Re: metamail (was Re: Replying to Hotmail users)
Andrei A. Voropaev <avorop <at> mail.ru>
2003-09-01 14:45:56 GMT
On Mon, Sep 01, 2003 at 10:04:12AM -0400, Derek Martin wrote:
> On Mon, Sep 01, 2003 at 09:37:26AM +0200, Andrei A. Voropaev wrote:
> > If you don't have it then try RFC 1524. Actually there's one good
> > point there. If options 'needsterminal' and 'copiousoutput' are
> > present in the same entry then most likely this is a mistake :)
> > Since copiousoutput means that interpreter will NOT need user
> > interaction and needsterminal implies that interpreter will need
> > user interaction :)
>
> I think that's wrong. copiousoutput implies exactly what it sounds
> like: there will be copious output, which may require, for example, a
> pager. That'll need user interaction... I refer you yet again to the
> mailcap man page.
Well. It might depend on implementation. But the idea is that if copiousoutput
is specified then caller knows that interpretator shall spit some text into
the stdout and let caller decide what to do with it. If needsterminal is
specified then caller knows that interpetator shall expect users input and
because of that caller must provide separate terminal for interpretator and
then simply wait till that finishes. Naturally caller will not receive any
text on stdout in this case because that text will show up in that new
terminal.
So in regards to pager etc this is already callers responsibility to provide
it and handle user interaction to it if it desires to do so. Normally separate
terminal does not provide pager functionality :) on the other hand if caller
is running in some terminal already then it can do paging in that same
terminal without starting another one.
Again, this is the idea how, I believe, the guys who wrote RFG1524 see it.
(Continue reading)