For the last couple of weeks I was exploring how to give our Enterprise 2.0 initiative a fillip. This included exploring the Media Wiki engine, the open source software which runs Wikipedia. We installed Media Wiki and configuring it to allow access to only logged in users. This involved some tweaking to the PHP code as well as Apache configuration. We also did a sample transfer of existing content to the Wiki. I will share the technical details of what we did in a later post.
This post is about 2 realizations I had. Both are counter-intuitive and negate some long held views on integrated environment and structured data. I can summarize it by saying that for Enterprise 2.0 to succeed …
- No Portal – Do not attempt unified access to everything – opt for loose coupling over tight integration
- No Structured Data – Do not try to structure knowledge and store in RDBMS – opt for unstructured or semi-structured content and for flexibility
Let me try to justify these statements.
What is the primary reason behind the success of the web? I think, it is the ability to have hyperlinks – the ability to attach a URI to a specific page and link it to anything and allow access from anywhere. It provides freedom, supports diversity and promotes innovation. Web 2.0 and by extension Enterprise 2.0 takes this one step further through the architecture of open participation. On the other hand, a portal …
“… provide a way for enterprises to provide a consistent look and feel with access control and procedures for multiple applications … (wikipedia)”
In an enterprise portal, the emphasis is on uniformity and control – which not only goes totally against the philosophy of Web 2.0, but also goes against the very grain of why web is successful.
So, my recommendation … if you are really interested in Enterprise 2.0, stay away from portal software.
No Structured Data
The conventional method to allow user to search knowledge repository was to classify pieces of knowledge, create a taxonomy and store it in an RDBMS. Unfortunately, this method is too rigid and is totally against emergence, which is key to success of Enterprise 2.0. Meanwhile, search technology and machine power seem to have advanced to a level where searching any amount of unstructured data looks quiet feasible.
So, it makes perfect sense to retain the knowledge in documents, web pages, blogs and wikis and let search engine handle it.
Do I hear you say “what took you so long to realize this?” Yes, I agree that I had been blind, but better late than never!