This looks like a lot of work and it looks like it duplicates as lot of the
work in the "submitting new proposals" section of instructions on our
website and in the standard.
We are getting a large number of *informal* suggestions for proposals that
are more or less clearly inappropriate and spend some amount of time on the
list dealing with them. However, neither UTC nor ISO/IEC JTC1/SC2/WG2 are
receiving a large number of inappropriate *formal* proposals. In fact, both
groups few proposals that don't have active support or involvement of
people active in either or both bodies - ad those people don't need an FAQ.
I submit that, while an interesting excercise, such a FAQ is a solution in
search of a problem. Having a FAQ will not do anything to keep enthusiasts
from storming in with their latest brain child - since it's the nature of
these informal suggestions that people never read the FAQ (any FAQ) before
hitting the send button. (OK, maybe some do).
At the same time, there's the risk that we maintain TWO sets of information
on the same topic ("what's an acceptable proposal"), with all the
maintenance issues and issues of which text is the binding one.
I suggest that we either not do a FAQ, or do a very simple one of a few
(three?) proposals that have definitely failed (or would definitely fail)
as illustrations, just to establish the idea that there are inappropriate
proposals, and then firmly point to the *official* document that sets out
the criteria for creating a *formal* proposal.
My favorite examples are
Klingon (or any of the Latin ciphers/ movie scripts)
Hexadecimal digits or 'Decimal separator'
Any proposal asking for the rearrangement/removal of characters
The decimal separator is a clear example of coding something that is
already encoded using a different model and coding a character purely by
function, which Unicode tends to be leery of, esp. if it can't be visually
distinguished from existing characters.
I would definitely NOT like to see a detailed typology of proposals in a
FAQ, with detailed script classifications etc.
At 12:59 PM 7/3/02 -0400, Suzanne M. Topping wrote:
>I realized that I should probably turn an off-list discussion back to
>the list, as it's illustrating an area of difficulty. (See the bottom of
>this note for a partial discussion of what writing systems could/would
>In the "appropriate use" FAQ entry, how the heck can we state what is
>and isn't a suitable writing system for inclusion? Fictional scripts in
>some cases would be considered, in other cases would not. Historical
>scripts would in some cases be considered, in other cases not. Can
>people from the review committee give me some hard and fast rules for
>when something is thrown out?
>Is there some way of listing or detailing the criteria in a way which
>potential readers could determine where theire script or character
>stands? It could perhaps be presented in a table or matrix form, so that
>people could look through the criteria and say "yep, it's fictional, yep
>people currently use it, no, there are no fonts yet" etc. Or maybe a
>decision tree would be better, where the criteria forks.
>I guess the first thing I need to collect is the criteria... Here are
>some starters to get the ball rolling:
>--Is this an entire script or additions to an existing script?
>--Is the script fictional?
>--Is the script in use? (as determined how???)
>--Does the character(s) already exist in some other part of the
>--Is there a compelling reason for including a characther which would
>normally not be considered, due to legacy support issues?
>--Is the character a precomposed ligature which can be encoded using a
>sequence of existing character (possibly joined by ZWJ's)?
>--Is the character a precomposed "accented character" which can be
>composed using an
>existing character and one or more existing combining diacritics?
>--Is the character a clone of an existing character whose sole purpose
>is making a *logical* differentiation from some existing characters
>(e.g., hex digits looking
>identical to existing characters "0..9" and "A...F"; or a symbol for
>looking identical to Latin "m")?
>--Is the chracter a clone of an existing character whose sole purpose is
>*graphical* differentiation from some existing characters (e.g., a
>letter "t", disunified from Russian on the basis that italics looks
>different in the two languages)?
>--Is the character really a presentation glyph for a shape that can be
>obtained using regular characters in conjunction with ZWJ or ZWNJ?
>Again, additions, suggestions, or other help appreciated,
>At 15:35 -0400 2002-07-02, Suzanne M. Topping wrote:
> >Apologies, I was sloppy with my phrasing; of course a script is
> >"written". What I really meant was that the script is in current use in
> >some sort of written form, as opposed to existing somewhere,
> >historically, without being in current use.
>Fiction is written within history. Jonathan Swift made up an alphabet
>once I think in Gulliver's Travels. But it exists only in that book.
>Tengwar is different.
> > > It has to be used by people who want to interchange data safely
> >> it. Tengwar meets this criterion. Klingon didn't.
> >Is there some sort of metric associated with the concept of "people who
> >want..."? Is there some sort of a threshhold of number of people?
>Certainly not. The number of users of Old Permic, for instance, is
>probably a few dozen or less -- specialists. That doesn't mean Old
>Permic doesn't deserve encoding. Numbers aren't really in the
This archive was generated by hypermail 2.1.2 : Wed Jul 03 2002 - 12:12:55 EDT