ontolog-forum
[Top] [All Lists]

Re: [ontolog-forum] Constructs, primitives, terms

To: ontolog-forum@xxxxxxxxxxxxxxxx
From: "John F. Sowa" <sowa@xxxxxxxxxxx>
Date: Sat, 17 Mar 2012 11:43:10 -0500
Message-id: <4F64BF1E.4030307@xxxxxxxxxxx>
On 3/17/2012 10:08 AM, John Bottoms wrote:
> Your rather brief retelling of the history indicates that you were not
> around in the early days of electronic publishing. But since you were at
> IBM you might have some insights into the publishing standards.
>
> In crafting SGML Charles Goldfarb undoubtedly must have run into what I
> call "the Von Neuman problem". That is, when you start turning math into
> something usable then you find that pieces are missing that prevent the
> application from working.    (01)

Back in 1973, Charlie Goldfarb was working down the hall from me at IBM.
He introduced me to GML, which I used until I retired from IBM.  The
acronym GML is formed from the names of the three developers -- the
"official" expansion to 'General Markup Language' was chosen to make
the acronym acceptable to IBM.    (02)

In 1983, I used GML to produce the camera-ready copy for my book
_Conceptual Structures_.  The publisher, Addison-Wesely, wanted me to
use the same fonts and formats that they had used for Chris Date's
book on databases.  GML in 1983 was sufficiently flexible that I could
reproduce every format exactly, except for one minor font, for which
I used a close substitution.    (03)

In fact, GML in 1983 gave me more precise control over fonts and
formatting than MS Word today.  After I left IBM, I switched to
using HTML for all my word processing.  I just type HTML tags,
which are almost identical to the GML tags I used for years.
For formatting, I load HTML into OpenOffice and make minor tweaks.    (04)

> Clearly SGML is built on FOL...    (05)

GML and SGML use only a tiny subset of FOL.  Their major contribution
is a very rich ontology that is designed for representing document
structures and formatting conventions.    (06)

> What do you see that was added to the SGML standard that was beyond
> the scope of FOL?    (07)

In general, anything and everything that has ever been implemented on
any digital computer can be specified in or translated to FOL.  The
hard work is to define the ontology for the application.    (08)

For formatting languages, it's usually convenient to use FOL at two
levels: an object level for representing the document content and
its formatting, and a metalevel for talking about and defining the
tags and names.  The XML namespaces, for example, would be specified
at the metalevel.    (09)

You can use different notations for the object level and the
metalevel.  You can also use the same syntax for both levels,
but use a different syntax and vocabulary for the metalevel.    (010)

In general, two levels of FOL (object level and metalevel) are
sufficient for most computer applications.  In some caes, you might
find it convenient to add a metametalevel of even metametametalevel.    (011)

John    (012)

_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/  
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/  
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/ 
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J    (013)

<Prev in Thread] Current Thread [Next in Thread>