CHI-WEB Archives

ACM SIGCHI WWW Human Factors (Open Discussion)


Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Paola Kathuria <[log in to unmask]>
Reply To:
Paola Kathuria <[log in to unmask]>
Fri, 16 Apr 2004 13:31:47 +0100
text/plain (101 lines)
Ramon Vazquez wrote:
> It seems to me that the worst usability example is in URLs. There's a
> lot of work to do on how the web, the internet and network works to make
> it more user friendly.

Is a telephone number such as +44 20 0555 2876 'user friendly'?

Telephones have been around much longer than the Internet but
if I were to call International Directory Enquiries and ask for
Dunkin Donuts they wouldn't be able to connect me without
first asking some questions.  They'd first ask for the country.
Okay, so I say "US". If I say, "the one in Springfield", this
still wouldn't be enough information as there are a lot of
towns called Springfield in the US, so I'd have to be asked
for a state.  Then the town, street, and so on. A (full)
telephone number therefore includes a country code, city code
and area code.

Like telephone numbers, host names (just the first part of a
URL) are made up of consecutive parts, each defining a
wider context. The host would be in the
maths department within the University of Arizona which is
an educational establishment within the US ( is the UK
equivalent of .edu).

I hope you would agree that host names are less obscure than
telephone numbers.

The suggestion of having "university of arizona" as the home
page URL would mean that each university would be limited
to just one web site, whereas universities often have a web
site for each department or faculty. How would a natural
language naming scheme work for the web site of the Maths
department of a university? Should I put the department
name before or after the university name?

And in a natural language naming scheme, what would be the
URL of the page on a specific course in a specific department
for a given year?  Would the URL still work if the positions
of words were changed (even though it would not make sense
in English)?  Would you expect all translations of the word
"university" in all languages to still work in natural
language URLs?

The existing naming schemes for telephone numbers and host
names are designed to be highly flexible, adaptable, precise
and extensible.  As is often the case, highly flexible systems
tend not to be simple. Trading off flexibility for simplicity
is common in computing. A good example of this is the
flexibility and power but steep learning curve one has with
a command-line interface compared to the simple but slow GUI
interfaces, often with very restricted functionality
(relatively speaking).

The dot naming standard has been around since the mid-1980s
and is, umm, somewhat fundamental to the Internet; it is
unlikely to change any time soon.

And domain names have to be efficient so that all the
computers on the Internet know how to pass on requests and
data.  Not every computer which routes information on the
Internet has a complete database that knows about every
host computer in the world; my local domain name server
will know what to do with a web request to a .edu domain
and will pass my request on to another computer.  The
request will eventually be passed onto a server that will
have details about and then one which will
know how to get to

The domain name system is a distributed database across
the world and is only made possible by the tree-like
structure of domain names (telephone exchanges work in
a similar way).  A natural language naming system for
host names could not be viable in terms of the amount
of space needed to store all the addresses and their

> Why isn't it possible to create URLs with special characters like
> apostrophe, @,  !,  comma, etc??

With some restrictions to make it possible for computers
to handle them efficiently, you can, although many of the
technical standards that make this possible are still just
coming into effect.

All Internet standards are published in documents called
RFCs which have always been available to the public.
Pertinent to this discussion is RFC 1035 - it is is
available from (and many other places).


        Tip of the Day: Email mailto:[log in to unmask]
               with any comments, questions or problems
     CHI-WEB: POSTINGS: mailto:[log in to unmask]
              MODERATORS: mailto:[log in to unmask]