> What is this list, and where can I get a hold of it?

Unicode: http://www.unicode.org/charts/

Thanks, kentlew. That is a huge selection!
Is there anykind of super unicode chart that contains all the others combined? I know Opentype can store 65,000 characters...

OK, this looks fairly close:

http://www.ssec.wisc.edu/~tomw/java/unicode.html

If you only want a list: http://unicode.org/Public/UNIDATA/NamesList.txt

Ryan, your "fairly close" list only goes up to Unicode U+FFFF. There are, quite literally, thousands of additional character assignments with yet even higher codes. Do not confuse the number of glyphs (not "characters") that can be stored into an Opentype file with the Unicode assigned to each glyph. They are separate things.
For all things Unicode, Unicode.org is the website to explore. Every other list is derivative.

Unicode, by the way, is the answer to your initial puzzling query "it's not in the ascii chart". "The" ASCII chart (a name so ill-defined that you are better off not using it) can only hold 256 characters. To circumvent that built-in limitation, as well as to finally set a world-wide standard to what character gets assigned what code, Unicode was devised. Current font design software may appear to "support" various ASCII-based constructions such as Code Pages, but in reality everything is done through Unicode.

Now Im curious if there has ever been a font that had all those characters. Even one without the asian glyphs and herioglyphs. And if there is nothing like that, what font DOES have the largest number of glyphs?

http://en.wikipedia.org/wiki/Unicode_typefaces#Comparison_of_fonts

http://typophile.com/node/89519

> http://en.wikipedia.org/wiki/Unicode_typefaces#Comparison_of_fonts

Why isn't Segoe anywhere in there?
BTW, [my] Tahoma doesn't have Armenian.

hhp

It's kind of hard to put a real number to that. The largest fonts, by far, are always going to be the Asian ones. I can't think of a "regular" font that literally has thousands of glyphs. Maybe Zapfino Pro just goes across that number, as it has lots of alternates and ligatures.

Arial Unicode MS and Lucida Sans Unicode are efforts to have a large, design-wise coherent set of glyphs. You might already have one of these, or both, in your system fonts. ("Design-wise coherent" was the intention, but there has been comments on Arial Unicode's. Search this forum, I'm sure it has been mentioned a few times.)

Code2000, by James Kass, is a serious attempt to contain all Unicode characters in the Basic Plane. Its official website has been taken over by someone else, so you have to Google for a download site -- it's freeware.
Code2000 is a huge font (8.4MB), but aiming for completeness rather than proper design. Lots of glyphs are "shared" between the same code points, some of which are perfectly acceptable ("Latin "A" and Greek "Alpha") and others where there seems to be some dispute. Code2000 contains 63,545 glyphs; but not every glyph is bound to a single character; the last 10,000-or-so glyphs are just there to make compound characters from, and should not be used on their own.

Since Unicode is layered (some might say, like an onion) -- additional Planes are put in separate fonts Code2001 and Code2002. A very worthy effort, except that Unicode is constantly expanding, and they are always adding more and more glyphs to the defined sets.

Then there are the so-called "Fallback" fonts -- an interesting category of their own. They do not define a single glyph for each separate code, but assign one per Unicode block. That way you always see something -- even if it's only a reminder you are "missing thus-and-thus". The canonical fallback font is "Last Resort", included in Mac OS X.

And then there is Ken Lunde's "UnicodeAll", which not as much concentrates on having correct glyphs (I think it only contains one ;-) ) but solely on stuffing as much as possible in the available table definitions in an Opentype font. Still -- font software ought to think it contains as much as 1,112,030 separate characters!

So I guess that makes Ken the winner!

Yeah, big or proper, this is the question.
Size matters, sure. But it’s not everything. In most cases, a font is big OR is well done.
Exceptions seem to be very rare, rather best kept secrets. For the least, the Wiki entry doesn’t tell us much about this.

Those who can boast with the highest number of characters are likely to be blamed for the most terrible glyphs.
Who is really caring for either goals?

Andreas, that's how I feel about kerning. If a font boasts "10,000 kerning pairs" I find myself thinking "perhaps the designer should re-do it".

.

I know, I said that

I can't think of a "regular" font that literally has thousands of glyphs. Maybe Zapfino Pro just goes across that number, as it has lots of alternates and ligatures.

but John Hudson's Gabriola reportedly has " 4500+ glyphs [..] another league as you may have noticed." Cor! Since it's not an Asian font, every single character must have at least a dozen alternatives!

Just a note: Code 2000 is useful, but I'm hard pressed to think of anything uglier. Beauty is in the eye of the beholder, I suppose.

As far as needed characters go, once they added U+1F4A9 in Plane1, I think they were done. Bring on the cat.

Quivira looks interesting, despite having an absolutley hideous looking website.
http://www.quivira-font.com/

I actually like DejaVu Sans Bold quite a lot
http://sourceforge.net/projects/dejavu/

The Armenian for one is unusable.

hhp

To me, the whole font looks pretty sorry -- a Big Mac font. YMMV

My font with the most glyphs has just over 1,200 glyphs. It includes Cyrillic and Greek as well as almost all Latin glyphs. A typical font for me is between 500 and 800 glyphs for Latin only support. My decisions on glyphs is based on languages rather than some numbers game. It boiled down to how many languages can I support without killing myself with labor. Some languages may only need a few more glyphs so I support them. Some require a great deal more time. Vietnamese needs many additional glyphs and stacking sets of diacritics which affects vertical metrics some so I have decided not to attempt Vietnamese at this late stage of my life.

.4500+ glyphs

:-)

5128

John, congratulations to Brill!

5500

Those who can boast with the highest number of characters are likely to be blamed for the most terrible glyphs.

.4500+ glyphs

5128

5500

;-)

Andreas, so you just added 1000 glyphs in the three hours? ;-)

hhp

Does anyone have opinions of Quiria or Dejavu Sans?

You can be assured that all of the Glyphs in Andreas and John's fonts are of the highest quality.

Except the interrobang. I make no guarantees for the quality of any interrobang.

Bang! Zoom! Alice ;-)

This is related, has there ever been a competent attempt at a font that contains the entire unicode standard in a relatively consistent style (where possible)?

Does anyone have opinions of Quiria

Yes. I dowloaded it this afternoon.

It's like a bad Times Roman.

Trashed it this afternoon, too.

Well, you asked...

Although far from the excellence of Andreas' and John's works, this font project sums to 6,416 glyphs covering the whole Latin, Cyrillic and Greek scripts (sorry, I made no updates in that thread for a while as I am 'out of type business' until the end of 2012).

The font should be available at some point of 2013. I uploaded a little sample (in two parts, here and here) and a not very good presentation here.

About other fonts, Quivira is just one of these Times-based adaptations, with no enough quality. DejaVu is better, but if you want a free one to build a character list, Gentium Plus is the best choice I know.

> a competent attempt at a font that contains the entire unicode standard in a relatively consistent style

Well, that very idea was actually the starting point of Andron (11 years ago), theoretically. The style had to be not only consistent but also rather classical, so that the font is suitable for large-scale scientific editing. “The entire unicode standard” is of course not feasible for one man in ten or twenty years. 90% of “the entire unicode standard” is Chinese, there you have the problem.

But there are other challenges beyond Chinese, which cannot be mastered by one alone. Above all, the vast range of Indian and Asian scripts. Many of them are highly sophisticated (as Latin or Chinese or Arabic, for instance), so you just can’t step in and make the glyphs of it, having no related background at all.

The classical type styles of many scripts seem to have some underlying common characteristic principles, which derive from the pen (or the brush). Latin, Arabic, Devanagari, Chinese … have their own respective ductus, but all those ductūs have more or less something in common. Would be worthwhile to explore how far you can go in “harmonizing” all that.
I’d love to study Tibetan, Devanagari, Arabic and Korean (and Chinese) for a while, and than start with Latin and Greek from point zero again … But for that I’d have to get imprisoned first, I suppose.

A world writing camp would be great, where masters from all scripts tutor the crowd how to hold the pen.

To me the "secret" to multi-script harmonization* lies
in the proper relative use of the Cartesian space, where
apparent size has to be balanced against the natural
bouma-formation needs of the given script. For example,
how to "translate" Latin's dominant x-height into other
scripts. Imposing the x-height on Arabic for example is
so hopeless that people have the sense to quickly give up,
but this undesirability is also true -just more subtle- in
the case of pretty much any other script, like Greek,
Cyrillic or Armenian. Human consciousness likes to line
things up; but our reading mechanism has deeper needs.

* At least for a text face.

In contrast, whether one uses geometry, chirography or
the "third way" (or some combination) in constructing the
"strokes" merely imparts an overall consistent style (which
to me is a more superficial/easy matter).

hhp

What if the project was not done by one man but many who have expertise in the various languages?

Sure, but they would have to agree on a basic philosophy.
Or have a tough boss. I've long admired benign monarchies. :-)

hhp

> What if the project was not done by one man but many who have expertise in the various languages?

– or rather: various scripts.
Yes, that might be the way to go.
I doubt if a ‘tough boss’ would be enough. Subtle skills and feeling would be required from all participants. And the ability to adjust …

The 'tough boss' approach seems to me counter productive if the whole point of involving multiple people is to overcome the difficulty of one person being able to master all the knowledge and skills necessary to designing for all the scripts. I've spent the last two years working with a team of six other people on a set of fonts supporting nine different Indian scripts, and it would have been pointless -- no matter how benign my monarchy -- having me lay down the law on how it should be done. We planned carefully. We analysed the client needs and target use environments in depth (the fact that the client then went and changed the environment after we'd done 95% of the work isn't our fault!). We researched options for the individual scripts, looking for ways to harmonise them to each other and to an existing Latin type that were in accordance with the individual writing and/or typographic traditions . We documented those technical issues that would inevitably force us to deviate from those traditions (notably, the difficulty of producing traditional bold designs for southern Indian scripts in an asymmetric rendering environment). We produced trial forms and tested them before proceeding to the main body of design work. And we discussed, discussed, discussed.

The ideal boss in this case wouldn't try to master
any of the individual scripts; he would make sure
those with mastery in a script engage in the right
attitude with respect to the others, and the whole.
It sounds like what you did actually. The best boss
is always a very good listener. But sometimes he
will have to say: "No, we need to do it this way."

For example he might guide the use of linguistic
frequency data in the determination of the ideal
vertical proportions* even if an individual in the
team does not believe in that sort of thing.

* http://themicrofoundry.com/ss_rome1.html
(Rightmost image)

hhp

I'd be quite interested to see some of the more exotic glyphs with a nice fat bold Helvetica feel to them. Is that even possible? Can you produce a "sans serif" FLOBT version of the middle eastern scripts?

So likely the idea of a font comprising the entire unicode standard in a unified style would be something very difficult to produce. This is quite disappointing. Many of the languages are similar enough to Latin, or are sub-Latin, where one designer can cover it, as is the case with Andron. It becomes a problem when you get into more "exotic" languages, I think. I'm guessing one could make less work from the CJK symbols by breaking the glyphs into a couple hundred main components and building the glyphs from there, although it seems that even still there would have to be adjustments to every glyph, since the stroke weights seem to vary greatly in order to achieve more unity in appearance between the complicated and simple glyphs.

Brian, typography as you know it is very Western (= Latin) biased. "Italics", for instance, has no real counterpart in a lot of "exotic" fonts. (I was surprised to find an Arabic Italic font with the latest InDesign -- I darenot vouch for its currency among arabists.)

A lot of CJK fonts *do* contain building blocks for more complicated glyphs. Then again, portions of Chinese characters, for example, lend themselves to this because of their natural design. I.e. if there were a script with just horizontal and vertical lines, a designer could just design one horizontal and one vertical line and then let software search things out.

"I can't think of a "regular" font that literally has thousands of glyphs."

A number have been cited. Pretty much all of Adobe's typefaces of recent years have 3000+ glyphs per font.

For the regular sort of typesetting I'm doing every day, I'm happy to even find a Greek set included when I need it ;-)

Minion Pro is my staple font -- wait, its Regular already has 1669 glyphs! (Plus a couple more white-space-onlies.)

"per font" -- do you mean per complete font set? Since Minion comes in three weights and two styles, for the entire font that adds up to about 6 * 1669 ≈ 10,000 glyphs. Not Even Including Swashes.

No, I meant per font. So yes, 10,000 glyphs for Minion, not including optical size variants. Throw those in and you get 40,000 glyphs.

My own Hypatia Sans (on which I had lots of help): 3,000+ glyphs x 12 fonts = ~36,000 glyphs

T