learning + networked society + dossiers + extra
home + what's new + index + comments + rss feed

One of the greatest feats of Apple has been to make computing more human and natural. Apple just published information about the upgrade of its operating system Mac OSX, named Tiger. Apple talk, amongst other things, of using metadata to improve searching on the Mac. The following short article is a suggestion how they could go further with metadata and make the Mac even better and more user-friendly.

The end of file names


File names are quite barbarian: if you have a lot of documents the coded names you use for files rapidly become incomprehensible. File names often give little of no indication of the contents, especially if someone else has created them. What if we were to do away with file names and replace them by something more useful – the title of the document, for example? How could that be done? By associating a document, whatever its format, with a metadata file in XML that indicates a number of standard fields about the document. These should include the title, the subtitle, the author, the source url (if it comes from the web), creation date, languages, authors rights, … Now some document formats contain part of this information, although it is not often used. The finder needs to compare the metadata of the document and the attached metadata file and automatically complete the latter if necessary. At the same time the user needs to be able to add or modify data when necessary. This is especially the case if you are a knowledge worker, like myself, and you work a lot with documents downloaded from the Web.

The beauty of such a system is that you can then adapt the finder to display (according to your preferences) the title of document rather than the file name. You should also be able to include the author’s names and the url, amongst other things. This means that when you look at a list of documents you immediately have meaningful information rather than a jumble of numbers and letters of file names. File names could then even be attributed automatically as they are no longer necessary to identify the file (except for the workings of the computer itself).

On a longer term basis, it ought to be possible to “tag” a text document in a standard compliant way, for example, by indicating which words within the text are the title, the author, the date, etc. rather in a similar way one highlights text in a word processing programme. One could even imagine that certain locked fields indicate questions related to authors rights.

Alan McCluskey, Saint-Blaise.

Share or comment
| More

learning + networked society + dossiers + extra
home + what's new + index + comments + rss feed

ISSN: 1664-834X Copyright © , Alan McCluskey,
Artwork & Novels: Secret Paths & PhotoBlog - LinkedIn: Portfolio - DIIGO: Links
Created: July 10th, 2004 - Last up-dated: July 10th, 2004