Re: [Maxima-discuss] Influencing the variable sort order

[Robert D. <rob...@gm...>]

Wolfgang, thanks for your interest in Maxima, I'm glad to hear that it
is useful in some way for practical engineering problems.

About variable ordering, it is not altogether easy to get variables
ordered in a way that seems obvious. For the most part, I accept the
default ordering and try not to worry too much about it. I know that
this is not a very satisfying approach.

That said, try declare(ε, constant) before anything else. I find that
makes ε go before other variables in expressions. Possibly declare(ε,
scalar) would have the same effect.

Something that may shed some light, here's a relevant passage in the
description of orderlessp:

     The canonical ordering of atoms (symbols, literal numbers, and
     strings) is the following.

     (integers and floats) precede (bigfloats) precede (declared
     constants) precede (strings) precede (declared scalars) precede
     (first argument to ‘orderless’) precedes ... precedes (last
     argument to ‘orderless’) precedes (other symbols) precede (last
     argument to ‘ordergreat’) precedes ... precedes (first argument to
     ‘ordergreat’) precedes (declared main variables)


Hope this helps,

Robert