restart# The result below seems natural: we were taught in school that exp
# being a bijective function we can get rid of it in the equality to
# solve and write simply x=Pi.
x = solve(exp(x)=exp(Pi), x)# But the solution method solve uses is not that natural (and I
# don't really understand it).
# infolevel[solve] := 10:
# x = solve(exp(x)=exp(Pi), x);
# Replacing now exp by some undefined function f produces a
# kind of "no-solution" answer: this seems quite normal because
# not knowing the properties of f one cannot simply get rid of it.
infolevel[solve] := 0:
x = solve(f(x)=f(Pi), x)# Finally replace f by a bijective function with no analytic expression.
s = solve(erf(x)=erf(Pi), x) assuming x::real# It would have seem reasonable for Maple to answer x=Pi, or
# at least it is what I would have done given the properties
# of the erf function.
#
# How can I "force" Maple to "simplify" it's RootOf result to get
# x=Pi?