[Linux-disciples] Catching URLs on the command line

[Stephen R Laniel]

On Sat, Jan 17, 2004 at 11:47:14AM -0500, Adam Kessel wrote:
> Very ununixy though. We had a conversation about this about three years
> ago, I think.

Huh. I don't remember this conversation. I'm sure it was
illuminating, though.

> My suggestion would be to have a one letter alias for sensible-browser
> and invoke URLs with that.

This is a good idea.

> There are a million things you might want to do with a URL; browsing to
> it is just one.  It strikes me as against the basic nature of the unix
> command line to have default actions associated with the complement of a
> predicate.  This is more something I would expect from Gnome and KDE as a
> higher level of abstraction from what's really going on.

I see your point, but maybe I've just been spending too much time
with Perl. There are defaults everywhere in Perl, and programmers
just learn to know what the default behavior is when a given
operator doesn't get passed a variable name (e.g., 'm/(.*)/;').
So perhaps not Unixy, but Perly?

I *do* remember a conversation we had some years ago in which you
proposed that there ought to be a 'Perl-Oriented Shell,' or
'POSH'. I still like that idea. Perhaps POSH would handle URLs
the way I want.

-- 
``I get the warm and fuzzies hearing about how standards-
  compliant Free software rocks. I don't get them from hearing
  about proprietary, non-standards-compliant software sucks.

  Make me warm and fuzzy, Shlomi.''
 -Peter Whysall on the linux-elitists mailing list,
  http://shorl.com/banafytrarere