From @mitvma.mit.edu:timbl@nxoc01.cern.ch Tue Jun 16 12:16 EDT 1992
Received: from MITVMA.MIT.EDU by zippy.lcs.mit.edu with SMTP
	id AA11671; Tue, 16 Jun 92 12:16:32 -0400
Received: from MITVMA.MIT.EDU by mitvma.mit.edu (IBM VM SMTP V2R2)
   with BSMTP id 1452; Tue, 16 Jun 92 12:14:58 EDT
Received: from CEARN.cern.ch by MITVMA.MIT.EDU (Mailer R2.08 R208004) with
 BSMTP id 6079; Tue, 16 Jun 92 12:14:57 EDT
Received: from CEARN by CEARN.cern.ch (Mailer R2.07B) with BSMTP id 4018; Tue,
 16 Jun 92 18:14:21 SET
Received: from dxmint.cern.ch by CEARN.cern.ch (IBM VM SMTP V2R1) with TCP;
   Tue, 16 Jun 92 18:14:19 SET
Received: by dxmint.cern.ch (dxcern) (5.57/3.14)
	id AA22981; Tue, 16 Jun 92 18:14:57 +0200
Received: by  nxoc01.cern.ch  (NeXT-1.0 (From Sendmail 5.52)/NeXT-2.0)
	id AA02880; Tue, 16 Jun 92 18:14:33 MET DST
Date: Tue, 16 Jun 92 18:14:33 MET DST
From: timbl@nxoc01.cern.ch (Tim Berners-Lee)
Message-Id: <9206161614.AA02880@ nxoc01.cern.ch >
To: timbl@zippy.lcs.mit.edu
Subject: From Robert
Status: R

>From cailliau@www2.cern.ch Tue Jun 16 18:01:56 1992
Return-Path: <cailliau@www2.cern.ch>
Received: from dxmint.cern.ch by  nxoc01.cern.ch  (NeXT-1.0 (From Sendmail
 5.52)/NeXT-2.0)
	id AA02840; Tue, 16 Jun 92 18:01:44 MET DST
Received: by dxmint.cern.ch (dxcern) (5.57/3.14)
	id AA22433; Tue, 16 Jun 92 18:01:56 +0200
Received: by  www2.cern.ch  (NeXT-1.0 (From Sendmail 5.52)/NeXT-2.0)
	id AA00408; Tue, 16 Jun 92 18:01:57 GMT+0100
Date: Tue, 16 Jun 92 18:01:57 GMT+0100
From: cailliau@www2.cern.ch
Message-Id: <9206161701.AA00408@ www2.cern.ch >
Received: by NeXT Mailer (1.63)
To: www@www2.cern.ch
Subject: ABC proposal
Status: R

Tim - Jean-Francois : Tony Osborne asked for more details, I have
filled them in, please comment and pay special attention to the
places marked with (??).


Project overview:
-----------------
WWW solves the problem of accessing information that is distributed
over a wide community of geographically dispersed collaborations by
unifying interaction with the incompatible systems on which the
information is maintained. The theoretical foundations are laid and
working versions of software have been released, showing its
viability and its utility in the daily work of our colleagues.
Try WWW by typing www on most of the CERN machines, or by doing
   telnet info.cern.ch
from elsewhere.

Salient features of the current release are:

- the information from many HEP sites is available on-line,
- all information is presented in the same way and is very easy
  to access,
- a wealth of non-HEP but academic information is available from
  existing sites that were integrated at little cost (WAIS, Gopher,
  Archie), these include a large amount of computer science related
  material, software repositories, news groups etc. useful to many
  workers at CERN.
- the information can be reached from anywhere in the world (provided
  there is telnet, which virtually all academic institutes now have),
- the software has been picked up by over 1000 sites,
- at least 20,000 documents are accessed per month, but we have no
  count of those used by downloaded browsers.

One "feature" is that all development and implementation is done
using Unix and its related tools. This is a disadvantage in that Unix
and TCP are not yet wide-spread tools in the HEP experiments
community, where VMS and DECnet still predominate, but it is an
advantage in that the trends are clear, the external experience we
can draw on is large, and we will be ready.


Purpose and overall scope:
--------------------------

- Be a tool for collaborative work, through which distributed
  experimental teams can compose and communicate. For the current HEP
  users, this will be possible only when we have integrated the
DECnet
  protocol (see below) and have a reasonable set of editing browsers.


- Disseminate information of a public nature to all users of the
  HEP community, (achieved)


- Provide an on-line documentation tool for experimental groups.

- Provide easy access for physicists to a wealth of information
  which is provided by other related disciplines - like computing,
  materials sciences,...(partly achieved, eg. by access to WAIS).


Current user communities:
--------------------------

Inside HEP: CERN, SLAC, NIKHEF, DESY; with experimental groups (for
their private data) being targeted as soon as we have access
protection and the DECnet protocol (both under development now).
WWW is now a part of the CERN library distribution tape.
The Library intends to make the Alice service available.

Outside HEP, WWW generates many projects in university campus
information systems, in-house webs for commercial companies etc.
The Scandinavian Libraries... (?? Copenhagen)
The Molecular Biology Laboratories (??)


New functionalities required:
-----------------------------

There many requests for improvements, all of which will increase the
function provided by WWW, but only a small fraction of which we can
implement with the current manpower. Here is a list of the more
important ones:

- Access protection for servers, to promote use of WWW by groups but
  prevent unauthorised access to their private information (being
  implemented),

- implementation of the HTTP protocol over DECnet, to promote use of
  of WWW by experiments (being implemented),

- A more generic server program skeleton for information providers
  inside HEP,


- Editing browsers (not just reading ones) for X window/Motif, Mac
and PC,
  so as to allow groups to put information in without need for a
NeXT,


- Format negotiation and conversion between server and browser,
  to allow presentation of a greater variety of documents
  (LaTeX, PostScript, ...),

- (??)


Manpower requirements and long-term support needs:
--------------------------------------------------
There are two staff on the team now:
TBL, doing Research&Development,
RC, doing Public Relations and Mac and part-time other duties for PT
group.

We need in addition:
  - at all times one Unix/C programmer-analyst,
  - at all times, one support person whose duty it is to provide
computing
    help to users who provide information to experimentsl groups
    (installation, documentation organising, feedback to us, ...),
  - a liaison-officer to do the PR, get feedback, design documents
and
    sub-webs, keep the basic HEP information accurate or advise those
	 who are providing data.


Time scales:
------------
- release of access authorisation for experiments: 92.12
- release of a DECnet version for experiments: 92.08
- if one programmer continued: basic multi-format document: 92.12
- one new editing browser per year per full-time programmer.


Commercial alternatives:
------------------------
We know of no commercial alternatives. As a sign, the special issue
of Electronic Networking on "Accessing Information on the Internet"
had only six articles. Of these, none was on commercial systems, one
was by us on WWW, and three others on Prospero, Archie and WAIS, two
of which are in fact accessible through WWW!

When will the editor exist for other platforms than NEXT:
---------------------------------------------------------
The Mac browser can be demonstrated but suffers from a residual bug,
probably locted in some of the external software used in it. There
are now two external people helping us there.

There is a team outside CERN working on an editing browser for X.(??)

There is no editing browser for dumb terminals, nor is it foreseen,
although it might be a good idea...


Notes:
------
Some people (Marseille) think it can be used for process control! (in
fact, NIKHEF have a joke server that returns the square root of any
number you give it)

The accelerator divisions have shown interest too.

                                 -=#=-


