Re[2]: Off topic: type (ME letter "Yogh")

[Emery Snyder emery.snyder@xxxxxxxx]

We might be able to combine our work. My spreadsheet doesn't have all
the XyWrite Speedo characters in it yet (there are a lot of them, and
my wrists started to hurt when I was typing them all in).

But it does have the character numbers from Win-ANSI (a.k.a. code page
1252), IBM code page 850, ISO 8859-1 (Latin 1: West European), 8859-2
(Latin 2: East European), 8859-3 (Latin 3: South European), 8859-7
(Greek) and Unicode for most of the Latin characters with diacritics
and the Greek characters.

I could add in 8859-4 (north European) and 8859-5 (Cyrillic) if people
want, since they're part of the Speedo set too.

I will add in 8859-9 (Turkish) because I use that sometimes, although
really 8859-3 handles most of it. I should probably add the
Windows-Turkish and Windows-Baltic character sets as well.

I can't really deal with 8859-6 (Arabic) or 8859-8 (Hebrew) because I
don't know any Semitic languages and they doesn't really interact with
my Xy documents.

8859-15 is really useful because it includes the euro symbol, and is
starting to be used as a standard for Europeans who don't use Icelandic.

---

Here's a nice page on the "alphabet soup", btw:
http://czyborra.com/charsets/iso8859.html

He also has a nice page on the code pages 437 and 850, which are very
relevant for XyWrite:
http://czyborra.com/charsets/codepages.html

It also has information on the Windows code pages, which are relevant
for those of us working in XyWin.

--

I often write documents which include text in German and French, and
need to cite authors with names including Turkish and Czech
characters. I tend to be using a TrueType or PostScript font. The best
solution has been to use one of the newer TrueType fonts which come in
the various Windows code pages. If I'm typing along in "Times New
Roman" and suddenly need to type a name with an r-hachek in it, I use
the font selector to switch to "Times New Roman CE" (same font, but
with the Central European character set), figure out what position
number my r-hachek is, and insert that number in the appropriate
place. The same thing works for the Turkish variant of Times New
Roman. (Or at least, all that worked fine in Windows 98; Windows 2000
is proving more problematic.)

(Oddly enough, the font selector utility in NotaBene (at least the
demos I've downloaded so far) doesn't show the variants of those fonts
with the different character sets; don't ask
me why. )

Of course, that means my documents are littered with markup for font
changes and character set changes, and while they print out OK,
printing to PDF files and passing the documents on to other people
both create problems. You can see why I'm enthusiastic about Unicode
-- I could use it to write entire documents without switching code
pages / character sets, and a person on the other end can figure out
what character I meant to show up at a given spot.

Note: Sorry about the pseudo-word "hachek". The "ch" there really
represents a 'c' with a hacek over it.

---

You can express any character in these things as a combination of two
numbers: character set and position number. I don't know that there's a
standard set of numbers to represent all the common character sets.

The "superscript three" character, for instance, is

Win-ANSI: 179
IBM codepage 850: 252
Xy Speedo: 906
8859-1: 179
8859-2: missing
8859-3: 179
8859-4: missing
8859-5: missing
8859-6: missing
8859-7: 179
8859-8: 179
8859-15: 179
Unicode 3: 179

-------

Query: does anyone know why the Speedo set contains some duplicate
characters?


On 7.2.2003, J. R. Fox wrote:
===8<==============Original message text===============
> Patricia M Godfrey wrote:

>> CHARSET has all the values and
>> the numbers of XyWrite's Speedo set. Beyond the first 255 (the Extended
>> ASCII set, dating from DOS days), of course, they are peculiar to XyW,
>> while Windows uses its own variation on the ANSI/ISO Latin set. (I'm not
>> yet perfectly clear how that relates to Unicode.) I am working on a table
>> (it's in a spreadsheet right now, so I can sort easily on any column)
>> that will list most of these characters and their codes in the various
>> systems. WordIMPerfect has its own scheme, which involves two numbers: a
>> characteristic (to borrow a term from logarithms) number indicating the
>> set, and a mantissa indicating the character's order in that set. In
>> ASCII/XYW, 192-199 are box-drawing charcters;

> When it is completed, that might be a good item to post here, if you are
> willing. Even if it is in a spreadsheet file, I expect there are ways we
> could convert it into HTML, which might be the most universally accessible
> format.

> Jordan

===8<===========End of original message text===========