unicase ???

dberlow's picture

Does anyone know, is the OT feature unicase, referring to u.c. forms made to join the l.c., l.c. forms made to join the U.C. both. i.e. is it ambiguous? or what?

Nick Shinn's picture

Hey, Unicode is awesome.
Complex typography is, er, complex -- as always.
But in general, it's all good.

dberlow's picture

All's I know, is that I have over 100 styles either in flight, on the flight deck or ready to move into the elevator, that have been ordered by customers. What customers want in these styles are a, e, m, n, and u, and sometimes t and r, drawn to the size of the H, accented to the t, and kerned to all high heaven. (No one, has ever asked for e the size of small cap E, or H the size of x...ever, but maybe it's because my customers are...justifiably valued). I've done some tests though, drawing and sending various forms, and what they are looking for is a special case to use for headlines, not l.c. or small caps, ever. You are the boss:
a. go to the end of the line and start registering ucbf, uppercase biforms.
use something else in the meantime (?), and then wait for updates to
ID that may or may not ever hardwire ucbf to show users "uppercase
biforms", scramble and explain until complete, or denied.
b. send clients feature named something generic, like special style feature #1556, explain endlessly, because changing would be messy.
c. appropriate the "Unicase" feature to a useful task. Non-ambiguously state to customers and other vendors that neither you nor your users will see anything but lowercase forms drawn to the size of the H when this feature is selected.

But the general feeling is, if this was implemented better in one of several ways it's be a lot less stupid a set of choices. If e.g. Features had tags and menu names, and apps just passed 'em through, and the market made the features market, not some clumsy bucket line, absolutely everyone would be happier.

Nick Shinn's picture

If e.g. Features had tags and menu names, and apps just passed ‘em through, and the market made the features market, not some clumsy bucket line, absolutely everyone would be happier.

Except the application designers, who can't resist adding faux features, which is probably why they are reluctant to just let the fonts be, and let users only get what's in the font.

That might have been a good idea for PC Word users when fonts only had 250 characters, and there were hardly any digital fonts with expert sets, but the situation is different now.

Would there be an outcry from InDesign users if the next version didn't enable them to create, for instance, superior figures for fonts that don't have 'em? They wouldn't even notice, with AdobeSans/AdobeSerif substitution.

Quark 7 still has a faux menu, and buries its OT menu. Aaaagh!

hrant's picture

> They wouldn’t even notice, with AdobeSans/AdobeSerif substitution.

A very good point - and one that applies to every area,
including smallcaps, maybe even so-called "true" italics.

hhp

Nick Shinn's picture

Right. If you're not bothered using a font with special features, and are happy with faux features, you're unlikely to be concerned whether your faux superior figure is from Times or AdobeSerif.

Paradoxically, you may be upset to get a true italic when you're expecting a sloped roman!

NB. The AdobeSans and AdobeSerif MM fonts I'm referencing are OS 9, stored in the Acrobat Reader 5 folder, and only have a basic character set -- no small caps.

hrant's picture

> If you’re not bothered using a font with special
> features, and are happy with faux features

The thing is, and my point is, most people don't even get that far! :-/
They don't even realize there's anything to be bothered/happy about.

> you may be upset to get a true italic when you’re expecting a sloped roman!

I actually know of one such case, and at the foundry level: when Joe Treacy subbed in a "true" italic for his Arrow family apparently his customers almost rioted. So now it has both.

hhp

John Hudson's picture

David,

c. appropriate the “Unicase” feature to a useful task. Non-ambiguously state to customers and other vendors that neither you nor your users will see anything but lowercase forms drawn to the size of the H when this feature is selected.

I don't think this is appropriating the Unicase feature at all: this is a perfectly valid use of the Unicase feature, as I stated at the beginning of this thread. It just so happens that the Unicase implementation illustrated in the feature tag description uses an Emigre typeface that matches uppercase forms to the lowercase x-height. But there's no reason at all not to do it the other way around. Go for it. If you want to make it a house style rule for Font Bureau that the Unicase feature will always imply this kind of implementation in your fonts, that's your prerogative.

Nick Shinn's picture

There is also the option to ignore OpenType features and have a separate common case font.
Some of the advantages:
* With InDesign, the font menu is easier to get at than the OpenType menu.
* A straightforward size relationship of "capitals" and "small" (cased unicase) -- having the two sizes is rather like optical scaling.

hrant's picture

> like optical scaling

?

hhp

Nick Shinn's picture

...Cased Unicase is like Caps with Small Caps, in that the two cases have identically shaped glyphs. So the larger size is lighter, more of a display treatment, while the smaller size has stronger features, more of a small text quality.

hrant's picture

You were saying... hack? :-)

hhp

Nick Shinn's picture

Caps with small caps is not a hack for optical scaling, as it is a genre of type which historically occurred before single-master (ie non-optically scaled) fonts.

dberlow's picture

"identical form in both cases, CcOoSsUuVvWwXxZz in Latin"

not u.

hrant's picture

> Caps with small caps is not a hack for optical scaling ...

Now my question mark kicks in again.
It's not a hack for what you seem to mean, but it very much is a hack when one uses any half-decent definition of "optical scaling" (a term you might consider appending to your New Year's resolution :-).

> identical form in both cases, CcOoSsUuVvWwXxZz in Latin

"Identical form" is way too strong. The most that can be said is
that they share a structure... but then you have to graple with the
paradox that typographic (unlike for example chirographic) forms
don't actually have structure (at least not the way we tend to see it).
It's an illusion.

hhp

Nick Shinn's picture

not u

Sure it's the same, because the right side vertical stroke is redundant, a grace note.
Futura, for instance.
Linguistically redundant, but typographically helpful nonetheless.

“Identical form” is way too strong.

Right. I'm referring to the underlying skeletal form.

it very much is a hack when one uses any half-decent definition of “optical scaling”

That's like saying a gramophone is a hack of a CD player.

hrant's picture

> the right side vertical stroke is redundant

That of course makes no sense, not least because it leads
straight to formal hell. Pray tell: how does one determine
what's redundant; what parts of what letters are redundant,
exactly; is the left stem of the "R" redundant - why [not]?

Nick, trust me, this direction leads to the greatest quagmire
in type design. Not somewhere an art director can be happy.

> I’m referring to the underlying skeletal form.

Except there isn't one. And if there is, you're conveniently
ignoring parts of it (like in the U/u) at whim anyway, so...

> That’s like saying a gramophone is a hack of a CD player.

?
Just add it to your Resolution and everything will be fine.

hhp

Nick Shinn's picture

formal hell

Trust me, it's the same character if the spelling's the same.
The Casing/Bicameral system works (ie delivers orthographic significance) if readers can recognize two separate classes of glyphs underlying the setting/writing. Size is the main indicator. What the actual glyphs are shaped like doesn't matter (they can even be the same shape), as long as the letters are recognizable, and categorizable by the principle of case. In this example, upper case is minuscule, and lower case is majuscule.

hrant's picture

More from previously:
> I’m referring to the underlying skeletal form.

The more one thinks there are skeletons under there,
the more one stays distant from true text face design.

The good way to think about typographic glyphs (the kind that designers never make by hand but users always read by eye) is in terms of notan, in terms of borders, not in terms of thicknesses being built up from skeletons.

> it’s the same character if the spelling’s the same.

But the spelling isn't the same. Your spelling there is Incorrect,
something that's possible to state in orthography but not typography.
The best evidence would come if you asked people -especially those
who don't know [of] me*- what that says. They might say... dunno...
"horizontal rant"?** But if you spelled it correcly they're quite
likely to know it's a proper name.

* Of which there still are a few I hope. ;-)

** And everybody knows that my rants are always vertical.

Also: Even if most people can read something spelled incorrectly
the way it was "supposed" to be spelled, that doesn't say anything
at all about orthography.

> Size is the main indicator.

That's crucially different than what you said before.

> What the actual glyphs are shaped like doesn’t matter

Maybe to some art directors.
To most other people they sure do matter.

--

But anyway you're digressing - maybe because you don't want to admit
that your stance on the U/u is wrong. Come on, it's not that hard.

Anyway.
What I was getting at above is that you can't make blanket statements about which parts of which characters (again, not the same as letters!) are "redundant" - certainly not without thinking quite deeply about it. Is the top-left "cornerness" of the UC "F" not "redundant"? If you make it round, it starts getting close to the lc form, so becoming part of your "UC/lc identical form" invention. And have you seen the lc "g" in the credits for Spongebob Squarepants? Is it really lc? In fact almost all the letters that are not in your List can be drawn quite easily in a way that they can be either UC or lc, so invalidating your U/u distinction. Futura has a lc "u" without a full right stem? Who cares. That says nothing about anything useful, like what the conventions are.

And when it comes to the U/u, instead of claiming the untenable "the lc form doesn't require a full right stem to be conventional"* you should instead simply point out that some entirely conventional fonts (like Perpetua) have an UC "U" with a full right stem. This converse is parallel to the truism that You Can't Prove Something Doesn't Exist.

* You might say that's not what you stated. But it's the only sensical thing that can come out of your line of thought, if you really think it through. Conventionality is really the key concept here - simply because if you loosen things up virtually anything is legible!

The fact is glyphs never fall neatly into any box, including UC/lc.
Your categorization holds about as much water as a lc "n". Without a full left stem.

hhp

Nick Shinn's picture

you don’t want to admit that your stance on the U/u is wrong.

Instead of seizing on any little thing you think you can "score points" on, try to look at the big picture and understand my full argument.

Characters don't change from Upper to Lower Case.
They may or may not look the same.
A is the same character as a, despite Unicode's expedient definition.
Casing is necessary to support the orthographically/grammatically meaningful "Titlecase", but it's immaterial whether the glyph shapes are very dissimilar (A and a), debatably dissimilar (U and u), or the same (S and s, etc.). As you put it:

The fact is glyphs never fall neatly into any box, including UC/lc.

hrant's picture

> try to look at the big picture

I always do, but since I know we disagree on it,
and since I harbor no hope of changing your mind,
the only thing left to discuss is the details, in the
hopes of helping third parties. And because talking
helps me think. And it's fun.

> Characters don’t change from Upper to Lower Case.

Incorrect.

> A is the same character as a

No, it's only the same letter.

> The fact is glyphs never fall neatly into any box, including UC/lc.

Glyphs, not letters.
And the point there was to explain how your stance on the U/u is wrong.

Your line of thought is only relevant in the context of convention.
In the context of convention, the lc "u" has a full right stem.

hhp

Nick Shinn's picture

Hrant, your nit-picking over U/u is avoiding dealing with the crux of my theory:

1. Titlecase is the only grammatically significant use of casing.

2. In terms of grammar, Titlecase is a form of Inflection. Whereas most inflection signifies a grammatical feature by altering the spelling of a word, Titlecase does so by upper-casing the initial character, and lower-casing the rest.

3. This grammatical signification does not happen at the level of character, but at the level of display (ie typography), primarily through size contrast.

4. While there are conventional groupings of majuscule, minuscule, and unicule glyphs into case pairings (primarily Upper & Lower, and Caps with Small Caps), there is no direct correlation between case and "-cule".

5. Unicase is a misnomer, as is common-case. This is demonstrated by the feasability of "Cased Unicase", and the fact that "Unicase" can be either upper-case or lower-case size. Perhaps "multicule" would be more accurate.

6. For basic typography, the present system of "characterizing" the difference betweeen upper and lower case has expediency, but it creates problems further along, e.g. in assigning the appropriate figures for a particular case.

hrant's picture

Avoiding it? The parts I thought were worth addressing,
I already have. But every new post of yours seems to
contain a new objectionable offshoot. Or three. :-)

The real avoidance Nick is you not admitting you're wrong about
including the U/u in that list. Nitpicking? On the contrary, that
fault has opened up a line of discussion much more interesting to
me than the original topic or the direction you want to take.

hhp

Nick Shinn's picture

The way the casing is handled in Futura, Bauhaus, Perpetua, Giovanni, etc. indicates that it's acceptable to handle the U/u character as a "unicule" -- i.e. the same structure in both upper and lower case.
As you say, it's possible to design upper and lower with more commonality than is usual, and I've done that in Morphica.
Like I said, as long as the reader can discern from the text what design system the typeface uses to differentiate upper and lower case, and as long as the glyphs are identifiable as characters, anything goes.

hrant's picture

> it’s acceptable to handle the U/u character as a “unicule”

But the point is that if/when that's acceptable, then a bunch of other letters (maybe even all of them) can be "unicule" just fine too. So your list is either pointless because raw legibility is very lenient, or it should not have the U/u because it codifies convention. Why is this so hard to understand/admit?

> as long as the reader can discern from the text what design system
> the typeface uses to differentiate upper and lower case, and as long
> as the glyphs are identifiable as characters, anything goes.

In display design. Text faces are a whole other ballgame.
And as long as you think Eunoia-Text merits its suffix you
won't be in that game.

hhp

Nick Shinn's picture

Text faces are a whole other ballgame.

For you, more of a hobby horse.
How's the Baskerville coming?

hrant's picture

There's the panic alarm going off in Shinnistan again...

Nick, you really think that admitting a mistake (it's not even a major
one really) looks worse than acting like you don't make any? Oh well.

hhp

Nick Shinn's picture

The reason I included U/u in my list of "unicles" is because of faces like Cocon, Futura, Giovanni, Perpetua, etc., etc., etc. The practice of using the same form of U/u for upper and lower case is widespread -- and done with both versions, with or without the right-side stroke extension. The casing of U/u is not like other bi-form characters such as N/n, where the typical lower case form may also be used in upper case, because there is a fundamental shape difference between N and n, and its majuscule form is not used in lower case. Simply put, the u's stroke extension is not significant in distinguishing majuscule from minuscule, and as such is redundant to the basic letter form.

dberlow's picture

"Simply put, the u’s stroke extension is not significant in distinguishing majuscule from minuscule, and as such is redundant to the basic letter form."

Oiy. Thou fiddleth around parts of characters while OT burns?

hrant's picture

> the u’s stroke extension is not significant
> in distinguishing majuscule from minuscule

You've got to be kidding. It's their main difference! And sometimes it can matter: picture "US" set in smallcaps in a sentence (in a font with the -traditional- x-height height smallcaps).

And that's just the biggest thing wrong in your new post - I won't get into the fundamental difference between "widespread" and "conventional" for one thing...

hhp

Nick Shinn's picture

Thou fiddleth

I ain't no emperor.
I do have the time to engage in banter on Typophile. But I wasn't aware that there was a role for independent foundries in developing OpenType, which is the preserve of MS and Adobe -- or layout interfaces, which is the preserve of MS, Adobe and Quark -- or Unicode, which is the preserve of scholars and programmers, not typographers and type designers like me who are at a far remove from the business establishment.

I have the impression that if the concerns of Font Bureau are a cry in the wilderness, I'm just a whisperer, not a fiddler.

It’s their main difference!

Exactly. But it is so slight and variable (in all three widespread conventions, i.e. both cases with extra stroke, both cases without extra stroke, and stroke only on lower case), that it does not create clearly defined, oppositional majuscule and minuscule forms. Compare the blatant contrast between majuscule/minuscule forms here, and see why I group U/u with the unicules:


Even if the U/u contrast is exaggerated, as in Scala Sans, it's very slight, less apparent than the alternate W/w forms, in fact.

hrant's picture

"Engage in banter"? A likely story...

> it does not create clearly defined,
> oppositional majuscule and minuscule forms.

Type design is all about subtlety (not least because perception, and certainly reading, is all about subtlety). When we worry about the smallest details that might affect the reader, you can't say that a full stem is irrelevant with a straight face. Not to mention that certainly the K/k, probably the Y/y, and possibly the P/p, are different to about the same degree; it's increasingly clear that the line you're drawing makes no sense (but that's what happens when you're more interested in winning an argument than seriously discussing type with a desire to grow, an ability to correct yourself).

hhp

Nick Shinn's picture

Hrant, I may be intransigent, but I would rather banter than trade personal insults.
Could you please remove the "psychological insight" from your last post?

hrant's picture

But of course it's entirely OK for you to call
me a "clueless idiot", mister "don't ask Gerry
about Greek type". You see, Nick, the only way
for people not to be able to hold stupid things you
say against you is if you formally retract them.
Be grateful you are being given the chance.
Otherwise, when you complain, you come off
as an unfair whiner.

Not that I will ever substitute cocktail-party civility for candor.

hhp

Nick Shinn's picture

unfair
When you say that I am not a text type designer (which you frequently insinuate into threads) it is candour. When I say you are a clueless idiot (which I did once, upon provocation by your "candour") it is stupid.

hrant's picture

Even when I use your own benchmarks of what constitutes a personal attack and what doesn't, you try to weasel out of it... Guess who else isn't a "real" text face designer: me! But I think I'm at least heading in the right direction, while you have yet to do that. Only a despot would hold that up as an unacceptable personal attack. Having an ego is one thing (I would know), demanding adulation is another. And how did Gerry "provoke" you? By calling your Greek something like "in the vein of misguided Greek designs of the 70s". Keep your ear on, Van Shinn.

BTW, if you think you were wrong to call me a "clueless idiot",
there's always the unthinkable recourse... Not that I care, since
it would basically be for your own benefit.

hhp

Nick Shinn's picture

And how did Gerry “provoke” you? By calling your Greek type something like “in the vein of the misguided Greek designs of the 70s”.

Worse, he compared it with Times Greek, the ultimate diss :-)
But I must protest that my "Don't ask Gerry about Greek" comment was made later, and not in anger, but jest -- a type designer who, from his techno/grunge oeuvre, didn't strike me as the calligraphic kind, wanted some pointers on Greek type design; it was in this context that I repeated Mr Leonidas' advice to me to get liquored and start writing with a soft pencil.

Only a despot would hold that up as an unacceptable personal attack.

That wasn't the consensus at Typophile last year, when OldNick, I believe, explained the principle of the ad hominem argument. It was generally agreed that the ad hominem argument should be avoided. You dissented. So I don't know why I'm arguing this again. Have the last word, if you wish.

dberlow's picture

"I have the impression that if the concerns of Font Bureau are a cry in the wilderness,"

Maybe so, but the wilderness is closin' in. Last week, the China.gov started the process of limiting the names of babies to glyphs in the first 15,000 of the std.. No more Chinese equivalents of DeShondra, Giselbert or Christian.

Syndicate content Syndicate content