🚀 go-pugleaf

On Saturday, in article
     <Snews.960817.085356.chris.1442@keris.demon.co.uk>
     chris@keris.demon.co.uk "Chris Croughton" wrote:

CC> In article <840195111snz@wash.demon.co.uk>
CC>            john@wash.demon.co.uk "John Washington" wrote:
CC> 
CC> >        Go grab that file.
CC> >        Go grab the whole of that site.
CC> 
CC> What do you mean by "that site"?  As I understand it you can't do a
CC> 'dir' on the remote site - or can you?  If you use a URL of (for

I'm not aware of being able to 'dir'.  Interesting if one
can.  I was speaking loosely, I guess I meant something like
"the maximal directed graph of links having the same initial
URL component".  Perhaps.

Admittedly there may be multiple 'sites' in this sense,
within a single author's files.  Indeed, it's rather
nice that it's easy to set up //joe.demon/work/ and
//joe.demon/play/, and conceal joe's multiple personality.

Strictly, I suppose one should also consider off site
links in this definition.  If //wash/A links to //keris/B,
which in turn links to //wash/C, even though there are
no links (direct or indirect) from A to C within //wash,
are A and C part of the same site, as far as depth of
search?

Frankly I don't really care, I think whatever way the code
happens to fall out, that's the way it's supposed to
work!

..
CC> >   In essence I think it boils down to six
CC> >different depth parameters, factored into
CC> >offsite/onsite, and html/images/other.
CC> 
CC> I get an 'or':
CC> 
CC>     (Page | Site)
CC>     HTML
CC>     images
CC>     binaries
CC> 
CC> as well as a depth for following HTML links.

Ooops, yes, I don't know why I imagined I could
have links from non-html files!  Thanks.

--
John Washington