eval(a, b) parsed as CallExpr instead of as EvalExpr about es-lab HOT 3 CLOSED

I just noticed that tomvc submitted a change intentionally to add a 
arguments.length 
=== 1 check.  What is the reasoning behind that?

Failing to get a satisfying answer from es-discuss, MarkM and I had another 
look at
this and decided it would make most sense to have the parser generate an 
EvalExpr no
matter the number of arguments passed to eval. It seemed like a strange 
restriction
for the parser to check the number of arguments. Furthermore, the spec does not
mention # of arguments when defining what a 'direct call to eval' means.

Change reverted in r136.

For reference, the relevant es-discuss thread is 
https://mail.mozilla.org/pipermail/es-
discuss/2010-January/010512.html .