From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Thu Mar 12 2009 - 08:33:07 CST
> De : unicode-bounce@unicode.org
> [mailto:unicode-bounce@unicode.org] De la part de Peter Constable
> Envoyé : mercredi 11 mars 2009 16:10
> À : Johannes Bergerhausen; unicode Unicode Discussion
> Objet : RE: India seeks Rupee status symbol
>
>
> When you say "the Indian writing tradition", would that be
> Devanagari? Gujarati? ...
>
> From one perspective, Latin has a convenient neutrality.
Latin is not very neutral because the letter R has an inherent
directionality and not mirrorable, for Indian users of the Arabic script
(notably Urdus). Other symbols also have the problem that they feature
design features that are significant to some of the Indian scripts.
So instead of looking at letters, or even digits, you should consider
symbols that are really neutral: mathematical symbols. India has a long
tradition for supporting mathematics and was the primary promoter of the
zero and the positional system for noting numbers in arbitrary ranges and
precisions.
That's why I think that a design based on a mathematical symbol, and that
features a self-mirroring glyph that can be displayed and interpreted the
same way in RTL and LTR contexts (not just with a numeric amount, but also
as a standalone symbol within text) would be a better winner.
My informal proposal (infinite symbol with double danda) symbolizes all
this. The association to the rupee may not seem evident, but the glyph by
itself would be enough to recognize it, and to associated it to some numeric
value by itself (note that the infinite symbol also states by itself
stability:
- there's nothing more valued than infinity except infinity itself.
- It's also associated to stability (because it is not bound and has indeed
no end) : it's meant to be used without any termination in time.
- the infinity symbol is already recognized internationaly (in almost all
books about mathematics).
- adding the double danda bar on top of it also reuses the concept of the
double bar for currency symbols : the symbol would be easily recnogized as a
currency even by those that have not been exposed to it before, when they
see it in some text near any numeric amount.
- the double danda bar opposes the concept of infinity and fixes an exact
end to a sentence, and for a currency amount it means a fixed and precise
term: you can't add more to it.
- using the two vertical bars of the dandas instead of horizontal bars
respects the traditions in Indic scripts for punctuations.
Note that existing currency symbols do not state if these bars should be
horizontal (euro, won, yen, pound, and even the Vietnamese dong) or vertical
(dollar/peso), or if there should be one or two bars (the dollar symbol is
ambiguous about this, and it is merely seen as a choice in font design,
adaptation to smaller font sizes or heavier boldness).
Another needed feature is that such symbols should be drawn with a limited
number of strokes or curves, for easier and fast drawing by hand : 3 is
almost standard, with only 1 curved line at most (there may be other glyphic
features for typed characters, like serifs, but these should not be
necessary for recognizing the symbol itself). My proposal respects this 1
closed curve and two strokes).
I also don't know what the "Indian National Language Script" ("INLS") means
in this competition, notably because India has a dozen of national languages
(possibly with only English and Hindi used widespread, but not universally
within the country).
The terminology speaks about a "language script" which is even more
confusive, given that neither Hindi or English are not written with their
own script. If the symbols aims to represent all the country, it has to be
acceptable by all communities, not just those using Hindi or English, or
even those using an official Indian language written with a Northern or
Southern Brahmic script (Bengali, Gudjrati, Tamil, Kannada...) or even
Tibetan, Buthanese and Nepalese. It should be usable without needing any
adaptation also in Urdu and Arabic (and probably also in Pashto and in some
Persian and Turkic communities).
If this means that the symbol should be immediately printable using common
existing fonts using some simply overlay layout (like the Euro symbol that
could be drawn as C plus an equal sign) then my proposal respects this: the
infinite symbol (U+221E) is already very common and you can easily overlay a
single or double vertical bar on top of it if you still don't have fonts
featuring the new symbol.
The requirement about the keyboard seems also a non issue for this project
and cannot be a factor influencing the choise for the winning glyph:
independantly of the symbol that will be chosen, it will need the same
single keyboard assignment, and new keyboards will also need to be built to
exhibit the chosen glyph. The assignment on a specific key will be dependant
on the various existing keyboard layouts for the various Indian scripts, and
the same work will be needed to adapt those keyboards to the new character,
and to update the keyboard driver so that they will generate the correct
code for the abstract character (when it will be encoded by Unicode and in
ISO 10646, something that the Indian government will have to sponsor to get
such assignment).
It will take some years to have it fully implemented, and for that reason
there will still remain the possibility to use paliative methods (like
overlaying multiple existing glyphs, or using other existing symbols or
abreviations, or the existing 3-letter ISO currency code). This will not be
braking the adoption by India for minting its national coins or printing its
bills (possibly with enve more advanced and complex design features like
ornaments, ike those seen on existing printed currencies for digits).
If the "INLS" requirement means that it should map unambiguously to an
existing standard, then I cannot see how the new unique symbol can cope with
this standard and maintain its distinct meaning.
Independantly of the chosen glyph, there will exist the need for a unique
abstract character with its own distinct encoding, and an update for
existing standards to support it (mainly ISCII if this is possible, and
Unicode/ISO/IEC 10646), but also in the repertoires of standard glyph names
(for PostScript fonts), plus the definition of compatibility mappings
(one-to-one or one-to-many) for other legacy standards that may still be
needed in India (but it will not be possible for these compatibility
mappings to preserve the unique identity if the target standard cannot be
updated to include the new abstract character). For the worst case (mapping
to 7-bit invariant subset of ISO 646), the only viable mapping will be to
use the existing ISO currency code (possibly within punctuation backets in
some contexts).
This archive was generated by hypermail 2.1.5 : Thu Mar 12 2009 - 08:36:13 CST