3.454 announcements (139)

Willard McCarty (MCCARTY@VM.EPAS.UTORONTO.CA)
Mon, 11 Sep 89 20:05:52 EDT


Humanist Discussion Group, Vol. 3, No. 454. Monday, 11 Sep 1989.


(1) Date: 11 September 1989 (29 lines)
From: Willard McCarty <MCCARTY@vm.epas.utoronto.ca>
Subject: SOVSET'

(2) Date: Mon, 11 Sep 1989 15:26:52 EDT (67 lines)
From: Tim Seid <ST401742@BROWNVM>
Subject: HyperCard paleography stacks (61 lines)

(3) Date: 11 September 1989 (19 lines)
From: Willard McCarty <MCCARTY@vm.epas.utoronto.ca>
Subject: Archives and Museum Informatics

(1) --------------------------------------------------------------------
Date: 11 September 1989
From: Willard McCarty <MCCARTY@vm.epas.utoronto.ca>
Subject: SOVSET'

Humanists involved with Soviet and Eastern European studies will likely
be interested in SOVSET', an electronic network dedicated to these
fields. (No, I don't know what the ' at the end of SOVSET' signifies,
but then I cannot read Cyrillic.)

SOVSET' provides conferencing (open or restricted), e-mail among
members, and a data library that contains a series of periodical
reports, some issued daily. The subject areas appear to be primarily in
political science.

SOVSET' has an hourly connect fee of $25 (in addition to communications
charges) but no membership fee nor minimum connect time.

SOVSET' is accessed through Compuserve, Tymnet, and Telenet. It has a
Bitnet address, but this is apparently used only for downloading files
whose existence you already know of. Perhaps some Humanist who has
experience with SOVSET' can supply details.

For more information, write the Center for Strategic and International
Studies, 1800 K Street N.W., Suite 400, Washington, D.C. 2006 U.S.A.;
(202) 887-0200; Bitnet: support@sovset.


Willard McCarty

(2) --------------------------------------------------------------70----
Date: Mon, 11 Sep 1989 15:26:52 EDT
From: Tim Seid <ST401742@BROWNVM>
Subject: HyperCard paleography stacks (61 lines)

I have submitted a series of HyperCard stacks I call Paleographer to be
made available on the HUMANIST file server (hcpaleo hqx is a binhexed
Stuffit file).

[Editor's intrusion: the file is indeed called HCPALEO HQX.]

Paleographer is intended to be an experimental prototype of a system for
the analysis of scanned images of ancient manuscripts. It is a fully
functional model with three basic procedures.

After the process of scanning the manuscript and pasteing the images
by sections onto the cards in the stack, the first step is to analyze
each character individually. The primary aspect of this step is the
circumscribing of the smallest rectangle possible around the
character. If the character is a distinct unit without the crowding
of adjacent characters, then this can be done automatically by
simply clicking on it. Since I wrote this procedure in LightSpeed
Pascal and made it an external command in HyperCard, the action takes
only a second. The modulus, modular area, density, and compacted
record of the pixels is then computed, all within a few seconds. If
the rectangle can not be done automatically, then I provide a manual
procedure in a close-up view with a dialog box for moving the sides
of the rectangle in and out.

When the data about the character has been recorded in a dialog box
(leaf#, side, column#, line#, character#), it is then stored in a
second HyperCard stack. The information on each character is stored
on a single card maintaining a link back to the card holding the
scanned image and to the exact coordinates on the screen. In the
data stack, each character can be viewed as a binary graph and compared
with another character. The comparison allows for the binary graphs
to be superimposed. The fonts I developed for this purpose show
where the two are different--one has +'s for on pixels and the other
has x's--and where there are the same--when they overlap they form
an asterisk. Another form of analysis is to compute the angle of
the widest and narrowest stroke by dragging along the stroke. (I
need an experienced paleograher to check this procedure.) There are
fields also for recorded the character or its transliteration, what
language it is, and if it is part of some "extratextual" unit such
as a colophon, title, gloss, scholia, commentary, etc (these have
not been fully implemented in my mark-up scheme).
The third step is to create a transcription using an SGML tagging
system. I provide the option of tagging to the level of character
or to the level of line.

I have written a number of external functions to speed up procedures
and have used pull-down and pop-up menus. There is a main stack for
keeping track of the projects in which each card represents one
manuscript under analysis with a link to the image stack and data
stack (<filname>.ms and <filename>.data). The system includes a
help stack with examples and a chance to try out some of the
procedures.

I am providing these stacks as a way to encourage further development
in this area. Besides the obvious problems in resolution and
other factors that have been recently discussed on HUMANIST, I
welcome suggestions and criticisms. The HyperTalk scripts are
open, but at various stages of commenting. Let me hear what you
think.

Timothy W. Seid
Religious Studies
Brown University
(3) --------------------------------------------------------------------
Date: 11 September 1989
From: Willard McCarty <MCCARTY@vm.epas.utoronto.ca>
Subject: Archives and Museum Informatics

I have just been sent a sample copy of _Archives and Museum Informatics_
(ISSN 1042-1467), a quarterly newsletter edited by David Bearman. Vol.
3.2 (Summer 1989) contains articles on the Computer Interchange of
Museum Information, an initiative of the Museum Computer Network; and
the use of MARC format by archivists. Regular features in this issue are:
a listing of conferences; a calendar of activities; publications; news;
software reviews; and a column on standards.

Subscription is available for $40 US. For more information write Lynn
Cox, Managing Editor, AMI, 5600 Northumberland St., Pittsburgh, PA 15217
USA; (412) 421-4638.


Willard McCarty