Re: A small change suggested

["Daniel J. Bernstein" <[e-mail]>]

In message <[e-mail]> you write:
> Well. I like the idea of "being liberal in what you accept, and
> restrictive of what you send out".

Blind adherence to that policy produces unnecessary complexity.
Consider, for instance, malloc(0) in ANSI C. As Plauger has pointed out,
callers are allowed to use malloc(0) so implementations have to do the
error checking, but implementations are allowed to return an error on
malloc(0) so callers can't take advantage of it. This is a lose-lose
compromise.

Anders is quite right: the server should be required to send uppercase,
and the client should not be required to understand lowercase.

(Of course, only the first change is necessary to fix this particular
interoperability problem with port 113---not that this is a concern,
according to our beloved group chairman. I have several more serious
incompatibilities up my sleeve [all things which have been discussed
before]; are we going to fix them one by one?)

> Yes, one would have to use a stricmp() or strcasecmp() (or whatever the
> case insensitive string compare is called) instead of strcmp().

What a shame that many of the machines running TAP don't support
new-fangled features like strcasecmp(). Life ain't so simple.

---Dan