Symbol for an upside down capital L, pointing to the right?
Asmus Freytag (t)
asmus-inc at ix.netcom.com
Tue Dec 29 07:09:13 CST 2015
On 12/29/2015 4:01 AM, Michael Everson wrote:
> On 28 Dec 2015, at 23:48, Asmus Freytag (t) <asmus-inc at ix.netcom.com> wrote:
>> Rather than engage in reflexive ad-hoc unification like this, it would be useful to find out why U+2142 was disunified from TOP RIGHT CORNER and any other symbols having two strokes at right angle with one of them pointing down.
> I think the letter-like symbols were added. Not “disunified”. The default state is not that “everything is already encoded”.
> 2142 and characters near it were added in Unicode 3.2, about the same time that the mathematical styled alphabets were added (3.1). I think you were involved with a lot of that Asmus.
I certainly was more active then.
If you follow the documents from the time, you'll find that U+2142 came
from a set of letterlike characters that were (together) part of a set
of mathematical characters being added. And, tellingly, it was not the
only L shape in the set.
As the goal was to cover existing sets in full, and the source was
entity or character sets, rather than "examples in print", the analysis
on some of the individual characters didn't go as deep as it has
recently with the addition of one-off extensions.
For example, the rationale for inclusion was the membership in one (or
more) of the sets; there was no independent verification that each of
the member of these sets would individually merit encoding, and no
examples of usage of individual characters were collected, the sets as
such being well-established.
Still, a classification was carried out, based on information of usage
available to various practitioners that were called as experts or were
expert contributors to the proposal - but again, without further
documentation for each individual character. The characters later
encoded at positions U+2142 and U+2143 where identified as "normal",
that is as variables (aka letter symbols) as opposed to operators,
delimiters or the like.
As the occur in the context of a turned capital G (U+2141) and inverted
capital Y (U+2144) (all sans-serif) their classification appears to be
well-motivated. From the latest stages of the proposal documents it's
not fully clear whether they were grouped in the source character sets,
or whether their presentation in the current order already reflects a
"sorting" of like characters for purposes of the final encoding.
However, the character sets investigated did include other symbols
(floors, ceilings) that were "L" shaped, and presumably the letter-like
symbols existed, from the start, in contrast. Together with the other
sans-serif characters, even without documentation where in mathematical
notation these are employed, makes it unlikely that their identification
should be questioned (Again, the proposers may well have had more
knowledge on usage than they documented, or some information may have
only been present in hardcopy form - at the time not an uncommon
Anyway, unless we have specific reason to doubt that the classification
of "normal" is indeed correct and that the shapes really are letters,
let's assume that they were correctly identified and encoded as such.
If we now have a putative mirror image of one of these symbols, we need
to know whether it is a letter, or some other symbol, perhaps an
operator or a delimiter. If neither, then we can exclude unification
with the corners, floors and ceilings, etc.
More information about the Unicode