Uppercase germandbls is coming to Unicode

twardoch's picture

Andreas Stötzner and the German DIN committee submitted a proposal to the ISO 10646 working group that uppercase ß (germandbls, eszett, sharp s) should be added to Unicode/ISO 10646.

U+1E9E is the envisioned codepoint.

The proposal can be viewed at:
http://std.dkuug.dk/jtc1/sc2/wg2/docs/N3227.pdf

It is important to note that according to the proposal, even after adding this character to Unicode, the standard uppercase mapping for "ß" will remain "SS". This encoding effort is not about changing existing application or even spelling rules -- it is simply an effort to encode a character to be used in an "alternate" spelling which some people use (and currently have problems with properly encoding the text). It is an observed fact that "uppercase ß" exists, even if the official rules don't envision it.

I believe it is an interesting effort, and it would be reasonable to discuss what the best possible shape for the new character would be.

Some links in German:

http://www.signographie.de/cms/signa_9.htm (published by Andreas Stötzner, I recommend reviewing all the PDFs published there.)

http://de.wikipedia.org/wiki/Versal-Eszett

Some German type designers posted some of their design proposals for an uppercase ß at:
http://www.typeforum.de/modules.php?op=modload&name=XForum&file=viewthre...

I find many of these design proposals structurally flawed -- they don’t look like uppercase letters. They look like lowercase letters enlarged to match uppercase. The graphical structure of the Roman uppercase is very different from lowercase. If one were to invent a new uppercase letter, it would have to stylistically match the Roman uppercase. If Unicode really decides to encode uppercase ß, type designers should imagine what the uppercase ß would have looked from the very beginning, rather than trying to work out of the existing lowercase ß form.

Note that the history of "ß" is somewhat surprising. The letter developed in a two-wise way: as a ligation of long s and round ("normal") s, and as a ligation of long s and z. The German language adopted unified spelling rules only in 1901. Before that, both in the middle ages and in the humanist period, German spelling differed much. For example, "Thor" and "Tor" were equal variants of spelling the word meaning "gate".

Sharp s was denoted by different writers differently (as ſs or ſz, which looked like ſʒ). The graphical shape of the ß ligature developed independently in these two ways.

This dichotomy still shows itself in a small minority practice of uppercasing ß as "SZ" rather than "SS". Incidentally, this practice is understandable for most German readers (though not actively practiced), i.e. "GROSZSTADT" or "MASZGEBLICH" is understandable as the uppercasing of Großstadt or maßgeblich.

See http://www.flickr.com/photos/adamt/490566363/ for an example.

One interesting issue is that in the 1996 spelling reform the status of ß as a single letter has been finally confirmed. In the previous spelling, the general rule was that short vowels are denoted by following them by doubled consonant letters while long vowels are followed by single consonant letters. So writing "met" always indicates a long "e:" while "mett" indicates a short "e".

In case of "s"/"ß", it was confusing. Following a vowel with a single "s" always denoted a long vowel, following a vowel with a doubled "ss" indicated a short vowel, but following a vowel with "ß" did not give clue whether the vowel was short or long. So "Ruß" was actually pronounced "ru:s" as if the "ß" stood for a single consonant letter, but "Nuß" was pronounced "nus" as if the "ß" stood for a doubled consonant letter.

The 1996 spelling removed this uncertainty by changing the spelling of all "ß" into "ss" when the preceding vowel was to be pronounced short. Today’s spelling of "Nuss" or "dass" underlines that the vowels are to be pronounced short.

The uppercasing of "ß" as "SS" but also as "SZ" defeats this clear rule. If I uppercase the word "Rußpartikel" into "RUSSPARTIKEL" or even "RUSZPARTIKEL", suddenly the natural way of pronouncing the "U" changes from short to long, so the reader is confused. The confusion is even bigger now, after the reform, because the special "undefined" treatment of "ß" no longer exists, so readers are used to "ß" being always treated as a single consonant letter, not as a ligature of a doubled consonant.

To remain logical, consistent and reader-friendly, "ß" needs (at some point) to assume a single graphemic shape in the uppercase.

I strongly feel that uppercasing "ß" as "SS" is now -- especially under the new rules -- a temporary anachronism. "ß" is a single CHARACTER (as per orthographic perception). It has functionally liberated itself from its historical background (which was a ligature of ſs or ſz).

Today, "ß" is no more a ligature of "ſs" than "ä" is a ligature of "ae". The transition process from "ae" to "ä" has been completed about 200 years ago, and the transition process between "ſs" to "ß" is happening now. Encoding the uppercase "ä" as "A ZWJ E" (or something like that) would make as little sense as encoding the uppercase "ß" as "S ZWJ S".

I believe that "SS" is an anachronic, still-in-use but slowly-to-vanish poor man’s solution to write the uppercase "ß". I believe that it should be an exciting task for type designers now to come up with a new form. In my opinion, this issue is definitely not one that is completely solved. We’re in the middle of a slow transition period for "ß". The 1996 reform started it and showed the direction.

I myself once had the idea that Scedilla (U+015E, Ş) would be most appropriate for denoting uppercase ß.

After all, Ş is historically an S with a subscribed z (that at this time looked like ʒ). Since ß is a ligature of either ſs or of ſʒ, uppercasing it as Sʒ, or, effectively, Ş, would historically make sense.

Using this notation, "Gauß" or "Roßberg" would be uppercased to "GAUŞ" or "ROŞBERG".

Similarly, the umlaut in "ä" or "ö" is historically a superscripted "e", so historically "ä" and "æ" are two different ligations of "ae", and "ö" and "œ" are two different ligations of "oe".

Since German readers are currently used to uppercasing ß as SS, i.e. they write "GAUSS" or "ROSSBERG", I even thought of a compromise: the SS remains doubled but for added distinctiveness, a subscribed z (i.e. a cedilla) is added after the first S. In other words, "Gauß" or "Roßberg" should be uppercased as "GAUŞS" or "ROŞSBERG".

Historically, this would make sense. The cedilla would here have a similar function to the trema in Spanish or French: "GAUŞS" would make clear that it comes from "Gauß" while "GAUSS" would make clear that it comes from "Gauss".

"ROŞSBERG" does not look very awkward to a German reader. The addition of a diacritic does not dramatically change the reading pattern but still adds a distinctive mark that is, indeed, needed. If I were to design a glyph that should go into U+1E9E, it would probably look like ŞS, or perhaps just SS, depending on the style of the typeface.

An alternative approach is to look at the existing uppercase-to-lowercase relations within the Latin alphabet and try to derive a shape for the uppercase ß which maintains the same relations.

In most of the middle ages and the period up until the 19th century, the long s ("ſ") and "f" were closely related, "f" being simply a "ſ" with a stroke going through. The same, very primitive graphic relation exists between the prototypic shapes of the Greek letters gamma (Γ) and digamma (Ϝ). Since the minuscule "f" always has been a "ſ" with a middle stroke, then the capital "F" might also be considered an uppercase "ſ" with a stroke going through. Of course an uppercase long s never existed, but this relation may be helpful when constructing the uppercase ß.

Because I think that *if* the Latin alphabet ever used or needed another capital S, the preferred shape could be that of a gamma (Γ). This is a simple, effective shape that maintains a stylistic relation to the lowercase long s that is typical of other uppercase-to-lowercase relations.

If we look at the relations between Aa Ee Ff Mm Pp, we will notice that sharp, edgy connections in the uppercase are related to more smooth, round connections in the lowercase. If "F" developed into "f" in a cursive hand, then it is very easy to imagine that a cursive rendition of the "Γ" shape might, indeed, look very much like "ſ".

This is an important observation when thinking about the shape of an uppercase "ß": I assert that the shape of uppercase "ß" must be "edgier" than the lowercase. In short, I think that the left part of uppercase ß should be "Γ".

What about the right part? Here, I would call to exploit the double origin of "ß", which developed paralelly as a ligature of "ſs" as well as of "ſz" (where the "z" historically used the "ʒ" shape, so "ſʒ").

These days, the lowercase "ß" is typically derived from the ligated form of "ſs". For visual dissimilation purposes -- to strongly set apart the lowercase and the (new) uppercase "ß" I would derive the uppercase "ß" from a ligation of the hypothetical uppercase "ſ" (i.e. "Γ") and the shape of "the other" origin of "ß", i.e. of the historical "Z" shape.

In short, I believe that the best graphical rendition of an uppercase "ß" would be be a well-designed ligature that incorporates these shapes: "ΓƷ"

I have made a small simulation using Garamond Premier (please excuse my poor drawing abilities):


http://www.twardoch.com/tmp/germandbls_garamond.png

The first line shows what the historical origin of ß looks like, i.e. long s followed by a round s. The second line shows the current shape of ß as we know it. The third line shows what a hypothetical uppercase long S might look like ("Γ"), which is just a mental exercise. The fourth line is my proposal for the uppercase ß shape.

Andreas Stötzner has proposed an elaborate document that tries to explore all possible combinations of drawing an uppercase ß:

http://www.signographie.de/cms/upload/pdf/Signa9_Kombinatorik_SZ_3.0.pdf

My proposal corresponds to the scheme A1-B2-C1, which I has the most "uppercase" appearance of all those presented there.

On a related matter, at the exhibition "Neue Baukunst. Berlin um 1800", which is on display at the Alte Nationalgalerie in Berlin until May 28, I have discovered a fantastic calligraphic lowercase "ß" shape, in which the "long s" part connects to the BOTTOM and not to the top of the following "short s". Please take a look:


http://www.flickr.com/photos/adamt/490547406/

This got my imagination going.

Regards,
Adam

ebensorkin's picture

Very Nice. I agree it isn't all worked out but the possibility of finding really nice forms is made abundantly clear with your example. I also agree that some added weight in the upper right - maybe a serif or a knob? - would make things feel more even. There is something about the diagonal that might feel more stonework like though ( less brushy ) and hence more trajanesqe. Also the diagonal feels strong and less swooshy.

ebensorkin's picture

duplicate

John Hudson's picture

Eben, yes, I agree about the diagonal. I think my quick Trajan test is too loose.

Rob O. Font's picture

"If one were to invent a new uppercase letter, it would have to stylistically match the Roman uppercase."

"Besides, isn't it fun to invent a new letter?"

hmm. It is serious business and usually only taken up in the interest of a large and pressing need, (over 150years since an alphabetic was needed, no?). It will be expensive, and time consuming if we ever have to do it, which I doubt. But if I had a 1/2 hour to spend on it...

Make us sweat. Besides, a simpler way to form an s in calligraphic writing, truncated to the lowercase f form in roman type with an s ligated on and then blown up to uppercase proportions is hardly inventive, confusable with existing glyphs and simply not Romanizable enough with it's "one-too-many" moves within the cap ht.

Cheers

hrant's picture

David, perfect for Valentine's.

Hey, what about this? WEI§ACKERS
:->

hhp

Rob O. Font's picture

"David, perfect for Valentine’s."
Oh, but you have not seen the serif design, that's for lovers of type.

For the others, there's always the 100% imagination-free proposal. :)

k.l.'s picture

[Rue Montmartre, Paris]

Mr Berlow, are you aware that your suggestion implies that in future the name of this café would be ‘croißant’?  :)

hrant's picture

Well, if they serve a croissant with a sausage in it...

hhp

John Hudson's picture

Well, if they serve a croissant with a sausage in it...

Hardly necessary to establish a Germanic flavour, given the croissant's Austrian origins. They were first baked in Vienna to celebrate the defeat of the besieging Ottoman army in 1683. They are crescent shaped as an insult to the Muslim invaders: 'We eat Turks for breakfast'.

Rob O. Font's picture

Ah yes, the old days when one could bake-n-eat the enemies of Jesus to the sound of the Lord's prayer. Make that a Vienna Sausage and Hrant'll take two — or maybe not. Me, I just think of it as a piece of bread with the butter built in.

"[Rue Montmartre, Paris]" From the looks of it, in this context, the sign says "Croissant", and assuming Paris is still in France and the French still speak French, (I didn't miss another Franco-German war Monday and Tuesday, did I?), then this sign can stay, thanks.

But more than that...I pine for the days when people knew how to make new letters: The migration of a lowercase to uppercase form is wrong in this proposal, isn't it? The migration from a long lowercase s, (invented for the lazy and rushed to improve the appearance of a difficult s to scratch), to an Uppercase F, is wrong isn't it? Following "historical precedent" from gravestones, sporadic bits of metal and folks with no apparent historical expertise or typographic experience is wrong in this letter, isn't it?

Giving righties something sinuous and sexy to draw "from their side", making a letter that looks something remotely like the sound, and following an historically correct path to a new and true uppercase letter — that cannot possibly succeed these days, I know. But making sense never goes out of style does it?

Cheers!

hrant's picture

I agree that you don't want to derive it from the lc, and you shouldn't let physical writability intrude (especially when you -consciously- exclude 15% of the population). But "virgin decipherability" does have to be ensured; you can just completely make something up. So what about forms that look like a fusing of two caps? As long as the width doesn't explode.

hhp

Rob O. Font's picture

"(especially when you -consciously- exclude 15% of the population)"

EEEyhrant, I'm a lefty, and so I exclude that 15% like, never, dude. I even sleep on that side in case of meteor strikes, dude.
I was just thinking how long it's been, like...2300, 2400 years? since you righties had an equal shot at an S,
and of how this long s came to be...Just trying to be fair.

Nick Shinn's picture


in which the “long s” part connects to the BOTTOM and not to the top of the following “short s”.

I surmise that the "following" short s was (conceptually) written first in the sample you show, That is the reason it connects where it does; -- the writer just had to be sure to leave a long enough width in the entry stroke to accomodate the subsequent long s crossing it.

By conceptually I mean that the sequence e_s_s was designed as one continuous stroke, at the rough stage, experimenting with a pencil perhaps, and then constructed with shorter pen/graver strokes that were less likely to create difficulties by "pushing" into the medium.

So I'd say that in formal calligraphy, where design and execution can be separate processes, the left-handed (with suitably adapted nibs) face barely more challenge than the right handed. Once-off handwriting is a different matter.

Is that a reasonable assessment, David?

hrant's picture

You're forgetting about ink not drying immediately...

Also, Gerrit Noordzij recommends that left-handers write vertically, with the paper rotated 90 degrees... Sure that's nuts (no matter what apologists like John H and Peter E say) but that's exactly what indicates the depth of the problem.

hhp

Rob O. Font's picture

Adam says? "The first line shows what the historical origin of ß looks like,"

The typographic origin, perhaps? but not the original origin...is it?

Nick says: "Is that a reasonable assessment, David?

I think we are looking at an engraving here and not calligraphy. So, in an n-step process, it's possible anything could be done, as opposed to a 1-step process like calligraphy, where what you see here would be impractical.

From the proposal, Ehmcke Antiqua shows a design with 10 uppercase letters as lowercase forms. Is this not an unusual typeface to show when one is trying to make the point they are making? lal.

The next page of examples, all three designs in the left column show faces with relatively large contrast between the ii space and o counter. I couldn't "prove" in type court that these are uppercase and not alternates or, that the compositions, (which place the specimen letter everywhere but with the uppercase) are showing me uppercase ß now could I?

All three designs in the left column are hideous as are many of the next set. And as I play the "whole library" ( i.e. everyone's), out on this form, it's a bad scene. Not the doing, though I'm sure that'll be hard for just the German founders, but the designing of some faces where it's just not going to "escape" the B without ugliness against the grain of the design.
Show me a geometric. Mistral please. Heaven help the font with a bigger top on the S.

Another example, in Adam's additional attempt, "This alternative treatment is analogical to the alternative treatment of the “J”, which may reside on the baseline or descend below it."

But why then would not the first stroke descend like the "J" and the second sit like the "S", i.e. the lowercase design? Afterall there is no "S" on this planet with a serif sittin' on the baseline like an "F"...is dare?

Cheers!

hrant's picture

> it’s a bad scene.

Hey, it's not easy making up a new letter. Even just a good open-bottom binocular "g" is hard to pull off. Part of it though is what one gets used to: people can live with something ugly for decades and not mind - literally. :-)

hhp

cuttlefish's picture

Has anybody heard any new developments about this whole deal here?
Does anyone care to add their thoughts?

Andreas Stötzner's picture

There are some new fonts around boasting with the new capital. Also offers for font-completion and a keyboard layout for German which enables you to simply type the capital Eszett – a suitable font assumed. See
this page for further information.
A:S

Miguel Sousa's picture

It's officially in Unicode v5.1.

Unicode 5.1.0 adds 1,624 newly encoded characters. These additions include characters required for Malayalam and Myanmar and important individual characters such as Latin capital sharp s for German. Version 5.1 extends support for languages in Africa, India, Indonesia, Myanmar, and Vietnam, with the addition of the Cham, Lepcha, Ol Chiki, Rejang, Saurashtra, Sundanese, and Vai scripts. Scholarly support includes important editorial punctuation marks, as well as the Carian, Lycian, and Lydian scripts, and the Phaistos disc symbols. Other new symbol sets include dominoes, Mahjong, dictionary punctuation marks, and math additions.
http://www.unicode.org/press/pr-5.1.html

Nick Shinn's picture

Woe betide any "monocase" character that thinks it can escape the Unicode agenda of bicameralization!
Are there any left?

There are several features of German typography that it would be more appropriate to include in a digital font, such as caps with lowered umlauts, and ch and ck digraphs.

solfeggio's picture

There are several features of German typography that it would be more appropriate to include in a digital font, such as caps with lowered umlauts, and ch and ck digraphs.

Maybe they were simply distracted by more pressing affairs? After all, there are significant improvements elsewhere in Unicode 5.1 that address greater communication needs: code points for checkers, dominoes, and mah-jong tiles! ;)

ralf h.'s picture

There are several features of German typography that it would be more appropriate to include in a digital font, such as caps with lowered umlauts, and ch and ck digraphs.

Maybe. But they wouln't need a Unicode point. OT feature access would be sufficient.

ebensorkin's picture

Re the ck & ch digraphs; if they really believe in OT then they could leave it to scripting. But maybe it's a backwards/trans system compatibility thing.

I am going to start a new thread regarding the do's & don'ts of the ck & ch. I don't understand them nearly well enough.

Miguel Sousa's picture

> Woe betide any “monocase” character that thinks it can escape the Unicode agenda of bicameralization! Are there any left?

I can think of two:
ĸ  kgreenlandic (U+0138)
ʼn  napostrophe (U+0149) -- Uppercase is ʼN (U+02BC, U+004E)

Thomas Phinney's picture

"Woe betide any “monocase” character that thinks it can escape the Unicode agenda of bicameralization!"

The agenda belongs to Germans or Germany, not Unicode.

Besides, I thought you were in favor of bicameralization. Aren't you the one who was bemoaning that oldstyle figures aren't encoded separately in Unicode?

"There are several features of German typography that it would be more appropriate to include in a digital font, such as caps with lowered umlauts, and ch and ck digraphs."

Perhaps. But people could do those things already (as I did in Hypatia Sans), while I'm not sure there was a good solution to the cap eszett until it got encoded.

Cheers,

T

Nick Shinn's picture

Are there any left?

ſ Long s

Perhaps. But people could do those things already (as I did in Hypatia Sans), while I’m not sure there was a good solution to the cap eszett until it got encoded.

True, but foundries will create strange glyphs that are characters in a Unicode group, as a basic font ingredient, before they will add other oddities that aren't Unicoded. The uppercase germandbls is, IMO, much stranger than lowered-umlaut caps and ck, ch and tz digraphs--these may be odd now, but they were much used not so long ago, so have a lot of period flavour. Of course, all that's necessary for ch and ck is for the typographer to kern them, but a discretionary ligature would be more convenient.

Aren’t you the one who was bemoaning that oldstyle figures aren’t encoded separately in Unicode?

No, I was wondering whether less encoding (i.e. not capitals) would have made things easier for typographic features such as figure styles. As for the problem of the vagueness surrounding the default figure style, layout applications should be able to calculate this, by checking to see which features the default style appears in.

The agenda belongs to Germans or Germany, not Unicode.

There are also a few characters in Greek and Cyrillic scripts.

Nick Shinn's picture

This is the model in the Unicode chart:


***

The Unicode face is Times, and here is a 1956 precedent of the glyph shown in the Unicode proposal, http://std.dkuug.dk/jtc1/sc2/wg2/docs/N3227.pdf in Times:


***

Here are some options I'm considering for a Modern treatment.
I don't particularly like the pointy-eared model, it seems counter-intuitive to the chirographic ductus of the antiqua.
The Times precedent (and there are other examples of it in the Unicode proposal) seems more typographically correct.
What do you think, should I follow the Unicode model, or my instinct?

ebensorkin's picture

I think the notan of the middle one is best. My eye travels most easily over it. Maybe with a quieter/smaller ball though so it sparkles a bit less. Not sure.

k.l.'s picture

Regarding proportions and evenness, I agree with Eben. Regarding details, except for the pointy ear, three is better.

I think that every single uppercase eszett which I have seen so far looks like comedy. Or tragedy. This new letter is supposed to resemble the lowercase eszett so it can be recognized as an eszett at all, and at the same time needs to fit into the visual language of the uppercase alphabet. Which should be hard to achieve since the lowercase eszett is a ligature of lowercase letters. It doesn't even matter if it combines long-s + s, or long-s + semicolon-like symbol. My eyes hurt.

dezcom's picture

My biggest concern is how it will be used or IF it will be used. If when case is changed, will the user be more annoyed to see a cap eszett or two cap SS in a row? Then there is the small cap issue. Which form will cause less anguish to German readers? I would think to put the new cap thing as an alternate but not default but I would hope Germans would chime in and say what they prefer a type designer do to make their lives easier?

ChrisL

Nick Shinn's picture

comedy. Or tragedy.

Don't forget romance. (Weißackers, see David's post above:-)

Nick Shinn's picture


Perhaps the theatrical effect Karsten decries is caused by the overtly "designy" quality of the new glyph.
So why not just give the majuscule character the same shape as the minuscule?
After all, there are many other letters of the alphabet which share their shape across cases: C, I, J, O, S, U, V, W, X, and Z.

k.l.'s picture

Yes, looks much better. Have you tried the uppercase S's serif instead of the ball?
In this case, no romance for me. Mr Berlow's post is really great since it highlights a couple of flaws in the arguments pro an uppercase eszett.

Mr Lozos, as I understand the unicode.org information, the uppercase eszett is there now but its existence does not change casing rules -- eszett becomes SS in all caps setting. Uppercase eszett is meant for 'special cases'.* Following this, it would be a mistake if features would switch from lowercase to uppercase or smallcap eszett.
The only thing you might do is
    sub whatever_this_letters_postscript_name_is by germandbls;
in the c2sc feature.

* The paragraph from the above-mentioned unicode.org link:
In particular, capital sharp s is intended for typographical representations of signage and uppercase titles, and other environments where users require the sharp s to be preserved in uppercase. Overall, such usage is rare. In contrast, standard German orthography uses the string "SS" as uppercase mapping for small sharp s. Thus, with the default Unicode casing operations, capital sharp s will lowercase to small sharp s, but not the reverse: small sharp s uppercases to "SS". In those instances where the reverse casing operation is needed, a tailored operation would be required.
[non-italics are mine]

dezcom's picture

Thanks, Karsten!

ChrisL

eliason's picture

So why not just give the majuscule character the same shape as the minuscule?

Because it's fun to invent new letters!

Nick Shinn's picture

Karsten, the modern "S" serif is a bit big for internal use.
And also a couple of "Dresdner" versions after Andreas.

k.l.'s picture

True. Now I remember why I omit serifs inside ...

paul d hunt's picture

I know this is going to sound very ethnocentric of me, but we have a saying in America: 'if it ain't broke, don't fix it.' i don't see the proposed uppercase eszett as an improvement over the double S, which seems Germans have not really had a problem with. Sure, give it it's own Unicode point, but fill it with what is already in common currency. just my 1p.

dezcom's picture

There must have been some amount of pressure from German speakerts to get it vto happen. What is the logic they used? Surely, there must be some reason behind it.

ChrisL

paul d hunt's picture

yes, what's the intention? and what does the standard uppercase mapping for “ß” will remain “SS” mean exactly?

Jos Buivenga's picture

... what does the standard uppercase mapping for “ß” will remain “SS” mean exactly?

That means that if you try to incorporate the capital ß into an OpenType case feature it will work fine but only in FontLab :-)

Nick Shinn's picture

feature case {
sub S S by uni1E9E;
} case;

Wouldn't that work?

(John: ...any software performing case mapping between the new uppercase ß and the lowercase ß characters must do so at a level independent of the Unicode character properties. This is quite possible, and if the uppercase ß character starts to become widely used I think we can expect to see it handled in such ways.)

Nick Shinn's picture


I think I'll go with number 2.
Number 1 is a bit dull. I mean, if you're going to have a new glyph available, why not get into the spirit of the thing?
Bur not too much, 3 & 4 with the spur are a bit flashy in this modern style. 5 & 6 are too odd -- foundries should try to stick to the pointy-ear Unicode model, because it has been thought out, and some consistency will help get the idea accepted. Or else why bother?
But perhaps such conformity isn't right, and now is the time for experimentation?

ralf h.'s picture

i don’t see the proposed uppercase eszett as an improvement over the double S, which seems Germans have not really had a problem with.

I am German and I had problems with it all the time! I can't write the name of my home town in caps. I can't set a book cover in small caps when the name of the author has an Eszett in it. (Meißner and Meissner are different names, so you can't set them both as MEISSNER.)

I don't need it in the case feature at the moment, because the replacement »ß to SS« will not go away for some time. It just needs to be in the font, whenever we need to set a proper name.

Nick,
don't take 1 or 5. They would be read as B. Your test word GROßE (»big«) would then mean »crude« (GROBE).

Ralf

Jos Buivenga's picture

Wouldn’t that work?



As Ralf already pointed out ... capital ß and SS are both used in all caps settings.
I tried and tried, but I couldn't script a proper ß / capital ß replacement.

Nick Shinn's picture

Thanks Jos, I understand now.

Would it be appropriate to include it as a discretionary ligature?
That would be convenient, although not technically correct--but as the casing rules for ß aren't logically consistent, why not?

Jos Buivenga's picture

I would prefer it as a contextual alternate ... in the context of the ß flanked by other caps. But when you –for example in InDesign– set your text to caps the capital ß changes into SS.

k.l.'s picture

J.B. -- That means that if you try to incorporate the capital ß into an OpenType case feature it will work fine but only in FontLab :-)
N.S. -- Wouldn't that work?
J.B. -- ... but I couldn't script a proper ß / capital ß replacement.

Technically, you could do the sustitution in case and smcp etc. But the Unicode information suggest that you should not. So no need to worry.
This uppercase eszett is an encoded character, so it is not advisable to make it the target of glyph substitution via any layout feature. A designer feeling tempted to insert it should use keyboard or Glyph Palette.

P.D.H. -- i don't see the proposed uppercase eszett as an improvement over the double S, which seems Germans have not really had a problem with.
Ch.L. -- There must have been some amount of pressure from German speakers to get it to happen.

My personal impression is that this "amount of pressure" actually stems from less than a handful of German typographers.

R.H. -- I am German and I had problems with it all the time!

Interesting. I never had problems setting my SCHLOSSTRASSE in all caps (before a bunch of half-hearted spelling reforms). And by current ortho/typo-graphic rules, Meißner and Meissner both are correctly represented as MEISSNER. If Herr & Frau Meißner should feel unhappy about it, your typographic education will have enabled you to emphasize the name by italics, larger size, another typeface, or not.

ralf h.'s picture

If Herr & Frau Meißner should feel unhappy about it, your typographic education will have enabled you to emphasize the name by italics, larger size, another typeface, or not.

Sure, there are workarounds. But isn't it bizarre that I should change the style of type setting, just because there is one name with an ß involved?
In books it is common practice to set names in small caps. But with German names involved I can't do it. And it's really not just about annoying Frau Meißner. It's about unambiguousness! Let's say a scientific text has a reference to a another book written by someone named MEISSNER. Is it Meißner or Meissner then?
I guess we can talk for hours why we need an fi ligature, because it's unacceptable that f and i would touch. But setting a proper name right doesn't matter?

paul d hunt's picture

It’s about unambiguousness!

more ethnocentrism: too bad we have so much of it in English!

Syndicate content Syndicate content