SML Data

I have been building my CMS for a while now. The idea of building a database came from noting that I have lots and lots of content that needs to be indexed. I'd like to put everything happening in my life so that I can possibility find hidden structure (emergence) within the structure. I believe that my multitudes of interests (the leaves) are preventing me to discover what I'm about (the tree).

Coming up with the database structure and figuring out an efficient way of accessing those data with unknown criteria was problematic. I'm not a database architect, most certainly lack the know-how to structure a scalable database architecture that can index everything about my life. So I decided to create my outsourced my CMS and database--free CMS database via user-generated content sites.

It works something like this:

• bookmarks: delicious
• music: acidplanet
• writing: blogger (or wordpress, haven't decided yet)
• photography: flickr
• portfolio: flickr
• flash 'experiments': still looking for a good database
• videos: you tube
• books: librarything
• book recommendations: amazon
• mails: most likely gmail
• news: probably twitter

Utilizing ugc sites to manage my content has many benefits

• My data are hosted on very safe environment, usually with very good backup systems, all for free
• Management of the content can be performed using tools specifically created for the management of those content.
• The CMS on those sites will continue to improve as the next generation comes, and I save on the need to improve my own administrative tools, which can become time consuming
• Accessing data on the services are easy through the their respectively constantly improving APIs
• The application to drive all the content become very easy to host--as the size required to host the large database is no longer an issue

Putting my data access API together is still at the works since the original idea from 2007--but the important fact is that it hasn't stopped me from content creation. One of the key problems in many startup is not having data to play with. Now that all of my 3rd-party content portals have enough data to play with, the development effort has been moving much rapidly since the original idea inception back in 2007.