From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <3A46DF64.46FD96A4@home.com>
From: Anthony Starks <ajstarks@home.com>
MIME-Version: 1.0
To: 9fans@cse.psu.edu
Subject: Re: [9fans] Re: Future of Plan9
References: <20001224021200.5227D199DD@mail.cse.psu.edu>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Date: Mon, 25 Dec 2000 00:47:16 -0500
Topicbox-Message-UUID: 3d418200-eac9-11e9-9e20-41e7f4b1d025

Russ Cox wrote:
> 
>         wouldn't a mothra port be the deal?
> 
> richard miller did this.
> i've tried mothra and charon, and
> neither satisfies me.  there's
> got to be a better way than
> the netscape and ie model,
> but i don't know what it looks like.

I wrote this in 1994 --- I think the ideas are still relevant, though
mostly unrealized. Perhaps we can see something in Plan 9...


.....
Date: Fri, 3 Jun 1994 21:29:11 -0400
From: Anthony Starks <anthony_starks@merck.com>
To: bug-chimera@cs.unlv.edu
Subject: The next generation of WWW browsers


Now that we have experience with the first generation of web browsers
exemplified by chimera and mosaic, I think we should think hard about
the design and implementation of WWW browsers:

The first generation is fine as is goes, but I think a tool building
approach that moves from monolithic browsers to browsers built from
small single-function modules would support richer browsing models than
we have now, and at the same time simplify the building and maintenance
of browsing software.

Consider:

 * a network module that takes a URL as input
   and then connects to network server, and outputs HTML.

 * a HTML module that takes HTML as input
   and renders it on some display. Touching a link could send the URL
   to the network module which would in turn emit more HTML to be
   rendered.

  * A control module that communicates with and coordinates
   the above modules for the purposes of navigation, printing, etc.

  * some communication method (IPC, pipes, sockets, etc) that
    is used to send messages between the modules.


With these modules it would be easy to build something like chimera or
mosaic, but, for example, consider an application that takes a list
URLs, and renders each one in its own display---this could easily and
quickly be built from the components above.

Also, since each HTML display is independent, one could touch several
links in succession and not have to wait for the synchronous display of
each.

I know that there are other (better) examples, but the point is that
the right collection of tools can make for powerful software that goes
beyond the function of a monolithic tool.

Why not apply this principle to the design of web browsers?


-ajs-