\ac clash
Will Robertson
As you know unicode-math currently defines its symbols based on the information from the STIX fonts glyph listing. Of these, ten have two-letter control sequences:
"000B1 \pm plus-or-minus sign
"02118 \wp weierstrass p
"02208 \in set membership, variant
"0220B \ni contains, variant
"02213 \mp minus-or-plus sign
"0223E \ac most positive [inverted lazy s]
"02240 \wr wreath product
"02260 \ne /ne /neq r: not equal
"0226A \ll much less than, type 2
"0226B \gg much greater than, type 2
I'm only worried about one of these -- \ac. As far as I can see from symbols.pdf, the rest are already standard. However, \ac doesn't seem to have been used before, and also clashes with a macro defined by the acronym.sty package.
I'd like to rename it to avoid this conflict since \ac doesn't seem to have been used before now as a math symbol. Any suggestions?
-- Will
Apostolos Syropoulos
Hi Barbara and others,
As you know unicode-math currently defines its symbols based on the information from the STIX fonts glyph listing. Of these, ten have two-letter control sequences:
Why haven't you used something closer to Unicode? After all the STIX fonts are by no means a standard!
Regards,
Apostolos Syropoulos
Khaled Hosny
Well, when it comes to math, they, mostly, are
--
Khaled Hosny
Arabic localiser and member of Arabeyes.org team
Free font developer
Ross Moore
On 26/06/2010, at 2:34 AM, Will Robertson wrote:
> Hi Barbara and others,
>
> As you know unicode-math currently defines its symbols based on the information from the STIX fonts glyph listing. Of these, ten have two-letter control sequences:
>
> "000B1 \pm plus-or-minus sign
> "02118 \wp weierstrass p
> "02208 \in set membership, variant
> "0220B \ni contains, variant
> "02213 \mp minus-or-plus sign
> "0223E \ac most positive [inverted lazy s]
What I'd like to know is what does 'ac' stand for,
in the context of this symbol?
Is it to do with AC-current?
--- that is, the symbol is somehow meant to
represent a sine wave.
But then why is it curled at the ends?
and how does this relate to "most positive" ?
What does "lazy s" actually mean, and why is
there a need for an inverted version in Unicode?
> "02240 \wr wreath product
> "02260 \ne /ne /neq r: not equal
> "0226A \ll much less than, type 2
> "0226B \gg much greater than, type 2
>
> I'm only worried about one of these -- \ac. As far as I can see from symbols.pdf, the rest are already standard. However, \ac doesn't seem to have been used before, and also clashes with a macro defined by the acronym.sty package.
I'd say that acronym.sty made a mistake using \ac
as a name. \acr or \acro would have been better.
But the package has been around for awhile now,
so changing it probably isn't going to be the best option.
>
> I'd like to rename it to avoid this conflict since \ac doesn't seem to have been used before now as a math symbol. Any suggestions?
I'd agree with changing it for math.
But to what?
This is why we need answers to my questions above.
Some possibilities, based upon the unicode description
are:
\invlazys
\mostpositive
but these should not be the only ones used, if there
is another use that corresponds to the \ac name.
This is because macro names should correspond to the
concept being used, not to the appearance of the
character itself.
>
> -- Will
Hope this helps,
Ross
------------------------------------------------------------------------
Ross Moore ross....@mq.edu.au
Mathematics Department office: E7A-419
Macquarie University tel: +61 (0)2 9850 8955
Sydney, Australia 2109 fax: +61 (0)2 9850 8114
------------------------------------------------------------------------
Barbara Beeton
i'm in san francisco, away from my files,
so i can't check everything until i get
back to providence. however, ...
As you know unicode-math currently defines its symbols based on the
information from the STIX fonts glyph listing. Of these, ten have
two-letter control sequences:
[...]
I'm only worried about one of these -- \ac. As far as I can see from
symbols.pdf, the rest are already standard. However, \ac doesn't seem to
have been used before, and also clashes with a macro defined by the
acronym.sty package.
I'd like to rename it to avoid this conflict since \ac doesn't seem to
have been used before now as a math symbol. Any suggestions?
the source of this symbol is the sgml
public entity set "isoamsb", but,
contrary to what this appears to indicate,
it didn't come from an ams source, as far
as i remember. the name "\ac" is marked
in my table as being assigned by taco
hoekwater in an early font implementation
of many of the "new" symbols. i don't
have an electronic list of the original
sgml public entity sets, only paper, so
it really will have to wait until i get
back to providence to see whether "ac"
was the name assigned there -- it may be;
those names are often very compact and
cryptic.
i'll get back to this after july 3.
but i do agree that a conflict is a bad
idea.
-- bb
Apostolos Syropoulos
Well, when it comes to math, they, mostly, are
> Why haven't you used something closer to Unicode? After all the STIX fonts are
> by no means a standard!
In this particular case, "mostly" means absolutely nothing: either it is a standard or
it is not one. After all, these are fonts not some sort of cult!
A.S.
Apostolos Syropoulos
GR-671 00 Xanthi, GREECE
Will Robertson
Apostolos Syropoulos
Can you send me a reference to the Unicode standard for how their glyphs should be named in a tex-like system?
What's this supposed to mean? To the best of my knowledge Unicode exists independent
from what goes on in the TeX world, while the opposite is obviously not true.
Ross Moore
On 25/06/2010, at 5:55 PM, Barbara Beeton wrote:
> I'd like to rename it to avoid this conflict since \ac doesn't seem to
> have been used before now as a math symbol. Any suggestions?
>
> the source of this symbol is the sgml
> public entity set "isoamsb", but,
> contrary to what this appears to indicate,
> it didn't come from an ams source, as far
> as i remember. the name "\ac" is marked
> in my table as being assigned by taco
> hoekwater in an early font implementation
> of many of the "new" symbols. i don't
> have an electronic list of the original
> sgml public entity sets, only paper, so
http://www.w3.org/TR/REC-MathML/chap6/ISOAMSB1.html
It starts out as:
>>> ISOAMSB Characters alphabetically
>>>
>>> Added Math Symbols: Binary Operators
>>>
>>> Character Name Code Description Group Alias
>>> ==================== ==== ========================================== ======= =========
>>> ac E207 most positive ISOAMSB
>>> acE E290 most positive, two lines below ISOAMSB
>>> amalg E251 amalgamation or coproduct ISOAMSB
>>> barvee 22BD bar, vee ISOAMSB
>>> Barwed 2306 logical and, double bar above ISOAMSB
>>>
and the following
http://www.orwell.ru/test/ISO/e?isoamsb
http://www.w3.org/TR/2009/REC-MathML2-20090303/isoamsb.html
give example glyphs:
>>> Unico. Chr Entity Description
>>> ------ --- -------- --------------
>>> 0x223E [∾] ac INVERTED LAZY S, ∾
>>> 0x223E [∾] acE INVERTED LAZY S with double underline (0x0333), ∾̳
>>> 0x2A3F [⨿] amalg AMALGAMATION OR COPRODUCT, ⨿
>>> 0x22BD [⊽] barvee NOR, ⊽
>>> Name Unicode Glyph Unicode Name Description Aliases
>>> ac 0223E
Will Robertson
Minor comment: considering there is also \sinewave, I think \ac is probably a bad name in itself. More below:
On 26/06/2010, at 8:37 AM, Ross Moore wrote:
> It would seem that it is a glyph, used for forming parts
> of different characters.
> So there needs to be a macro name to access that glyph,
> which should be different to the name used for the character
> as the glyph by itself with \mathbin spacing.
>
> e.g.
> \ac@glyph refers to U+0223E as a \mathord
> (or whatever code-point in a legacy font)
>
> \invlazys refers to U+0223E as a \mathbin
> (or whatever code-point in a legacy font)
>
> \acE@entity
> can be defined as an internal macro which constructs
> a representation of the entity either as
> U+0223E U+00333 when the font has appropriate glyphs;
> or as a box-construction placing an '=' sign or
> a pair of lines beneath what \ac@glyph supplies.
>
> Then a package might well choose to define:
> \let\ac\invlazys
> \let\acE\acE@entity
I don't disagree with these ideas here but I wonder at the efficiency. Is memory cheap enough that we can tie up ~3000 rarely-used control sequences/mathchars?
The point with \acE is more troubling to me. If there are a number of math symbols that are used as "glyphs" but only constructed of composite glyphs, unicode-math should really be defining control sequences so they can be conveniently used. Where would I go about looking for such things?
-- Will
P.S. Anyone coming to TUG, I'll be in the lobby of the Sir Francis Drake for pretty much the whole weekend (working when I should be sightseeing, I suppose). Come say hi :)