I remember reading something online about this a while back and thinking it was as pointless as proposing uppercase forms of fi, fl, and ff. But maybe not. I don't read, write, or speak German, but I like your solution best of all the ones I've seen so far.

Mark,

it is not like uppercase forms for fi or fl. The lack of "uppercase ß" means that it is uppercased to "SS", but "ss" is also uppercased to "SS". So "SS" is an ambiguous representation of two different lowercase combinations. This is particularly important in personal names: "Peter Weiß" and "Peter Weiss" are two distinct names and one may not be substituted by the other. But in uppercase, it's only "PETER WEISS" which is ambiguous because it does not exactly tell you what name it actually stands for.

It is a bit like the French practice of removing accents over vowels when uppercasing. This practice was introduced when typewriters came into use simply because typewriters did not have separate uppercase accented letters. These days, it's a bad ambiguous anachronic practice which is no longer necessary but still practiced by many French people who are used to the typewriter customs and actually think you HAVE TO remove the accents when uppercasing.

This also reminds me of another practice, i.e. replacing long dashes by single or (perhaps) double hyphens.

As I have explained, the method of uppercasing "ß" as "SS" is in my eyes an anachronism. It is still the majority practice according to official German spelling rules, but I think this is largely a technological limitation. If well-designed uppercase ß exists in fonts, and there is a Unicode codepoint for it, *and* (at some point) it is mapped to the German keyboard layout in Windows and Mac OS X, some people will *happily* start using it.

"Popular use" of "ß" in the middle of all-uppercase text suggests that there is an authentic need for that character. I believe we're still years ahead of widespread adoption, but I think type designers should do their best now to come up with appropriate forms, or at least once the character has been officially encoded.

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

Adam

Oh boy, another letter to add to Agamemnon, but I like it! You'll see it in my next revision.

BTW, in a font, the appropriate glyphname for the uppercase ß would be germandbls.case or germandbls.cap but not Germandbls! Should the character be encoded in Unicode under the proposed codepoint, a better glyphname would be uni1E9E, but I wouldn’t start using that glyphname yet.

In an OpenType font, you might want to encode the glyph as a "stylistic case variant". To do that, do the following:

1. Create an indentical duplicate of the lowercase germandbls glyph named germandbls.ss01.
2. Design your uppercase ß and name it germandbls.case.
3. In the OpenType Layout feature definition code for the salt and ss01 features, include the substitution sub germandbls by germandbls.ss01.
4. In the OpenType Layout feature definition code for the case feature, include the substitution sub germandbls.ss01 by germandbls.case.

If your font contains smallcaps, then in addition:

1. Provide a glyph named germandbls.smcp which looks like two smallcap-S letters (the "traditional" way of making a smallcap ß).
2. Design your "new" ("true") smallcap-S and name it germandbls.smcp_ss01.
3. In the OpenType Layout feature definition code for the smcp feature, include the substitution sub germandbls by germandbls.smcp.
4. In the OpenType Layout feature definition code for the salt and ss01 features, include the substitution sub germandbls.smcp by germandbls.smcp_ss01.
5. Make sure that the feature definitions for salt and ss01 are ordered below the definition of smcp.
6. If uppercase ß gets encoded in Unicode and your font has a glyph with it named uni1E9E, then in the OpenType Layout feature definition code for the c2sc feature, include the substitution sub uni1E9E by germandbls.smcp_ss01. It that case, you may also consider using the name uni1E9E.smcp instead of germandbls.smcp_ss01.

It just occured to me that another approach to the uppercase ß might also be feasible (again, sorry for the imperfect color of the strokes):

This alternative treatment is analogical to the alternative treatment of the "J", which may reside on the baseline or descend below it.

Adam

Does SS ever begin a word? I would guess not, in which case there wouldn't seem to be a need for it in a formal script. (Except perhaps to accommodate people who insist on setting formal scripts in all caps...?)

No words or names ever start with a "ß" so you're right, the character would not be necessary in a formal script.

A.

Rats.

This isn't the euro currency symbol. I wouldn't expect to see most font producers integrating an uppercase ß into their products anytime soon.

You can see a lot of stuff in Germany, where people put in a lowercase-ß when using All Caps, like ROßMANN and such. There is surely a need for it, especially for names. It's different, if you called ie Heike Rossmann or Heike Roßmann. Written in All Caps, it is impossible to know, if it's Rossmann or Roßmann.
But the Versal-ß has no common known form yet. If there would be a typeface with Versal-ß included, would i, as a designer, use this Versal-ß or the common "SS"? I think, i would use the "SS" to play save. Ok, it really depends on how obvious is the connection between the Versal-ß and the common lowercase ß.

» 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.« (A.Tw.)

NO! No, no, no.
We need to distinguish the ORIGIN OF THE CHARACTER from the LETTERSHAPES.
This has been mixed up by everybody for far too long.

The origin of the ß *as a character* is: a long s with *some* curl or swash or stroke attached, in order to mark it being not an 'ordinary [long] s'. That attachment became randomly written like the ezh- (or 3-) shape by time (the same we find with other Latin letters for abbr., e.g. q).
Therefor the letter became named “eszett”, the common mind thus just describing what he sees, regardless of the fact that the letter (the character) was NOT derived from long-s and z functionally or phonetically or in the sense of a typographic ligature. Note that up today “Eszett” is also named “Dreierles-Es” or “Scharfes Es” which indicates the very idea of ß as a character of its own.
The earliest sample of a “ß” in an upright roman-style printed German text (1667) is definitely no long-s_s nor a longs_z (ezh-shape). It’s something else.

The “ß is a s_s-ligature” theory was induced by the italic long-s_s ligature later becoming (ab-)used to represent the German ß. Today the glyphs we use in Roman and Italic for ß are actually two different letters (or glyphs) by origin.

A tricky history, admittedly.

But thanks Adam, for your elaborate explanations. Sorting out what the capital ß might look like this should be the level to work on it.

By the way, you’ll find it notable that we (searching for Signa 9) actually testified two medieval examples for the capital long S, one of them Gamma-shaped.

It says here that it’s official. The uppercase “ß” is coming:
http://www.fontblog.de/das-versale-ss-rueckt-naeher-bauanleitung

There’s also a guide how to build an uppercase “ß”:
http://fontshop.de/pdf/versales_eszett.pdf

All in german, of course ;-)

Thank you Adam for this very elaborate and profound thesis of ways to find a design for a new UC Germandbls. This is the best thread i have read on this topic so far. There are very hot discussions in Germany about wether or not it makes sense to include this character.

As nice as it would be to *invent* a new character (i am convinced of your examples of the Gamma-S ligature), i find the S-cedilla followed by S solution very plausible.

My concern is: if the knowledge of the inclusion of an UC Germandbls into Unicode is spread everywhere i fear that some commission will determine what that character has to look like. The S-cedilla proposal is very easy and fast to determine as standard (and it would follow the convention to capitalize *ß* by *SS*). Also there would be no need to update fonts. This could be taken up into the orthography (and the Duden) immediately.

That’s my pragmatic view. A new character might come up in a disaster as it happened with the Euro (if some commission sets up the design guides) or it will take a very long time until there is a form for the new character that will be widely accepted. I absolutely agree that a new character has to follow the forms and origins of the Roman alphabet (which are the preceding Greek letters).
It also could all end up in a design competition (hope not!)

Thank you Andreas for an update of this character’s tricky history.

If this world would be populated by typographers only, S-cedilla would certainly do.

Sorry gents, but this consideration was under discussion already 1903 and again 1956.
Forget about it. The common German reader-writer Herr WEIßMANN will *never accept a S-cedilla in his name.
And as to the Euro-sign: yes, the commission did a dreadful job, like commission tend to do. But today, since the character is “free”, we have fonts with lots of fairly well-designed Euro’s.

Well, i doubt that Herr WEIßMANN even cares if he wears a ß in his all caps setting. Only typographers do.

If you want to do something for the general public, forget it. How many people in Germany think the ß doesn’t exist anymore anyway? They thought it was banned with the last spelling reform.

I think the UC-ß should be designed with the above mentioned roots in mind and not what Herr WEIßMANN likes. He was never used to see anything else than the ß in his name. But there are more important spelling situations where this character changes meaning (Maße and Masse, e.g.). So please forget the WEIßMANNS and the like. It doesn’t change the meaning of their name nor does it change its intonation.
My two cents.

Jürgen,

> It doesn’t change the meaning of their name nor does
> it change its intonation.

Of course it does. Herr ROSSBERG (i.e. Rossberg) is pronounced "rosberg" and Herr SCHUSS (i.e. Schuss) is pronounced "ʃus", while Herr ROßBERG (i.e. Roßberg) is pronounced "ro:sberg" and Herr SCHUß (i.e. Schuß) is pronounced "ʃu:s".

Adam

Got me. Accepted. Yet, i don’t think that names are the foremost concern.
In your examples it’s only about intonation but not meaning. Roß or Ross is generally short and it means horse. A ro:s doesn‘t exist, as far as i know. Could be short for Rose which is a rose, hence it would be ROSBERG.

I was only saying that in names it occurs to me that there is a favour towards a UC-ß only for aesthetic reasons. To denote different pronounciation a S-cedilla would do perfectly well. I am not against a new letter, i only try to be pragmatic and the cedilla does make sense in the historical and formal context.

Adam, superb contribution to typographic discourse.

Personally I'm very happy to hear about this Unicode proposal,
and even if most people never take advantage of it a slot for the
UC eszet would be a great thing to have.

In fact in general I feel we need more characters, not fewer.

hhp

I start with the caution that I was not involved in the implementation of the German spelling reform and that I am not an expert in this matter. However, having been a tutor at the German spelling reform publishing house DUDEN for nearly a quarter of a century, I occassionally had discussions with the editors of the spelling reform rule books.

I am under the impression that both Mr. Twardoch and Mr. Stötzner missed the crucial point concerning the s rules. The "long/short"-vowel rule as stated by Mr. Twardoch does not reflect the full rule, because the full rule distinguishes between surd (voiceless) and sonant (voiced) s sounds, i.e. the spelling depends on whether the s sound is surd or sonant.

The IPA chart was quoted by me in another context on my website

http://www.sanskritweb.net/deutsch/ipa_sans.pdf

On page 2 of this PDF, in the top chart "consonants", row "fricatives", the surd s sounds are to the left and the sonant s sounds are to the right.

In the "alveolar" column, the surd s is "s" as IPA symbol, and the sonant s is "z" as IPA symbol.

While some languages, e.g. Sanskrit, do not have sonant (voiced) s sounds, German (and also English) have both voiceless and voiced s sounds. Therefore, in such words as e.g. "Ruß" and "Nuß", we must examine two things:

Firstly, is the preceding vowel sound short or long?
Secondly, is the following s sound sonant or surd?

Mr. Twardoch considered the first question and omitted the second question. However the full rule states (I translate "DUDEN", 24th edition 2006, page 94):

"For the voiceless s sound after long vowel or diphthong, the spelling is "ß", provided that the s sound in all inflectional forms remains voiceless and no other consonant follows."

Since proper names (see e.g. English "Smith, Smyth") often disregard spelling rules, the matter is highly tricky, and I doubt whether the Unicode committee ever examined what sound (voiceless, voiced) is to be defined by the proposed "uppercase ß", considering that neither Mr. Twardoch nor Mr. Stötzner mention the tricky phonetical implications stated above.

This is how we do it. The graphic can be downloaded as eps too.
http://www.germantype.com/cms/upload/data/GTF_versales_sz.zip

--www.germantype.com--

Adam, I echo Hrant's comment on quality of discussion. I found your writing lucid and really interesting. Thanks for the time you have taken on this.

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”

Yes but if I want my C and c to look similar, and my S and s, and my O and o, etc why would I want my U/c ß to be different in design derivation from its lower case counterpart? i.e. if you're going to use s for derivation of ß then shouldn't you be using S for derivation of u/c ß? Is visual dissimilation (between upper and lower cases at least) necessary in a font? (It probably depends what the font's going to be used for which is a whole different discussion.)

Andreas: your pdf is excellent. It appeals to me that there are people that are willing to do lots of very hard work and then generously share the results with all for free.

My personal view is that the Γ approach makes it look like an uppercase character (whatever that looks like - it's very subjective) more so than the long s/inverted U approach. That is not to say that I'll use the Γ approach if it doesn't suit the rest of the font. The u/c ß is certainly not going to be a fundamental character which shapes the rest of the font.

But nobody's seriously going to tell me which actual design to use in my own font? Are they?

You've surely got to do what suits the font best and, as with all fonts, there are going to be some u/c ß's which are totally different from others because there are fonts which themselves are totally different. This reflects the diversity of man and the wisdom of his Creator.

I don't pretend to know anything about the history of type, derivation of characters especially German ones, German pronunciation and the like. Nevertheless, a bigger principle is at stake:

Individual 'interpretation' of character forms is surely one of the things that makes a font great. That's probably why some of us were so hacked off by being told what the euro symbol should look like - in every font! The thought that a designer must, in any measure, comply is abhorrent to me.

Provided my font can be 'read' (whatever that word means - one person reads very differently from the next), then it can look like anything I please, can't it?

I guess the real art is designing a font that can be read well by as many different types of reader. And unfortunately, when it comes to the real art is designing a font, I'm well out of my depth.

You know, if I had a German castle, I'd print all ten pages of Andreas' pdf and hang them side by side in one of the long rooms.

> if I want my C and c to look similar, and my S and s, and my O and o, etc

Well, you shouldn't. :-)
I mean for text.

> if you’re going to use s for derivation of ß then
> shouldn’t you be using S for derivation of u/c ß?

Derivation should be a means, not an end.

> nobody’s seriously going to tell me which actual
> design to use in my own font? Are they?

Some people will. Ignore them.

> one person reads very differently from the next

Actually, during immersion, "normal" adults all read about the same way.

hhp

> Yes but if I want my C and c to look similar,
> and my S and s, and my O and o, etc
> why would I want my U/c ß to be different
> in design derivation from its lower case counterpart?

The handwriting practice for ß by contemporary writers allows both "IS" and "I3" structures. The S shape is more flowy, cursive, written in nature than the 3 shape, which is more constructed, drawn. The uppercase-vs-lowercase contrast is exactly about that: drawn vs. written. Pretty much all of Latin uppercase has very simple, basic forms stitched together, which are easy to construct and draw. "S" is the by far most complex form within the basic scheme.

Combining a complex "long s" skeleton with an equally complex "S" skeleton is an overkill for a letter. It starts looking like a horribly complex character such as "&", which practically no-one can write properly.

I assert that letters must be simple. Uppercase letters must be simpler in skeleton than lowercase, because that is their very nature. Also, uppercase forms should be solid, strong, monumental in character.

I believe that the "monumental skeleton" of uppercase ß I propose:

http://typophile.com/files/germandbls_garamond2_5760.png

fulfils the criteria better than the "Dresdner skeleton" that Andreas Seidel and Andreas Stötzner prefer:

http://typophile.com/files/sz_birth_6221.gif

The Dresdner form lacks something sturdy, defined, at their top-left. It simply looks like it is half-way between uppercase and lowercase.

> Is visual dissimilation (between upper and
> lower cases at least) necessary in a font?

Of course, within a reasonable range.

A.

Ps. Edit: I meant "Dresdner form", not "Leipziger form", i.e. I meant the form that Andreas Stötzner included in his DIN/ISO/Unicode proposal (first link in my first posting).

Uli,

Unicode does not need to deal with phonetics. The uppercase ß character is for people who wish to unmistakengly represent the lowercase ß in an uppercase context. There are defined rules for the use of lowercase ß, and there is no need to explain them in the Unicode standard or anywhere near that context.

The phonetic rules I described in my original posting are, of course, simplified. I did not deal with "s" that is pronounced like the English "z" (sonor). In fact, I did not mention single "s" at all, because in German, the rules of its use are actually fairly complex. What I wrote about is the difference between "ß" and "ss", which both stand for a voiceless "s" sound.

A.

I think there is room for both the "Monumental" and "Leipziger" forms of the character, as one shape or the other will be more readable and graphically complementary within a given font, but it seems this is for the Germans to decide which one, if not both, is acceptable. (How, I don't know.)

What would a blackletter cap ß look like?

> letters must be simple.

It's very easy to read that in a detrimental way. I think the ideal is that letters are no simpler than readers can handle (which too often they are, like in the lc "el") otherwise you're reducing differentiation for no reason. Sure, maybe Chinese is too complex, but to me Latin is clearly too simple, and when given an opportunity to add a character to it we should diversify, not solidify.

As for what structure I prefer, I do tend to agree that your structure is better than that of Seidel & Stötzner, but Jason is right that the actual design on hand will always have a lot to do with it.

hhp

Just to stay on the right terms like Mr. Stötzner has "invented" it.
Leipziger shape = S / Dresdner shape = Z (З).

To the "problem" of the upper left corner, (make it true uppercase like a Γ or more round) is not a significant issue for me. Both approaches can work!
To me, Adams favorite shape is a derivate of the Dresdner shape like we have done it for our GTF typefaces.
I like the Leipziger shape too. It depends on the nature of the project which one I would prefer. The Leipziger shape is more complicated and usefull for some delicate titlings or for more calligraphic styles. The Dresdner shape I prefer for everyday typefaces.

--astype.de--

>What would a blackletter cap ß look like?
Blackletter don't work* in All Caps and there is no beginning "ß" same as in Serif and Sansfaces. That's why, there is/was no Versal-ß for blackletter "possible".

*and never intended to work

There is no there is no.
Just picture a document that explains this cap-eszet business, set in blackletter.

hhp

Poms is right, since the ONLY word what has a tradition in gothic or fraktur typefaces (in German languages) to set in uppercase is the word for god/lord "GOTT" and "HERR". But if you need the latest Hells Angels golf club tattoo, then of course, you need an uppercase blackletter sharp s. :-)

--astype.de--

@hrant and cuttlefish
I try to show you the "why it (especially in Frakur) was not intended" later, in another thread, cause i don't want to hijack this one.

Edit:
Not another thread :-)
You see why it was not intended. Even in Schwabacher or Neu-Gotische Schriften made in the beginning of the 1930s it doesn't work.

I love intent. But as a means, not a goal.

Saying a blackletter cap-eszet should't be
made does not befit a good type designer.

hhp

This has certainly given me a lot more insight on how to treat the character in my unicase face.

I am well aware that it is not right to set blackletter (and especially Fraktur) in all caps, and that no word in German begins with ß, but people will try to do either anyway. What is the harm in acommodating them? Let's leave that conflict of capitalization to the copy editors and tattooists. If they want to use the letters we design, let them do what they wish.

Yet, i don’t think that names are the foremost concern.

Oh, no, they are! I come from a town called Pößneck and we have a battle for decades how it should be pronounced and how it should be written in uppercase texts. It literally splits a whole town.

Of course it does. Herr ROSSBERG (i.e. Rossberg) is pronounced “rosberg” ... while Herr ROßBERG (i.e. Roßberg) is pronounced “ro:sberg”

No, the reform of 1996 doesn't apply to names. On my businesscard is says »Roßbach and Herrmann GbR« and it is still (and will always be) pronounced »Rossbach«.

In our business the correct spelling of names is the biggest concern in the capital sharp s debate. Just think of all the bibliographies in books. We usually like to set names in small caps, but names have to keep their Eszett, because Weiß and Weiss are different names.

For the public this is a solution for a problem that most of the people aren't even aware of. Again: Herr Roßbach from Pößneck will always write ROßBACH and PÖßNECK to keep the proper names. So now we offer him a way to do it with letters that are typographically correct.

About the shape:
I don't think we should define the one and only way of designing the capital sharp s. There are different solutions and over time one will eventually prevail.

CORRECTION

The glyph prefered by the germantype-colleagues and me is the *Dresden glyph*, not the "Leipziger form". The Dresden glyph was first proposed by a man from Dresden in 1955.
It does not help to mix things up.

Actually having a descender on it (at least in fonts that have a descending
"J" and/or "Q") seems quite smart, as it helps differentiate it from the "B".

hhp

Although the tail makes it easier to cofuse with Beta. That said; the tail appeals to my eye in big way and (obviously subjectively) helps me resolve what I am seeing as an Eszett much faster.

For more on my beta comment see here

http://en.wikipedia.org/wiki/Eszett

Although the tail makes it easier to cofuse with Beta.

You would expect a greek lowercase letter in an uppercase German proper name!? ;-)

Folks have used ß as a substitute for beta for as long as computers have had the character in their fonts, for such things as marking pre-release version numbers and such. Why anyone would want to uppercase their version number marker, I don't know, but if it extends utility, I don't have a problem with maintaining a passing resemblance.

>You would expect a greek lowercase letter in an uppercase German proper name!? ;-)

Why not? Most readers on the planet know neither German nor Greek. Yet we expect them to know the difference between betas and Eszetts. When I was doing corporate design work in Washington, DC, we had to disguise a lot of the companies who we researched. So instead of publishing a report on Acme, Inc., we called them Beta Corporation, or something like that. Readers of our reports knew who we were talking about, but we didn't name names without permission.

Anyway, the authors I worked with always thought that it would be clever for the first letter of Beta Corporation's name to be an actual beta, rather than a b. But their fonts did not have betas in them, or they didn't know how to input a beta. So they just write ß (ßeta Corporation). Most of them thought that the ß actually was a beta anyway. Good luck trying to educate them to see around that one ;-)

As a teen, I myself used to think the eszet was a beta. But I think in context (not least the context that somebody reading German probably knows German...) it's not a problem.

hhp

So they just write ßeta Corporation.

That's a completely different thing. Expectations play a big role in reading texts. So when you write ßeta you suggest this way of reading.
Of course people could be confused with ß vs. β when placed as a single letter. But in this discussion of a capital Eszett in proper names this really isn't an issue.

I think Ralph's point about context is a good one. Also, the sharp angle of the 1st loop does a good job at quickly clueing you in. And having a descender is a nice way to keep distiction with the UC B.

My hat is off to Adam like eveeryone else for introducing the thread and his highly lucid text on the subject, but to my eye the samples he gave feel cyrillic rather than latin which Ralph & Andreas' do not. And I think German aught to feel mostly latin - even in a glyph such as this. I do like the fact that he showed some descenders though.

> I think German aught to feel mostly latin

Only after it feels fully German.

hhp

Only after it feels fully German.

In the sense that a 'new' glyph should be 'on the team' and not stand out particularly much when used. Then yes. Entirely. In the sense that it should extend naturally from the historical German culture of letters. Again yes.

In the sense that it's 'Germaness' as an essence needs somehow to measured - then no.

Actually, I had been wondering if anybody would confuse my statement as being anti-blackletter. It wasn't. (There is a false but long held historical/cultural dichotomy of Roman/Latin vs. Blackletter/Fraktur) But actually what I meant by Latin was this visual/cultural continuity and the reality of intercultural pentration via loan words etc.

And in fact one great way to give consideration to a form might be to try to write it out in the various blackletter styles: Textura, Bastarda, Schwabacher, Fraktur, Burgundica, etc, to see what happens. BTW, I am also not suggesting that direct continuity of form from the Sans serif to the Blackletter would be desirable either. Looking at Fette Fraktur's C,L, and S certainly suggest the opposite might be true.

When I think about this letter it seems like the top 3 purposes would be :

- Monuments in a Trajanesque mould,
- Signs in a Sans Serif style,
- and things like Pickle Jars ( packaging )

... Admittedly tattoos and the back windows of pickup trucks might be candidates as well but I am less worried about those applications.

At Hrant's request, here's the text of a message regarding the uppercase eszett that I sent to the ATypI discussion list. I should note that I originally opposed the encoding of the new character in Unicode, arguing that it could be implemented as a ligature of SS, with a plain text distinction made using the Combing Grapheme Joiner control character. I still think that good arguments can be made in favour of such an encoding -- for instance, it enables documents to be backwards compatible with older fonts that do not support the uppercase eszett glyph --, but on the whole I've accepted Andreas Stötzner and, especially, Asmus Freytag's reasoning in favour of the proposed encoding.

Anyway, Hrant thought the following was a good synopsis of the matter, offered here as a contribution to the understanding of how we came to be where we are: wondering what this thing should look like.

A few comments about the proposed uppercase eszett character:

This has been proposed to ISO 10646 (and hence Unicode) by DIN, so the likelihood of it being accepted for encoding is very high. Since it is a German character, with no implications for other languages, no one feels particularly inclined to vote against a DIN proposal. If the Germans determine that they want this character, they can have it.

So why do they want it? Basically, they want it because at least some people are using it and it is considered better to have an unambiguous encoding distinct from SS. The uppercase eszett is not a new invention: it has been around for at least a hundred years. It has been used on-and-off by some publishers -- notably on the cover of the DDR edition of Duden -- and although it has never been enshrined in the official norms of German orthography it persists. As Asmus Freytag wrote on the Unicode discussion list: it is remarkable that something without official sanction should have persisted so long in such a rule-bound culture.

There can be little doubt that at least some of the users of the uppercase eszett have employed it in the interests of alphabet reform, i.e. they want to see it become standard. Others may have used it only as a display solution for a particular piece of typography. For whatever reasons, some Germans have found a use or a need for it, in preference to SS.

And let's be clear that the casing rule ß = SS is really messy for a nominally bicameral alphabet. Although ß began life as a ligature, it has long since ceased to function as one: it has distinct semantics from the sequence of two letters that it used to represent. It is a distinct character. The idea that it can be a distinct character in lowercase but not in uppercase is perverse, and causes major headaches in German text processing. The simple fact that case conversions in German are not roundtripable without dictionary support is crazy. Even if it were decided that the uppercase form of ß should always look like SS, a good argument could be made for encoding this as a distinct uppercase character. However...

...that's not how German is encoded, and the DIN proposal for uppercase ß is explicit that the new character should not interfere with existing standards for German text. So this means that ß still = SS, and the Unicode casing rules for German remain in place. The new uppercase ß character will exist alongside these rules. What this means is that 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.

The uppercase ß encoding is an interesting instance of a character being made available outside of the norms of encoding for the language in which it might be used. This might sound silly, but it allows this persistent heterographic letter to function in modern text encoding and display environments, and in so doing puts a lot of control in the hands of users to determine its future. It may fall by the wayside -- one more little used Unicode codepoint -- or it may, in time, become formally recognised and made part of the standard orthography.

At first glance, it may seem contrary that, on the one hand, the Germans have regularised and simplified the use of ß -- suggesting, at least to Bruno, that it is on the way out --, but on the other hand providing at least some recognition for an uppercase ß. But it is easy to see how these may be complementary moves: providing a distinct uppercase equivalent to a lowercase character regularises an anomaly in a bicameral alphabet and simplifies case conversion.

Monuments in a Trajanesque mould...

http://www.tiro.com/John/TrajanEszett.gif

This is an illustration I made during discussions of this topic on the Unicode list. I decided to put the uppercase eszett to the Trajan Test. In this instance, I was also interested to see how convincingly a form could be created that explicitly referenced the uppercase S for the right side. I'm not completely satisfied with the results, and in general prefer the models I have seen with the diagonal at the top right, but some people liked it.

Thanks John.

For the Trajan cap eszett, I think it's decent, but what I might try is
adding an explicit serif/terminal of some sort at the top-right corner.

hhp

Thanks John.

For the Trajan cap eszett, I think it's decent, but what I might try is
adding an explicit serif/terminal of some sort at the top-right corner.

hhp

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.

duplicate

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

"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

David, perfect for Valentine's.

Hey, what about this? WEI§ACKERS
:->

hhp

"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. :)

[Rue Montmartre, Paris]

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

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

hhp

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'.

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!

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

"(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.

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?

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

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!

> 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

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

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

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

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.

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! ;)

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.

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.

> 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)

"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

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.

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?

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.

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.

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

comedy. Or tragedy.

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

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]

Thanks, Karsten!

ChrisL

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

Because it's fun to invent new letters!

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

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

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.

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

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

... 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 :-)

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.)

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

Wouldn’t that work?

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?

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.

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.

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?

It’s about unambiguousness!

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

A designer feeling tempted to insert it should use keyboard or Glyph Palette.

I value your opinion, but Ralf also has a point. I think it would be nice if the designer could choose to do it manually or via an OpenType feature.

Fortunately, I think, there are organizations around in Germany who make decisions about proper and improper German writing. The Unicode Consortium is not one of these organizations. IIRC, the most recent issue of DUDEN states that ß should be written as SS in strings of capital letters.

With a cap ß in Unicode, you could program a font to make substitutions like Meißner to MEIßNER (with cap ß… not in this font yet, hehe), but it still would not be correct German. The correct way is still MEISSNER. Now, because software has changed, perhaps a future DUDEN will make a spelling change, too. But it has not yet done so. And I think that they will want to hear from more than just a few concerned typographers, too, before they change the language's rules. Who knows?

So, when you build your fonts, if you put in a capital ß (who is really going to start doing this? Adobe? Microsoft? Monotype? Linotype? FontFont?), it requires a bit of code, because when the word Meißner is highlighted in InDesign and made all caps, the change should still be to MEISSNER. Since there is no cap ß key on any German keyboard, the cap ß will only have accessible via the glyph pallette or a stylistic set, contextual alternate, etc.

but it still would not be correct German. The correct way is still MEISSNER.

Dan, that's not true. According to the DUDEN proper names should keep their Eszett when set in caps.
Aus Gründen der Eindeutigkeit. (»for the sake of unambiguousness«)
They use the example »HEINZ GROßE« (set with a lowercase Eszett of course)

Ralf

Good heavens! Figure out how to program it correctly then, I guess.

I am very sad.

Ralf, you have to provide the full quote to do justice to the Duden. ;-)

It’s merely a marginal note to the general ‘ß → SS’ rule and reads:
In Dokumenten kann bei Namen aus Gründen der Eindeutigkeit auch bei Großbuchstaben das ß verwendet werden

For the sake of unambiguousness, the ß may also be used for names, when set in capital letters, within documents.
[emphasis on the various qualificatory remarks by me]

Figure out how to program it correctly then, I guess.

Setting German text could never be done automatically. In blackletter you need to care about ch, ck and so on and in modern texts you still have to break automatic ligatures (fi/fl/ffl/ffi and so on) across the parts of compound words.

If you do want to use the ß in a caps environment you should bear in mind that most ß's are of ascenders height and do not have the same stem width as regular caps.

This uppercase eszett is an encoded character, so it is not advisable to make it the target of glyph substitution via any layout feature.

There are some precedents, for ordinals, basic fractions, and the fi and fl ligatures.
What could go wrong if:

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

Although the cap ß is supposed to be a fully-fledged upper case letter, it isn't treated as such by the casing rules.
Isn't it reasonable to assume that typographers will think of it as a ligature?

Although the cap ß is supposed to be a fully-fledged upper case letter, it isn’t treated as such by the casing rules.

but it's not (not really). at least not according to the DUDEN, as explained by Florian above.

There are some precedents, for ordinals, basic fractions, and the fi and fl ligatures.

if you look at the behaviour for Arno Pro, you will see that none of these substitutions fall into Adobe's current "best practices" and have effectively been deprecated.

R.H. -- But setting a proper name right doesn't matter?

Again, there are well-established rules for eszett case-mapping. If you do not know or like these rules, it is your problem.

R.H. -- 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.

Not sure if this was before you started studying design, but maybe you remember that it took some time until InDesign finally supported the current orthographically correct case-mapping from eszett to SS. I would not like to miss that.

Again, there are well-established rules for eszett case-mapping.

So what? Something can be »well-established« and still be wrong.

As a summary:

(1)  There is a new Unicode codepoint for uppercase eszett.
(2)  There are orthographic rules for case mapping. And it is not the Uncode Consortium's business to change these. This means, the eszett-to-SS case mapping is left untouched, as pointed out explicitly in the UC's information. What is allowed is mapping uppercase-eszett to lowercase-eszett.
(3)  If anyone plans to add this letter to fonts, the only questions left are:
What does it look like? Up to the designer.
How to name it? Maybe 'uni1E9E'.
How to deal with it in features? Add "sub uni1E9E by germandbls.sc;"* in 'c2sc' to address the uppercase-to-lowercase mapping mentioned in (2). No coverage in other features though, following current best practice to not substitute encoded glyphs with encoded glyphs.

This doesn't have much to do with opinion but with currently valid orthography and best practices as regards font production.

So my question is, what is the point you are trying to make? Of course you can spell and case as you like, but the result would be your "private Sprache", nothing that should concern anyone else.

Sorry for repeating things already said in previous posts.

* Or whatever your smallcaps suffix is. If you use two sets of smallcaps (one mapping from lowercase, the other from uppercase) then it would be something like "sub uni1E9E by uni1E9E.c2sc;".

[Edited. Originally located between Ralf Herrmann 9.Apr.2008 3.33am and dezcom 9.Apr.2008 8.11am.
Corrected post and added a note. Hell, thanks Nick! Another typo ...]

"...How to deal with it in features? Add “sub uni1E9E by germandbls;” in ’c2sc’ to address the uppercase-to-lowercase mapping mentioned in (2). No coverage in other features though..."

Thanks for that succinct clarification, Karsten. I now at least have a way to deal with the nuts and bolts of this issue as a non-German speaker and will happily allow others to debate the pros and cons of inclusion as a codepoint. :-)

ChrisL

My point is that it is not true that there were no problems at all. Your statement sounded as if a handful of people just made them up. The problems are known for at least a 100 years. Ambiguous proper names for hundrets of German cities and tens of thousand Germans are simply not acceptable in modern printing and data processing. That's pretty obvious and not just my own strange opinion.

Now that the Unicode point is official, it's up to the type designers and graphic designers to support it or not. Since the old case mapping is still in effect, everyone who doesn't like the capital sharp s has no need to worry. You can leave your OT features as they are and you can go on and set German texts as always.
As someone who would like to use, I'm perfectly fine if I have to fish it out of the glyph palette whenever I have to set a German proper name with an ß.

Ambiguous proper names for hundrets of German cities and tens of thousand Germans are simply not acceptable in modern printing and data processing. That’s pretty obvious and not just my own strange opinion.

of course it's an opinion. the question of what is acceptable will always be based upon opinions. the fact in this statement is that there are ambiguities that may or may not be acceptable depending on who you're talking to.

Karsten kindly reminded me that the concern of most people here is probably not the discussion about the pros and cons of this character, but how it affects OpenType features programming. The answer is simple: not at all. Put the character in the font and name it »uni1E9E«. That's all!

“sub uni1E9E by germandbls;” in ’c2sc’

Why not: “sub uni1E9E by germandbls.smcp;” in ’c2sc’?
(Where germandbls.smcp is a small-cap version of the new character.)

Won't this decompose correctly in plain text?

How to deal with it in features? Add “sub uni1E9E by germandbls.sc;”* in ’c2sc’

even better would be "sub uni1E9E by uni1E9E.sc"

Not really, because the character should decompose to ß in plain text.

A problem with "uni1E9E.sc" would occur if:

1. The original plain text is U&lc, with ß.
2. This is capitalized in a layout program with an "all caps" command.
3. Double-S (the capitalization of ß) is manually replaced by uni1E9E.
4. The text is changed to "All Small Caps", and output to pdf.

Now, a search will not recognize words which are composed of lower case characters, with uni1E9E mixed in.

At least, that's my present understanding of the protocol.

P.D.H., yes, in both cases wich I mentioned it should be “sub uni1E9E by uni1E9E.sc;”, with "uni1E9E" before the smallcap suffix. I must be sleeping today.

Note that the Unicode example gylphs are just informative, gylphs in fonts actually might look considerably different.

1½–2 years ago I posted in a german type forum a shape I personally liked (and for fonts where the Uppercas J has a descender, it's the one I'd ever use in my types; personal preference though):

I dislike the left part's "turned U" shape in most other versions and prefer the ſ~f — Γ~F analogy. Clearly preferring the З form of Z (mostly used in Fraktur, but sometimes found in Antiqua as well).

Of the four blue forms I prefer the third.

I'm just trying to give inspiration, of course, (this is clearly only one option of many), and am hoping that you find it useful or interesting at least. The nicety of type design is, that you can find so many different solutions for a given problem.

More than just a pumped up B: Germany celebrates recognition of the letter ß

It looks like ISO has finally ruled on the Eszett, but that article in the Guardian doesn't really say anything about the uppercase form. Does anyone know if a preferred approach was part of ISO's decision?

That article is soooooo superficial, conflating issues of ß and its capital version.

--Szabolcs

Exactly. It didn't really say much of anything useful, Whig is why I was hoping someone else
may have dredged up any more info.

It seems that most readers of fontblog.de agree that the ẞ form that I proposed at the top of this thread (a ΓƷ ligature, so to say, with a pointed upper-left corner) is superior to the form with the round upper-left corner — which to many still looks like an enlarged lowercase letter.

Also, it might be a good idea to give ẞ the proportions of a "wide" capital letter (M, W) due to its "double" nature.

Regards,
Adam

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 film izle solved. We’re in the middle of a slow transition period for “ß”. The 1996 reform started it and showed the direction.

>I believe that it should be an exciting task for type designers now to come up with a new form.

Absolutely! Without exception every time a bunch of crackpots come up with a new character, that four or five academics and their dog might use once in a blue moon, there’s dancing in the streets!

Thank you, Si. That was wonderful! I don't know how I could enjoy Typophile without you :-)

Plenty of other uppercase letters look like enlarged lowercase counterparts, such as C, O, S, V, W, X, Z and others depending on the style. Why should this be taboo with regard to an uppercase longs_s or longs_z ligature? Is it because of insisting that the uc long S be based on the Gamma rather than something more S-like, such as the integral symbol?

Plenty of other uppercase letters look like enlarged lowercase counterparts, such as C, O, S, V, W, X, Z

I disagree. The letters v w x z are reduced versions of the uppercase rather than the other way round. Definitely historically speaking, and also stylistically imho. C O S are neutral in this respect, I would say.

> Plenty of other uppercase letters look like enlarged lowercase counterparts

A sad reality that we should not amplify.

hhp

I have played around a bit:

Like others, I have come to the conclusion that it absolutely must have a top left corner, otherwise it will look like a lowercase letter.

> Plenty of other uppercase letters look like enlarged lowercase counterparts

A sad reality that we should not amplify.

Sounds like the seeds of an interesting Type Battle.

Tim,

handsome work. I like the one that you used in the top row (the one incorporated into the word "großes"). But my favorite is number 2 in the bottom row (the one with the edge on the top): thanks to the edgy roof, there is no way to mistake it for a "B". Also, it's easier and more natural to draw by hand (one-two-three straight strokes plus an S shape). The forms with the half-rounded top are somewhat unnatural and clunky to draw by hand -- the top part requires an "unfinished", "interrupted" hand movement. Also, the form (number 2 in the bottom row) has an unmistakengly "capital" flavor.

A.

Adam: "...there is no way to mistake it for a “B”..."

LOL...there is no way to mistake it for an 'S' for that matter.

Why not make an "uppercase" long and short 's' look more like an FG ligature, or a PQ ligature, or a pie chart! ;)

Cheers!

David,

There is no need to mistake it for an "S" just like there is no need to mistake a "ß" for an "s". ẞ should look like an uppercase form of ß. That's it.

Like others, I have come to the conclusion that it absolutely must have a top left corner, otherwise it will look like a lowercase letter.

Like others, I have come to the opposite conclusion.
IMO a top left corner makes the letter too busy, accentuating its "ligaturishness" rather than its integrity as a discrete character.
In all your examples, the lower case "s" is clearly visible--it would be better to more fully integrate it into the letter form.
Don't the better lower-case "ß" designs follow the principle of camouflaging the "s" component?

Nick, I agree that there's too much "ligaturishness" overall, and that that's bad. But it's actually not the top-left corner doing that - how could the corner there be causing the overt visibility of the [second] "s"? I think the corner is helpful; but Adam's preferred form is not good enough.

hhp

Adam: "ẞ should look like an uppercase form of ß"
If we all agree on that, then what's the 3/4 of a lowercase 's' doing there?

Hrant: "...how could the corner there be affecting the overt visibility of the [second] “s”?"
By being there?

Cheers!

David,

have you come up with a better solution than the ones shown here?
If not, I won't accept your criticism.
First, show us your solution and then you can criticise the others.

Nick,

Don’t the better lower-case “ß” designs follow the principle of camouflaging the “s” component?

Do you mean the beta-shaped ß? I certainly don't like it but I guess that's a matter of taste. There are some very good fonts with the ligature/visible s model. Btw, wasn't this type first proposed by Tschichold?

Do you mean the beta-shaped ß?

Not necessarily. Just look at the Garamond ß in the first post of this thread.
The shape of the curved side of the ß has an integrity of form, it is a complete "line of beauty" in its own right, with a central angle which has a perpendicular quality that is part of the curve shape, not merely the joint between the "f" and "s" components.

Here's how I would do it, using for my examples Garamond Premiere Pro, Bodoni Bold BT and Futura Book BT:

My design decisions are based on the following thoughts:

- I think the cap ß should take its cues from the lowercase ß as much as possible -- they should look like they are related stylistically, since they are part of the same typeface, and people are already familiar with the lowercase form.

- I don't think the ligated (Tschichold) form can be translated to an uppercase form, as much as I like it in the lowercase form. It's too complex and relies on the space above the s, which caps don't have. This leaves the "B" form.

- However, I tried to get a hint of the ligated form in the Garamond example.

- In the lowercase it helps that the ß doesn't look like any other lowercase letters. But, as a capital, I don't think the "B"-ness of the ß can be avoided by introducing novel forms to its construction. My solution, to help distinguish the cap ß from the cap B, is to open up the interior space and make it a bit wider

- The tail on the bottom right takes a form similar to the lowercase form, but in scale with the other capitals. The terminal on the cap J usually works.

- I used a corner shape in the upper left because it makes it look more like a cap and to help distinguish it from the beta.

- Some of the other proposed designs I think look too busy, novel or exotic, calling attention to themselves and failing to harmonize well with the other letters. I also think a lot of them work for some faces better than others. I hope to overcome this by sticking with elements of letters and forms that are already present in the typeface and keeping the general form simple and straightforward.

Here's how I might do it with Trajan:

Nick,
now I see what you mean. The ß from Garamond is a real beauty – maybe because it is the golden mean between the "beta" and the "ligature" models?

Mark,
interesting comments. Before I read them I thought, is he trying to make the UC ß as similar to the B as possible? But maybe it's just a matter of getting used to it. After all, we distinguish many letters by relatively small differences.
That letter remains a difficult thing to design and I have the feeling we might never find really convincing solutions.

I think I made the Futura version too narrow. I think this works better:

After all, we distinguish many letters by relatively small differences.

Which ones?
There are letters like E/F or P/R which share large parts, but they still have a significant difference in there general skeleton, and this difference is maintained and clearly visible in any size or any kind of type style. But all of Marks designs use the skeleton of the B with a "strange thing" going on at the bottom. I read nothing but Bs even if I try to see a cap Eszett.
Tim's designs are not the ones I would favour, but they clearly solve the problem. I am not tempted at all to read GROBES.

This thread has a sort of surreal vein to it… :=)

I tried to work out my solution for Neoritmo, and I am satisfied with it. I must say I probably cheated, since Neoritmo, from its conception, already incorporates what "technical people" would call a sort of "disntegration/integration of bicameral logic", and it has an uppercase H which is still rooted in the lowercase. A beta of Neoritmo can be seen here: http://www.cannibal.gr/multimedia_sub.aspx?cat=3&sub=7&prj=38&xyz=742&la...
I will try to post my "Double S" ASAP…

I read nothing but Bs even if I try to see a cap Eszett.

I was afraid you'd say that. :-)

Depending on the font (Georgia's not bad, because of the old style figures), many of these are hard to distinguish without context: Il1, O0o, rnm, 5S, 2Zz. In some cases, the only differences are a little hook or serif or corner instead of a curve or a slight gap or difference in height or width.

Context helps a lot. Even the current ß, if used in place of a B, reads as a B.

It's problematic to introduce a new form to the alphabet. Reading depends on familiarity of forms. If a novel form is introduced, it's natural to try to interpret it as a letter we already know. (A lot of logos use this to great effect.) In cases where there are two words, such as große and grobe, it may take more time to get used to, but I suspect that this is an edge case and unusual, and that most words that contain ß do not have such an evil twin and would be interpreted correctly due to context, given time to become familiar with the new form (even mine).

In any case, as someone who doesn't read German natively, I do feel at a disadvantage with this problem. I'm just trying to apply what I know to the problem and making some observations. Part of my motivation for speaking up is selfish: Some of the proposed designs look very awkward to me and I would hate to have to be the one to adapt them into existing faces.

I must admit that most of the designs Mark posted look like broken B glyphs or R's with a wayward tail. I think as designers, we work on glyphs and have a knowledge of what is the essence of that glyph since we have been seeing it or drawing it all of our lives. The trouble comes when we define a new glyph that as of yet has no agreed upon essence in form, It is simply not in our repertoire yet. I guess we are a few hundred years late to the game :-) Instead of depicting the essence by giving it form, we are defining the essence without knowing the form. It is like a tailor who knows how to fit a suit of an old customer, having done it many times. He may get a call from a brand new customer but gets no details of size or measurements or taste in fabric or color, he only know that he is a different size than all of his other customers. Would our tailor create the perfect fit sight-unseen on the first try?

ChrisL

ABCDEFGHIJKLMNOPQRSTUVWXYZ pretty clearly expresses itself as dominated by second-side terminal logic. So, when something is expected to be, by the sound unvoiced though it may be in a reader's mind, first-sided terminal logically, I think it is pretty important that it is, just to keep der dear reader downset. That a barless F looks more like an uppercase character than a J, is interesting, and may be appropriate in some places, but maybe just easiest way to make something look uppercase, and is not specific enough. Most who have experimented here have gravitated to serif designs, the ones with little feet things attached which makes it blend in all nice, but... I think in humanist sans, and of the sans experiments here, I think few approach success. So, relying on the stored biological memory of the original, pre-laziness lowercase form, and the space-saving intent of its creators, I go that way — clearly not a B, Beta, F3 ligature, SS, with no confusion at the waist and no gigantic open counter and with first-sided terminal logic.

Cheers!

Cagri, could you please me point to those two medieval examples of capital long S?

Teşekkürler:
Szabolcs

Drawing, you know, is the first refuge of the confused.
@David: why?

Beware of people who (want to) make history without knowing it.
[…]If you don’t have a proper understanding of what you are supposed to design, what are you designing then?
@Karsten: most important remarks. Altaira is also right: see historical precedents, if the form has in some cases been written in "more uppercase" form (there is no fixed distinction about the two, save the one matured in typographic tradition).

A great specific example here is the ampersand. It comes from “e” + “t”. Whoop-di-doo. When you take that too seriously you end up with a form that causes reading problems for many people*, like in Poppl Laudatio (which I otherwise adore). A well-designed ampersand is based on what people need; what it used to be is just fodder for pedantic discussion.
@Hrant: Downright pretentious, because the "and-per-se-and" is, and actually remains an "et".
I would have had problems to understand its being an [e] just *before* someone told me it was an "et", which – as an Italian – comes as a perfect logical succession, or better as the same thing.
It has happened some people asked me to change Rotis' "Et" into a more common [&] form, but that's because people in Italy still think of the [&] solely as a "commercial e" (in Italian it's called "e commerciale"), and they do not know it's an "et", out of ignorance.

Read both these entries:
http://it.wikipedia.org/wiki/Ampersand (Italian)
http://en.wikipedia.org/wiki/Ampersand#Usage (English-International)

As much as Wikipedia is nonsense, these entries show how the evolution happens by a combination of planning and usage, despite all the thinking it may be done before.
The world may have become "anglocentric" (in terms of shared tongue), but – as you see – this happens out of necessity, otherwise we wouldn't be allowed to be here to talk together in a shared experience…

Now, my big question (since it is still not clear to me): is the [ß] a simple double s? Of course not, since it's a "sharp" z-sound. So, why people keep using it as a substitute for "ss"?
Besides, how an "ss" ligature should be drawn?
And [AE] or [OE]? They are not ligatures, they address sounds, right?

Followers of this thread may be interested in an article on the history of ß that appeared a year ago on the Typefoundry blog: Esszet or ß. Interestingly, there is a reference to this Typophile thread in the Sources section of the article -- is this an example of circular reasoning? (Smile.)
 
         ~~~~~~~~~~~~~~~~~
         When going from A to Z,
         I often end up At Oz.

Any takers?

Nope. This is anything+J.

(The bottom bowl of the lowercase eszett is too wide. I think you can move it toward the left stem -- there is no reason why the internal distance is wider than that between eszett and e, or e and n.)

It looks that there will be some spotty font support for this character in Windows 7: http://blogs.msdn.com/michkap/archive/2009/07/28/9850675.aspx

I thought it was interesting that at the "crit session" at TypeCon, the panel of experts seemed pretty skeptical of the whole idea of the glyph.

Yes, I was in the hot seat at the time when the wrath of the Fonting Dutchman fell upon me for including it. I take their comments on this with a grain of salt. There is a unicode point for it and there were enough German speakers supporting it to make it accepted as a legit glyph. Even though it will rarely be used, I just include it for those who need it. I don't care to reopen the battle Teutonic Knights. I will let the Germans who want it have it, and those who don't can just ignore it like the florin.

ChrisL

I dont see, into now, samples of this uppercase "germandbls" in calligraphic styles. Someone here can me show this kind of samples? Where I can find?

Hi Paulo. This article from Andreas Stötzner shows samples for script Cap Eszett. You may find two samples at the bottom right of this PDF.

Additional sources of research:
http://www.spiegel.de/kultur/literatur/0,1518,539266,00.html
http://www.signographie.de/cms/front_content.php?idart=136
http://www.signographie.de/cms/upload/pdf/Signa9_ZurDresdnerForm_1.0.pdf
http://www.signographie.de/cms/upload/pdf/Signa9_Kombinatorik_SZ_3.0.pdf

All these links are found from Signographie.de. Andreas Stötzner is the great specialist on this character and also the one who proposed it to Unicode. You may also note there are different historical forms according to German region the shape comes from. For me, Dresdner is the most interesting one.

Thanks, Freiberger. I am thinking to add (althought I dont know where...) this character in my new Penabico font. Thanks ever. abraços gaudérios.

So, Freiberger and people, I create my version of the Uppercase germandbls, to my Penabico font. What you think about the aesthetic solution

by the way, I have a topic forum in the Calligraphic forum about the first tests with Penabico in deutsche language texts (with Lowercase germandbls, but without the Uppercase) - here is http://typophile.com/node/74389

and, a last question. How I add the Uppercase germandbls at FontLab. Just create a 1E9E unicode gliph?

Yes, the Unicode codepoint should be 1E9E and the glyph name should be "uni1E9E".

Thanks Adam.

Paulo, my favorite is the first Eszett, although all them are beautiful. Anyway, I think the design with rounded upper may cause confusion with B. A straight upper (the Ezh-like form) probably would be preferable to avoid confusion. Hope to hear some German typophilers about this.

@ Paulo

No word begins with an Eszett. So I wonder, if your script Eszett is intended to be used in the begin of words.

Arno, you are right. The uppercase Eszett can be used only at the middle of the words. I show here my lowercase versions - you can see the lowercase eszett in german testing in the pdfs here http://typophile.com/node/74389 -

Unfortunately, Penabico (like many copperplate scripts) dont work good using only the versals, but only using versals followed by lowercases. The use of all uppercase letters to write words need by avoided. In the case of the uppercase germandbls is really a problem. But, I am working in this. This is a new experience for me, in any case.

Freiberger, you can be right. So, I will design a version (the Ezh-like form) to try using your advice. However, in this font, the B and Eszett are very distincts, like you can see here.

In time : the design of this font is from my wife, Iza W. I am only the art creation director and the programmer at the FontLab.

Since the character ſ is an alternate form of the lowercaſe S, one could claim that it has no business in an uppercase character at all. And so perhaps the uppercase ß (for use by people who wish to use this character, which traditionally has no uppercase form, and so words that contained ß and which could alternately be spelled with ss, for which ß is, as pointed out, a ſs ligature - and, therefore, in printing, the word Große becomes GROSSE when in all caps, and current Unicode mappings reproduce this - although since no Unicode code point was previously defined for "uppercase ß", this means information is lost, which can cause problems when changing back to lowercase, it's not as if an uppercase ß codepoint just prints as two S glyphs) ought to be derived from SS - just print the two uppercase S's close together and overlapping.

As long as they're normal curved S shapes from the font in question, and not in any way... runic... a certain disturbing comparison ought not to arise as an issue.

Here is what I am thinking of:

It looks like the uppercase of the things the lowercase eszet is made from, rather than like an eszet turned into an ahistorical character that sort of looks like a capital letter. Of course, as I'm neither German nor a professional type designer, I can't claim to be qualified to make a suggestion to be taken seriously... but, then, this is such an unusual notion that perhaps anyone can play.

It looks very dark and calls too much attention to itself. It also reminds me of the Section

It is just a rough sketch on my part, though. I see that at least one person here did already suggest something based on the capital S but in a different way (elongated middle bar, top and bottom curves including a pointed cusp to suggest two 'S's). The middle stroke of the two 'S's could always be made lighter.

But you are right that something with two parallel strokes separated by a small space "calls too much attention to itself", as it looks more like a logo than a letter, even if the "too dark" part is easily solved.

The other problem is that, as far as anyone who cares about this is concerned, the train has long ago left the station.

Freiberger, concerning the Ezh-like form, like you advice me, I did make these variations. But, I pretend add to the font all the six versions (three earlyers plus these ones) I think

@paulow

To me as a German reader, I still see them as Bs.
It doesn't matter, whether your real B look different in this typeface. Reading is too automatic to make such comparisons "on the fly".

But Nick Shinn has already solved it outstandingly:

Thanks, Ralf, very clear the distinction between the shapes. I think now I understand.

Paulo, I missed your question. Anyway, Ralf already give an answer –BTW, much better than I could. Hope to see Penabico released soon.

Thanks, Freiberger. I think which, maybe, Penabico will be released to public in few days, althought I will work for many months making updates, into the final project to be complete. I am attaching here a new design to the Germandbls uppercase, following the advice from Ralf (thanks, Ralf). I know which the shapes are not good yet, but, the overall shape is running to the Ralph/Nick Shinn concept, what you think?

Paulo, can we see this in some kind of context? It isn't possible to judge these shapes without seeing them next to other letters to compare proportions and fit.

I suggest going back through this discussion and find some sample German words involving the sharp-s to set in all caps.

Yes, John. Here is a sample with two words. A great problem is which this kind of font is projected to works with uses of uppercases followed by lowercases. The dozens of versals variations works well in this way, but are confused writing all in uppercases.

UPDATE: In 2010, the German Federal Agency for Cartography and Geodesy (“Bundesamt für Kartographie und Geodäsie”) published the 5th revised edition of the document “Toponymic guidelines for map and other editors for international use Federal Republic of Germany”.

Section 1.1.3 of this document says:

1.1.3 Special letter ẞ ß
The special letter ß (“strong s”) existed hitherto only as small letter (minuscule). In 2007 DIN and ISO accepted the capital letter ẞ which is rendered on position 1E9E of the Unicode character tables. In official spelling it must not be substituted by any other letter combination, as e.g. SS, ss. In the alphabetical order ẞ, ß is treated like SS, ss. Since a typographical implementation of the upper case letter ẞ in the various character fonts will take some time, it may be temporarily substituted by SS, ss.
NEW: Due to the new regulation of German spelling the letter ß is after a short (stressed) vowel now replaced by ss. The letter ß remains after a long vowel or a diphthong.

Examples for names of geographical regions:
— now: Hassberge (short and stressed vowel a), until now: Haßberge
— will continue: Meißner (diphthong preceding ß), MEIẞNER
— will continue: Großer Feldberg (long vowel o preceding ß), GROẞER FELDBERG

Example for an officially approved name of a municipality
— will continue: Haßfurt, HAẞFURT (though a short and stressed vowel a is preceding ß)

This is, to my knowledge, the first official German government body that not only approves but actually prescribes the use of uppercase ß (U+1E9E, ẞ), and only accepts the use of SS “temporarily” (“since a typographical implementation of the upper case letter ẞ in the various character fonts will take some time”).

Regards,
Adam

the capital letter ẞ which is rendered on position 1E9E of the Unicode character tables. In official spelling it must not be substituted by any other letter combination, as e.g. SS, ss.

This is a little bit astonishingly put. That an ẞ “must not” be subsitutet (»darf nicht«) would mean that WEISSENFELS is not acceptable because WEIẞENFELS is right. In practical terms, both versions are right.

what i really want to know is when are uppercase Latin ligatures coming to Unicode? i need my FFI!

The day after the Rapture, at the earliest. IOW, it's the PUA purgatory for them.

Andreas,

well, I think this body (permament council for geographical names, Ständiger Ausschuss für geographische Namen (StAGN), at the Federal Agency for Cartography and Geodesy) is concerned with uniquely identifable names, so there are no ambiguities. For example, the name of the town I used to live in is officially "Frankfurt (Oder)". In the past, there were various spellings ("Frankfurt (Oder)", "Frankfurt/Oder", "Frankfurt-an-der-Oder", "Frankfurt a.d. Oder" etc.), but it seems that the Germans like to have everything normalized, so, well, they also decided to normalize the geographical names.

Which I think makes sense. After all, these days geographical names need to be stored in databases, processed digitally etc., so it's useful to have *one official* spelling of a geographical name.

A.

what i really want to know is when are uppercase Latin ligatures coming to Unicode? i need my FFI!

Not sure it that was supposed to be an argument or a joke, but just for the record:
The ß does not belong to ligatures like fi and fl, it belongs to ligatures like w, which (might have evolved from a ligature but) are a an official letter of the alphabet with a distinct phonetic function. And this is no matter of opinion.

At least we have the double lambda :-)

Paul, you could present the case for FFI to Unicode, using René Chalet's work in the 1960s as an important precedent.

As Ralf said though, there's a difference between
basic members of an alphabet versus ligatures.

hhp

where can i find the ‘official’ version of the German alphabet?

Here is the official and mandatory, most recent version of the German orthography from 2006, defined by the Council for German Orthography:
Rules: http://rechtschreibrat.ids-mannheim.de/download/regeln2006.pdf
Words: http://rechtschreibrat.ids-mannheim.de/download/woerterverzeichnis2006.pdf

Page 15 lists the official German letters:
a b c d e f g h i j k l m n o p q r s t u v w x y z ä ö ü ß
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z Ä Ö Ü

I also must stress that no one should use the term ligature as something intrinsically different than normal letters of the alphabet. This misinterpretation must be based on the fact that we usually just think of the typical f-ligatures.
But “ligature” just refers to a formal connection of some sort:
From Middle English, from Middle French, from Late Latin ligātura < Latin ligātus, past participle of ligāre (“to tie, bind”). http://en.wiktionary.org/wiki/ligature

A ligature can serve a typographic function (fi, fl, ...), a decorative function (ct, ...) or it can be an official letter of the alphabet like for example w, œ, æ, … and ß.

Pragmatically speaking, I think it's confusing to call something like the ß a ligature.

hhp

Three observations on German special glyphs:

1) fi, fl, ff, ffl, ffi, and st were grandfathered into Unicode as Alphabetic Presentation Forms, but not fz and tz.

2) German ligatures that aren't: ch and ck.

3) Ä, Ö and Ü — with lowered accents in many mid-20th century Antiqua fonts.

For more information on the phonetic importance of the letter ß also check out Adam’s post here:
http://typophile.com/node/82806#comment-467600

3) Ä, Ö and Ü — with lowered accents in many mid-20th century Antiqua fonts.

By the way: The German umlauts had the same development from a lowercase only connection to a full letter of the alphabet. (ß is just a little bit behind, because in blackletter there was no need for it, because blackletter wasn't set in uppercase.)

Here is a book from 1774. The e had moved on top of a, o, and u:

But there was no uppercase version yet: (Ueberlegenheit instead of Überlegenheit)

This stayed this way for some time, but the uppercase umlauts still emerged, because like Adam said: it’s the only logical thing to do. Now we are at the same point regarding the ß. It just took a little longer.

Ralf,

not only good examples but also some beautiful shots of them, thanks! :) Do you have them somewhere in high-res? I'd love to use them in a presentation if you don't mind (crediting you of course).