Okay, so that's a little overdramatic, but not by much.
I'm remixing
Lawrence Lessig's
Free Culture by editing a
nice html version to correct the links (and add a ton that weren't there in the first place). The problem, of course, is that the Web addresses suck at being permanent -- you can link to a perfectly valid address on Monday, and have the webmonkeys at wherever it's hosted muck around with the page structure on Friday, and suddenly the dead links are
everywhere. The horror, the horror, etc.
Professor Lessig addressed this issue by having a dedicated
page of links that he mentions in his footnotes, which he can then change at will when they go down. While this works just fine if you're working off of the
dead tree version where you have to drag yourself to an internet connection to look up links anyway, it makes for yet another page to click through if you're reading it off of a screen.
Enter
PURLs --
persistent uniform resource locators. These doodads are basically an easy way to set up addresses that never go bad (well, so long as somebody is there to keep them up to date). Say I wanted to link to a page in my document that has the potential to change its position every once in a while (and they all do); for example:
http://foo.bar/foodomain/foopage.htm
I take this address and plug it into a
purl resolver (basically, the server where the PURLs you create will live), which spits out a PURL that looks like this:
http://purl.oclc.org/NET/foopage
When I put the purl in my document, all it will do is bounce whoever clicks on it to
http://foo.bar/foodomain/foopage.htm
pretty much invisibly unless the user takes a look at the target. Now, say the page I'm linking to suddenly changes to:
http://foo.bar/foodomain2/foopage.htm
If I put the direct link into my document, it's now broken. If I've distributed the document far and wide -- say, as a pdf or an ebook -- all those are broken, too, and the users will have to come back and download a corrected version. A royal pain, to say the least.
On the other hand, if I put the purl in there instead, all I have to do is go find that purl on the purl resolver and modify it to point to the page's new location. All my documents now point to the right place, and my readers (except for the one who pointed it out) will never know it was broken. In combination with the
Wayback Machine, this will even work for most things that mysteriously disappear from servers in the dead of night.
I realize that this is just another version of the
standard redirect, but I (in all my technically incompetent glory) don't have a server to handle my redirects, much less the expertise to set one up. The purl resolver also does a wide variety of neat tricks that I would have a great deal of difficulty replicating on my own -- it makes it much easier to keep track of a large number of redirects, invite others to help me keep my links valid, and provides the ability to construct partial redirects (more on that later, possibly).
On the issue of persistence, I have some misgivings about putting all of my links in one basket, as well, but I figure since the resolver that I'm using is associated with the
OCLC (who are teh r0xxers, by the way), it should be okay.
To date, I've finished converting all the pre-existing links to PURLs, and will start working on hunting down some auxiliary ones as soon as my workload slacks off a bit. You can check my progress
here. If you want to help me out with maintaining the purl links, please go to
purl.org,
register for an account, and then send me an email at sgrice (pada) gmail dot com. ("Pada" is Malay for "at." Curse you, spamming types.)
P.S. I think Lessig might be the only law professor in the world with fangirls.
P.P.S. I'm also on the lookout for a nice stable place to put this version. If you know of one, please let me know.
[via the
O'Reilly XML Blog]
