Mick Moignard's Blog

I followed a link from Ed Brill to a blog post about Sharepoint.  Interesting discussion of how Sharepoint more often than not fails to hit the spot, and why.   How the implementation, and more importantly, the expectations of that implementation are what really matters.  Being sure that the organisation understands what it thinks it's it's there to do, and how that often fails to translate into reality. And how the costs can just go out of the window.  To be fair, quite a number of the criticisms and issues can equally apply to Quickr too, because most of them are not about the technology per se, but how enterprises totally fail to understand what they are trying to achieve and why.

There's a bit of a buzz going on about this blog post too; Mike Gotta has some perspectives, as does Oliver Marks.  

One major issue that I see is the one about how you want to hold and disseminate content - and I feel very strongly that the answer to this comes from how the intended audience will want to get at that content.  I don't think that they want to have to deal with files, and the applications that work with files, such as Office and Symphony.  I think they want to be able to find the content and view the content without having to worry or deal with how that content is stored.

For example, if I want to find something out, my natural reaction is to reach for a browser and type a search into the search bar, and then browse the links returned; each of which, more often than not, opens a page that I can review, read or discard.  The target of that search on the web is the most complete set of knowledge that there is.  Or I do the same thing in Lotus Notes, but with a much smaller context - a single database or a domain search - but in the same way, I get a list of potential results as links to pages. And Lotus Notes opens those pages when I click the link, just as the browser does.

But if I do a search that involves a file-based medium such as the shared drives available to my Windows file system , or Quickr libraries. I get a list of files.  The file system offers no  meta-data that expands on the files, and so I have to guess at which files may help me. And then I have to wait while some application loads to show me the content - and that application then has to be guided to find the search results.

And that means that information stored in files is, to be blunt, lost.  It's of little consequence or use to anyone who doesn't intimately know that it's there.  To my mind, that, ultimately, is the big issue that such implementations fail to address; it's not how you store the stuff that you should pay attention to, but how the audience that you expect - and possibly ones you don't expect - will want and expect to find and use the material.