From: Cameron Kaiser <spectre@stockholm.ptloma.edu>
Message-Id: <200105211614.JAA10382@stockholm.ptloma.edu>
Subject: [gopher] Re: web->gopher
In-Reply-To: <20010521090703.C15451@siva.heatdeath.org> from "em@nuel" at "May
 21, 1 09:07:03 am"
To: gopher@complete.org
Date: Mon, 21 May 2001 09:14:56 -0700 (PDT)
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 8bit
Sender: gopher-bounce@complete.org
Errors-to: gopher-bounce@complete.org
Precedence: bulk
Reply-to: gopher@complete.org


> > *and* :-) One tiny suggestion -- maybe make it turn "GET /" selectors into
> > real HTTP URLs?
> 
> I used to do that, but I removed it after deciding that it's contrary to
> the spec.  My interpretation is that the 'h' type denotes an HTML
> document stored on a gopher server.  According to the spec, selectors
> have no meaning.  There's nothing stopping 'GET /whatever.html' from
> being a selector for a gopher server, so turning it into an HTTP request
> is incorrect.
> 
> I create web links by creating an HTML document on the gopher server
> that redirects the user to the destination web site.  It works well with
> everything.
> 
> I don't handle HTML documents correctly yet.  Web->gopher should go
> through and re-write all the URLs so that relative links work.

I'll buy your theory -- yeah, my main reason for concern was that images
didn't show up right if they were relative URLs.

One other more practical concern is that some clever d00d might use
heatdeath as an HTTP proxy, which might be a problem if it became
popularised in terms of bandwidth. Using real URLs would not stop this
completely, of course, but it wouldn't make doing it so obvious.

-- 
----------------------------- personal page: http://www.armory.com/~spectre/ --
 Cameron Kaiser, Point Loma Nazarene University * ckaiser@stockholm.ptloma.edu
-- "I'd love to go out with you, but I'm rethreading my toothbrush bristles." -