Uppercase germandbls is coming to Unicode

Primary tabs

297 posts / 0 new
Last post
Jos Buivenga's picture
Offline
Joined: 19 Nov 2005 - 5:11pm
0

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.

Dan Reynolds's picture
Offline
Joined: 20 Jul 2002 - 11:00am
0

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.

Typography.Guru's picture
Offline
Joined: 8 May 2003 - 2:39pm
0

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

Dan Reynolds's picture
Offline
Joined: 20 Jul 2002 - 11:00am
0

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

I am very sad.

Florian Hardwig's picture
Offline
Joined: 18 Feb 2007 - 6:41am
0

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]

Typography.Guru's picture
Offline
Joined: 8 May 2003 - 2:39pm
0

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.

Jos Buivenga's picture
Offline
Joined: 19 Nov 2005 - 5:11pm
0

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.

Nick Shinn's picture
Offline
Joined: 8 Jul 2003 - 11:00am
0

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?

paul d hunt's picture
Offline
Joined: 5 May 2005 - 8:44pm
0

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.

Karsten Luecke's picture
Offline
Joined: 6 Aug 2005 - 8:41am
0

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.

Typography.Guru's picture
Offline
Joined: 8 May 2003 - 2:39pm
0

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

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

Karsten Luecke's picture
Offline
Joined: 6 Aug 2005 - 8:41am
0

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 ...]

Chris Lozos's picture
Offline
Joined: 25 Feb 2004 - 11:00am
0

"...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

Typography.Guru's picture
Offline
Joined: 8 May 2003 - 2:39pm
0

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.

paul d hunt's picture
Offline
Joined: 5 May 2005 - 8:44pm
0

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.

Typography.Guru's picture
Offline
Joined: 8 May 2003 - 2:39pm
0

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!

Nick Shinn's picture
Offline
Joined: 8 Jul 2003 - 11:00am
0

“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?

paul d hunt's picture
Offline
Joined: 5 May 2005 - 8:44pm
0

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

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

Nick Shinn's picture
Offline
Joined: 8 Jul 2003 - 11:00am
0

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.

Karsten Luecke's picture
Offline
Joined: 6 Aug 2005 - 8:41am
0

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.

A. Szabolcs Sz.'s picture
Offline
Joined: 3 Jan 2008 - 1:30pm
0

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.

Dan Rhatigan's picture
Offline
Joined: 1 Aug 2004 - 11:00am
0

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?

A. Szabolcs Sz.'s picture
Offline
Joined: 3 Jan 2008 - 1:30pm
0

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

--Szabolcs

Dan Rhatigan's picture
Offline
Joined: 1 Aug 2004 - 11:00am
0

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.

Adam Twardoch's picture
Offline
Joined: 3 Dec 2002 - 7:36pm
0

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

Cristiano's picture
Offline
Joined: 11 Nov 2008 - 4:26pm
0

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.

Simon Daniels's picture
Offline
Joined: 11 Apr 2002 - 6:37pm
0

>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!

Dan Reynolds's picture
Offline
Joined: 20 Jul 2002 - 11:00am
0

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

Jason Pagura's picture
Offline
Joined: 10 Sep 2006 - 6:19pm
0

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?

Tim Ahrens's picture
Offline
Joined: 28 Sep 2004 - 9:15am
0

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.

Hrant H Papazian's picture
Joined: 3 May 2000 - 11:00am
0

> Plenty of other uppercase letters look like enlarged lowercase counterparts

A sad reality that we should not amplify.

hhp

Tim Ahrens's picture
Offline
Joined: 28 Sep 2004 - 9:15am
0

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.

Craig Eliason's picture
Offline
Joined: 19 Mar 2004 - 1:44pm
0

> 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.

Adam Twardoch's picture
Offline
Joined: 3 Dec 2002 - 7:36pm
0

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.

David Berlow's picture
Offline
Joined: 19 Jul 2004 - 6:31pm
0

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!

Adam Twardoch's picture
Offline
Joined: 3 Dec 2002 - 7:36pm
0

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.

Nick Shinn's picture
Offline
Joined: 8 Jul 2003 - 11:00am
0

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?

Hrant H Papazian's picture
Joined: 3 May 2000 - 11:00am
0

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

David Berlow's picture
Offline
Joined: 19 Jul 2004 - 6:31pm
0

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!

Tim Ahrens's picture
Offline
Joined: 28 Sep 2004 - 9:15am
0

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.

Tim Ahrens's picture
Offline
Joined: 28 Sep 2004 - 9:15am
0

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?

Nick Shinn's picture
Offline
Joined: 8 Jul 2003 - 11:00am
0

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.

Mark Simonson's picture
Offline
Joined: 3 Dec 2001 - 11:00am
0

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.

Mark Simonson's picture
Offline
Joined: 3 Dec 2001 - 11:00am
0

Here's how I might do it with Trajan:

Tim Ahrens's picture
Offline
Joined: 28 Sep 2004 - 9:15am
0

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.

Mark Simonson's picture
Offline
Joined: 3 Dec 2001 - 11:00am
0

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

Typography.Guru's picture
Offline
Joined: 8 May 2003 - 2:39pm
0

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.

Claudio Piccinini's picture
Joined: 11 Jan 2003 - 9:32am
0

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…

Mark Simonson's picture
Offline
Joined: 3 Dec 2001 - 11:00am
0

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.

Chris Lozos's picture
Offline
Joined: 25 Feb 2004 - 11:00am
0

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