Plain Text Hypermedia

linking to web pages has become a universal way of pointing to relevant information. in most cases, it is done by linking to a complete page. but html supports also supports links which point to just a fragment of the html page, for example by pointing to .../netdret/#vcard, which is only the vcard part of my home page. the part after the # is called a fragment identifier, because it, surprise surprise, identifies a fragment of the referenced resource.

html fragments are not heavily used in web links, mostly because they can only point to fragments that have an identity, which must have been assigned by the author of the html page (using something like <p id="bio">). current browsers do not do a good job at telling you which fragment potentially could be pointed to, they hide that information from users, so that you actually have to look at the source to figure out which ids are available in an html page.

surprisingly, and even though nobody ever uses it, pdf supports fragment identifiers, too, .../wilde-ht2005-textfrag.pdf#page=2 identifies the second page of a paper about fragment identifiers. so, fragment identifiers are useful, because they allow you to point to something more specific than just a complete resource.

we are currently working on a way to identify fragments within plain text files, which lets you point to fragment identified by character or line counts. there are three main scenarios where this can be useful:

1. if you have a bigger plain text file and want to be more specific (look at this part of the document), fragment identifiers allow you do to that.
2. if you want to point to a particular location in a plain text file (insert your text here), fragment identifiers allow you to do this as well.
3. in a more elaborate scenario, you can even attach links to plain text files by storing the links individually and associating them with a fragment of the plain text file; this effectively lets you create hypermedia using plain text files.

of course, this will only work once browsers start supporting the new standard, which may take some time. but fragment identifiers degrade gracefully (one of the main principles of web architecture), and the latest version of amaya already supports this feature. so if you look at .../ht05-abstract.txt#line=5,10 using the latest version of amaya, you will actually see a fragment of the plain text file being highlighted. pretty cool.