How to make Angstrom in Latex?
Louis Vosloo
> It's not "Angstrom", it's "Ångström". I.e, the first letter is
> capital "Å", not "A" or "A$^\circ" or anything such. Then penultimate
> letter is "ö", not "o". If you are using a modern 8-bit TeX you just
> enter them as such. If not, they are entered as \Aa{} and \"o
> respectively
> (but they do not produce the typographically correct glyph unless
> you use the ec fonts!). These two distinct letters are part of the
You mean they are somehow `wrong' in Adobe Times, Helvetica and
Courier for example? Or any of the other 90,000 fonts in Type 1
format?
Donald Arseneau
>> (but they do not produce the typographically correct glyph unless
>> you use the ec fonts!). These two distinct letters are part of the
>
>You mean they are somehow `wrong' in Adobe Times, Helvetica and
>Courier for example? Or any of the other 90,000 fonts in Type 1
>format?
90,000 fonts can't be wrong! Just ask Bo's brother, Text Ascii.
Donald Arseneau as...@triumf.ca
Bo Thidé
>
> Bo Thidé wrote:
>
> > It's not "Angstrom", it's "Ångström". I.e, the first letter is
> > capital "Å", not "A" or "A$^\circ" or anything such. Then penultimate
> > letter is "ö", not "o". If you are using a modern 8-bit TeX you just
> > respectively
> > (but they do not produce the typographically correct glyph unless
> > you use the ec fonts!). These two distinct letters are part of the
>
> You mean they are somehow `wrong' in Adobe Times, Helvetica and
> Courier for example? Or any of the other 90,000 fonts in Type 1
> format?
It's not the font, its the rendering. The Swedish (*UNACCENTED*, mind
you) letter Å is something entirely different than an A with a circle
operator on top! All the professional fonts I have checked have a
proper, single-glyph Å. In many Swedish fonts, the "A"-part of the Å
glyph is substantially smaller (lower) than the A letter in order to
make the
entire glyph for the letter Å be of equal height as the height of the
letter
A (by itself).
The problem occurs if you think of this Å as a combination of the letter
A
(with which it has just as little relationship as Q has with O, say) and
a
"small circle". The Å is a letter in its own right, just as Q and O
are.
Another example:
Think of the English word 'coordinate'. In old, high-quality
British dictionaries this is written as 'co\"ordinate'. Here, the
double-dot-accent '\"o' means an 'o' with a trema (compare the trema in
the French word for Chistmas: 'noël') to emphasise that the 'oo'
combination should be prononunced as 'oh-oh' rather than as the 'oo' in
'cool'. The glyph produced with this '\"o' construct in the word
'co\"ordinate' is *NOT* the same glyph as the Swedish (or German and
Finnish) 'ö'! Good typography should make a clear distinction between
the two. Just as the 'ö' in turn is different from the Hungarian
umlaut-o,
a difference that also should be visible in the typography (and indeed
is in TeX).
This 'ö' versus 'o'-with-trema-accent ambiguity becomes interesting in
the Swedish spelling of the capital of South Korea. It is usually
written Söul. But from this it is not clear whether the 'ö' is really
the Swedish letter 'ö'
or an 'o' with a trema accent. In the first case the Swedish
pronounciation
would be "Siroul" (in a kind of British way; in order not to open yet
another
can of worms I refrain from trying to represent this in the
international
phonetic alphabet ... :-> ), whereas in the last case, the
pronounciation
should be (again in a kind of British way) "Soh-oul".
There is nothing, in principle, that prevents you from putting a trema
(two dots) atop the 'ö' despite the fact that it already has two
dot-like things above its 'o' part! Generally speaking, the trema
should be bigger than the
two dot-like things that are an integral part of the (again,
*UNACCENTED*)
'ö'. Unless you exercise extreme care, the combining of an 'A' with a
ring
to produce an 'Å' or an 'o' and two dots to produce an 'ö', will only
get them *APPROXIMATELY* correct. This is true for all the "extra"
*UNACCENTED* Swedish letters "å, ä, ö, Å, Ä and Ö.
The bottom line is that in LaTeX you should enter the Ångström symbol as
\AA{}
(or as Å, if you use latin1 input encoding) and not as \r{A}, A$^\circ$
or any such "combination".
Bo
--
^ Bo Thidé (Ass Prof)------------http://www.wavegroup.irfu.se/~bt
|I| Dept. of Space and Plasma Physics, SE-755 91 Uppsala, Sweden.
|R| Office: Villavägen 3. Phone: (0)18-4717269. Fax: (0)18-554917
/|F|\ Mobile Phone: 0705-613670 Home Phone: [+46] 18-554184
~~U~~ E-Mail: mailto:b...@plasma.uu.se Mobile: mailto:cal...@irfu.se
Bo Thidé
>
> In article <35AD6AC1...@YandY.com>, Louis Vosloo <sup...@YandY.com> writes...
> >> you use the ec fonts!). These two distinct letters are part of the
> >
> >You mean they are somehow `wrong' in Adobe Times, Helvetica and
> >Courier for example? Or any of the other 90,000 fonts in Type 1
> >format?
>
>
> Donald Arseneau as...@triumf.ca
I am *NOT*, repeat *NOT* trying to say that the fonts are wrong!!!!
I am trying to be serious here and try my best to draw people's
attention to the fact that the "Angstrom" symbol should be typeset as a
*CORRECTLY* rendered Swedish Å (the guy was a predecessor of mine in
physics,
for God's sake!), and not as a glyph for the letter A plus the
glyph of another letter or symbol ('o', \circ or whatever) on top of it!
David Carlisle
> The bottom line is that in LaTeX you should enter the Ångström symbol as
> \AA{}
I repeat, putting in spurious braces will remove any kerns specified
in the font. So use \AA ngström not \AA{}ngström.
David
Bo Thidé
I was talking about the single-glyph '\AA ngström' *symbol*.As such, kerning
is less important and either \AA or \AA{}
will do (there are som practical advantages with the latter).
You say: "use \AA ngström'". That's the correct way of entering
his name ("Ångström" is an even better alternative, IMHO). But
this *name* can never be used as the physics *symbol*. But I
don't think you mean that.
Bo
PS. Those of you who are interested can compare the rendering
(would the English term "cut" or "cutting" be OK here? I am only
familiar with the Swedish typography vocabulary and there
it's called "skärning") of the Å in the Berling Roman font
at Adobe's own site
http://w1000.mv.us.adobe.com/type/browser/F/P_221/F_BERM-10005000.htmlwith)
with that at the site
http://www.software.net/pages/adobe/example.htm?BERM10005000Berling-Roman
where software.net claims the font to be published by Adobe Type Faces
(whatever that means).
Then you tell us all which of these two, quite different, glyphs
is supposed to be a "real" Å and which is something else...
--
^ Bo Thidé (Director of Research)-----http://www.wavegroup.irfu.se/~bt
|I| Swedish Institute of Space Physics (IRFU), S-75591 Uppsala, Sweden
|R| Office Phone: [+46] (0)18-303671 Office Fax: [+46] (0)18-403100
/|F|\ Wireless Cellular Phone: 0705-613670 Home Phone: [+46] (0)18-554184
~~U~~ E-Mail: mailto:b...@irfu.se Wireless e-mail: mailto:cal...@irfu.se
Bo Thidé
> with that at the site
>
Sorry, Netscape cut-and-paste goofed. The correct URL
ishttp://w1000.mv.us.adobe.com/type/browser/F/P_221/F_BERM-10005000.html
Louis Vosloo
> I am *NOT*, repeat *NOT* trying to say that the fonts are wrong!!!!
> I am trying to be serious here and try my best to draw people's
> attention to the fact that the "Angstrom" symbol should be typeset as a
> *CORRECTLY* rendered Swedish Å (the guy was a predecessor of mine in
> physics, for God's sake!), and not as a glyph for the letter A plus the
> glyph of another letter or symbol ('o', \circ or whatever) on top of it!
There are two entirely separate issues:
(1) The Angstrom character is not an `accented' A.
(2) The glyph to represent this character should have an attached ring.
IMHO (2) does not follow from automatically from (1).
When I look at some fonts installed on my system, I see all kinds of
variations. For example, the ring is attached in Lucida Console but not
attached in Lucida Sans Unicode. These are both, of course, fonts from
Bigelow & Holmes. The difference is that Lucida Console lives under
severe constraints for on screen use and has `scrunched down' accented
upper case characters. It also has an attached ring in the Angstrom
symbol perhaps because there is no space for a detached ring.
Arial has an attached ring, Arial Black has a detached ring.
Times-Roman has a detached ring, while Times New Roman has an attached ring.
Charter BT, MT Coronet, Adobe Courier, IBM Courier, MonoType Courier,
Helvetica 95 Black, Helvetica Light, Kepler, Marigold,
Sanvito Roman have detached rings.
Lucida Bright, Lucida Casual, Lucida Handwriting, and Lucida Calligraphy
MT Corsiva, and Utopia have attached rings.
The curious mix indicates either that typographers are uninformed,
or that they consider the attachment of the ring a design issue, not
one of semantics.
I am not trying to be argumentative. I want to know! So it would be
great if we could also hear from other people on this topic,
particularly people from Sweden, and particularly from desgners of type.
Gary Munch
http://wwww.adobe.com/type/browser/F/P_221/F_BERM-10005000.html
for the larger ring, grazing the tip of A; (I'd make it this way, now. . .)
>Charter BT, MT Coronet, Adobe Courier, IBM Courier, MonoType Courier,
>Helvetica 95 Black, Helvetica Light, Kepler, Marigold,
>Sanvito Roman have detached rings.
>
>Lucida Bright, Lucida Casual, Lucida Handwriting, and Lucida Calligraphy
>MT Corsiva, and Utopia have attached rings.
>
>The curious mix indicates either that typographers are uninformed,
>or that they consider the attachment of the ring a design issue, not
>one of semantics.
>
>I am not trying to be argumentative. I want to know! So it would be
>great if we could also hear from other people on this topic,
>particularly people from Sweden, and particularly from desgners of type.
I had thought the matter one that at last had been resolved by our new
technology; to whit, that now we can set any character with any height
(within some limits, of course) and that the issue of the ring over A --
which has always been the tallest of the accented* capitals -- could be
solved in a way that more closely represents what the scribes who invented
the thingie did; add a nice circle to the top of an A. (The å is of course
best with a floating ring, eh?)
The solutions of the past, in type, were often to -- for not just Å Aring
-- shorten the base cap. For many designers, this is a painful distortion
of the letter; being able now to use the full character, with the full
accent, or diacritic, or fully composed glyph, is a good thing.
Which do your prefer, then, as a reader of Swedish; that the letter be
shrunk to fit the other cap+accent height, or full height and full ring?
*Accented in the sense that there is an addition to the base glyph; with
acknowledgment that the full accented letter is a letter in itself, and not
a mere variant, and is a character in the alphabet along with the
unaccented character.
--
Gary Munch
gmu...@pipeline.com
http://members.aol.com/munchfonts/
Type & Type Design
Bo Thidé
>
> In article <35AD6AC1...@YandY.com>, Louis Vosloo <sup...@YandY.com> writes...
> >> (but they do not produce the typographically correct glyph unless
> >> you use the ec fonts!). These two distinct letters are part of the
> >
> >You mean they are somehow `wrong' in Adobe Times, Helvetica and
> >Courier for example? Or any of the other 90,000 fonts in Type 1
> >format?
>
> 90,000 fonts can't be wrong! Just ask Bo's brother, Text Ascii.
>
> Donald Arseneau as...@triumf.ca
Well, Don, since you seem to be the world's leading expert on
Scandinavian
fonts, maybe you, or your brother, Font Expert II, can tell me which of
the
two very different renditions of the <A-circle> glyphs in the Berling
Roman font at
http://w1000.mv.us.adobe.com/type/browser/F/P_221/F_BERM-10005000.html
and
http://www.software.net/pages/adobe/example.htm?BERM10005000Berling-Roman
is a (relatively) correct rendition of the Scandinavian letter "Å" and
which is the entirely different "A-with-a-circle-above-it" or even a
nonsense
character...
Bo Thidé
> Bo Thidé wrote:
>
> > I am *NOT*, repeat *NOT* trying to say that the fonts are wrong!!!!
>
> > I am trying to be serious here and try my best to draw people's
> > attention to the fact that the "Angstrom" symbol should be typeset as a
> > *CORRECTLY* rendered Swedish Å (the guy was a predecessor of mine in
> > physics, for God's sake!), and not as a glyph for the letter A plus the
> > glyph of another letter or symbol ('o', \circ or whatever) on top of it!
>
> There are two entirely separate issues:
>
> (1) The Angstrom character is not an `accented' A.
> (2) The glyph to represent this character should have an attached ring.
>
> IMHO (2) does not follow from automatically from (1).
>
> When I look at some fonts installed on my system, I see all kinds of
> variations. For example, the ring is attached in Lucida Console but not
> attached in Lucida Sans Unicode. These are both, of course, fonts from
> Bigelow & Holmes. The difference is that Lucida Console lives under
> severe constraints for on screen use and has `scrunched down' accented
> upper case characters. It also has an attached ring in the Angstrom
> symbol perhaps because there is no space for a detached ring.
>
> Arial has an attached ring, Arial Black has a detached ring.
>
> Times-Roman has a detached ring, while Times New Roman has an attached ring.
>
> Charter BT, MT Coronet, Adobe Courier, IBM Courier, MonoType Courier,
> Helvetica 95 Black, Helvetica Light, Kepler, Marigold,
> Sanvito Roman have detached rings.
>
> Lucida Bright, Lucida Casual, Lucida Handwriting, and Lucida Calligraphy
> MT Corsiva, and Utopia have attached rings.
>
> The curious mix indicates either that typographers are uninformed,
> or that they consider the attachment of the ring a design issue, not
> one of semantics.
>
> I am not trying to be argumentative. I want to know! So it would be
> great if we could also hear from other people on this topic,
> particularly people from Sweden, and particularly from desgners of type.
>
>
You are right that 2) does not automatically follow from 1).
But:
a) The Ångström symbol is the first letter in Anders Ångström's
second (family) name. (Incidentally, the first letter in his first name
is an "A", an entirely different letter.)
b) Hence, the glyph representing the Ångström unit and the glyph that is
used for representing the first letter in his family name must the one
and the same identical glyph! This is no different from the fact that
the symbol the SI unit of electric current is "A" just because it is
the first letter in the name "Ampère"). Everybody would react if
somebody instead used the glyph for \Lambda as the symbol for the (SI)
unit of electric current, just because the glyph of the Greek letter
\Lambda "looks almost the same" as the glyph of the letter A in the
particular font in use!
Now, this Å glyph, as you point out, varies from font to font, and even between
shapes of the same font. However, the important thing that *in one and the same
font* you cannot use one glyph for the first letter of Ångström's name, and *a
different one* for the Ångström symbol!
If you are not aware of this subtlety, you can make such mistakes. And
since I see such mistakes being made over and over again, I wanted to
be as precise as I could be (I am not an expert font maker, but I have
typographical expertise in the family whom I often consult). One
very common mistake is that people don't use the Å (as in Ångström)
glyph for his unit, but instead something different, for instance the glyph
A plus another "small circle" glyph (o, or \circ or...) on top of
it. Even an untrained Scandinavian eye will spot this irritating
difference immediately. And rule number one in scientific typography
is to not irritate.
Furthermore, my experience tells me that *in the one and the same font
and shape*, the glyph that represents the Scandinavian letter Å (again
as in Ångström) typically has the "circle" part on top of the "A" part
placed at a different height above the "A" part and with a different shape than
an "A" with a circle operator on top.
What would you say if, in the same vein, somebody consistently used the glyph
for "O" plus the glyph for "\tilde" to produce the capital "Q"? I am sure you
wouldn't call the result "professional, high-quality typography". This example
is not entirely academic because, just as the letter "Å" is not used reguarly
in common English, the letter "Q" is not ised regularly in common Swedish.
Neither is W or Z, by the way.
I hope I presented my arguments in a reasonably clear way. If not, let's
continue off-line.
Bo
--
Gernot Katzer
I think this thread is very interesting, providing information
about subtleties of typography (if we were not interested in such
subtleties, we probably shouldn't use TeX at all).
Yet I think Bo is mistaken when he writes:
> What would you say if, in the same vein, somebody consistently used the glyph
> for "O" plus the glyph for "\tilde" to produce the capital "Q"? I am sure you
> wouldn't call the result "professional, high-quality typography". This example
> is not entirely academic because, just as the letter "Å" is not used reguarly
> in common English, the letter "Q" is not ised regularly in common Swedish.
> Neither is W or Z, by the way.
Your point seems to be that A and \AA\ are different letters which just
resemble each other by chance. Historically, this isn't true: At a point
in language history, it was considered neccessary to distinguish between
two types of `A', since one of them had been shifted in articulation.
So, the relation between a and \aa\ is similar to that of u and v or
i and j: The Romans used only u and u and pronounced the glyph
according to whether the next letter was a consonant (->u,i) or a vowel
(->v,j). Later, two symbols were invented to have unambigous symbols for
all four different sounds.
Anyway, this doesn't change much of Bo's line of arguing.
One question has not been asked nor answered: If I want to have a
graphic distinction between Swedish \AA\ and some ringed A character
I have invented in the transcription of some exotic language, how
can I do so in T1 encoding? It is my understanding that all
\r A, \AA and (if appropriate input encoding is selected) Å
give the same glyph, \char"C1. How can I get A-with-ring?
Gernot
Chris Maden
> Well, Don, since you seem to be the world's leading expert on
> Scandinavian fonts, maybe you, or your brother, Font Expert II, can
> tell me which of the two very different renditions of the <A-circle>
> glyphs in the Berling Roman font at
> http://w1000.mv.us.adobe.com/type/browser/F/P_221/F_BERM-10005000.html
> and
>
> http://www.software.net/pages/adobe/example.htm?BERM10005000Berling-Roman
> is a (relatively) correct rendition of the Scandinavian letter "Å"
> and which is the entirely different "A-with-a-circle-above-it" or
> even a nonsense character...
I don't know if they were intended to be two different glyphs. A
number of things about the graphic at the latter site, especially the
quality of the ring and the mis-centering of the circumflex over the I
and the umlaut over the U, indicate to me that it was made somewhat
clumsily with blindly combining diacritic "dead key" characters, while
the latter used actual accented-character glyphs.
I'd never heard that there was a text ring-accented A, as distinct
from A-with-ring, but I don't speak any Scandinavian languages. I did
spend three years of my career discussing weirdnesses of various
linguistic tricks with a Scandinavian font expert, and he never
mentioned it. Maybe he took it for granted.
-Chris
--
<!NOTATION SGML.Geek PUBLIC "-//Anonymous//NOTATION SGML Geek//EN">
<!ENTITY crism PUBLIC "-//O'Reilly//NONSGML Christopher R. Maden//EN"
"<URL>http://www.oreilly.com/people/staff/crism/ <TEL>+1.617.499.7487
<USMAIL>90 Sherman Street, Cambridge, MA 02140 USA" NDATA SGML.Geek>
Louis Vosloo
> One question has not been asked nor answered: If I want to have a
> graphic distinction between Swedish \AA\ and some ringed A character
> I have invented in the transcription of some exotic language, how
> can I do so in T1 encoding? It is my understanding that all
> \r A, \AA and (if appropriate input encoding is selected) Å
> give the same glyph, \char"C1. How can I get A-with-ring?
I guess you can't separate them using T1 encoding.
Just like there are not separate glyphs for the
German Umlaute versus accented characters in other
languages using dieresis. You'd have to explicitly
construct one or the other by composing base and accent
in order to get different spacing or some other difference
in visual appearance.
Fortunately there does not appear to any use of Aring in
any language distinct from the Aring glyph in Swedish
(Yes, you can it something else if you like, but the
standard PS glyph name is Aring, just like the standard
PS glyph name for `a' with Umlaut is `adieresis,' even if you can
argue that this is not a character with a `diacritical' mark).
And if you are using A with ring to indicate a quaternion in
math then you are introducing an unneccessary ambiguity that
can be easily avoided by using lets say Lambda with ring instead.
David Carlisle
> I have invented in the transcription of some exotic language, how
> can I do so in T1 encoding? It is my understanding that all
> \r A, \AA and (if appropriate input encoding is selected) Å
> give the same glyph, \char"C1. How can I get A-with-ring?
see the \realring example posted earlier in this thread
David
Bo Thidé
>
> Bo Thidé wrote:
>
> about subtleties of typography (if we were not interested in such
> subtleties, we probably shouldn't use TeX at all).
>
> Yet I think Bo is mistaken when he writes:
>
> > for "O" plus the glyph for "\tilde" to produce the capital "Q"? I am sure you
> > wouldn't call the result "professional, high-quality typography". This example
> > is not entirely academic because, just as the letter "Å" is not used reguarly
> > in common English, the letter "Q" is not ised regularly in common Swedish.
> > Neither is W or Z, by the way.
>
> resemble each other by chance. Historically, this isn't true: At a point
> in language history, it was considered neccessary to distinguish between
> two types of `A', since one of them had been shifted in articulation.
>
> So, the relation between a and \aa\ is similar to that of u and v or
> i and j: The Romans used only u and u and pronounced the glyph
> according to whether the next letter was a consonant (->u,i) or a vowel
> (->v,j). Later, two symbols were invented to have unambigous symbols for
> all four different sounds.
You are absolutely right. And I think I am right too. In fact, if you
go back in this thread, you will see that in one of my earlier postings
I did say that A and Å are connected *historicically* (in fact the Å
started
out as "Ao"). But since a few hundred years or more, å ä and ö are in
Swedish considered unique, distinct, unaccented letters in their own
right (which, reportedly, is *not* the case in German, even if
typographically ä and ö should be identical in the same font in the two
languages; å is not used in German -- yes, it *is* complicated!).
Therefore, å ä and ö (but neither é nor è) have the status of full
members of the Swedish alphabet (which, hence, consists of 29 letters,
and not only 26 as the English alphabet). It is quite possible, as
recognized by Unicode 2.1, to accent any of å å ö Å Ä and Ö with a
grave, tilde etc.
So, this historic fact aside, I think my O<-->Q example is relevant or,
at least, illustrative; I could, as you allude to, have taken i and j
instead,
to show how unprofessional it is to use the glyph for one letter (O/i)
and combine it with another glyph (~/comma - or similar) to construct a
combined glyph for another letter in the alphabet (Q/j); you get the
point. Yet, this
is exactly what Don Knuth did when he designed Å in cmr! From
plain.tex:
\def\AA{\leavevmode\setbox0\hbox{h}\dimen@\ht0\advance\dimen@-1ex%
\rlap{\raise.67\dimen@\hbox{\char'27}}A}
This is nothing but a terrible hack which makes you ask yourself
what the heck Å has to do with h?? And pity the fact that Knuth,
despite
his Norwegian(?) ancestry, did not realise the need for a unique glyph
for the Scandinavian letter Å, when he did so for the Scandinavian
letter Æ
(\AE)! My guess is that Don Knuth did, just as seem to be the case still
for
many "TeXperts" still seem to so, believe that Å is an accented A
(corroborated by the fact that \aa in plain TeX is defined as
\accent`27a). It's not.
And with Knuth's above defintion of \AA, what you do if you want to use
Å as a math symbol when in the cmmi font \char'27 is not a "small
circle" but a \nu?
(A Russian physics professor and a colleague of mine once visiting
Sweden wanted to do so, rather than using a Greek symbol, in the
viewgraphs for his seminar to show his gratitude to the Royal Swedish
Academy who paid for his visit :-) -- I had to help him with a
\mbox{\textit{}} box or something.)
>
> Anyway, this doesn't change much of Bo's line of arguing.
>
> One question has not been asked nor answered: If I want to have a
> graphic distinction between Swedish \AA\ and some ringed A character
> I have invented in the transcription of some exotic language, how
> can I do so in T1 encoding? It is my understanding that all
> \r A, \AA and (if appropriate input encoding is selected) Å
> give the same glyph, \char"C1. How can I get A-with-ring?
In an earlier posting in this thread, I included an example from David
Carlisle which demonstrated this. In the latest June 1, 1998 release
of LaTeX, a new acctent, \mathring, has been added. According to the
LaTeX News, Issue 9, June 1998, this is a math mode version of the
ring accent which is available in text mode with the command \r. When
I try it, \r{A} looks more like an Å (even if the "ring" is of different
shape) whereas, in my opinion, it should have been identical to
$\mathring{\mathrm{A}}$ (assuming that \mathrm and \textrm fonts are
identical). While the simple solution for a circle-accented-A in text
mode, r\{A}, was at hand and would have felt very natural, it seems
that the design team screwed up. Perhaps the next LaTeX release will
get it right?
My answer to your explicit question is therefore: either do as in David
Carlisle's example or use $\mathring{\mathrm{A}}$ (and make sure you
have the same text and math roman fonts).
Bo
--
David Carlisle
> While the simple solution for a circle-accented-A in text
> mode, r\{A}, was at hand and would have felt very natural, it seems
> that the design team screwed up. Perhaps the next LaTeX release will
> get it right?
Perhaps we screwed up, perhaps not.
Look at it this way, the fact that the \r accent is designed to special case
a and A and give the Swedish letters means that 99% of the time it produces
a better result. As the Swedish letter is what is wanted (even if the
author didn't know that).
Granted there are circumstances when you want a ring accent on an A
that is not this letter. In math mode you have that by default.
In text mode you don't but since it takes exactly one line of code to
declare yourself a ring accent that does not special case a and just
always puts a ring over the supplied letter, I don't think this is
much of a hardship.
David
Bo Thidé
>
> "Bo Thidé" <b...@plasma.uu.se> writes:
>
> > Well, Don, since you seem to be the world's leading expert on
> > Scandinavian fonts, maybe you, or your brother, Font Expert II, can
> > tell me which of the two very different renditions of the <A-circle>
> > glyphs in the Berling Roman font at
> > http://w1000.mv.us.adobe.com/type/browser/F/P_221/F_BERM-10005000.html
> > and
> >
> > http://www.software.net/pages/adobe/example.htm?BERM10005000Berling-Roman
> > is a (relatively) correct rendition of the Scandinavian letter "Å"
> > and which is the entirely different "A-with-a-circle-above-it" or
> > even a nonsense character...
>
> I don't know if they were intended to be two different glyphs. A
> number of things about the graphic at the latter site, especially the
> quality of the ring and the mis-centering of the circumflex over the I
> and the umlaut over the U, indicate to me that it was made somewhat
> clumsily with blindly combining diacritic "dead key" characters, while
> the latter used actual accented-character glyphs.
>
> I'd never heard that there was a text ring-accented A, as distinct
> from A-with-ring, but I don't speak any Scandinavian languages. I did
> spend three years of my career discussing weirdnesses of various
> linguistic tricks with a Scandinavian font expert, and he never
> mentioned it. Maybe he took it for granted.
>
A "ring-accented-A" can appear in any language (it is not uncommon in
dictionaries, and is also used in scienctific literature related to
phonetics, linguistics, and natural sciences), but the cruical point is
that this "ring-accented-A" has nothing to do with the Swedish letter Å
which -- to connect back to the subject of this thread -- is the only
acceptable symbol for the "Angstrom" (or, rather "Ångström") unit!
Louis Vosloo
< ... continued tirade about \AA and Knuth deleted ... >
I think you may be upset about something that is basically a
misunderstanding, already hinted to in my earlier response.
The fact that the Angstrom CHARACTER is distinct from the
letter `A' - or the letter `A' with some accent - has little
to do with how the GLYPH Angstrom is constructed.
The GLYPH could be either based on a superposition of an `A'
(or a reduced size `A') with a `ring' - either derived from
an accent or not - either overlapping the `A' or not -
or it could be something designed independently (although
it is hard to see how to avoid starting somehow with something
that is at least vaguely `A'-like).
CHARACTERS have to do with meaning, GLYPHS with shape.
The mapping is not one-to-one. In fact, one character
can be represented by more than one glyph (e.g. different
forms of `a' in roman and italic writing), and one glyph
can represent more than one character (e.g. Dbar/Dcroat and Eth).
In some languages the CHARACTER `odieresis' is treated
as a `separate' character, while in others it is considered
as an `o' with a diacritical mark. Yet the GLYPH for odieresis
may in each case be constructed by combining the GLYPH for `o'
and that for `dieresis' --- possibly with some slightly
different spacing, but maybe not even that.
Louis Vosloo
> A "ring-accented-A" can appear in any language (it is not uncommon in
> dictionaries, and is also used in scienctific literature related to
> phonetics, linguistics, and natural sciences), but the cruical point is
> that this "ring-accented-A" has nothing to do with the Swedish letter Å
> which -- to connect back to the subject of this thread -- is the only
> acceptable symbol for the "Angstrom" (or, rather "Ångström") unit!
OK then, what language uses an `A' accented with a `ring'?
Louis Vosloo
Yes, the Angstrom character is a distinct from `A'
Yes, the Angstrom glyph should be designed to look right.
The typographer decides what it should look like.
Hopefully well informed by historical information,
examples, common useage, proper style for the font
in question and subject to constraints of line leading.
The fact that the Angstrom character is not related to the
character `A' does not imply that the Angstrom glyph should
bear no relationship to the glyph `A' It also does not
imply that the ring should be attached.
If some language were to use an A `accented' by ring (and
none that I know of do) then presumably one would want to
design that glyph properly also. There is however no
law that says it cannot look exactly the same as Angstrom.
Witness the glyph Croat Dbar, which looks the same as
the glyph Icelandic Eth. Yet the corresponding characters
bear no relationship whatever.
Yes, fonts should have ready-made accented characters,
and most of the 90,000 fonts rumoured to exist in Type 1
format do. Unfortunately Computer Modern does not.
I think the only thing you have not said that might be
helpful is some comment on useage. How much is an attached
ring used versus a detatched ring in Angstrom glyphs
in real Swedish writing?
Louis.
Bo Thidé
>
> > While the simple solution for a circle-accented-A in text
> > mode, r\{A}, was at hand and would have felt very natural, it seems
> > that the design team screwed up. Perhaps the next LaTeX release will
> > get it right?
>
>
> Look at it this way, the fact that the \r accent is designed to special case
> a and A and give the Swedish letters means that 99% of the time it produces
> a better result. As the Swedish letter is what is wanted (even if the
> author didn't know that).
>
> Granted there are circumstances when you want a ring accent on an A
> that is not this letter. In math mode you have that by default.
> In text mode you don't but since it takes exactly one line of code to
> declare yourself a ring accent that does not special case a and just
> always puts a ring over the supplied letter, I don't think this is
> much of a hardship.
>
> David
It's not the hardship. All along, I wanted to explain that the correct
"Angstrom" symbol is the letter Å. The combination of the completely
different letter A with a circle is likely not to produce the
correct symbol, particularly in fonts which have a separate, single glyph
for the Å letter designed in from the beginning (good fonts to be
used for printing science and/or Scandinavian books do).
Bo Thidé
>
> Bo Thidé wrote:
>
>
> I think you may be upset about something that is basically a
> misunderstanding, already hinted to in my earlier response.
> The fact that the Angstrom CHARACTER is distinct from the
> letter `A' - or the letter `A' with some accent - has little
> to do with how the GLYPH Angstrom is constructed.
>
> The GLYPH could be either based on a superposition of an `A'
> (or a reduced size `A') with a `ring' - either derived from
> an accent or not - either overlapping the `A' or not -
> or it could be something designed independently (although
> it is hard to see how to avoid starting somehow with something
> that is at least vaguely `A'-like).
The only person in the world who knows how to do this correctly,
is the font maker him-/herself. The probability that somebody else
who superposes an A with some kind of "ring" will succeed in
producing the same Å glyph as the font maker desinged is quite small.
See the Berling example I alluded to in earlier postings.
If the "superposition technique" would be preferable over using
the Å itself, why stop at superposing only an A and a ring? Why not use
a capital lambda, an em-dash, and a ring to produce the Å?
Whether the font maker started from an A or a capital Lambda when he/she
desinged Å or not is uninteresting since there's no point in trying
to redesign this Å. The font maker already designed it for you.
>
> CHARACTERS have to do with meaning, GLYPHS with shape.
There's no disagreement here. The point is that for any given font
and shape there exists only one typgraphically correct GLYPH for the
Å CHARACTER and hence for the Ångström symbol.
In certain fonts, Å may look like a ring-accented A, but in many
other fonts these two are different. It all depends on the font design
(see the example at www.plasma.uu.se/parad.jpg; for this font
it seems almost impossible to create the Ö from a superposition
of the O and something else). An author/publisher who needs both an Å
and a "ring-accented A" symbol should therefore, in the interest of clarity,
choose a font where these two are easy to discriminate from each
other and make sure he/she don't mix them up. In particular, he/she
should make sure to use the Å to denote the Ångström symbol.
Johann Gerell
On Mon, 20 Jul 1998, Louis Vosloo wrote:
> I think the only thing you have not said that might be
> helpful is some comment on useage. How much is an attached
> ring used versus a detatched ring in Angstrom glyphs
> in real Swedish writing?
Hello,
I have during my sixteen years in our educational palaces (!) _never_
seen a swede hand-write the discussed letter as an A with an
_attached_ ring. When the kids learn to write, the ring is just put
above the A (as in _detached_), because the study-books tell them
that's the way it should be.
As for printed stuff, such as books, the situation might be different
- I haven't thought much about this earlier.
#------------------------------------------#
J o h a n n G e r e l l
phone: 46+(0)90-198750
email: joge...@acc.umu.se
web : http://www.acc.umu.se/~jogerell
#------------------------------------------#