As many of you may know, I've been working at establishing The Court, a new weblog at Osgoode Hall Law School that focuses on the work of the Supreme Court. In the course of editing contributions I've come to realize how handy it would be to be able to link over to a designated paragraph in an SCC judgment, para numbers being the points of internal reference for online judgments since there are no page numbers, of course.
Unfortunately, in the official judgments available online, the paragraph numbers are not also html internal anchors in the way that they are in LexisNexis/Quicklaw.
Everyone is familiar with the URL hyperlink by now: this is the way in which you make a chosen piece of link text become the open end of a magical space-time wormhole that leads directly to elsewhere. Thus, for example, the phrase Slaw, when surrounded by these tags <a href="http://www.slaw.ca">Slaw</a> becomes the hyperlink Slaw. If you want to link to a particular spot inside a web page, rather than to its beginning, there has to be an internal or named anchor (the "a" in "<a href", or hypertext reference). So, if I wanted you to be able to jump to the second paragraph in this post, I would put there an anchor I could then link to, e.g. <a name="para2"></a>, with some link text such as second para. All of which is to show you how easy it is to structure an internal linking point.
A couple of years ago some bloggers mooted having a named anchor on every paragraph with a link to it appearing as a # or a ¶ at the end. These were called purple numbers or plinks, for paragraph (or purple) links. They didn't catch on for various reasons, though you can see them in action on Tim Bray's site, Ongoing, when you mouse over a paragraph. The idea was to let you direct your readers to a particular paragraph in someone else's material that you found interesting or otherwise noteworthy. Now, while this precision might be overkill for simple blog entries, it's practically of the essence for judicial decisions.
(Although, interestingly, the United States Supreme Court doesn't number the paragraphs in its online judgments. The judgments released by the court itself are in PDF, where there is pagination and it is possible to create links to interior pages of PDF files; and those in HTML — on FindLaw, for example, have neither paragraph numbers nor pagination. I find the Canadian approach superior here.)
One objection might be that it would add to the "weight" of a page, i.e. the size of a file, making it slower to load. I doubt that the addition of small amounts of code would have this effect. And besides, the LexUM judgments (and those at CanLII, which are identical) are so full of "cruft" (unnecessary code) that LexUM clearly can't be specially concerned about file size. (The judgments are essentially Microsoft Word files saved in Word as HTML documents, a notoriously ugly form of coding.)
Although it's always a pain to have to modify a system, I think it would be relatively easy for LexUM to incorporate internal anchors at paragraph numbers. Indeed, I should imagine that a simple script might enable them to modify all the cases in their database to do that for past judgments as well. It would be a valuable improvemen.
Simon,
I too would love to see internal paragraph linking references in web delivered judgments. Even if it was done on a going forward basis it would be more useful than no plinks at all. I am sure that the 'wired' judiciary would appreciate this as well.
Imagine a judgment from a provincial jurisdiction with a plink after each quote.
Mind boggling usefulness!
Cheers, Shaunna
Usability of legal documents (information) – and today for many users of the documents it means usability on the web – seems to be an area that little attention is paid. In our country (Estonia), legal acts have been put on the web, but with low real usability (no tables of contents, no links to critically important related law, search facility that allows only exact match searches of legal act titles), the much-touted e-government in reality is much less glamourous. Some people have tries to alleviate the problems by writing scripts. It works, but only to some extent. Governements have to start thinking about usability.
Personally, I'd like to see FF or IE will come up with a software solution that allowed page positioning for any link – perhaps expressed as a percentage of the page length. The link could go to "page.htm#%=95.6" to land at the 95.6% mark of the document's length (or close to the end of the page). And maybe browsers could display this deep linking information within the status bar at the bottom of the browser?
Hey! that's not bad! … Copyright 2007. Steve Matthews. All rights reserved. Send royalty cheques to stm@cwilson.com. :-)
Steve, that's a good idea. I've been searching for a way to do that now, but to no avail. There's a favelet that will identify all the ID's in a document, but that doesn't get you much further. I think your notion of striking at a point down the page might be a good way to do it, perhaps using em's rather than percentages, or, ideally, number of characters past the first — if there were an easy way to calculate that, and there probably is — so that it wouldn't matter if you were on a browser or a PDA etc.
Anyone out there who can come up with this?
an interesting possibility, but not portable. Going 95.6% down the page at CanLII will give you a different location than at the DOJ for the same document… or in a PDF…
Michael, the concept would be a software solution, and not document/content driven. We could deep link into a webpage without using an in-page anchor. More to do with the software resolving link destination positioning, rather than the original content & context of Simon's post.
Also, as we've noted before, this type of positioning is already possible in PDF documents.
Speaking of pdf, a recent press release (Adobe to Release PDF for Industry Standardization) may be of interest to people.
Assuming an original text can somehow be turned into nice xml, it shouldn't be too difficult to use xsl to transform it into either html or pdf with anchors inserted in appropriate places, for example, one for each paragraph.
As to getting nice xml out of standard wordprocessing software, the ODF and OOXML saga is relevant.
[...] Crud, yes, but 'cruft'? Another word that was new to me this week, found in The Guardian: 'Fill databases of valuable customer information with rubbish, to let the valuable names hide among the cruft.' My big Collins dictionary (old-tech, paper and a weight), the British National Corpus and Cobuild all drew a blank (apart from the inevitable dog show), but Webcorp led first to a site which, in February 2007, still felt it needed to use quotation marks and explain the word in brackets as 'unnecessary code'. Then the links led to the ever-useful Wikipedia, where we learn: 'In hacker jargon, cruft describes areas of something which are badly designed, poorly implemented or redundant'. There's an interesting suggested etymology from Harvard University's Cruft Laboratory: 'if the place filled with useless machinery is called Cruft Hall, the machinery itself must be cruft'. The links goes on, as the page reminds me of the term backronym – something for another time? [...]