Tricki

This expression:

results in this HTML markup:

Notice that the actual TeX is present in the (value of the) alt attribute. That's what we call "semantic" in the HTML standards world and that's a Good Thing.

However, it could be even a little bit better. Have a look at the way wordpress.com renders this:

$(1) \sum_{j=1}^{i}{j}$

<img src="http://s1.wordpress.com/latex.php?latex=%281%29+%5Csum_%7Bj%3D1%7D%5E%7Bi%7D%7Bj%7D&amp;bg=000000&amp;fg=bfbfbf&amp;s=0" alt="(1) \sum_{j=1}^{i}{j}" title="(1) \sum_{j=1}^{i}{j}" class="latex">

Notice that the TeX code is present not just in the alt attribute (value) but also in the src link itself.

The value of this is more than aesthetic. By replacing the random number (is that a hash of the TeX?) in the link, with the actual TeX, automated tools gain the ability to spot identical expressions. Yes they can do it now via the alt attributes, but operating on the src links seems much more orthodox.

Oh hang on…

From inspecting the two instances of Euler's identity above I see that they both get the same src link so indeed it does appear that it is probably a hash of the TeX code. Can someone confirm. And also can someone please publish the hasing algorithm?