RE: [External] : Re: cond construct for situation when a variable is t

[Drew Adams <drew.adams@oracle.com>]

> >> Right, because one can set the letter t to something which
> >> would then screw up the idea that t means
> >> a truth condition.
> >
> > (setq t 42)
> >
> > Debugger entered--Lisp error: (setting-constant t)
> >   signal(setting-constant (t))
> >   (progn (signal 'setting-constant (list 't)))
> >   (progn (progn (signal 'setting-constant (list 't))))
> >   elisp--eval-last-sexp(nil)
> >   eval-last-sexp(nil)
> >   funcall-interactively(eval-last-sexp nil)
> >   command-execute(eval-last-sexp)
> >
> > And `t' is not a "letter" here; it's a _symbol_ whose name
> > is "t".
> 
> Okay, so they don't allow that anymore. Good, I guess.

Anymore?  It's true at least as far back as Emacs 20:

Signaling: (setting-constant t)
  (setq t 42)
  eval((setq t 42))
  eval-last-sexp((4))
  call-interactively(eval-last-sexp)

You can't even try to get around it by first
unsetting `t':

(makunbound 't)

Debugger entered--Lisp error: (setting-constant t)
  makunbound(t)
  (progn (makunbound 't))
  elisp--eval-last-sexp((4))
  eval-last-sexp((4))
  funcall-interactively(eval-last-sexp (4))
  command-execute(eval-last-sexp)

> Or maybe I did it in another way than using `setq', I say this
> because I remember it very well that suddenly nothing worked.

`M-x report-emacs-bug'