Samsung had increased its share of the U.S. market to 22% from 15% and LG rose to 21% from 16%. RIM's BlackBerry devices comprised 8% of the market, enabling the company to take the No. 4 slot from Nokia Corp., the global market leader.

JavaOne

Soft phones in Java !

I remember the days when people were wondering about Java's survival.

Today it is used in some of the coolest, most disruptive telecommunications technologies.

VoIP-info.org has a list of these technologies which include some soft phones in Java. I'd be interested in knowing if anyone has tried them. Some samples are: Ubiquity's SIP User Agent, NIST's JsPhone and iptel.org's BonePhone.

Has any of you used any of these phones?

SIP Roles and Activities

Session Initiation Protocol (SIP) specification (RFC 3261) describes four main types of roles. The user agent client (UAC), the user agent server (UAS), the redirect servers and proxies. There are also many activities defined such as registration, querying for capabilities, dialogs, initiation of a session, modification of an existing session and termination of a session.

Many commercial and non-commercial operating systems are now embedding the SIP protocol stack. These are often accompanies by APIs to implement any one of the various roles in any one of the activities described in RFC 3261. The trick then becomes to conceive of applications. Many of the authors of RFC 3261 have done exactly that. For example, you can take a look at Henning Schulzrinne's many papers in this area, describing various use models explored in his labs at Columbia University.

SIP is about to become the de facto session initiation and control signaling protocol dominating the converged world of Internet-based communications technologies and services such as VoIP. (On the related subject of VoIP, see the recent post by Richard Giles on VoIP on Java. There's of course much more that can be said about VoIP and SIP.)

In the JCP, there are several SIP-related JSRs. These are JSR 32 (JAIN SIP API Specification), JSR 116 (SIP Servlet API) and JSR 125 (JAIN SIP Lite).

SIP -- Main Purpose

The main purpose behind RFC 3261, Session Initiation Protocol (SIP), is to to set up, modify and tear down multi-media communications sessions on the internet.

An important feature of SIP is its addressing model:

To: E. Schroedinger <sip:schro@quantum-heaven.com>
From: Werner Heisenberg <sip:heisi@uncertainty-zone.com>

This addressing model is opaque for internet routing and direct access to user end-points. (If they were just e-mail addresses, we sould know how to address the SMTP servers.) Somehow SIP addresses need to be associated with actual end-point addresses on the network. IP addresses of users need to be obtained. That's where the rest of SIP comes in. Once you have these features, we can start talking about the use of SIP in VoIP, IP mobility, IMS and more.

[Names for addresses are inspired by examples in "SIP: Understanding the session initiation protocol" by Alan B. Johnson]