Symbolic Substitution

This page is validated XHTML 1.1 (and announces itself as such): if you doubt me, click the button in its footer. It includes a DOCTYPE complication which, if Ian is to be believed (for he taught me it), should cause various custom character entities to be displayed as appropriate code-points of Unicode. The first column below informally describes the symbol that should be displayed in the other two columns; in the middle column, I use the DOCTYPE's custom character entity; in the right column the code to which the DOCTYPE header says it should be mapped. These two columns (unless Ian was wrong) should thus coincide (they do in Presto).

This stuff matters because:

• The half-way decent author wants anyone reading the document – even if using a different tool than the author used to write it, particularly when actually seeing the raw text of the file – to be able to make sense of the text. Any theoretical physicist shall understand that ℏ means Planck's constant, but it is far from obvious what ℏ means (unless seen via an HTML-aware tool, or by a reader who has personally memorised the entire Unicode mapping); and, when I'm looking at someone else's source file in the wrong encoding (because, oddly enough, the world of computers is full of ways for the downstream program to not know things about a downloaded file, notwithstanding that they were communicated with the download), some random byte-sequence seldom means anything to me, whereas ℏ (in ASCII) is very clear indeed.
• Some symbols have differing meanings in diverse contexts; for example, the code-point known as "∧" that logicians use for and is also used by linear algrebraists as an antisymmetric tensor product operator. It is desirable that the author have the freedom to distinguish the distinct uses of such a symbol, e.g. using ∧ for the logicians' meaning and ∧ for the linear algebraist's meaning, if only so that software – such as proof verification tools – has some chance of knowing which meaning to attach to the symbol. As ever, the crucial point here is that semantics should not be the slave of presentation. My understanding was that XHTML supported this (and it wasn't just an oddity of Presto).

Add to this the existing aliasing – for example, &#295; = ħ is another h-bar but not Planck's constant, &#8463; = ℏ – and you can see why someone who cares about the semantics of a document (as opposed to what it looks like, which is all some authoring tools <spit/> can believe might matter, thanks to being WYSIWYG, a.k.a. What You See Is All You've Got) might actually care about this stuff. I would like the blind (and smart 'bots) to have a decent chance of knowing what I mean, without having to guess based on what they've been told about how things might appear.

It would seem that this does, in fact, work (mostly) in the WebKit-based browser on my tablet; but that the browser lacks needed fonts for Planck's constant and the function composition symbol.

Written by Eddy.