[IAEP] Article on "Why files need to die"

[Chris Leonard]

On Fri, Jul 15, 2011 at 8:32 AM, Dave Bauer <dave at solutiongrove.com> wrote:
>
> I think search is the answer. There's no reason why a hierarchal
> categorization can't be one of the wants to access information, but it
> certainly isn't the only way. I used recent documents feature very
> often, I usually search for downloads in my browser instead of opening
> the folder where they are automatically stored. I send myself email to
> my gmail account for anything I want to remember (except passwords)
> and rely on the search feature to find it again.


I agree with some of the sentiments tha Dave expresses.  My expereince
is in running enterprise systems for pharmaceutical research.  In that
industry documents are created by large teams of people all referring
to a compound, or group of related compounds which have code names
(that often change through-out the process of development).  Often the
most important element relating two documents is a common bit of
chemical structure (e.g. a pyrrolidine ring).  This differs from the
chronological sequencing on the Journal, but bears some relationship
as a criteria other than files/folders as key organizational
principals.

Hierarchical files systems fall short of having sufficient metadata to
enable full retrieval of all relevant documents in that setting.
There are software companies pursuing Web 2.0 enabled approaches to
enterprise document / content management, in particular, my good
friends at ArtusLabs developed a system where you could find documents
/ spreadhssets, etc. (of nearly any type) by performing a chemical
substructure search (e.g. find me all documents that contain a
chemical structure with a pyrrolidine ring).

Full disclosure - I am a member of the Scientific Advisory Board of
ArtusLabs and have benefited financially from that relationship, most
recently when Perkin Elmer bought them out.

cjl