PCP Web Site

Brian Gaines (gaines@fsc.cpsc.ucalgary.ca)
Sat, 14 Jan 1995 10:59:26 -0700

Following recent discussions, we have added a separate PCP page to the KSI
web site. It has:-

A brief introduction to PCP (largely Kelly quotes)

A link from the intro to an HTML version of the paper on Kelly's
Psychological Geometry that we wrote for "New Psychologist". This
attempted to place Kelly historically in the growth of cognitive
science and show how his geometry is equivalent to the intensional
logics used in knowledge representation. It has a good set of refs
for a newcomer.

A note on David Nightingale's PCP list server with details of how to
join, and hypertext links to the mail archives.

A link to other PCP-related web sites. This currently has no entries
and just asks everyone to send us their url's to put in the page.

A link to PCP ftp and gopher sites. This currently has links to archives
of Gaines/Shaw and Boose/Bradshaw papers. It again requests url's for
other sites.

A note on NAPCN and how to join.

A note on EPCA and how to join. This was taken from some material that
Mildred had on file (Devi, let us know if the note should be amended).

A note on the Barcelona meeting.

The nice thing about WWW is that is very open and egalitarian. We don't
intend this to be THE PCP page but rather one of many, all cross-linked.
Any PCP group can put up material on the Internet which reflects its own
approach and applications.

So, the next step is for everyone to send us the url's of any material
they have up so that this page can be linked to others.

In the short term, those who do not yet have web sites can send us
a note of any organizations or meetings that should be added to this page.
We can also add papers in electronic format to the archive (we have already
done this for the Boose/Bradshaw material sent to us by John Boose).

The simplest way to send electronic material is to ftp it to
ftp://ksi.cpsc.ucalgary.ca/incoming
and mail one of us that it is there and should be put into the archives)

In the long term, it is best if as many groups as possible maintain their
own WWW/ftp/gopher sites. The Internet is a nice arena in which to publish
what you like with no heavy hand of publishing bureaucracies!

The PCP page can be accessed through the KSI url below or, directly, as:-

http://ksi.cpsc.ucalgary.ca/PCP/PCP.html

For those setting up web sites for the first time, Brian wrote a long 'how to'
paper on supporting scientific communities on the Internet for a multimedia
conference. It has all the details on formats, conversions, tools, what not
to do, etc. It is on the KSI ftp site in Word RTF and PostScript formats:

ftp://ksi.cpsc.ucalgary.ca/KSI/RTF/CMMC94.RTF.Z
ftp://ksi.cpsc.ucalgary.ca/KSI/ps/CMMC94.ps.Z

This paper also addresses the issue raised in another message to this group
of putting up interactive programs in the client-server environment of the
web. It is simply done. However, the current interface available through
HTML level 2 forms is very restrictive. We have added HTTP server code to
a development version of RepGrid and used it experimentally to allow grids
to be entered in WWW forms, sent to the server, FOCUSed, PRINCOMed or INDUCTed,
and the results sent back as a gif image in a WWW document. We are also
experimenting with elicitation through forms but it is a bit clunky compared
with specially designed interfaces.

We haven't made these systems public yet because they are not multi-threading
and can't handle more than one request at a time. When they are we'll send
out a note. It will be interesting to see how such WWW PCP programs are
used.

We are designing the systems such that they can be used by other programs
as services, as well as by end users. This is an exciting aspect of
the open client-server environment that the web offers. It supports world-wide
collaborative research. For example, one group can put up an analysis program,
e.g. a principal component agent, and another group can put up an elicitation
system that uses that agent for analysis. There is tremendous scope for
loosely-coupled community research.

Best wishes, Brian & Mildred

Brian Gaines Knowledge Science Institute, University of Calgary
gaines@cpsc.ucalgary.ca Calgary, Alberta, Canada T2N 1N4
Mildred Shaw http://ksi.cpsc.ucalgary.ca/KSI/KSI.html
mildred@cpsc.ucalgary.ca tel: 403-220-5901 fax: 403-284-4707

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%