Re: [PATCH] * etc/NEWS: Document incompatibilities introduced by record types.

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

  > I'm just really puzzled by all the strong reactions for a problem that's
  > so% hypothetical, 

It is not hypothetical.  I am sure people will define such types in
their code, and get screwed.

                      and really very similar to lots of other "risks" that
  > we don't care about at all.

The similarity is only at an abstract level.

People understand that redefining primitive functions will break
things.

Redefining 'apply' in a careless way will break things very fast.  I
just tried (fset 'apply 'list), and then kill and yank did not work.
There is no danger people will develop Lisp packages which do this
and release them believing they work correctly.

If you define 'integer' as a record type, it will break some things,
but you might not notice them.  That makes it a more dangerous trap.
You could release a Lisp package which does this, not realizing that
it will cause problems for other users.

Since we agree that defining these symbols as record types should not
be done, let's make it easier for programmers to avoid doing that.

-- 
Dr Richard Stallman
President, Free Software Foundation (https://gnu.org, https://fsf.org)
Internet Hall-of-Famer (https://internethalloffame.org)
Skype: No way! See https://stallman.org/skype.html.