Computer technology

Sun Jun 6 12:56:32 UTC 1999

>I am also investigating this and related problems with
>Cyrillic - my colleagues and I at Exeter are trying to
>launch a project to create our own CALL exercises for ab
>initio students. The font issue seems to get in the way for
>all sorts of things (particularly trying to get
>multilingual support for networked keyboard drivers...).

In theory, this shouldn't be so hard. A properly configured browser on a PC
with Windows98 and its Cyrillic support, or a Mac with the Cyrillic
Language Kit, should be able to show web pages regardless of which of the
four major encodings (Code Page 1251, MacCyrillic, ISO 8859-5, KOI8) the
page was written in. If the character set is declared in the <HEAD> section
of HTML code (<META HTTP-EQUIV="Content-Type" CONTENT="text/html;
charset=windows-1251">), the browser will adjust itself automatically to
show Cyrillic. You could try one of my test pages for Bulgarian CALL:
http://www.hf.uio.no/east/bulg/mat/annot.html

However, the different browsers on the different platforms behave in weird
and mysterious ways that are understood by few people on this planet. There
is, for instance, the so-called "Netscape burp", which manifests itself in
Netscape drawing a page with a declaration like the one above, TWICE, and
the second time, it neglects to draw anything that is put into the HTML
code by Javascript. This is a major handicap, as Javascript is a powerful
language that is cross-platform and well suited for developing CALL. I have
tried to overcome the problem of keyboard drivers by laying out letters at
the bottom of the page in a keyboard pattern, so that the student can click
on the letters to enter words into the input form, but the Javascript code
gets devoured in the Netscape burp. Anything you make should be tested at
least on both the major browsers on both the major platforms to make sure
it works.

For Cyrillic pages that seem to be unreadable even on a properly configured
browser, I have found two reasons so far. One is that the page has been
made with some "WISIWYG" Web page creation program that automatically
translates high ASCII (where the Cyrillic characters are) into HTML
entities for the accented letters and symbols in the high numbers of ISO
8859-1. If you look at the HTML code for such a page, it is full of HTML
entities like "†" etc. The other reason is that the programmer of
the page has specified that the fonts to be used in displaying the pages
should be, say "Arial, Helvetica" - which would look fine on the
programmer's computer with Cyrillic Arial or Helvetica, but like alphabet
soup on a computer with Cyrillic fonts with different names. These are easy
to fix, though: just specify in your browser's preferences: "Use my fonts,
overriding page-specified fonts".

-- Kjetil Raa Hauge, U. of Oslo. Phone +47/22856710, fax +47/22854140
-- (this msg sent from home, phone +47/67148424)