From: Asmus Freytag (email@example.com)
Date: Tue Apr 05 2011 - 23:02:51 CDT
On 4/5/2011 7:18 PM, Mark E. Shoulson wrote:
> On 04/05/2011 09:31 PM, Asmus Freytag wrote:
>> Even when the underlying objects are identical (or "unifiable")
>> doesn't mean it follows that it's appropriate to unify different
>> representations of them on another layer (the writing layer).
>> Characters are an abstraction for the purpose of writing, and not
>> entities that directly represent real-world objects.
>> This fact alone would suffice to convince me that the decision to
>> encode any playing card symbols was carried out on an insufficiently
>> thought through basis and that one is best off abandoning the
>> existing symbols as "mistakes" (or compatibility characters that map
>> to other character set implementers "mistakes".)
> I don't know; I'm pretty sure I've seen playing-card images (and not
> ♠A) used in plain-text contexts. Yes, ideally I should find some
> examples and scan them for you. But encoding the cards does not
> strike me immediately as a mistake. And given those, the major arcana
> are defensible.
You can argue this on a lot of different levels.
My beef is with the *unification *implicit in the original encoded set.
I think it is on very questionable grounds (for the reasons I've given).
But that argument does not address the underlying question of whether or
not plain text symbols for playing cards are appropriate as such.
Before using the existing set (with it's decidedly questionable choice
for unification) as a precedent for adding additional characters, it
would make a lot of sense to sit back and find out what the actual user
experience has been with these already encoded characters. Now that they
exist, it should be possible to search for evidence of actual (and
non-contrived) usage of these characters in plain text.
If almost no usage is attested, then it would be a bad idea to proceed
with an extension at the *present* time. Also,it might be possible to
ascertain whether any working and widespread implementation of the
unification implicit in the encoding can be found (in other words,
whether the fonts people actually use *do* have the glyph variants, or
documents *do* have the markup to cause the right variants to be selected).
If there's no such implementation, then the unification should be
considered "untested". Adding characters to a set with "untested"
unification is also something that can wait.
There simply isn't any urgency here, unless the existing characters can
be shown to have been widely adopted and fully implemented (with the
variants available *and* in use).
> Not so sure about the hatching squares, though. Much as I love
> blazonry, I can't really picture them in a truly plain-text setting.
This archive was generated by hypermail 2.1.5 : Tue Apr 05 2011 - 23:06:53 CDT