This is Part1 of a two piece post about my dotCom Idea about overhauling and web2.0-iszing the traditional Contact database.
One of the reasons I am posting it is purely because you can already implement my idea using existing free online tools and services.
So the value proposition of the idea is effectively zero.
[I will cover how to implement it in the second post]
All dot com ideas are solutions to a problem that irks the original developer.
Yahoo exists because there were no indexes on the early web.
Google exists because most early search engines sucked.
My idea stems from the folowing easy to understand list of facts;
- Contact databases suck!
They really do.
And the reason they suck is that they have basically not changed in the last 70 years!
Ok. They went digital, but they are nothing more than electronic versions of the traditional rolodex.
What has changed is the type and amount of information we associate with our contacts.
Something that contact databases have not kept apace of.
A pertinent example is that I oftentimes use IRC. Now, some of the users of the channels I frequent I do not know their real names [IRL] but I still want to associate information, email, websites, birthdays and the like, with them.
With a traditional contact system you end up with a mess.
In fact it is easier to use post-it notes or a Moleskine to keep such information. [Or even, gasp, simple text-files!]
so I have spent a lot of my spare time trying to work out what a contact database system for web2.0 would actually look like.
How would it tie together the disperate pieces of information that reside in my mind when I think of a contact?
The answer to this is tags. LOTS of tags!
Tags would allow me to give a basic context to the disperate pieces of information.
For each “contact” there would be an unspecified number of “data entries”.
Each “data entry” would consist of “a chunk of data” and a “user defined tag” to help identify what that data represents.
For instance, using the traditional Contact paradigm, there would be a “name” tag with the persons name contained in the “chunk of data”.
I immediately noticed that there has to be a default tag. One used for displaying, and thus indexing, on a list page.
But I decided to make that tag flexible.
i.e. YOU choose the “Chunk of data” that you are using to represent the contact.
You would be able to do pretty-neat searches across the dataset. How many people have the “IRC_name” tag. Whos country is “UK” etc.
Then I stumbled on the idea of having “Meta-tags”.
This is where the “chunk of data” is a set of tags itself!
So say you have “FOAF” as the “chunk of data” tag, but the “chunk of data” contains the normal FOAF relationships, “friend knows” etc.
So with this second layer of tagging I sat down and drew out the database table structure I would need to support it.
I even started playing around with databases, and installed postgress on my Laptop.
But then, today actually, I realised that you can do almost all of this in free and easy to use tools on the web!
… cont in part 2