I'm in a room with at the old WebCT headquarters near Boston for the Common Cartridge work. The combination of this effort, the Interactive Learning Environments and the SOA working groups holds some real promise. More on this later with some thoughts on real world applications...

Now for the guts of the meeting.

The IMS team requested an audience with BB management to allow for it's members to pose questions on the recent lawsuit against D2L.
IMS requested that the meting not be recorded and as such I have not included any direct quotes from the interaction.

I wanted to comment on some interesting highlights from that exchange.

BB was represented by the Matt Small, Chris Vento and Jan Day although Matt fielded most of the questions as might be expected as the legal counsel.

One of the first questions directly from IMS was for Matt Small to define the scope of the patent. Matt Small described the complexity of the legal process and stated that he could not provide specifics but gave examples as he has done in previous interviews.
Matt then described the use of simple websites used by faculty to deliver courses on a one-off basis until Blackboard arrived on the scene with an 'enterprise' CMS that introduced the concept of roles for course delivery. This statement raised a few eyebrows around the room.

The example he gave was to describe the use case of a system where a user might have multiple roles across multiple courses.
This was particularly unfortunate since his example is vague and he challenges that the FUD in the community is due to a misreading of the patent and the claim.

One of the questions posed to Matt Small from IMS was the issue of the IMS IPR policy. IMS has it's own IPR policy where members are requested to raise awareness of any existing patents that they know may be in conflict with the work of the group at the beginning of each conference call and meeting.
It was asked of Chris Vento, why he (or any other BB members) had not raised these patents before in those calls.
Chris restated Matts point that this is a difficult and complicated issue. Both Matt Small and Chris Vento reiterated the view that it would take detailed analysis of each spec and it's work to assess if they were infringing on the patent(s).

An excellent point raised from a member was the fact that the IMS Enterprise Spec includes the facility to import and export users and roles across a course structure. If someone was to build such a tool (I know quite a few) then they would be at risk on infringement.

Matt Small was very clear that there was no plan to take action against the standards/specifications groups or the open source CMS projects, namely Moodle and Sakai. The problem is that these efforts are supported by commercial companies that support and promote their success in partnership and those companies should be worried as they represent revenue threats for BB. If small and medium players are removed from the competition, then there is no IMS or uptake of open source projects since there will be no support services available to drive adoption.

An interesting point Matt made was that BB and D2L will argue the issue in a court of law and the exact definition will be determined as a result.

Matt reiterated that BB does not want to stifle innovation in the industry. By placing us in a position where we won't know the exact scope of the patent until after a lengthy legal debate, they have done exactly that. While we wait for a court in Texas to decide what use case definition, analysis model or physical domain model we must avoid using in building systems for supporting education, we might as well sell sandwiches.

I never saw thet film with Forest Gump stuck in an airport but sometimes I feel like I live in Newark International. I'm supposed to be presenting an Identity workshop at the annual QuestNet event in Brisbane, Australia and the friendly Continental representative has just informed us that the plane needs an engine overhaul and it would only take an hour.


My connection in LA is looking even tigher now but please, please, spend more than an hour for an engine overhaul.

Segue is a little known gem. It is a PHP based LMS built at Middlebury College and used in many others. It offers a host of the standard course features you'd expect but is much more flexible than any other I've seen. In particular I like the web 2.0 models of built in blogging and tagging as standard. There is also a wide selection of templates to allow non-programmers to change the entire look and feel of their site from a simple blog tool to a full blown Course Management System.
You can try the hosted version at Education Commons or go directly to the project website.
The Segue team built the first PHP OKI OSID implementation in project Harmoni which is what is used in Segue to access Digital Repositories.
In todays OKI webcast we kicked off the cross language Java / PHP OKI effort to investigate the XML-RPC bindings. Initially the scope of the project is to develop a proof of concept for existing consumers and providers using a mix of Java and PHP. The list so far includes:

• VUE II
• Segue
• Moodle
• Harvest Road Hive

Project plans and this weeks meeting podcast soon to be posted on Education Commons.

This post on emedia news blog is the first public reference we've had to Education Commons and I'm pleased to get a healthy 'wait & see' message. Scope creep is always a tough issue to manage on any project but the expansion into 'Entertainment' is a welcome addition as the discussion of gaming and learning must include television and next generation media. These are some of the topics passionately discussed in the Digital Hollywood University Project where I like raising the question of the game engine as an LMS - why not?

Ok Ok... I've been lazy.....and pretty busy in my defense...I've been working on a huge project for the development of next generation teaching and learning environments for schools in the UK.

Imagine a school where a teacher walks into a classroom with thin plasma screens on the walls and inserts their java Id card into a wireless laptop, immediately the walls render posters and images relevant to the class that's about to start. The students enter and find themselves a wireless laptop and insert their cards which authenticates them to their individual portal learning environment. The teacher then progresses through the blended learning session moderating the online sessions and group work.

I'll be posting more on our next generation learning architecture later...

I'm listening to a great pod cast on from EdTechTalk on a proposed wikipedia for K-12 text books. In the world of educational technology podcasts this is not a new topic of discussion although there are some great ideas. This podcast struck me on 2 fronts, firstly the idea of print vs. digital resource in teaching and they format being used for this discussion. I'll get to the latter in a sec.

As a systems designer I try to model the concept in my head. I start with the notion that if you treat the student as an actor and the system boundary as the learning environment. Assuming the text book is issued to the student as part of the curriculum for a course they are enrolled, it should simple be represented as a non-digital resource with a relationship to the course that the student can access.

So what are the properties of a text book? Well, it's static and non-personal. It's also hugely expensive as was pointed out by all the discussion members but for the practical purposes of modeling that is somewhat irrelevant.

The wikipedia book is another beast completely. It is non-static and could be made to be personalized based on historical context of the learning environment in which the student has navigated the digital resources. It also can have complex relationships to the course, program, institution and existing resources (digital and non-digital).

The problem presents itself when you add assessment. If you map assessment targets to to the teaching resources you add a great deal of complexity to the system design. Assuming you want to assess all students using a static framework, we'll leave our personalized assessment, then you have to link the assessment targets to a some constant reference in the teaching resources.

Now onto my second point, What a fantastic format for a podcast! The session was a planned discussion using a combination of moderated online chatrooms and voice conference call with conversations bubbling up from the chat rooms and invited into the audio. I had planned for a similar format on the upcoming education commons and I can now see how well it works . Thank You. Truly Inspirational.

For more on the wikipedia project see http://Education Bridges.org

I'm attending the Digital Hollywood event here in Santa Monica as part of sharing best practices in gaming and education.
First observation is the lack of free wireless! no further comment.
Second observation is the buzz around the gaming and Hollywood coming together and wondering what the future holds.
Not surprisingly one of the biggest topics of discussion is Digital Rights management (DRM) and what that means for the publishing industry. Exactly the same concerns we see in the education publishing world but that is only part of the problem.
I recently worked on a solution for an education authority to offer material through set top boxes using Video on Demand (VOD) and IPTV. The obvious challenge was defining the business rules to be implemented through one of the many 'open standards'. Like most design problems the there are multiple technical solutions to the problem but the real issue is defining the problem correctly in the first place. Accurately describing the business domain is the most difficult and time consuming part of any project.
Developing a solution for education requires us to go further and look at the interaction models of the media instead of just the distribution of media. Getting the assets to the user is a very small part of the system. Nearly all learning models require some level of interaction and the biggest challenge of a successful interactive learning platform is the flexibility of that 2-way interaction.
This conference has shown that Hollywood is still focussed on the protection of content and has not yet moved to the services around the content. The user is still the un-invoved user - being broadcast to and less involved. Personally I prefer the challenge of the Learning Management System (LMS) co-ordination.
I have found more fruitful conversations with the smaller interactive simulation companies that understand the technical challenges that these interactions pose on a larger scale. Most of these discussion have gone down the route of 'we have some cool interaction technology but how do we scale?'.
There is, however, a very interesting move to the gaming sector by Hollywood and more recently TV broadcasters. The obvious link is for publishers to re-use their assets like the spin off games from blockbusters. For me the next step is community building allowing re-purposing of those assets based on 'chunked' content. They could learn a great deal from our work in large scale teaching and learning systems.

Community building is a difficult thing to do. I've been involved in two developing communities over the past months. SAKAI and Open Solaris and they are both fascinating. I have managed many large scale development projects with huge numbers of developers both on site and in distributed environments. I am a big supporter of community development and have used it for many years on my development projects. Having been involved in these high profile projects and a few on Java.Net have helped me refine the development process I'll use on my next project.
I believe there is a place for open source, community source and bespoke purpose-built software where they seek to solve different problems. More thoughts on this later.
I have been blown away by the success of Open Solaris. In only 2 days there is a massive community and growing very quickly. The guys have done a great job in putting the right tools in place to support virtual collaboration.
The coolest thing about the Open Solaris community is the Sun Studio 10 compiler to build the OS. It rocks and is blazing fast.
I can't wait to see how this evolves...build your own OS and then build your own eLearning apps....now that's choice.
I'm off on a much needed vacation now but when I come back I'll be setting up SAKAI and Moodle on Open Solaris.
Stay tuned for build instructions appearing on www.sun.com/edu

After a long 4 days at the Community Source Conference for SAKAI, OSPI and uPortal, I'm taking a short break and heading home for a day of painting before returning for the uPortal part of the conference where I'll be presenting some ideas on uPortal and Identity.
On Thursday I kicked off the first face-to-face meeting in my new role as Chair of the SAKAI Performance Working Group where we discussed a process for capturing lessons learned from the two big production deployments of SAKAI a Michigan and Indiana. This information will be captured in the collaboration site and factored into what we hope will be useful information to other schools seeking to deploy SAKAI as well as hints to inform the architecture team to address manageability, maintainability and other non-functional systemic requirements as well as performance characteristics.
As usual the real work gets done in the hallways where I managed to catch up with with some of my UK colleagues doing great work on the JISC projects and the awesome research at the CARET organization.

I've just installed Moodle on my Solaris x86 workstation after hearing about the performance problems some teachers were having running it on their Mac OS X server. It seems MySQL craters at over 15 connections on the new version 10.5 so they are forced to deploy onto linux in production.
After compiling PHP with the enable-fastCGI switch it's just a matter of configuring the JES Webserver for the PHP modules - similar to Apache.
You could of course use the Apache webserver bundled with your free download of Solaris x86.
Here it is running on my W2100z

Full configuration instructions coming shortly.
Now to setup authentication to the JES Directory Server...

Sooner that I though but here it is...

e-Books to be delivered on a Sony PSP along with magazine subscriptions. With an e-Book reader students can sync their PSP through a USB cable or wireless bridge to the school LMS host and download the latest book lists for their course.


The PSP has a memory stick for persisting state enabling the book reader to use personal bookmarks, notes and track progress is the developers choose to take advantage of it.
Their progress and interaction could then be uploaded to the school or university host system to record their progress against their e-Portfolio and/or student record system.

Now all we need is the browser for the perfect thin client device we know will be in every kids schoolbag along with their ipod.

There is already a collection of designer add-ons for the Play Station Portable looking to follow the hugely successful ipod accessories market. The most interesting for me is this...

There are currently no games or packages available to take advantage of the keyboard but that hasn't stopped Level 3 from making it. They know something we don't but it's not surprising really. The PSP has a WiFi card, a USB port and most importantly the ability for system updates from Sony. Look out for a browser 'update' soon and bingo I you've got a thin client we device in every school kids hands. I had been working on a small device Learning Solution using the ill fated Sharp Zaurus but the PSP is sure to have a longer life span and worth investing the time.

Scott Leslie's excellent EdTechPost blog recently highlighted the press release from WebCT on the IMS Tools Interoperability Special Interest Group.

I participate in this working group with Blackboard and the University of Wiskonsin who were not mentioned in the press release. I'm actually in a working meeting with the team now...

I was disappointed by this press release so early on in our development as it has already lead to the hype I was seeking to avoid by ensuring this group was not labeled a 'specification' and put in the same category as IMS Content Packaging or Metadata. Those specs are well grounded and tested and I have built several systems that use them for interoperability with other Learning Management Systems.

It is so important to manage expectations in the education technology space.

This is not the 'wonder pill' of systems interoperability. WebCT, Blackboard, Perceptions (Question Mark), the Samigo project, Wiskonsin and Sun are all working hard to develop some scenarios using web services as one mechanism to show how you might launch an external tool from your LMS.
We are having to work around constraints imposed on us by existing system interfaces and limited resources. I am a strong advocate of good architecture foundations and open systems.
This is an integration project with a very simple context and Systems Integration is no simple thing. I spent may 'droid' years at Andersens gluing systems together and now focus much of my efforts on 'real world' SOA... I won't start on the SOA hype - more later...

This is only the first step and we have a long way to go. I'll keep you posted on our progress...

Why is it so hard to be original?
I can understand the huge number of spin off accessories from smaller companies riding off the back of Apple success with the iPod...cases, speakers, fm transmitters, underwear...

This is really desperate for HP to seek a larger role in iPod mania where they offer a plastic mould to fit around your ipod dock with an 'HP' logo on it.
I like my HP printer. Please make more but I don't need an ipod dock on it...

I remember my toughest assignment at University was having to write an essay on defining the meaning of quality. At the time it was in the context of the US car manufacturing industry and their adoption of the Japanese 'Total Quality' techniques which have now evolved into the Six Sigma quality improvement approach pioneered at Motorola now adopted at most organizations including Sun.

Study: Few bugs in MySQL database | CNET News.com

What is a bug anyway?

In a services contract for a bespoke software system it is a mismatch between requirements and the delivered software. More often than not it's a mismatch between expectations and the delivered software because the requirements were so poorly documented or the functional domain was not clear to the team. You could say unexpected outcomes are bugs. This might be true of uncaught exceptions. Generally, poorly written code that leads to memory leaks, race conditions or is just plain ugly is also called buggy.

If the software is not delivered to contract as is the case for open source then how do you measure it's quality? I don't know the environment that the MySQL software was tested and am not familiar with the company that performed the tests. Did the team develop a tight test harness to verify the documented functional requirements for the RDBMS? My guess is yes - to a level but there was more automated code profiling and checking for code cleanliness and security holes.

Don't get me wrong, I think MySQL is an excellent piece of engineering but I object to the often touted claim that open source software has a higher 'quality' than commercial products simply because there are more eyes reviewing it. Quality is ensured by a structured process of managing expected outcomes to delivery and that requires more than releasing your code for review. I agree that more eyes is a good thing not just to catch 'bad or ugly code' but also to get as many possible unique solutions to a problem is mostly a good thing - too many chefs and all that. Any good architect encourages peer review on their project, open source and commercial.

The only bug free software in the world is our old friend Hello World. Its functional design and expected outcome is well documented in every programming 101 book in history. With my expectations set I am assured Hello World is of excellent quality.

Of course someone could claim they expected "Hello World!" instead of "Hello World". Doh!!! My perfect software is riddled with bugs and 100% failure rate !!

If you haven't heard, Solaris is now Free so dust off that old pc in the closet and bring it back to life with Solaris 10 for x86.
I just rebuilt an old pentium III server running RedHat Fedora Core 3 and it flies. Boot time was cut in half and shut down in 5 seconds.

The most awesome developer platform out there. True resource virtualization with zones and DTrace for hardcore performance tuning. Performance programming isn't what it used to be cheap memory and the availability of rapid application development tools. For those of us old enough to remember how to optimize memory allocation and read a stack trace DTrace is a huge step forward in program traceability.