Re: Tildes on vowels

From: William Overington (WOverington@ngo.globalnet.co.uk)
Date: Wed Aug 14 2002 - 06:42:42 EDT


Philipp Reichmuth wrote as follows.

>Hello, William,
>
>This is sort of lengthy once more. Forgive me and put me in your score
>files. :-)

What please is a score file?

>Note that asking Microsoft to have Notepad support courtyard codes is
>a lot more work and a lot less likely to succeed than working with
>custom XML files. Especially on older hardware, for obvious reasons
>(if new software is needed, it will probably require newer hardware,
>too)

Well, actually I was not thinking in terms of large companies taking up
courtyard codes, though they are welcome to do so if they so wish.

What I had in mind was individuals and smaller companies who might write and
publish on the web desktop publishing packages and electronic book authoring
packages and the like.

>Having to work around a software vendor's decision to use U+XXXX from
>the PUA to mean, say, "BOLD" when XXXX is in the PUA and I need to
>process it for my *private* use, is not trivial. It puts everybody
>right back into pre-Unicode times.

That is why I have asked if end users might kindly agree not to use the
U+F300 to U+F3FF block of the Private Use Area for other allocations if they
could possibly avoid it, suggesting that perhaps they might feel that they
would benefit from that as well.

It is possible that such agreements can work. For example, since the
publication of code points in the Private Use Area for the symbols of the
Phaistos Disk I have decided not to clash them with any code point
allocations of my own. No one asked me to do that, I did it for my own
interest in the Phaistos Disk. Later I realized that it also avoids any
font developer needing to decide whether to include the Phaistos Disk
symbols or some of my allocations in a given font, so that is potentially
another advantage for my research of my having avoided making any
allocations in the same places in the Private Use Area as the Phaistos Disk
code points are placed.

I know that using Private Use Area allocations is not mainstream, yet those
people who choose to use Private Use Area allocations are able to act in
their own interests by taking into account uses of the Private Use Area by
others if they so wish, notwithstanding that they are not in any way obliged
to do so.

>With glyphs in the PUA, the problem is less bad than with control
>characters. Putting control characters in the private use PUA and
>expecting

I am not expecting them to do so. I have published the code points and if
people choose to use them so as to provide a consistent encoding format
which is independent of any one manufacturer, they are welcome to do so.
There is no question whatsoever of me expecting people to take up my ideas.

>others to adhere to them is dangerous and, to some extend,
>disrespectful of other people's need of private characters in the PUA.

Well, as I am not expecting people to adhere to them, then it cannot be
disrespectful. Publishing code point allocations on my own initiative
merely provides the opportunity for an agreement to use particular code
points for particular meanings to be wider than just two people as the
sender and receiver of a particular document. This is not an expecting,
requiring and disrespectful approach, it is simply enthusiasm for pushing
the envelope of what is possible if some people voluntarily agree to some
particular code points having particular meanings in relation to only
certain uses of those code points.

>It might be a good idea to move your Courtyard Codes to the Corporate
>Use PUA subarea, partly because in your suggestion that everybody else
>should share the codes, you are beginning to act like a software
>vendor, and partly because it saves other people hassle.

Well, as far as I am aware, by putting courtyard codes in U+F300 through to
U+F3FF I had effectively done that, while also allowing space above them for
manufacturers of software. I chose U+F300 through to U+F3FF so as firstly
not to clash with any part of U+E000 through to U+EFFF, secondly so as not
to jam up the top part of the Corporate Use Subarea and thirdly so as not to
clash with what is, in effect, the Symbols font Subarea. Indeed, I only
became aware of the location of that latter area from a posting in this
list. As the Unicode Technical Committee is considering rewording the
section of Chapter 13 of the Unicode specification about the Private Use
Area, I feel that it would be helpful if the fact that symbol fonts use that
area of the Private Use Area could be stated in the book. I realize that
there may well be concerns that such a mentioning might be seen as endorsing
that usage, yet the alternative is that people using Unicode are not able to
find the location of the area used by symbol fonts easily. A form of
wording which states that readers might like to know that that code range is
often used for symbol fonts and may choose to use or ignore that use as they
think fit would be better than just not mentioning the matter, as not
mentioning the matter could lead to problems for new end users.

>Especially given that no software whatsoever supports the codes, and
>if it did, one would have to work with custom application software
>(that knows that U+XXXX means "BOLD") and/or with special
>Courtyard-Code-compatible fonts that know about Golden Ligatures, of
>which there are none in existence today.

The golden ligatures collection is of character glyphs and is a separate
thing from courtyard codes which are mostly control codes for formatting and
markup.

As far as no fonts or software being available today, well that might well
be true. However, that might change. Please know that I am not expecting a
major reorientation of the software industry, it is just a matter that if
someone does like to make a font which includes, in a Unicode compatible
manner, whole precomposed ligature glyphs which are accessible directly from
a code point, whether for ligatures such as ct or ch or for long s b or for
ppe, then there is available a set of code point allocations, which, while
not standard, are perhaps more likely to be consistent than any other set of
code point allocations which might otherwise be used. As for writing
software which recognizes courtyard codes, well maybe people might use
courtyard codes in computing generally, and if so, good, yet a primary
reason for introducing them is for using them in educational software
packages to be broadcast on digital television channels throughout the world
using the DVB-MHP (Digital Video Broadcasting - Multimedia Home Platform)
system. DVB-MHP uses Java and Java uses Unicode, so DVB-MHP programs which
are broadcast use Unicode. These telesoftware programs are a specialist
niche in broadcasting, yet, though I say it myself, telesoftware is an
extremely powerful computational technique and hopefully in the next few
years will begin to fulfil its potential. There is much to be done, yet it
is an exciting field and Unicode is a key feature in being able to use it
effectively throughout the world.

Thank you for your help.

William Overington

14 August 2002



This archive was generated by hypermail 2.1.2 : Wed Aug 14 2002 - 05:09:11 EDT