COMMENT ⊗   VALID 00003 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	.CB TARGETS FOR STANDARDIZATION
C00006 00003		1. fonts and texts containing fonts
C00007 ENDMK
C⊗;
.CB TARGETS FOR STANDARDIZATION

.item←0
#. Introduction

	Many of the benefits of intercomputer communication have so far
been lost because there aren't standard ways of representing documents,
describing files, representing messages, representing business
communications and even representing certain aspects of computer
programs.  This paper describes some areas that will benefit from
adoption of standards and offers some suggestions for what these
standards should be like.  A major new idea is that it will be
possible to avoid traumatic changes in the way many things are done
by introducing %2meta-standards%1, i.e. standard ways of describing
what is done and programs that can use the meta-standards to interact
with programs and data bases that continue to do things in the old
way.

	Our shocking tale begins with the fact that many organizations
have developed ways of preparing high quality multi-font documents on
computers and "typesetting" them for reproduction.  The source documents
and the output of document compilers can be transmitted by ARPAnet and
other networks, by the dial-up telephone network, or on magnetic tape from
one installation to another.  Once transmitted, they can rarely be
printed or even viewed on terminals in their typeset glory.  As more
systems are developed by organizations for their own use or for sale,
the situation is getting worse.
Some of the developers recognize the problem, but others, especially
the vendors of high quality word-processing equipment regard each
system they sell as an independent means of producing paper and haven't
even considered the problem of communicating editable documents
electronically.  One can always transmit paper copies or images
from paper to paper by telecopiers, but the results are not
electronically filable, are not searchable, indexable or modifiable
by computer.

	A brief listing of some systems I know about may illustrate
the nature of the problem.

	1. In the early 1970s Larry Tesler, then at the Stanford
Artificial Intelligence Laboratory, Developed PUB, a document compiler
that accepted source files with commands for formatting, automatic
numbering of parts of the document, making footnotes, indexes and
tables of contents and for designating any fonts that might be
available.  PUB is used at Stanford, Carnegie-Mellon, and M.I.T.,
and many hundreds of PUB source documents exist.
	1. fonts and texts containing fonts

	2. documents in computers

	3. program libraries

	4. data base description

	5. business communication

	6. programming languages

	7. dialnet