Hi!

"Russell Penney" <russell.penney at tincanct.com> wrote:
> Sorry Goran, but I really don't agree.

No problem, I think the whole mess is confusing so I wouldn't expect
people to easily agree. :)

[SNIP of quotes from RFCs that you wasn't elaborated on]
> > So what did I mean? :) Well, I think I meant that a *file URL* is "a
> > weird thing" that can't really be treated in practice based on the RFCs
> > for URIs. The tools do lots of "smart things" in order to handle a whole
> > range of false file URLs.
> 
> I don't think a *file URL* is a weird thing but something that is

I find them weird because they are so poorly defined IMHO. And since
file paths are so clearly not portable over different filesystems
(volume names, filename lengths, allowed chars etc etc).

Also all the BIG projects do a lot of trickery to make them work even
when they are not following the standard.

For example, a file URL should ALWAYS have two slashes after the scheme
name and should ALWAYS include the hostname/"localhost" or "" - which
leads to these three variants:

file:///
file://myhost.domain.com/
file://localhost/

That is at least how I interpret it.

> translatable to all file systems (and I do mean all, the examples in the RFC
> are for a VAX system!). Also there is nothing stopping Squeak defining a new
> scheme for mac files or unix files or zip files. Any scheme name that starts
> with "x-" is considered experimental.
> 
> Russell

Let me just point to others feeling the same confusion:

	http://curl.haxx.se/mail/lib-2003-08/0078.html

regards, Göran

PS. If anyone has a specific, concrete problem with how I rewrote
FileUrl then I am happy to discuss it!