Tom The Architect

Technology and other Interesting Stuff

Month: May, 2006

Experience Continuity

I have been thinking about how companies build systems. In my experience the companies focus on the things like business continuity and leave for last (if they are lucky) experience continuity. The systems are reliable, transactional, scalable and redundant but yet from the outside the system seems a bit off. The systems are build to reflect the desires of the company and in many cases the customer perspective is the lip stick on the pig. In the end who cares if the system is sound and provides for business continuity, if it fails to provide experience continuity. Doing the wrong thing the right way is still doing the wrong thing. Experience continuity is the starting point and the rest will take care of it self.

More than a product offer, It’s attention efficiencies and products

Traditional companies (manufacturers or any one in the classic supply chain) will be faced with competition for their customers attention. A company today strives to spend the majority of its attention on its differentiating value. That means GE doesn’t think how it buys pencils is a differentiating value. The attention that GE wishes to spend on those non-differentiating items is small and getting ever smaller. The funny thing is the attention efficiency GE seeks from its suppliers is in some cases the efficiency GE’s own customers are seeking.

So what does a company like GE do to meet the customers desires, it offers its customers the opportunity to spend less time managing the domain that GE’s products and services are applied in. For example, GE may build custom fuses for power plants but that is not the real value. The real value is ensuring that a fuse failure doesn’t take the plants production capability off line. What GE could provide the customer is an attention efficiency around monitoring a plant’s distribution network, coupled with a highly integrated logistics capability, and a ready supply of power distribution products. GE becomes the Tivo of power distribution, identify the parameters of need, set it and forget it. The attention that GE set free can be applied elsewhere. That attention efficiency doesn’t mean it costs less, though it should. It means that the company trades unskilled dollars for talent that can be used to improve revenue making operations.

More and more companies are going to be required to deliver more than a product, they will have to deliver a product wrapped in an attention efficiency.

The GE example is just for illustrative purposes.

Attention efficiency

I think that the attention economy is best served talking about Attention efficiency.

Attention efficiency is the ratio of useful, valuable, or effective output to the total attention input in any system or systems.

For example, the attention efficiency of television changed a great deal when Tivo arrived on the scene, reducing the required input to get the same value. To deliver on the attention economy is to deliver attention efficiency. If managing Tivo consumes the same amount of attention as the eliminated commercials then Tivo has done nothing to deliver attention efficiency (not the best example, but its late). So as attention becomes the measure of value then we must understand what burdens our applications and businesses place on the attention of our customer / users / observers.

Today we lack any real attention efficiency due to the fact that we are required to interact in so many different silos. I like to think that we will move the online experience (offline experience will slowly disappear) from one of silos that force attention inefficiencies, to a paradigm of situational awareness that is driven by creating attentional efficiencies based on our individual values.

It is critical to point out that the folks who think about attention are themselves early adopters and forward thinkers, that is not the case for everybody else. So our success is dependant on folks that don’t think about attention but value it and desire simple means to deliver more effective output from their attention investments.

The definition of efficiency courtesy of

List of languages I have programmed in

A list of languages I have had to code in. The list is in no particular order.

  • Perl
  • Java
  • Ruby
  • Basic
  • Fortran
  • PL/1
  • Visual Basic

Most Popular Books at Java one: Update

.flickr-photo { border: solid 2px #000000; }
.flickr-yourcomment { }
.flickr-frame { text-align: left; padding: 3px; }
.flickr-caption { font-size: 0.8em; margin-top: 0px; }

I thought I would update the list of best selling books at JavaOne. It was sad to see Ruby for Rails drop off the list. Looks like AJAX was a popular topic, taking 6 of the spots on the list.


Most Popular Books at Java one, originally uploaded by TomC.

JavaOne Highlight: Meeting Adam and Jamie from Mythbusters

.flickr-photo { border: solid 2px #000000; }
.flickr-yourcomment { }
.flickr-frame { text-align: left; padding: 3px; }
.flickr-caption { font-size: 0.8em; margin-top: 0px; }

The After Dark party at Java one featured the Mythbusters. I really enjoy the show, so getting to meet them for a brief moment was way cool. They really liked the winning t-shirt launcher. The party was cool loud pounding urban techno with a cool light show.

James Gosling should let Adam and Jamie build a t-shirt cannon for next year. The entrants would then compete against Adam and Jamie for longest throw and most radical design.

At the Mythbusters Toxic waste Bar at the party at JavaOne, originally uploaded by TomC.

The winning t-shirt launcher

The Mythbusters

The most popular books from JavaOne day 1

.flickr-photo { border: solid 2px #000000; }
.flickr-yourcomment { }
.flickr-frame { text-align: left; padding: 3px; }
.flickr-caption { font-size: 0.8em; margin-top: 0px; }

It is funny at JavaOne to see the 8th most popular book. I was in sessions where people were asking what Ruby on Rails was.

JavaOne: Fireside Chat w/ commentary

Summary (Not in any particular order):

  • Get rid of AWT (Gosling)
  • Operator over loading (good (Gosling) and bad (Hamilton))
  • Simplicity, Yes
  • Participate in the JCP
  • Tell your vendors to improve their Java support
  • Netbeans
  • class loaders hard
  • Schedule Builder yeah what ever
  • US mobile infrastructure sucks
  • Fewer wireless providers could be good
  • AJAX use Java Server faces
  • Uh Time for Beer

My commentary:

Ok, this was the first time I attended the fireside chat. Problem number 1, the session should really be called the fireside question and gripe (bidirectional) session. Problem number 2, the panel seemed to me to be a bit aloof. If the question seemed a bit off or on a topic a bit off someone would mumble some remark followed up by Graham Hamilton PR techno speak (he does it quite well). Problem number 3, the answers for the most part were not really candid, I guess that happens when you are a VP Muckety Muck at a publicly traded company gotta watch what you say. And for the most part I think it was Graham Hamilton did most of the talking.

There were some interesting questions, there were a few about simplicity and ease of use. The responses from the panel were okay, but they lacked something. James Gosling talked about netbeans wrapping all the complexity which makes sense given the state of Java. It is many things to many people and that is never a good thing, someone is always left wanting. From the remarks I think Sun gets the ease of use issue, the problem is that there isn’t a great deal they can do.

Graham Hamilton urged everyone to get involved in the JCP process and work to make it better. Now I am not familiar with the JCP process, but I am familiar with standards bodies. In the end it comes down to vendors working very hard to defend there products and their investments. The customer is not a prime consideration, because there aren’t a lot of choices for the customer out there. I have seen first hand all the big vendors fight over standards, there aren’t many vendors who don’t have standards blood on their hands. So, I guess the call by Graham Hamilton rings a bit hollow for me.

There were some questions about mobile and why there are so many issues around the distribution of applications. One question about VMs for PDAs came up and the response from Sun was fair. The PDA phone market are converging and Sun has significant penetration and support on the phone platforms. The question about why its so hard to distribute applications lead to a round of telco bashing and rightly so. The comment from the panel that seemed to be wishful was that consolidation on the wireless market into a few companies might help the mobile application and data market.

OK, when has a consolidation ever been good for the customers of a regulated business like telephone industry. The consolidation of carriers in the wireless arena in my opinion will only continue to limit third party mobile application distribution. The telecoms want to maintain complete control over the vertical market, and even with big player like Google things like tiered network pricing could be used to limit third party access to the handset.

Finally, the schedule builder was mentioned at least twice. The topic was brushed off by the panel, a panel who admittedly does not use the tool (there’s a problem right there). I understand the schedule builder is a weak topic, but the brush off stuck in my craw. I spent a couple of hours dealing with the tool, time I could have been using for something a bit more productive. People pay almost 2k plus expenses to attend, then take time to attend the fireside chat only to have the questions about the schedule builder to be brushed off. Let me see if I can make the suggestion more palatable, How about a contest to see who can build a conference scheduling tool that also throws t-shirts.

6 million missing Shopping carts

In the Los Angeles area last year a minimum of 6.2 million shopping carts went missing according to California Shopping Cart Retrieval corp and Hernandez Cart Service Inc. (via Harper’s). I find this number amazing. I wonder if the number of shopping carts that go missing are reflective of the number of homeless. I know scrap steel has gone up but where do these carts go. I also find it funny that there are companies specializing in shopping cart recovery. These companies in my mind are like Dog The Bounty Hunter for shopping carts. It just goes to show that there are niche businesses everywhere. This problem does beg for a technology solution like RFID or something. 6.2 million WOW, I wonder what the recovery cost is to the stores.

Get Harper’s Magazine RSS feed here
Harper’s Magazine

Comparing programming languages in Google Trends

I thought I would try out Google Trends and compare popular programming languages. It appears that Java is still the big dog. Ruby came out of nowhere according to Google Trends. Click on the image to get a larger version.

Google Trends
Originally uploaded by TomC