The JP Allen Blog

I’ll be presenting my short paper on “Three Strategies for Open Source Deployment:  Substitution, Innovation, and Knowledge Reuse” at the Open Source Systems 2010 conference this May.

I wrote the paper because I see a number of organizations (including my own dear University) using what I call a substitution strategy for open source:  rip out existing proprietary software, and replace it with a ‘free’ open source equivalent.  That strategy has advantages, but it ignores many of the unique benefits of open source use.  I classify these unique advantages into two types:  an increased rate of innovation inside organizations (innovation), and an increased rate of innovation sharing between organizations (knowledge reuse).

In certain situations, such as my San Francisco local government study, I’d argue that the smarter open source strategy would be innovation, not substitution.  Focus your open source efforts on new deployments for unmet organizational needs, and let them grow.  Don’t spend all your time trying to replace existing proprietary software that ‘works’.

Update on the open source ‘miracle’ paper:  It has been accepted at the HICSS 2010 conference track on Open Movements:  FLOSS, Open Content, Open Access and Open Communities.  The paper now features the more sedate title “Open Source Deployment at the City and County of San Francisco:  From Cost Reduction to Rapid Innovation”.  The reviewers were enthusiastic, but still a bit skeptical that open source really could deliver such rapid innovation, and customer delight, under very challenging circumstances.  Hopefully other researchers will take up the challenge to help prove me right (or wrong).

A version of the open source innovation paper is available here, but the copyright now belongs to the IEEE.  Let’s hope they use it wisely.

For most of June, I was a visiting researcher in the Informatics division of the Swedish Business School, Örebro University.  Thanks to the group’s rigorous, precise daily schedule of lunch times and coffee breaks (fika), I had the structure I needed to be productive!

I gave a talk on my open source deployment research in San Francisco (talk title:  “Open Source Deployment at the City and County of San Francisco:  From Cost Reduction to Rapid Innovation”).  Open Source (or öppen kod, as they say) attracted good interest, particularly with the eGovernment angle.  I also served as an external examiner on Ella Kolkowska’s licentiate thesis, which focused on a values-based approach to IS security.  As Örebro, along with many other European universities, makes the transition to a more American-style system of rewards based on publications, strategies for cracking the right journals were another hot topic.

I’ll miss having the intellectual stimulation of Karin, Ella, Fredrik, Johan P, Johan A, Kai, Nena, Hannu, and the entire group on the rest of our sabbatical world tour.  Thanks Örebro!

In partnership with the Department of Technology at the City and County of San Francisco, I’ve been studying the deployment of new applications using open source.  We looked at three projects (including the recoverysf.org site), and found the people involved reporting results that seemed almost too good to be true:  working systems developed and deployed in record time, at very little cost, that increase the skills and importance of locally-employed IT talent.  It seems like an ‘IS miracle’, especially in these tough times.

While cost-cutting is often seen as the big selling point for open source, faster ‘time-to-market’ for new applications is the focus of our results.  To quote one of our respondents:

“Now, you can stand up that site in 8 hours, four hours, have it done by the
end of the day.  It flips the customer out.  Really?  I thought it would take 2
months.  Being responsive is huge.  Technology changes fast, but business
requirements change faster.”

For example, there’s no way the recoverysf.org site would have been built in less than three weeks without using an open source platform (WordPress), and without a new technology mindset.  It’s a mindset where the IT department sees itself as offering great new products for its diverse customers, rather than trying to control and constrain them all the time.

The paper with our preliminary research results is currently under review.  Please contact me if you’d like a draft.  Special thanks to the City’s CTO Blair Adams, USF MBA Dave Geller, and all of the study participants.

The open source platforms I admire the most, and are most useful for business, are the ones with the largest communities behind them.  For open business platforms, it’s not just the contributions to the core software that matter.  It’s the number of extensions (or modules) and the number of themes (or styles) that’s critical.  Having many extensions and themes to choose from give business users the best of both worlds:  a standard software package, and lots of easy customization possibilities.

I presented some preliminary research on open source communities for business platforms at OSCOMM 2009, the first international workshop on open source communities, held after OSS 2009 in Skövde, Sweden.  Our data shows that the best supported platform with the highest number of community-contributed extensions is WordPress, followed by Joomla, phpBB, MediaWiki, and Drupal.  Moodle, SugarCRM, Elgg, Magento, and Gallery are the next five, with not much in the way of community contributions after that.  Only award-winning open source software packages that a business user would directly interact with were included.

The paper and slides are available at the OSCOMM program website.  The paper is called “Community Building for Open Source Business Applications:  The Core-Extensions-Theme Pattern”.

I gave two Instructional Technology talks recently, one formal and one spontaneous.

The formal one was a presentation to our USF Wikis and Blogs group on “Pro Blogging for Dorks Academics”. I talked about the main tradeoffs I see for academic blogging:

• The personal vs. professional balance
• The person vs. topic focus

I also talked about why academic blogging has been valuable for me:

• Self-expression – the usual reason to blog, but looking back over a year’s worth of posts, categories, topics and tags gives you a new perspective on your interests. Writing practice doesn’t hurt either.
• Republishing – I’ve had at least one example of a conference paper that was translated and published in a journal, after it was found on the blog.
• Connections – I’ve met people with similar interests, both academic and professional, that I wouldn’t have found otherwise. My biggest success was finding the designer of a web site I had used in a teaching case, with zero additional work on my part. I posted my case, and within 24 hours the designer, who I had never met before, left a comment with great behind-the-scenes information about the site.
• Media – I’ve had reporters find me through my blog, which can be good or annoying. I showed a recent example where my analytics data told me a reporter had spent almost half an hour on my blog before contacting me for an interview. That information gave me confidence this person was serious, and worth giving some time to.
• Explaining publications – hey, academic publications don’t make sense to a lot of people basically everyone, including people who might be able to use your findings. Blog posts give me the chance to explain publications and presentations in somewhat normal language.
• The living CV – I’m not sure if this is good or not, but I seem to do many things that wouldn’t make it to my academic CV.  The academic blog is a place to capture those. It’s great for those end-of-year reviews, promotion cases, and to quickly introduce others to my work and interests.

Two days later, I gave a spontaneous show-and-tell on my Moodle site (university.jpedia.org) that I use instead of our official Blackboard product. USF is considering Moodle as an alternative, but it will be another classic example of selling the unfamiliar benefits of open source to an institution that has spent serious time and energy on the proprietary path. In industry, open source can sneak in the back door on new projects, and gradually take over from within. Ripping out the existing system is a tougher sell, without some vision of the long-term innovation benefits.

Our teaching case, using WordPress as a simple content management system for small business, has been nominated for the Best Teaching Case Award at the 2008 WITS Technology Instruction in Business Curriculum Competition.  Nice!

The teaching module, “Instant Websites:  Using WordPress as a Content Management System”, is now available:

• The “Instant Websites” Teaching Case Executive Summary
• The “Instant Websites” Instruction Module
• The “Instant Websites” Teaching Note

Slides from my talk on “Web 2.0, Open Source, and the Mass Production of Knowledge:  Why Collective Platforms Might Hold the Key to Understanding a Knowledge-Based Economy” are now available.

Thanks to the USF Faculty Development Committee for supporting my research this summer.

While WordCamp 2008 attendees were likely impressed with the huge number of page views (6.5 billion per year – roughly one for every person on the planet) and monthly unique visitors (up to 160 million per month) being racked up by wordpress.com, I was focused on a different number.

2,604,288. That’s the number of people running WordPress blogging software on their own websites, with their own web hosting. You’d think that only a hard-core techie fringe would choose to pay for their own web hosting, and deal with the geekiness of it all, when they can get WordPress for free on wordpress.com.  But, as of this morning, 3,870,299 blogs were running on wordpress.com.  That’s a close race.

In other words, the do-it-yourself web crowd is looking mainstream, not fringe.

The one-click install revolution on web hosts has made this possible.  The amount of software/web services power at your disposal with today’s inexpensive web hosting is ridiculous.  Take a look at a typical menu of open source software choices (this one is from Simple Scripts).  Blogs, wikis, forums, serious content management, e-commerce, CRM…often the best software in its category.  We know people are using install scripts, because of the growing number of blogs that are launching with slightly out-of-date versions of WordPress.  (Script services are often behind the latest version, one of the downsides of using one-click installs vs. slapping it together by hand.)

Not all is perfect in one-click install land.  Upgrades and backups are nowhere near as painless as getting started.  But it’s been good enough to compete with free, and it keeps hope alive for a more open web future:  not everything has to happen through Google, Yahoo!, MSN or even wordpress.com.

I’ve developed a teaching module that helps students start to create a simple business web site using WordPress.  The students launch a new site on a web host via an install script, come up with a simple category structure, and download/upload a new theme.

As an example of a business WordPress site, I use nextbusnews.com.  NextBus is the groovy technology that tells me real-time how late my next MUNI bus will be (more details on how NextBus uses WordPress as a simple content management system here).

It amazes me that only 36% of US small businesses with net access have a web site (as reported in the Wall St. Journal last week).  This is 2008, not 1998!

Is there an opportunity for WordPress to become a kind of generic small business solution?  Business sites can be done now, of course, with some tweaking and geeking.  But, following the analogy from Stephen O’Grady’s talk at WordCamp on Saturday, perhaps someone needs to build a company on top of WordPress, in the same way that Google builds its services on top of open source software.  A small business website service built with WordPress, but where 99% of the users don’t even know what WordPress is?  Edublogs for small business, but maybe without even using the term ‘blog’?  Is this a good idea?  Is somebody doing this?  In the meantime, we think there are lots of good reasons to teach students about open source business platforms and basic content management via WordPress.

(I’m going to wait on an official release of this teaching module until after I hear from reviewers at the WITS 2008 Technology Instruction in Business Curriculum Competition.)