From: William Overington (WOverington@ngo.globalnet.co.uk)
Date: Mon Jun 02 2003 - 13:19:10 EDT
Peter Constable wrote as follows.
>William Overington wrote on 06/02/2003 01:06:25 AM:
>> I am wondering whether the range from U+F200 through to U+F2FF is being
>> by anyone for anything.
>This is a nonsense question. It should never matter to person A whether
>others are using particular PUA codepoints *unless* person A needs to
>interchange with person B, in which case A and B need to agree on that
>range if A intends to use it in interchaning with B.
Suppose person Ai and person Bi are both people with an interest in the
texts which contain these particular rare extinct ligatures.and wish to
exchange documents, which they have keyed themselves, over the internet and
view them using a package such as, say, Microsoft WordPad. I use Ai and Bi
to mean some particular pair of persons A and B. You wrote never, so one
counter example will disprove the generality of your claim. Neither Ai nor
Bi has facilities to make fonts, so they need to rely on having a font made
by a third party. They have a better chance of having a font to use if the
characters are added into an existing font which already has many other
characters in it, such as the basic latin alphabet and punctuation, so that
only the rare extinct latin letters represent special drawing work, rather
than the whole font. So, if they look at fonts such as, for example,
Code2000, Gentium and Junicode and observe which Private Use Area code
points are already in use within that font, then choose code points for the
rare extinct latin letters which code points are not used in the fonts at
which they look, then the chances of getting their chosen characters
implemented in those fonts will be increased.
For example, considering my own Quest text font. If Ai and Bi choose to
place their characters in the U+E7.. block or the U+EB.. block, then I would
not implement them in Quest text. However, if they place them in the U+F2..
block, then I might well try to have a go at adding them in. I recognize
that the lettering style of Quest text might not be appropriate to those
characters and Quest text might not be liked as a display face by Ai and Bi,
yet please allow me some latitude in this as I am trying to explain my
thoughts without speculating about the thoughts of some other person who
produces a font which might have a face design considered more appropriate
to the particular application.
So, bearing in mind my knowledge of some uses of the Private Use Area I
thought that the U+F2.. block looks prima facie reasonable, in that it
avoids code points used for Tengwar, for Phaistos Disc, for Ewellic, for
golden ligatures and courtyard codes, while also avoiding the very top end
of the Private Use Area. So, instead of simply sending a private email
response I posted to the mailing list in the hope that the readers of this
forum might like to help the process along of helping the gentleman be able
to use those rare extinct latin letters which interest him, in a practical
>Your question seems to be assuming the community of Unicode users at large
can share agreements on PUA assignments,
Well, surely they can if they choose to do so. Please note that I am not
saying should, must, will or whatever: you used the word "can" and I answer
> and in response I'd say that effectively you must assume that every last
PUA codepoint is being used by somebody somewhere.
I accept that that assumption needs to be made in generalized theoretical
considerations, yet in a practical situation of trying to get a few special
characters added into one or more existing fonts, it is highly relevant to
know which code points are already in use and which are not already in use
in a selection of fonts as that information can then be used so as to devise
a Private Use Area encoding scheme for the desired characters which has a
higher chance of being implemented.
> (And I can assure you that somebody has their own usefor F200..F2FF.)
Well, unless it is a secret or confidential it would be helpful if you could
please say what it is, as that information could be used to consider whether
a font needing both collections of characters would be likely to be needed
for one particular document produced by an end user.
2 June 2003
This archive was generated by hypermail 2.1.5 : Mon Jun 02 2003 - 14:10:23 EDT