Re: Lynx

Lou Montulli (montulli@ukanaix.cc.ukans.edu)
Wed, 9 Dec 92 14:03:47 CST


From: montulli@ukanaix.cc.ukans.edu (Lou Montulli)
Message-Id: <9212092003.AA24547@ukanaix.cc.ukans.edu>
Subject: Re: Lynx
To: timbl@nxoc01.cern.ch
Date: Wed, 9 Dec 92 14:03:47 CST
Cc: www-talk@nxoc01.cern.ch
In-Reply-To: <9212090926.AA01030@www3.cern.ch>; from "Tim Berners-Lee" at Dec 9, 92 10:26 am

> 
> Lynx looks very nice -- but is it WWW?  From the comments about the syntax
> it doesn't seem to be HTML.  It has Gopher access built in, but there is no  
> access from KUfacts.cc.ukans.edu to the web.

No it is not WWW, but you are right that it should not be difficult to
add html compatibility.  Maybe I'm wrong but it seems that the majority
of the resources that I see available are in Gopher format.  Our 
concentration has been more towards the user interface than the protocal.
A fatal flaw perhaps?  Adding html will correct those problems because
obviously a great deal of though has gone into its implementation.
If we had know about WWW before beginning work on this project, chances
are we would have used WWW and just rewritten the curses client to look
halfway reasonable.
> 
> These guys have done some good work on hooking into existing services.
>The emphasis seems to be on building in extra bits and peices(rexec, hytelnet, 
> etc) into the browser rather than making gateways.  In the long run I think  
> this approach will get too heavy on browsers. For hytelnet, for example, a
> gateway is more efficient than building stuff very specific to one information  
Hytelnet compatability was not added as an after thought.  The markup 
language began compatible with hytelnet and evolved from there.  Our 
focus in designing the language was to make it easily understood by
non-programmers and I think we have achieved that, but this limits
its functionality, again we are back to needing html.

>providing application into the browsers.
> 
>The screen management is neat. I wonder whether we could persuade them to make  
>it W3 compatible?  (And their data with it?)  The user interface is quick and  
>simple. I missed "home" and "back", "next" and "previous" commands of the www  

If you wish to retreat from a document you simply have to push the left arrow.
The user controls were adopted from hytelnet, very simple and user freindly.
Right arrow - follow a link
Left arrow  - retreat form a link
up arrow    - next link
down arrow  - previous link

>line mode interface: one has always to go through the history page. But that  
> keeps it simpler. I was glad I had more than 24 lines 9and the program  
> recognised the fact) when the history list started to grow.
> 
> (Anyone want to make an W3 gateway for hytelnet?)

Will this gateway re-write the links inside hytelnet documents to match
html?

> 
> I am sure they could parse HTML with very little effort, and join the WWW club.
> The existing text at KUfacts would all fit into the <PRE> format I think.
> Their <!RLOGIN@library.host.edu -user=libcat> would be www's
> <a href=rlogin://libcat@library.host.edu> I assume.(This is not very general  
> of course as not all systems support rlogin).

Rlogin code can be included in the client for systems without native Rlogin.
This is a very useful feature, is it unreasonable to include the code?

:lou
-- 
  **************************************************************************
  *           T H E   U N I V E R S I T Y   O F   K A N S A S              *
  *      Lou  MONTULLI @ Ukanvax.bitnet                Nothing difficult,  *
  *                      Kuhub.cc.ukans.edu               is ever easy!    *
  *  UNIX,               Ukanaix.cc.ukans.edu       ACS Computing Services *
  *   have more fun!     Ukanvm.cc.ukans.edu           Lawrence, KS 66044  *
  **************************************************************************