The problem with blog footnotes

February 7, 2010

Here is something that has just occurred to me (courtesy of seeing an example of it): footnotes are hard to do well in blogs, and may need actual software support if you want them to be completely correct.

The conventional way of doing footnotes in HTML is to use fragment URLs and anchors, with the footnote text at the bottom of the entry and your choice of footnote markers in the main text. But, like anything involving anchors, this means that you need to come up with unique anchor names.

On one level this is no problem; you can just use 'fn:1', 'fn:2', and so on. But on another level this is a problem for blogs, because blog entries are repeatedly aggregated together with each other on web pages. When you put multiple footnote-using entries on the same HTML page, you need all of their anchors to be unique; you are not likely to get this if you use 'fn:1' style anchors. (This is especially pernicious once you start considering syndication feeds and 'planets', that put content from multiple blogs on the same HTML page.)

You can just punt on the issue and say 'well, it's up to the author to come up with unique anchor text (ideally globally unique text)', but in practice people won't always do this and this is equivalent to having non-functional footnote links under some circumstances.

Admittedly, I suspect that most people won't really care about all of this, and will be perfectly happy using 'fn:1' style links and having them not work. Regardless of whether the actual links work, your intent is likely to be pretty easy for users to follow.

(And who knows, maybe the proper implementation of footnotes in blog entries is pop-up alt text, like xkcd famously does on the comics images. Alternately, footnotes are a printed thing that are not appropriate in HTML.)

Comments on this page:

From at 2010-02-07 10:04:11:

For blogs, I'm not convinced footnotes are relevant. In print, I generally see footnotes use to give citations. On a webpage, that's just as easily accomplished with a direct link to the source in many cases.

Ben Cotton

By nothings at 2010-02-07 11:44:56:

These days when writing long HTML documents I've ended up using a pretty arbitrary convention: I provided an unlinked numbered footnote in the text, and then the footnote itself appears as the very next paragraph after the current one (in <small></small>).

This is kind of totally wrong compared to how they work on the page in books; but then again the division of books into pages is itself arbitrary (and notoriously problematic with long footnotes), and the feeling when you get when clicking a footnote teleports you a long way down in a document, even if there's a back link, is enough to make me feel like this is actually a better way, not a cheap hack.

(Well, maybe they should still be linked, I'm just too lazy to bother.)

By cks at 2010-02-08 12:40:37:

I should have clarified this: I'm thinking of footnotes used as asides or extra commentary, as opposed to footnotes-as-citations. Citations can just be direct links on the web and everyone is used to that usage, so there's no problem with handling them.

(Indeed, the explicit footnotes I've seen in blog entries tend to be of the aside form, eg here and here.)

From at 2010-02-09 01:59:05:

Those examples are just bloggers being wanky. Don't Use Footnotes In Blogs!

By cks at 2010-02-09 16:41:04:

Well, my view is that you need some way to handle asides; footnotes are one of the traditional approaches. I don't think that well done footnotes are necessarily worse than any of the other approaches.

(Embedding things in parenthetical asides is the one that I use (or abuse), but it can somewhat take over the main text, as I periodically demonstrate.)

From at 2010-03-16 20:45:40:

I think these asides are a moderately futile attempt at emulating print output with HTML. While it's an interesting presentation hack, it feels like something that defeats the entire point of wildly interconnected and hyperlinked text. This way instead of being able to keep reading with the "flow" of the text itself:

iPhones and iPods Touch in use today: 50-60 million. Apple reports that it has sold over 70 million, but I’m assuming that not every device sold is still in use.

One has to jump around just to get at a small bit of text that is interestingly related to what one is currently reading. This feels wrong, but then again it's just my own personal way of looking at it.

Then there's the case of the hyperlink footnote. When one sees text like:

iPhones sold in the first year (Q3-4, FY 2007): about 1.4 million(2).

and then has to scroll down to the end of the page to see that the footnote merely lists a URL. I often prefer seeing the same text written with the link right there in the middle of the text it refers to:

iPhones sold in the first year (Q3-4, FY 2007): about 1.4 million.

I don't think it makes sense to try to emulate a printed page of output by using fancy tricks. The web is a vastly different medium. It seems both pointless and wrong to try to shoehorn it into a peg that looks like a printed page; because it really isn't exactly what a printed page is.

-- Giorgos Keramidas

Written on 07 February 2010.
« Why a laptop is not likely to be my primary machine any time soon
A thought on deliberately slow disaster recovery »

Page tools: View Source, View Normal, Add Comment.
Login: Password:
Atom Syndication: Recent Comments.

Last modified: Sun Feb 7 03:06:17 2010
This dinky wiki is brought to you by the Insane Hackers Guild, Python sub-branch.