Re: Nitpick on 2a23

On Sat, 31 Jul 1999, Daniel Stenberg <Daniel.Stenberg_at_sth.frontec.se> wrote:
>
> On Fri, 30 Jul 1999, Eric Stewart wrote:
> >
> > On an http: address that's linkified in a message, if it's at the
> > end of a sentence with a trailing '.', it puts that in the link.
>
> And how is hypermail supposed to know when the dot is actually included
> in the URL or not?
>
> One could argue if not the most common case is when the dot is not in the
> URL, but there still is no 100% working way to know.

You probably have no idea how many countless hours at the IETF I sat through meetings with people griping about this very issue...

I'll admit to not having read the RFC(s) on URLs lately, possibly never in the case of the one(s) actually approved by the IESG, but there was, at one time, recommended practices within the first URL RFC that a properly formatted URL included within a text documents should be wrapped in angle brackets. And, of course, URLs pointing to directories/indexed collections should always end with a trailing forward slash. E.g.

   <http://www.cni.org/>

Shortcuts, such as <www.cni.org> and <http://www.cni.org> are malformed URLs, even if popular web browsers do support these strings as input at the interface level.

Of course, try to get people to do this these rules when they are keying in their email...

Nevertheless, I think it is fairly safe to assume any URL ending in .html., .htm., .shtml., etc. can safely have the trailing period stripped. I have never encountered one instance where this would not be the case.

Personally, I'd favor a more aggressive approach in this case... I think stripping the trailing period will be right 99 out of 100 times.

-- 

   Craig A. Summerhill, Systems Coordinator and Program Officer
   Coalition for Networked Information
   21 Dupont Circle, N.W., Washington, D.C.   20036
   Internet: craig_at_cni.org   AT&Tnet (202) 296-5098