Re: builtins, changes et al

Re: builtins, changes et al

[Alan Watson]

> I don't like the idea of a builtin read
> with sh semantics too much. I'd prefer a more general solution,
> some kind of printf - scanf pair which is _not_ builtin. It's already
> part of the GNU shellutils.

The people who have expressed a desire for a built in read (including
myself) have normally used performance as their justification (in the
same way that echo is justified).  Your suggestion does nothing for
that.

I feel that the man page for and implementation of a shell level scanf
will be horrid.

builtins, changes et al

[malte]

Why not discuss changes and/or additions to rc until blue-in-the-face?
I for myself am very grateful for the reponses I received in return to
my last proposal 'cause if I had already implemented it, I'd have to
rework the code now. I don't think this forum should evolve into some
kind of beta test group.

And about reusability, libsh.a (librc.a in this case): As someone else
already stated, there is no easy way to ensure that your shell scripts
uses the same version of some not builtin feature, making it difficult
to pass them on to others. The multics approach stems from a time where
networking was a thing of the future. Who can assure that
gnuerks@macrohard.com.orion won't receive the necessary library module
lightyears after my script?

Now back to the real world, I don't like the idea of a builtin read
with sh semantics too much. I'd prefer a more general solution,
some kind of printf - scanf pair which is _not_ builtin. It's already
part of the GNU shellutils.

Malte