Florin sign (ƒ)

Mark Simonson

I think the descender is what made the florin attractive to hijackers, not so much the slant.

Hrant Հրանդ Փափազեան Papazian

Each character is ideally its own thing.

I used the buy the argument that the florin should be slanted even in a upright font, but now I favor Andreas's leftmost form.

Nick Shinn

What are superior and inferior currency symbols used for?

This little superior dollar helped justify the line.

John Savard

Denis Moyogo Jacquerye said:

In Togo, the same character is used as a letter Ƒ/ƒ in the orthography of the Ewe language. If you plan to support that language the glyph must be a proper letter (proper ascender and descender, upright in upright, distinguishable from f in italic, etc.). What’s interesting is when one tries to support both cases, the letter and the currency symbol, what do you leave as the main glyph, how do you access the other(s)?

Are you sure that they don't have distinct Unicode codepoints?

Ah, the problem is that the Dutch florin symbol is absent from the Currency Symbols section, so that people are using the letter for it.

I have now found that there is already a proposal before the Unicode Consortium to remedy this, and it even mentions the Éwé language.

Thomas Phinney

The problem is that Unicode kinda goofed on the florin/hooked f character (IMO). I think a disambiguation proposal would be reasonable.

The hooked f needs an upright form and an italic form.
The florin should always be italic, even in an upright font.

There isn't a clean way of solving this, IMO. Basically if you want to support Ewe, you need to keep your ƒ upright in upright fonts. If you want the font to look right for more general use, you need to keep ƒ italic in upright fonts.

Andreas Stötzner

Thomas Phinney:
> The hooked f needs an upright form and an italic form.
Right.

> The florin should always be italic, even in an upright font.
Not neccessarily. Although it comes very often in italic fashion by tradition, it works also as an upright glyph. The downswing part is more important than the italic slant.
The florin also occured in blackletter typefaces where it was upright – and worked.
(sorry I have no historic samples at hand right now)
It is similar with the £ sign: needs to get a scripty shaping but *can* be upright the same time. But what really startles me if the ƒ glyph does not reach ascender height like in this font seen here.

Paul van der Laan

The /florin – as with other currency symbols – needs to work with numbers and therefore I usually design it differently than italic /f. Especially when it needs to work with lining numbers I make the descending part shorter. Also I might design the proportions wider, and place the horizontal bar on a different height than x-height.

Here is an example of /florin versus /f from AudiType Bold Italic:

James Puckett

Perhaps the florin is used for f-stop because it harmonizes with the numbers better than a letter would. I honestly doubt many designers would learn the location of florin and begin using it accidentally.

Craig Eliason

Mark Simonson said:

I think the descender is what made the florin attractive to hijackers, not so much the slant.

Though when separated from the number with a diagonal slash (as is the convention I think), the florin's slope fits well.

Michel Boyer

Here is what is written on my old Nikon 105mm lens.

The letter f stands for the focal length, 105mm, and is neither slanted nor hooked;

 1:2.5 means that the diameter of the largest aperture is the focal length divided by 2.5, which can be written in TeX as the mathematical formula $f/2.5$ giving this in the mathematical Fourier font with mathematical spacing:

(The largest aperture has thus a diameter of 105mm/2.5 = 42 mm.) 

What I deduce is that, in modern parlance, the letter f is U+1D453 since it appears in a mathematical formula. On the other hand, there are mathematical fonts used with LaTeX for which $f$ is not hooked (and on top of that, the Euler $f$ not even slanted, which I personally find troubling). So I see no more reason for  the f number to be hooked than the fact that the italic f is often hooked (plus, of course, personal taste which, in my opinion, also applies for other uses of $f$ in mathematical formulas).

Nina Stössinger

James Puckett said:

Perhaps the florin is used for f-stop because it harmonizes with the numbers better than a letter would. I honestly doubt many designers would learn the location of florin and begin using it accidentally.

As Kent mentioned above, ƒ is accessed perhaps a little too easily on the Mac keyboard under option-f, so there’s not much learning involved if you’re on a Mac, as many designers are. I suspect that is related to the pre-OSX file-naming convention (I do remember folders named Somethingsomething ƒ), I wonder why else it would be there, but I’d guess it’s not too uncommon that people stumble on it and use it as basically an “f” that looks nice and different and, yes, ideally works with numbers, and now we have to consciously support this usage? I’m not sure.

tl;dr the ƒ is the Arial of Unicode characters, it’s not really quite right but people use it because it’s there? ¯\_(¨)_/¯

Hrant Հրանդ Փափազեան Papazian

Coz cursive is purdy.

Chris Lozos

I wonder if it works in passwords?

Thomas Phinney

Chris Lozos said:

I wonder if it works in passwords?

Depends on what character sets the password can handle. It's not in ASCII (nor WinANSI, a.k.a. codepage 1252). If it is a Unicode password field then you are good....

Craig Eliason

The florin (or at least a hooked and sloped f) was conventional for photo apertures long before Macs were around. (This is from a 1920 journal.)

Michel Boyer

What you call a hooked and sloped f is what is called an italic f in Caslon, Ronaldson (very popular at the end of the 19th century), Times, Times new roman and many fonts that were or are still commonly used.

In the sample above, it seems indeed that there was a change in font to get the desired shape but I don't find a one item sample convincing.

Craig Eliason

You're right, it's an italic f (I was able to track down an example of an italic f in the same journal and it looks like the same sort). But then I'd be curious to know when, if ever, the florin existed as an independent type sort distinct from f in the pre-digital world. 

Kent Lew

when, if ever, the florin existed as an independent type sort distinct from f in the pre-digital world.

I wouldn’t be surprised if it never did. It is possible that it was only distinguished when Apple encoded it as 0xC4 in the MacOSRoman charset.

Michel Boyer

And what is said about f with hook in https://en.wikipedia.org/wiki/Ƒ as being used to represent mathematical functions (note the “citation needed”) is sheere nonsense. There is no such a thing as a function symbol. The small letters f, g etc are indeed commonly used to denote functions but writing

is like writing

def f(x):<br>&nbsp; &nbsp;return 1.5*x + 2<br>

in Python and the letter f in the math formula is again nothing but U+1D453, it is not U+0192 or any other special character.

On the left is an example of the letter f used as a function in Jean Dieudonné, La géométrie des groupes classiques, Berlin, Springer, 1955 (in the series “Grundlehren der mathematischen Wissenschaften" that was particularly well edited). 

On the right hand side is how the same formula comes out with the mathptmx LaTeX package, based on Nimbus, a Times like font that I think was used as base for the Stix fonts, indirectly commissioned by the American Mathematical Society, the American Physical Society, IEEE and some publishers and which would correspond to current American standards. I am not so sure the right-hand side is much better than the left hand side. I like the way the two levels of parentheses come out in the 1955 Springer Verlag version.


By the way, is there any math font where U+1D453 is not just the italic f with a different spacing? Of course, the letter u needs some special care to be different from nu, but I have never seen any recommendation about U+1D453.

Nick Shinn

A short compendium of doppelgängers

Pieces of foundry type that did double duty, rotated 180°:

• Comma and Left Quote mark

• Left and Right Parentheses and Square Brackets

• Question/Exclam and Questiondown/Exclamdown

• In Baskerville, £ and Italic J 

Without rotation:

• Comma and Quotesinglebase

• In some typefaces, 1 and l

• Hyphen and En dash*

Many mathematical symbols, often taken from Greek, and:

• x and Multiply* 

• En dash and Minus

• Colon and Ratio

• Primes: Inch/Feet marks and Minutes/Seconds (of arc, time)

And now, notoriously, single right quote and apostrophe.

* at the typographer’s discretion

attar

Michel Boyer said:

By the way, is there any math font where U+1D453 is not just the italic f with a different spacing? Of course, the letter u needs some special care to be different from nu, but I have never seen any recommendation about U+1D453.

Lewis by Alexandre Saumier Demers (unreleased) has a special math italic (see bottom-left of that picture).

Michel Boyer

Adrien Tétar said: Lewis by Alexandre Saumier Demers (unreleased) has a special math italic (see bottom-left of that picture).  

For the record

Thanks.
Michel

PS I also found the link http://typemedia2014.com/lewis which is worth having a look at.

Nick Shinn

Usage in some old photography manuals: just plain roman “f”.

The Rolleiflex Book, 1932: f4
However, the usual mention in this manual was just “stop 4”, only rarely using f.

Pocket Leica Book, 1952: f/4

The Camera, Time-Life, 1970, f/4

From which I deduce that the practice of using a fancy “f” is relatively recent.

Michel Boyer

The table https://commons.wikimedia.org/wiki/File:Diaphragm_Numbers.gif is from 1899 and is all upright except for some  f / both in the header and in a footnote. On the other hand, my guide for the Canon SD400 uses a plain upright sans serif f. However I doubt that a book on geometrical optics would ever use an upright font.

Nick Shinn

Right.
According to wiki usage was varied up to 1961 when the ASA (American Standards Association) fixed it as hooked-italic f. That would firmly associate it with the mathematical variable “f”, represented by a serifed character (as Michel noted), which traditionally all had descenders with “hooks”, except for a few Linotype fonts where the hook was chopped off.

Michel Boyer

The recommendation the wiki refers to is superseded. The one that replaced it too. I could find no current recommendation. I looked at some current documentation for Leica and Hasselblad and they either use a straight f or no f at all. All the Canon guides I looked at use a straight sans serif f, same straight f in a March 2016 Nikon lens brochure. I saw one straight sans in the text of a Zeiss brochure when saying "at an aperture of f/1.4" but for the lenses, no f at all: "ZEISS Otus 1.4/55".

My conclusion is that the "f stop symbol" does not exist. There is a letter f that represents the focal length. When used in a physics book (geometrical optics is physics) it occurs in lots of formulas and then follows the rules of mathematical typesetting. That is all.

PS By the way, the Zeiss 1.4/55 is written mistakenly. It should be 55/1.4   since f=55 and the largest aperture is f/1.4 = 55/1.4

Mark Simonson

This whole thing reminds me of the way Apple II fanatics would always type it "Apple ][" because that's how the logo looked on the case. I also remember seeing Apple III rendered as "Apple ///". 

Nick Shinn

Nice work, Michel!
It’s a myth.
But there is one f stop symbol…a logo: 
http://www.fstopimages.com

Denis Moyogo Jacquerye

Denis Moyogo Jacquerye said:

but must be in italic to be used as the florin symbol: ƒ5.

Revisting this assumption, it is false in fact. The upright f with hook was sometimes used, even if to a lesser degree, for the florin symbol.

See for example the upright f with hook on the cover of the Algemeen Dagblad, 10 September 1980:

Or in an advert in De Stem, 6 August 1986

Andreas Stötzner

good example. Note that the £ in the above specimen is also upright, although ‘scriptish’ (but not cursive).