Built-in variables not properly secured after redefinition trial

[Former user]

Originally reported on Google Code with ID 135

What steps will reproduce the problem?
1. 1+2   [3]
2. ans   [3]
3. ans=5 [correct error message]
4. ans   [0]

What is the expected output?
[3]

If we use pi instead, in the end [5] is the result, but 3.14... from the
second usage.

A similar situation happens when doing something like

   1+2   [3]
   sin=9 [correct error message]
   ans   [9] (or 0, can't remember)

Reported by helder.pereira.correia on 2008-01-10 15:41:02

[Former user]

``` Fixed in trunk (revision 971), will be available in 0.10. ```

Reported by `helder.pereira.correia` on 2008-01-10 15:42:29 - Status changed: `Fixed`