Let's end document centric thinking and move to functionality based thinking

I started this post a while ago, but today I had a brief discussion about content management so that pushed me to finish off this rant and post it:



Why is it that we are still dealing with Word .doc files and WordPerfect files and Excel files...

Shouldn't we be dealing with a help document file, or a resume file, or a numbers comparison file?

I got to thinking about Microsoft's next big release and how the operating system will include/leverage the database engine so you don't have to worry as much about where things are stored (be it files or contact info or emails etc) and you just need to state what type of information it is and then later on you will be able to retrieve it.

I have had a project idea on my to-do list for a while where I combine my Outlook emails & contacts with my doc files in my folders and all of my bookmarks and have one way of getting at that information versus me trying to remember - so did I save that attachment to a folder or is it still in an email or was it even an attachment or was it maybe just on somebody's web-page?



We should have to care about where the information is, we just need to know what it was to find it again.



Recently I was given to files to review from an organization. I got confused because they gave an xml file, a Word doc file, and a pdf file with essentially the same info. Mainly because they didn't know who would have what application to read the information.

While that was a nice thought for them to do that, it exemplifies the problem that nowadays too many companies waste time repeating the same functionality.



Do I care if company A stores their data in a more elegant storage structure than company B? Not a rat's ass.

Do I ignore company C if their files are 20% bigger than company D? I might not care if it was 2000% bigger - I would probably focus on the functionality of the application and how easy it is to use above all else.



So the next question is: how can company's use the same file for different purposes?

I am not seeing the world as one mother of an xml file. Some scholar can take this the next step and toss the file concept and we apply the principle of getting at data through interfaces to a generic data store. But for now different information can be stored in different files.

What I am saying is that we have Excel spreadsheets for storing numbers, and then somebody copies a page of those numbers into another doc because other users are not allowed to change the file and they don't know how to use Excel to get at the data they want. Somebody else exports a graph and makes it into a pdf for easy viewing.

And there is a different spreadsheet for each department and each month of the year.

And there is an accompanying Word doc that describes how to use the file.



How come we don't just have one file that has defined interfaces on what data the user wants to see and how the user is going to see that data.



Shouldn't a software company's focus be on the functionality, not how they store the data?

Gone are the days that it was company secrets how the information gets stored, complex file formats that prevent competitors to get at that information.



Nowadays:

1) if a company wants to block anybody from getting at that information - they can. Security algorithms easily solve the challenge of hiding the information

2) a company's strengths is normally how well they SHARE the data, with other systems and how things can get automatically updated in or out. the exact opposite worry of how to hide data so nobody can see it

3) the company strengths is in how to make sure the end user can use the data easily.



If one person should only see a part of the data, that is possible.

If another person should only be able to update part of the data, that is possible.

If another person wants to see a different view of the data, that is possible.

If another person wants to update the data from a different data source, that is possible.

If another person wants to quickly/automatically jump to a section of the data, that is possible.

If another person wants to include other references to data, that is possible.

If another person has a lot of data to include but wants to compress it for storage reasons, that is possible.



So if all of this is possible, and just about all companies can do ALL of these things - why is it that we still use different file types for different reasons?