From: Jukka K. Korpela (jkorpela@cs.tut.fi)
Date: Tue Jul 25 2006 - 01:17:50 CDT
On Tue, 25 Jul 2006, Philippe Verdy wrote:
>>> It seems to me that the character as proposed is a symbol found as a text
>>> element in a wide variety of printed and on-line sources.
>>
>> By the way, could someone comment on the use of such symbols in
>> right-to-left scripts? It seems to me that the proposed shape postulates
>> left-to-right and top-to-bottom writing direction. This is one of the
>> reasons why it looks like an iconic image rather than something definable
>> as character.
>
> Why do you think that? Couldn't the symbol be marked as mirrorable, and
> sensitive to the writing direction, like the parenthese punctuation
> signs?
It could, but I think such issues haven't been considered; that's way I
raised them. The abstract idea of an external link can be conveyed in
different ways, and images with a square and north-east-pointing arrow
looks like an iconic image related to particular writing directions.
> The exact form of the arrow is also variable (sometimes thick,
> sometimes thin, sometimes multicolored, sometimes outlined, sometimes
> not...); what is really encoded is a abstract meaning.
Unicode does not encode abstract meanings but characters, which sometimes
have a very specific meaning, sometimes a very broad range of meanings
(or, at the extreme, just their shape as the meaning) - and usually
something in between.
As noted in the discussion, _essentially different_ graphic symbols are
used to present the abstract meaning of an external link. Some people
think that one particular class of such symbols can be recognized as a
character and should be encoded as a character. However, I would say that
this would mean that the Unicode Standard would _create_ a new character
rather than encode a character.
For example, Google image search for "external link icon",
http://images.google.com/images?q=%22external+link+icon%22
shows a rather diverse collection of images. It's of course just a small
sample (with some false hits as usual).
-- Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/
This archive was generated by hypermail 2.1.5 : Tue Jul 25 2006 - 01:23:59 CDT