Windows File System CMS

By Deane Barker on January 5, 2004

Anyone an expert on the WinXP file system?: Phillip Greenspun proposes using the Windows file system as a database for organizing pictures, instead of some complicated structure that depends on other software.

I’m thinking of writing a tutorial on how to use the Windows XP file system as a photo database. […] the Windows file system seems likely to outlive many of us […] if you create a document using Notebook and click right on the file icon then select Properties you get to a page where a “Summary” tab is an option. You now have the opportunity to edit Title, Subject, Author, Category, Keywords, and Comments fields. These fields are searched if you use the built-in Windows file search mechanism and look for “a word or a phrase in a file”

I’m very much on the same wavelength as Greenspun here, and I’ve actually pursued this a bit. Using the Indexing Service you can get a lot of information about files, and it’s not that hard to re-purpose that meta into something else.

Furthermore, using WMI, you can capture events from the file system — have a script execute whenever a file is added to a certain folder, for instance, or when a file is modified or deleted. There’s a lot of untapped potential in the Windows file system. With enough time, you could make a heck of a content management system out of nothing but XML files in a directory structure.

You can get even more bizarre by using hidden text files to store “properties.” If an “object” or “record” is analogous to a directory, you can store text files in a hidden folder as “fields.” The content of the text files are the values of those fields. Indexing Services will take queries that can make this work. Yes, it’s kludge, but when the core thing you’re trying to organize are files themselves, this may end up being less of a headache than managing them in other ways. Remember that the file system is a…well, a system to manage files.

If you like the idea of expanded meta information for files, It’s the Content is a good system that allows you to define your own meta in an environment that’s an extension of the Windows file system. Not that expensive, either.



  1. That part above storing the value of “fields” in text files was way off — why not just a hidden XML file? A lot of search engines can index XML files so you can do XPath queries into them.

    Thus, you could have a directory which is a “record.” An XML file (a “manifest”) would have all the meta about the record which could be searched, reported, etc. A simple Web interface could let you browse the directory structure and modify the manifest.

    For that matter, you wouldn’t even need an XML file. Have a text file to which you add a bunch of meta using the Properties dialog in Windows. Thus, even an end user could manage it.

Comments are closed. If you have something you really want to say, email and we‘ll get it added for you.