Re: symbol-at-point

[martin rudalics]

> Try C-x v ~ 1.35 to see the code before my change (or try C-x v = to
> see the diff between 1.35 and 1.36).  You'll see that the previous code
> returned a symbol, not a string.  

Silly me.  I stubbornly decided the previous code returned a string
built from word and symbol constituents.  Apologies to Reiner who
sent me the appropriate entry.