On Mon, 23 Sep 96 22:58 BST-1 Mike Holderness wrote:
> Of course, if librarians and not physicists had invented
> HTTP, it'd be called LURVE (you have 30 seconds to expand
> the acronym*) or SAUCE or something...
HTTP is a precise technical designation of an internet protocol
rather than the more appealing but less accurate LURVE.
And surely if librarians had designed URLs they would have been
in some totally logical but totally impenetrable decimal code?
On Tue, 24 Sep 1996 09:18:49 +0100 (BST) B.Naylor wrote:
> It's becoming more blindingly obvious by the day, as the web
expands and
> becomes more complex, that the system and structure of URLs is
impossibly
> complicated and, above all, unmemorable, and, in the long run,
cannot bear the
> burden of use being imposed on it. Shouldn't we be ditching it
and starting
> again before it gets too late - or at least relegating it to
the background,
> like machine code, where it needn't bother most of us?
Does anyone have any suggestions, however tentative?
The mention of machine code is interesting because a URL is
already a higher level representation of 'machine code'
addressing. The server identifier: //www.name.domain/ is a
human-friendly representation of a numerical IP address and
(usually implied) port number. The folder(directory) and file
name is a representation of a series of physical disk addresses.
The internet techies are trying to devise a system of URNs -
Uniform Resource Names. The intention is that these will be
location independent. This may or may not improve the
memorability of web page names, but it will require a worldwide
arrangement of name servers which can translate a URN into a URL
(or whatever) which gives the precise physical location of the
page. This obviously requires pages to be registered with the
name servers and reregistered when they are moved - extra work.
I am not clear what Bernard means when he says that the "system
and structure of URLs cannot bear the burden of use being imposed
on it". Technically it would appear to be capable of addressing
an unlimited amount of data. Or more precisely, any amount that
can be addressed by a URL of a length that can be handled by
browsers/servers.
If he means that URLs are becoming increasingly unmemorable then
it is impossible to disagree with him. This is hardly suprising
with there being at least "30 million pages found on 275,600
servers" (quoted from today's Alta Vista home page).
As for the URLs which are displayed in TV adverts, program
titles, credits, etc. the convergence of TVs and network
browsers/computers should make it possible to store a displayed
URL as a bookmark by pressing a button (Possibly from a Teletext
page rather than the actual TV picture).
The 'human memorability' of a web page address is irrelevant if
it is delivered in bookmarkable electronic form.
Peter.
==============================================================
Peter W Duncanson E-mail: [log in to unmask]
Principal Analyst Fax: (+44) 1232 230592
Computing Services Phone: (+44) 1232 335375
The Queen's University Of Belfast
==============================================================
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|