I recently used arrows in a image-heavy document to point to illustrations (rather than ‘above left’, ‘above’ etc.) I needed them to work with Meta Plus Book and wanted something a little bolder than the body copy at about em height and to the 8 major points of the compass – in the end I used Wingdings, each arrow with slight positive and negative baseline shift, but would have preferred something a little lighter, they were mainly adjacent to non-aligning numerals but were in the main flow of the text so had to integrate with upper and lower case characters as well. I seem to remember Foundry Gridnik had some useful arrows.

Tim

I've done the same thing as Tim, using arrows to indicate which cutline goes with which image. I've always built the arrows in Illustrator and imported them. Apart from Pi and dingbat fonts, I wouldn't consider looking for arrows in my font sets.

So in your use you were looking for something that aligned to text... yes? Interesting. Thanks!

I used the font Geometric Arrows for years. Like this:

saturday july 12 -> el rey theatre

I now use the above combination (tweaked a bit) - but yes, I have a use for arrows.

Hi Eben,

I am a fan of arrows. I am constantly making custom arrows (mostly for web designs these days). They are usually used for links or navigation clarity. It would be nice to have a set of arrows for different purposes. Often I set the size of the to the same x-height as its related typeface. But, its all about the style of the arrow. It needs to match the design aesthetic that is being created for the site as a whole. That is why multiple arrow styles would be great. I know I ask a lot . . .

Shu, Paul, more interesting still! What about multiple arrow forms within the same 'style'? That seems more likely in a single font. Unless you build an arrow only font...

Do you want to do that Shu?

Maybe I can post an examples of the 'standard' forms in Unicode later.

Arrows are used extensively in mathematical typesetting. Generally boring, thin arrows that balance with other math symbols. You need left, right, up and down, and angled would be nice too. An extension that will attach to the left and rights is good (the em dash will do if it fits) because sometimes you need to place a formula over (and under?) an arrow, and thus need to grow it.

http://www.myfonts.com/fonts/storm/sebastian-arrows/light-light/charmap....

Just noticed that, fittingly, Sebastian has an arrows supplement in a variety of heights and for each weight.

Tim

I'm with Paul. I've used them this same way.

I use InDesign's arrow-maker, which contains all the traditional arrow designs I need.

Cheers!

Given David's comments it sounds like the primary reason to include arrows ( to the extent that there is one ) would be to support specific math functions... All other uses are essentially clip art* related rather than typographic in nature. I can imagine that an arrow which matches the chosen font might be good to have as well but I suspect that custom arrows will be used more in part because the arrows in a font are harder to access ( I guess you could create a font in the Opentype set called 'symbols'....) but mostly because I think graphic designers will usually want something impossible to anticipate. In any event it probably isn't going to pay to sweat this very often even if you are just supporting Math use.

* This probably sounds more dismissive & pejorative than I intended...

James, Shu, Paul - What arrows would you want to match within a type face? I am guessing 12, 3, 6 & 9 o'clock and the points in between. Anything else?

Tim, nice find.

Don, would you expand on your comments?

For what it's worth, here are some arrows that I have designed recently. I try to harmonise the arrow with the typeface, and now (after Feijoa) align them to the cap dimensions. I don't italicise arrows. There are also 45° arrows as well.

I also do the opentype calt trick by subbing -> with arrowright or -^ with arrowdown for instance. Quite handy!

—K

Nice tips Kris! Thanks!

Have you done any math-arrows?

I also thought I would post this:

so that we could be specific.

BTW are these all the code points that are relevant or are their others lurking around?

Christ, look at all those arrows! I haven't done any Math arrows, I am not even too sure what constitutes a 'complete' Math glyphset. That would be interesting to know. What do people setting Math actually use & need?

—K

Just noticed that, fittingly, Sebastian has an arrows supplement in a variety of heights and for each weight.

Many of Storm's OpenType fonts contain these arrows. I haven't had a chance to use them yet, but it's nice to know they're in there. Also, Andrij Shevchenko's Zion Train includes arrows.

My sense is that some of these arrows are probably Math, some for Boxes ( this side up [ 21EB? ]), some for electrical /construction work [ 21DE? 21AF?], some for maps [21DF?] etc. But this is conjecture only. It would be great to Know? Maybe Adam Knows...

Kris, what code points do you use for arrows? Or is it all a glyph sub thing? If so, that would be cool too frankly!

My guess is also that Sebastian's arrows are on , ABCDEF etc. Are Storm's?

hmmm, the arrows in sebastian pro are accessed via the 'ornm' feature and are mapped to letters: i'm guessing the same as set up in the arrow supplement font. i'm adding a small set of arrows (8 in all) to a project i'm currently working on and using the Unicode points u+2190 to u+2191.

Isn't this a waste of time?
In many years of advertising and design work, the only arrows I ever used came on the end of line rules in layout software, not in fonts.
Perhaps there is a place for them in math fonts or dingbat fonts, but not in general purpose fonts.
However, I note the above postings of Paul and James mentioning what is a genuine usage -- but surely there's only one arrow that's needed for that, the basic left to right?

but surely there’s only one arrow that’s needed for that, the basic left to right?

Don has not told us. We need more info.

I have suspended my judgement about this being a waste of time - but I suspect that you are correct about the real utility of this. Still, I am interested to see if my assumption holds up - or not.

And in the meantime I have been very interested to hear what designers and type makers have had to say.

Maybe Shu & I can make a world beating Arrow font in the future...

In terms of math fonts, the code page Eben listed looks comprehensive. The AMS (American Mathematical Society) financed a math set (2 fonts) to supplement the basic math characters provided in TeX (3 fonts). The AMS font that holds the arrows is here. http://www.micropress-inc.com/fonts/brmath/brf10.htm#font

The basic math characters are in Adobe Lucida Math font (among others) http://www.adobe.com/type/browser/F/LUCM/F_LUCM-10005002.html

Thanks Don!

"Maybe Shu & I can make a world beating Arrow font in the future..."

I'm all aquiver.

Don—so those dashes & squiggles & lines & wacky brackets in this font and in this font are all that would be needed to set complex math?
—K

I don't know how you will see these, but look at what's included in the new Vista font Cambria; it's packed with math operators as well as a lot of arrows. This doesn't answer Eben's question of course.

This doesn’t answer Eben’s question of course.

No, but it's interesting. Thanks!

I’m all aquiver.

Yeah that's kinda what I meant too. I guess I didn't supply the needed emoticon... {Shrug} Still, maybe Shu can sell it to Veer. Although I agree, it probably does make more sense as a clip art project though - if at all.

all that would be needed

Ha! Is that all? ;-) Actually it wouldn't be that hard to make - but you would want to have a math guy/gal to take em for a spin I think... And I doubt that it is a super hot market. That said I don't LOVE what I am seeing.

It certainly looks like some of my guesses about use were quite wrong!

As kris illustrated, it would be nice to have complementary set of arrows with fonts for use in wayfinding signage, at least covering codepoints 2190–2199. When I was a signcutter I had to use the arrow generator included in FlexiSign, which was adequte enough, but always produced rather generic results. They served as directional pointers, but never really harmonized with the type.

at least covering codepoints 2190–2199

These are the eight I think Kent was advocating for.

There's ten of them actually, counting the horizontal and vertical double-headed arrows in the middle of the sequence.

Kris

Yep, all that is used in math. But that does not mean that it is all you would need to set math. Mathematicians are notorious for trying to come up with new symbols and such. But those are the basics that will cover the majority of math setting.

When a researcher comes up with a new formula, he/she often wants a special character to represent a certain important variable. First used are Greek characters ... you will be familiar with the large Sigma used for summation, and pi used for that number.

They also will use script characters (l is common to avoid confusion with 1, although a good designer should have chosen fonts where there is no confusion), Hebrew characters, Fracktur characters, and the outlined characters shown.

The symbols are variations of basic symbols. For instance, the less than or equal to sign can have one or two lines under it, or one or two curved lines, each meaning different things to different mathematicians. I'd tell you what they mean, but I can't. I just typeset the stuff (back then). I didn't understand it.

it would be nice to have complementary set of arrows with fonts for use in wayfinding signage,

It would be nice to have a lot of things in fonts. I have recently been persuaded to include several historic Cyrillic characters in a retail font, why, I don't know, as there are probably 2 or 3 people in the world who might use such a thing, and they are not font buyers (I don't expect those who persuaded me to include the characters to buy the fonts either) --however, it is entertaining for a type designer to investigate such novelties, and we are "completists" by nature.

The WGL4 which includes arrows and boxes does so, I believe, because it covers four Unicode groupings, one of which includes arrows and boxes, for no particular reason, other than they might as well be grouped somewhere.

What is the correlation between what is useful to font buyers and what is in fonts?
This has been cocked up digitally from day one, as you can tell just by looking at your keyboard.

These days, wayfinding systems don't use serif fonts, conventional wisdom being that these are less legible and therefore unsuitable, so Kris, why bother? Why not add something equally useless, such as a unicase or Bold Italic Small Caps?

If your typeface really is designed for a specific use, then some special characters which address that usage can act as a "gimmick" to draw attention to that and position the typeface in the marketplace/collective conscious. Otherwise, they are a distraction and a dilution, and may even be counter-productive by suggesting inappropriate usage.

The astrological symbols in Josh Darden's Corundum were spot on, a genuine novelty in a font (quaint ct and st ligs have been done), but appropriate to the genre. So they serve a purpose, even if those who buy the font don't use them.

These days, wayfinding systems don’t use serif fonts, conventional wisdom being that these are less legible and therefore unsuitable, so Kris, why bother?

Basic NSEW arrows whose dimensions and stroke widths match the rest of the font are surely nice to have, rather than having to use an approximation from a dingbat font. Especially in a multi-weight family.

Also, assuming the kind of arrow made from 3 straight lines meeting at a point (surely there's a term for this?) the terminals might be square, diagonal or round to better match. More esoteric display faces (I'm thinking dot matrix) would also benefit.

This variant of the arrow is part of the Arrival font. There were almost 30 different versions in this series and I used all the arrow directions. But as mentioned, wayfinding is a special use situation.

http://i6.photobucket.com/albums/y234/jupiterboy/3N.png

“Maybe Shu & I can make a world beating Arrow font in the future...”

David - I am in! We'll talk. I'm sure I can squeeze in a few minutes during TypeCon. Sometime after that party and sometime before that other party. I'll have to check with my people...

There’s ten of them actually

Whoops. Yep. It must have been the single headed variety that Kent had in mind.

Nick, Elliot, nice points!

These days, wayfinding systems don’t use serif fonts, conventional wisdom

I have seen more & more shifting away from that over the last 10 years although in general that does seem to be the dominant outlook

James, your example makes me think that for the 8 single directions it might be nice to offer the arrow in a circle as a discretionary alt.

The thing about arrows is that in general they are pretty easy to make in the sense that unlike a b to d ( or n to m even) where you really do have to think a bit more a left arrow really is just a reverse of the right. Fancy Arrows with 3 shafts, flat or v shaped backs, or double heads & so on are something else again as are arrows in a monospace.

Also, Hunting around for the arrows not covered by Don's link to Math Arrows I found these:

U+21DE ⇞ upwards arrow with double stroke So 1.1
aka page up

U+21DF ⇟ downwards arrow with double stroke So 1.1
aka page down

U+21EA ⇪ upwards white arrow from bar So 1.1
aka caps lock

U+21EB ⇫ upwards white arrow on pedestal So 3.0
aka level 2 lock

U+21EC ⇬ upwards white arrow on pedestal with horizontal bar So 3.0
aka caps lock

U+21ED ⇭ upwards white arrow on pedestal with vertical bar So 3.0
aka numerics lock

U+21EE ⇮ upwards white double arrow So 3.0
aka level 3 select

U+21EF ⇯ upwards white double arrow on pedestal So 3.0
aka level 3 lock

U+21F0 ⇰ rightwards white arrow from wall So ON
aka group lock

U+21F1 ⇱ north west arrow to corner So ON
aka home

U+21F2 ⇲ south east arrow to corner So ON
aka end

U+21F3 ⇳ up down white arrow So ON
aka scrolling

U+232B ⌫ erase to the left So ON 1.1
aka delete to the left key

U+2326 ⌦ erase to the right So ON 1.1
aka delete to the right key

U+23CE ⏎ return symbol So ON 3.2
* may be shown with either hollow or filled glyph
ref U+21B5 ↵ downwards arrow with corner leftwards (Arrows)

U+23CF ⏏ eject symbol So ON 4.0
* UI symbol to eject media

There's a party over here.

There's a party over there...

Shu, ... After all having spent so much time with type folk isn't it time for a font of your own? ;-)

Maybe. Maybe. There are only so many hours in the day. I am sometimes shocked at my own schedule, especially during the ramp up the TypeCon. Now that I think of it, back when I was in school I took a intensive type design class with Max Kisman.
I did create an experimental "font" made from cutting out letters in latex stretching them out with pins and the digitizing them in FontLab. Oh, those were the days. Experimental design just for the fun of it.

Actually, I think it was that class when Max first suggested I volunteer for TypeCon. The rest is history...

Type Design - flickr style

A lot of this lost me but if you included arrows in a font and they matched the vibe - there's a good chance I would use them for design, not math. I use this combination all the time: ->

Also my needs are simple - a forward pointing arrow would do it.

pbc

There are situations where matching a decorative theme is more important than strict legibility in wayfinding, e.g., amusement park signage. If anything, Victorian wood types dominate in many parks, and those look quite dismal with generic arrows.

But yes, even there in more utilitarian areas off the main public paths they revert to a common sans face for directional signs.

These days, wayfinding systems don’t use serif fonts, conventional wisdom being that these are less legible and therefore unsuitable, so Kris, why bother?

As a practising graphic designer & typographer I find matching arrows extremely useful. They also don't take very long to draw, so I find it appropriate to include them. Why would you assume that arrows are only useful for wayfinding?

Why not add something equally useless, such as a unicase or Bold Italic Small Caps?

I also find Bold Italic SC useful, although the occasions to use them are rare. If you're going to do something, might as well do it properly!

Why would you assume that arrows are only useful for wayfinding?

Because I never use them myself, and because I never come across them in use!

I could be mistaken, but will only be convinced if a few real-world examples are posted.

The printer's fist is, I suspect, more useful, (the Calgary Typophiles have one in their recent poster -- see their thread).

If you’re going to do something, might as well do it properly!

My sentiments too, but it's a question of priorities.

Nick, a printers fist is a pointing finger - right?

Also I have more code for us.

These arrows are not computer input symbols or math symbols. They are dingbat arrows. A whole new code area...

I am putting them in just so the record is more complete, to potentially help avoid confusion - some of these seem pretty close to the math symbols; and of course for the those “completists” lurking among us.

Dingbats block
➔ U+2794: HEAVY WIDE-HEADED RIGHTWARDS ARROW
➘ U+2798: HEAVY SOUTH EAST ARROW
➙ U+2799: HEAVY RIGHTWARDS ARROW
➚ U+279A: HEAVY NORTH EAST ARROW
➛ U+279B: DRAFTING POINT RIGHTWARDS ARROW
➜ U+279C: HEAVY ROUND-TIPPED RIGHTWARDS ARROW
➝ U+279D: TRIANGLE-HEADED RIGHTWARDS ARROW
➞ U+279E: HEAVY TRIANGLE-HEADED RIGHTWARDS ARROW
➟ U+279F: DASHED TRIANGLE-HEADED RIGHTWARDS ARROW
➠ U+27A0: HEAVY DASHED TRIANGLE-HEADED RIGHTWARDS ARROW
➡ U+27A1: BLACK RIGHTWARDS ARROW
➢ U+27A2: THREE-D TOP-LIGHTED RIGHTWARDS ARROWHEAD
➣ U+27A3: THREE-D BOTTOM-LIGHTED RIGHTWARDS ARROWHEAD
➤ U+27A4: BLACK RIGHTWARDS ARROWHEAD
➥ U+27A5: HEAVY BLACK CURVED DOWNWARDS AND RIGHTWARDS ARROW
➦ U+27A6: HEAVY BLACK CURVED UPWARDS AND RIGHTWARDS ARROW
➧ U+27A7: SQUAT BLACK RIGHTWARDS ARROW
➨ U+27A8: HEAVY CONCAVE-POINTED BLACK RIGHTWARDS ARROW
➩ U+27A9: RIGHT-SHADED WHITE RIGHTWARDS ARROW
➪ U+27AA: LEFT-SHADED WHITE RIGHTWARDS ARROW
➫ U+27AB: BACK-TILTED SHADOWED WHITE RIGHTWARDS ARROW
➬ U+27AC: FRONT-TILTED SHADOWED WHITE RIGHTWARDS ARROW
➭ U+27AD: HEAVY LOWER RIGHT-SHADOWED WHITE RIGHTWARDS ARROW
➮ U+27AE: HEAVY UPPER RIGHT-SHADOWED WHITE RIGHTWARDS ARROW
➯ U+27AF: NOTCHED LOWER RIGHT-SHADOWED WHITE RIGHTWARDS ARROW
➱ U+27B1: NOTCHED UPPER RIGHT-SHADOWED WHITE RIGHTWARDS ARROW
➲ U+27B2: CIRCLED HEAVY WHITE RIGHTWARDS ARROW
➳ U+27B3: WHITE-FEATHERED RIGHTWARDS ARROW
➴ U+27B4: BLACK-FEATHERED SOUTH EAST ARROW
➵ U+27B5: BLACK-FEATHERED RIGHTWARDS ARROW
➶ U+27B6: BLACK-FEATHERED NORTH EAST ARROW
➷ U+27B7: HEAVY BLACK-FEATHERED SOUTH EAST ARROW
➸ U+27B8: HEAVY BLACK-FEATHERED RIGHTWARDS ARROW
➹ U+27B9: HEAVY BLACK-FEATHERED NORTH EAST ARROW
➺ U+27BA: TEARDROP-BARBED RIGHTWARDS ARROW
➻ U+27BB: HEAVY TEARDROP-SHANKED RIGHTWARDS ARROW
➼ U+27BC: WEDGE-TAILED RIGHTWARDS ARROW
➽ U+27BD: HEAVY WEDGE-TAILED RIGHTWARDS ARROW
➾ U+27BE: OPEN-OUTLINED RIGHTWARDS ARROW

Why not add something equally useless, such as a unicase or Bold Italic Small Caps?

Nick mate what is so useful about a ten style monospace font in this case?? or that trendy outline font of yours??

-----------------------
Chopper Reid says "Harden the **** up".

Pay attention Muzzer, we're talking characters not typefaces -- as the thread title might indicate.

But in reply to your question, my types appeal to a discriminating enough niche as it is, no point in narrowing it still further.

It does seem like a good idea to put one arrow in a font, though, a companion for the printer's fist (pointing hand).

For your reference, this image is from an invitation for an event featuring
the work of Daniel Bozhkov.

http://i6.photobucket.com/albums/y234/jupiterboy/Arrow.jpg

http://www.danielbozhkov.com/

Pay attention Muzzer, we’re talking characters not typefaces — as the thread title might indicate.

I am mate! ears are pricked up and all, although the eyes are a bit buggeresd. I just find it funny that you question including arrows in a font when the idea of most entire fonts are questionable!! probably a bit sideways for this thread.

-----------------------
Chopper Reid says "Harden the **** up".

>it’s a question of priorities.

Cyrus Highsmith at last year's TypeCon critique session mentioned--to Chris Lozos--his opinion that it is more worth your time to do new typefaces than to try to make them very complete as far as languages, symbols, etc.

Muzz, "questionable" has nothing to do with it.
Someone, somewhere is licensing fonts you consider questionable, because there's no accounting for taste.

The issue for the foundry is this -- does the group of people who may licence a particular font overlap with the group of people who use arrows in their designs?

It's hard to say, because people use fonts in unexpected ways. But for me this thread has been informative, and at the moment I am considering adding one left-to-right arrow to my neverending typeface.

Thanks for the links James & Shu.

Thanks for clarifying about the 'fist' Nick.

I have learned a heck of a lot too. I am having a go at 10+ arrows for the heck of it. And a pointing hand as well. And some suns, and moons. And an umbrella. oh dear...

I expect that Cyrus is right.

Cyrus Highsmith at last year’s TypeCon critique session mentioned—to Chris Lozos—his opinion that it is more worth your time to do new typefaces than to try to make them very complete as far as languages, symbols, etc.

is that what you got out of it, chris? i thought he and matthew were saying that it's important to settle on your base design before expanding your typeface. but i might be wrong.

>And an umbrella. oh dear...

…hedera too.

Tim

What's the code point for hedera?

I did major in maths, and I'm a type enthusiast, but I frankly don't know that much about mathematical typesetting. So from a user perspective, I will try to share my impressions of which arrow symbols are actually used in mathematics, though with an algebraist bias. I don't have that many maths books with me right now as I keep them at work, so I might be missing a few.

Most mathematical typesetting I've seen require no more than the basic single arrow ('→' shows up everywhere including maps and limits) and double arrows ('⇒' denotes logical implication). Also fairly common are the arrows with bars at one end ('↦' is used to denote that an element is mapped to something). Don't forget, all these will have corresponding slashed versions (i.e. '↛') to denote negation.

In lectures, I've seen a double-headed arrow ('↠') used for surjection and an arrow with a tail ('↣') used for injection. I've seen very few typeset examples of those, though.

In handwriting, arrows are sometimes used like diacritical marks to denote that the letters symbolize vectors ('⃗' over a dotless 'i', for example). In typesetting, however, vectors are indicated by using bold letters, i.e. i.

I really love Kris's arrows designed to harmonize with the text. But for mathematical typesetting, simple geometric arrows generally suffice. I think a dingbat font with arrows in different weights and styles designed to be combined with a large range of typefaces would be enough to do the job.

Hederae: 2766, 2767, and 2619, for downward, right pointing, and left pointing, respectively

I don't know why the leftward one is out of sequence, but the printers fists follow it at 261A–261F.

The main reason I have been playing with fonts these last years is to get all the symbols I need for mathematical texts and I would never expect a general purpose font to contain them all. If you want to have a look at symbols used in LaTeX, have a look at the pdf text, MnSymbol -- A Math Symbol Font. Those symbols are meant to be used with the font Minion Pro. The otf files for MnSymbol are here in the CTAN Archives and complement the MinionPro package (public domain).

Michel

under my um-ber-ella-ella-ella-ey-ey-ey

As a somewhat related side note, I thought that people might find some interest in this article on the history of the printer's fist, or "manicule":
http://www.livesandletters.ac.uk/papers/FOR_2005_04_002.html

There are sixteen more arrows in the range 27F0-27FF.

I am into mathematical typesetting, but am unsure how I feel about arrows. Especially for the longer and wacky ones, they feel more like graphics than glyphs, or, to be more precisely, occupy the hinterland, neither exactly one nor the other. Certainly for things like commutativity diagrams, I'd much rather see them synthesized based on the geometry of the diagram rather than just pulled out of a font file like so much digitized 18th century metal. I think these guys had the right idea. (also see Dynamic optical scaling and variable-sized characters).

For non-mathematical work, I think others have pointed out fonts designed for wayfinding. There was one that was proprietary for a museum for a while, then released more broadly a couple years or so back, but I can't remember which one it was. Whitney doesn't seem like it has arrows, but I remember arrows from the specimens. Ah well.

> I think these guys had the right idea
Thanks for the link, and I agree with you about those long arrows in diagrams (be they mathematical or not).

As for the arrows in the range 2190--21FF, and on a more practical basis, if an undergrad student of mine decides to write a mathematics homework using a basic text editor, he will certainly need the characters 2192 (for functions), 21D2 (for "implies"), 21D4 (for "if and only if"); character 21CF (for "does not imply") and 21CE (for "is not equivalent to") would be useful, as well as 21A6 ("maps to" as in "x maps to y"). Less importantly, characters 21A0 ("maps onto" or "maps surjectively" or "epimorphism") and 21AA ("maps 1--1 into" or "maps injectively" or "monomorphism") might serve. Off course that student would also need other mathematical symbols that are not arrows. If that student uses Word, there is an equation editor where all the basic characters can be selected.

More in line with Raph's comment, if I use the amsart (American Mathematical Society) LaTeX package for articles, when I want a "if and only if", I type \iff and what comes out is a nice thing made out of two glyphs that look like one: a double left arrow and a double right arrow glued together.

In case anyone is interested, I put together an OTF of most of the TeX math symbols, reencoded as Unicode: ghr10.otf. I'm not sure I got everything 100% right, but I do believe that all of the arrows are in there. The math symbols in the TeX font library are scattered all over the place, with the most common symbols in the Computer Modern text fonts, a great many symbols in the CM Math range, and a bunch more in the AMS extensions.

Incidentally, ghr10 is for the Roman (upright) versions of glyphs. There's a separate ghmi10.otf for the Math Italic range, and that's where the lowercase Greek lives. TeX, of course, doesn't have italic arrows, so I'm only posting this second file here for the thread's broader purpose of discussing math glyphs.

raph: thanks for posting those files, I had been looking for an example math font file, and hadn't been able to grab a few off of my dad's machine yet (between LaTeX, Maple, and Mathmatica I think he's got a few). I haven't seen any truly indepth tutorials, but I did find this PDF of a presentation that goes over at least some of the design issues: http://www.typoma.com/publ/20040430-bachotex.pdf

«El futuro es una línea tan fina que apenas nos damos cuenta de pintarla nosotros mismos». (La Luz Oscura, por Javier Guerrero)

I always find myself wishing a full arrow set was available in every font. for web work i often use > or < to point, but have no up down options with the exception of ^… which sucks. Not even talking nicely designed… just having anything resembling arrows that can be added to xml docs for content management.
_____
www.choppingblock.com
www.chopshopstore.com
www.wanderingspace.net
www.grodzki.com

For those interested in mathematical typesetting, there's a pre-conference workshop on math fonts at Microsoft before TypeCon: http://www.typecon.com/workshops.html

And I'm just putting the finishing touches to a booklet on the subject, which goes to print in the morning. If all goes to plan, we'll have a complimentary copy for every TypeCon attendee.

A couple of illustrations from the booklet showing arrows in mathematics

And here are arrows used to represent a transform (here a rotation) on quantum bits implemented as polarized photons (used in quantum cryptography)

John, wow, that's great! :-)

Thanks Michel!

I just installed the MnSymbol package, and used it with teTeX to try. Here is now a table giving the symbol from MnSymbol with its unicode number, and its corresponding possible usage to represent qubits. The text font is Utopia regular

Nice. Thanks, Michel.

Speaking of polarisation of photons, did you know that dung beetles navigate by referencing the polarisation patterns in moonlight? So even these arrows are a kind of way-finding system. :)

No, I did not know and just had a look at Dung beetles on Wiki. Thanks!

Has anyone mentioned arrows 21DB (triple arrow) and 21DD (\leadsto).

First 21DD; it is used in a standard reference on Algorithms (Cormen et al.) to denote a path in a graph, i.e. a way to get from one vertex to another by following arrows.

About 21DB; in the mathematical theory of n-categories, there are arrows, double arrows, triple arrows etc. up to arrows with n lines and a pointed end. Arrows go from objects (like vertices) to objects, double arrows from arrows to arrows, and you imagine the rest. Here is a (commutative) diagram taken from An introduction to n-categories

And it is said by the author that this has links with General Relativity and Quantum Cosmology...

Michel

Generally speaking, an arrow or other symbol should be easier to work with in a design as a piece of vector art.

links with General Relativity and Quantum Cosmology..

...not to mention a General Theory of Readability.

And I recognize it immediately as the mystical map of creation in the Kabbalah :)

^ ; )

With apologies...

There is indeed just one triple arrow in the diagram, labeled M, in the middle; in his article Baez seems to miss u21DB for, though he uses lots of double arrows u21D2, when comes the time to speak of M he does not write

Michel

[edit] > With apologies...

Your variaton on the diagram reminds me of pottery figurines from prehistory or early antiquity.

@russelim:Generally speaking, an arrow or other symbol should be easier to work with in a design as a piece of vector art.

That rather depends on the tool you're using. Arrow diagrams occupy a nether region which is neither pure text layout as in book paragraphs nor pure graphic design like drawing a portrait in Illustrator. In this way, they are somewhat similar to tables, just a little different graphically. You can with some pain fake tables using paragraphs and tabs (that's basically what the tab function is for, historically at least), and you can also make tables in Illustrator, but both approaches involve serious compromises. The serious problem with treating it as pure graphics is that it becomes very difficult to do simple operations like adding a row or column, or even changing the size of one already existing.

Since the problem is so thorny, there are several different approaches. TeX pretty much holds down the "text" end of the spectrum, using a complex ASCII markup that is actually a relatively powerful programming language. Since the user base of TeX tends to be computer scientists, mathematicians, physicists, and statisticians, this actually works out pretty well.

But I definitely agree, if you're a graphic designer without deep semantic knowledge of what the diagram means, and the ability to code that in the typesetting system's programming language, then you are better off pretending it's pure graphics, and just sacrifice some of the fluidity in being able to edit and adapt it.

I say tables and arrow diagrams fairly interchangeably here. There are important differences between the two, but the structure of the problem is similar, and, in the case of the chemistry diagram posted by John above, it's clear from the grid alignment that the tool used to create the diagram was in fact some form of table.

@raph . TeX pretty much holds down the “text” end of the spectrum

But Baez article above was written in LaTeX (based on TeX) and his graphics are .eps (the source is available from this link when you click on "other formats"); and the LaTeX package graphics allows importing pdf files, png files, etc in pdflatex. Finally, if you look on arXiv, and in particular quant-ph, you can see from the source files that many papers written in LaTeX come with figures in external .eps files. We are thus not limited to the "text" end of the spectrum, we are free to decide what suits us best.

For my take on the font vs. graphic debate, you need to know that at the time TeX was being developed 1980 it was very difficult and expensive to get graphics into a page. In book production in that time of galleys of phototype, it was nearly impossible, without huge cost.

Thus TeX was built to allow any characters to be built into fonts, and Knuth even invented a program for creating fonts called Metafont.

As Russel says, today graphics are a completely viable alternative. In fact, when I include math in my equations today in web-based courses, I use graphics, due to the inconsistent handling of math by different browsers. A gif will always show up. (I do feel bad that my pages are less accessible to the handicapped, as the alt tags only say "math formula").

Michel, that's great stuff not least because it explains the context for an arrow that has 3 shafts.