[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ih] FTP Design

On 7/3/2012 12:25 PM, Tony Finch wrote:
> Dave Crocker <dhc2 at dcrocker.net> wrote:
>> I'm going to claim that the innovations in the URL construct were:
>>     1.  extensible declaration of service mechanism (http, ftp, ...)
>>     2.  rigid requirement for domain name, to specify the place for evaluating
>> the rest of the string
>>     3.  essentially no constrains on the rest of the string.
> But, gopher's links had all of those except the first.

My immediate reaction is "yes, but..." but the reality is that I don't 
want us to get caught up with this particular tar baby.  For example, as 
my reference to the email address model demonstrates, #2 and #3 had 
occurred previously.  (But as I said, let's not pursue this issue.

Mostly because...

> What multiprotocol user agents were there before the www?

This strikes me as a deeper and more troubling point:  the failure to 
distinguish between an integrated service architecture, versus 
user-level integration of otherwise-independent services.

What makes this a more interesting topic is that the design of the URL 
meta-construct has essentially permitted the Web to coopt all other 
independent work.  While that's useful, it also misleading in terms of 
protocol and format innovation and development.

That is, the ability of a URL to encompass independent services is 
really the benefit I meant to point to, with the the list of 3 points 
(above.)  But the result has also been confusion about architectural 

That's why I usually distinguish the Web as object, protocol and 
reference:  html, http, url.  But the web isn't gopher or email or ftp, 
even though URLs can refer to them.  And yes, I realize that that is a 
more constrained view than most people have.

> A couple of points I want to agree with and expand on:
>> However it was easier to set up a gopher site than a web site, because the web
>> required specialized documents while gopher ran on text.  There was serious
>> competition between the two.
> It took a few years before the web became a hypermedia system (embedded
> images and so forth) which was when it became really distinct from its

As I recall, images could be included from the earliest that I started 
learning html.  That wasn't 1990, but I'm pretty sure it was before 1994.

Also as I recall, the only initial document that was supported was 
.html.  That is, raw .txt files, for example, were /not/ supported.  And 
it was a pain.

> predecessors. It also helped that web browsers were good gopher clients.

cf, above, about architectural confusion.  The fact that a user agent 
can encompass multiple services does not mean that the services have 
been integrated.  So the UA is useful, but being useful isn't the same 
as being integrated.

> And there were the botched commercialization attempts by the web's
> competitors.


>> Another major design difference was that gopher provided no useful information
>> until you reached the leaf, whereas the web could produce an 'interesting'
>> document with every click.  That is, the Web permitted a far sexier
>> experience, of course.
> The difference between hypertext/hypermedia and a catalogue.

mumble.  maybe.  in any event, certainly an important difference.


  Dave Crocker
  Brandenburg InternetWorking