Blogging, Wikis and Snippets - A new approach to PKM
We all know what blogs and Wikis are, but what the heck is PKM? The
acronym means "Personal Knowledge Management", and it describes a new
category of software, bigger and more dynamic and interesting than old
fashioned PIMs. I have my own PKM dream, and that's why snippets are
important.
For those rare souls who don't know what I'm talking about, some
definitions. Despite being relatively new in the web scene, blogs are
very popular now, and Wikis, albeit a little less known, are also hot
with the hardcore Net community. Both are exceptionally easy to use
tools to manage information. Blogs allows for non-structured, personal
information to be published chronologically, which is the logical
approach for something done for we humans to use; that's the "my
diary" of the Net age. Wikis are perfect to structure arbitrary
networks of information in a fairly straighforward way; they allow any
user to write about concepts, and the relate them between each other,
doing things that no physical card filers could ever do. You can jump
from info to info, define new terms, and relate everything with a
minimum of syntax required.
Having said that, what is a snippet? Well, a snippet is just a small
piece of information. A blog entry can be a snippet. A Wiki entry can
also be a snippet. Anything is a snippet. The snippet is the smallets
block of information in my dream's PKM.
What I have in mind is a fairly flexible framework, where snippets of
information can be stored and structured in different ways, depending
on the way you like it to be. Basic to the system is the snippet
store, where all snippets are kept. To keep things easy, the design of
the system must no be particular about the nature of the store. I'm
thinking for now about using the filesystem, pretty much like simple
Wikis do. One file per snippet, that's all.
The snippet server just takes a snippet, formats it accordingly and
presents it. It's a small HTTP server, one that knows at first very
little about the snippets. And that's where things start to get
interesting.
Every snippet has some little metainformation about it. Some of it can
be inferred indirectly by looking at the snippet itself. For example,
file extension or contents can store information about the type of the
snippet. For example, text files can be identified by the extension
(.TXT, .HTML, and so on); the first line of the text file can also be
used to infer some of the information; graphical images can be
detected by the file extension, so.GIF or .JPG files are easily
recognizable. When the snippet server is about to serve a snippet, it
looks at the snippet type, and than act accordingly.
Up to this point, nothing is different. But snippets can be related to
each other using, and classified, forming different things depending
on the way you look at it. Snippets can be used to keep a blog, using
chronological metainformation to order them. Snippets can also be
structured as a Wiki; in this case, the snippets themselves can
cointain the internal links, or crossreferences, needed to keep the
Wiki working.
Depending how do you refer information in the Snippet server, you can
have either a Blog, or a Wiki - or both! That's the idea, but I'm
still working out the details... But I hope to be able to have
something working fast!
