User blog comment:Hyp cos/Analysis - BEAF, FGH and SGH (part 3)/@comment-24509095-20140626092059/@comment-5150073-20140626121635

In fact, there is a stronger problem. Consider the rule:

$$f_{\alpha}(n) = f_{\alpha[n]}(n)$$.

If it's defined like so, it turns out that $$\alpha = \alpha[n]$$.

There's a trick which allows to avoid this problem. In all our rules, write T(a) = b instead of a = b.