On Wed, Jan 12, 2005 at 05:24:50PM +0100, Hubert Feyrer wrote:
> On Wed, 12 Jan 2005, Ben Harris wrote:
> >[ Incidentally, if I were doing this, I'd make seq an alternate front-end
> > for jot, like od is for hexdump. ]
> 
> Either way round, I'd prefer to have the code in tree only once than 
> twice (->"bloat").

Oh, please.  I'm sure we have other things in tree that others (or
even yourself) could consider as bloat.  I don't consider ~450
lines of C bloat.

Could it be rewritten as a alternate front-end for jot?  Probably.
But it would bloat jot.  They aren't exactly identical.  In glancing
at the jot code, I'm not sure how much code could/would actually
be shared.  Each approaches the problem domain (printing sequential
numbers) differently.  I'm not convinced forcing one to look like
the other is a good fit (or idea).

I was just offering up something that I have done and working.
The version I have currently passes the seq tests in GNU
coreutils.  It also seems to act like plan-9 seq.

--
Brian