A proposal for new version of Greek glyphs for DejaVu Serif

vank's picture

Hi everybody,
I suspect i don't need to say much about the open source DejaVu project. The link speaks for itself. "The" DejaVu "fonts are free as in freedom, and may thus freely be embedded".

What i 'm posting below is a recent attempt to design Greek glyphs for the serif version from scratch, initially for my very personal use (i guess i liked too much typesetting in the Latin version). The idea of including these glyphs in the original DejaVu Serif is currently under consideration by the team running the project, so, your feedback on the way the glyphs look and any ideas for their potential improvement are more than welcomed.

DejaVuSerif.pdf48.72 KB
rcc's picture

[comments elided]

hrant's picture

Funny, I was gearing up to soften things up when I read that last bit!
Before I forget: RCC, what's your actual name?

My problem is I'm not even close to being an expert in Greek...
But I can offer the following, based on my general ideas about type:
1) It's too latinized. Too many serifs for one thing, and I'm guessing you've matched up the vertical proportions WRT the Latin. Please read my article in the 2005 edition of Hyphen (the Greek typographic journal). Or for starters: http://www.themicrofoundry.com/ss_rome1.html _
Specifically in the lc for example, the eta and chi are fully Latinized. They're not Greeks, they're tourists.
2) There is a general malformation of curves. A quick look at the alpha indicates there's a lot of work to be done...

All this said, I think with research, thought, sensitivity and persistence you can and will arrive at something good enough. And try to be happy even for the negative opinions: it means people care. Also, you have something very valuable going for you: nativity. The fact that you've been (I presume) exposed to Greek since childhood means you have the invaluable but elusive cultural sensitivity in you - you just have to get to it, by effort, but also by knowing what elements from Latin type to throw away instead of blindly replicate.

BTW, I noticed that* DejaVu has an Armenian component.
I'd be curious to see that, and there I could help you guys for real.

* I'm asssuming it's the "Arev" set mentioned in that wiki - which is
pretty clever, since it's the reverse of "Vera", and it means "sun"!


Rhythmus.be's picture

vank — Really not bad at all. And the fact that you're designing for an OpenSource project has my full sympathy. I also strongly support your choice in favour of letter shapes that comply to what a typographical typeface should be, i.e. that what some consider "latinocentric". That choice is seconded by authorities in type design such as Jan van Krimpen and Eric Gill. However, some suggestions:

Α: cross bar too low
Β: too wide; cross bar too low
Γ: too wide (= arm too long)
Δ: foot stroke too thin (match it with the left slanted stroke)
Ε: cross bar too high
Η: cross bar too high
Θ: cross bar too high; serifs too big (long), which clutters the counter
Ξ: cross bar too high and too thin, serifs to big (cfr Θ)
Ρ: (bowl) too wide
Σ: the diagonals shouldn't have the same thickness; either the upper one is the thicker (following the translation of the pen), or the lower one (as in Z)
γ: you should choose between these alternates (I prefer the first one)
ζ: you should choose between these alternates (I prefer the second one)
η: perhaps making a descender from the second stroke would be one concession you could make to the misolatinists (and would better match the μ)
ν: you should choose between these alternates (I prefer the second one)
ξ: you should choose between these alternates (I prefer the first one while is the most latinised)
ς: you should choose between these alternates (I prefer the first one)
υ: you should choose between these alternates (I prefer the second one)
ΰ: acutus too low in regard to the trema

Did you plan to add polytonic glyphs as well? (Having been a classical philologist, I deem this ablosutely necessary for a Greek font.)

crossgrove's picture

Don't call Eric Gill an authority in type design. He was nothing of the sort. He was a lettercarver with occasional forays into type design. I don't think he was the designer of the Greek sets to his own types.

Vangelis, you might want to get Gerry Leonidas' document on Greek Type Design. It's very thorough in explaining various conventions of Greek type design. Though you might want more sources, this is a valuable document for what you are doing.

rcc's picture

[comments elided]

hrant's picture

> authorities in type design such as Jan van Krimpen and Eric Gill.

Two people who made some of the most horrid Greek type ever seen.
And I really like Gill!

Latinization is generally dysfunctional, and almost always disrespectful.
Especially in a text face, superficial formal harmony is for little children.

> Unwind a bit.

I'm sorry if I seemed wound up - I wasn't.
I really am just curious to know your name.


vank's picture

...had a bad day "rcc"?
I can take all the criticism in the world, I LOVE constructive criticism. But i can hardly see any real criticism on your posts nor do i feel that you essentially offered any real input. I shall therefore not consider these posts any more seriously than
you considered mine. Just for the record, without being a "self-respecting typographer", but definitely considering myself a self-respecting reader, you need to be reminded that we do need to typeset other texts as well rather than...the Bible or the Battle of Marathon, in Greece. Regardless of your nationality then, your stay for as much as a quarter of a century in the country should have offered you a couple of looks on some newspapers at least. Garamond's or Didot's designs were wonderful, the Slimbach and the Greek Font Society's revivals are magnificent, still there are people who need to typeset texts that simply do not look OK with the "retrospective touches" typefaces like these have. Oh, and by the way, my sincere apologies for hurting your eyes.

Dear Hrant, Ludwig M. Solzen and crossgrove sincere thanks for your comments. Now, that's real food for thought. Indeed, i still have a lot of work on the curves and lots of research to carry out. To be honest, i was expecting to hear that it looks too latinized and was more than sure that the remarks about the latin "chi" and "eta" would immediately pop-up. I've kept these glyphs like that after some consideration though. I'm not sure whether it's just a trend or whether the real reasons behind it are more practical (e.g. saving up some space and simplifying letterforms when typesetting in very small sizes etc.), but chi's and eta's without descenders are something that more and more appears in contemporary Greek typefaces (a big percentage of typefaces designed and used inside Greece); you might want to consider that. Also, you might not be aware of that or like it, but another "trend" is a latinized zeta or final sigma and I'm not speaking about just a couple of typefaces for specific uses here. I'm speaking about typefaces used by the biggest newspaper/magazine publishing companies. The only contemporary typeface that managed to be a real exception to all these "trends" without looking "old-fashioned" (don't misunderstand my point here, please) is, to my eyes, P. Bilak's "Fedra". Now, it was after some thought that i chose my pathway; i wanted to attempt a design that, while it might not be so academically consistent, it might actually turn out usable by people who simply don't want the classical touches of already existing, beautiful serif typefaces, some of which are already offered as freeware. Moreover, to my eyes, an overall too curvy and "calligraphic", too-"traditional" (e.g. Garamond-like since "rcc" mentioned it) design of the Greek glyphs could easily end up leading DejaVu to a "typeface-within-a-typeface" problem, namely, Greek glyphs side by side with Latin glyphs looking as if they actually belonged to different typefaces. A too-Latinized version, as you nicely pointed out for my design, is on the other edge, though. And upon this i should do some more research indeed.

For the Armenian component of the DejaVu you might want to contact the team running the project?

All the best,


PS: Dear Ludwig, yes, polytonic support is going to be added as soon as the glyphs forms are finalized.

Rhythmus.be's picture

The concept of latinisation is wrong. Western script originates from its Greek ancestry, which in its turn derives from the Phoenician alphabet, which in turn, &c. Alphabetical script forms one insoluble whole, as opposed to syllabic or pictographic scripts. The differences between the diverse sets (Latin, Greek, Cyrillic) are in the letterforms, their skeletons that is, not in the way they are stylistically rendered. This derives from the tools and the intended use.

Printed scripts are formalised because of the way in which they were and are produced, and because of their use with composing and printing machines. Printing was invented in Western Europe (forget about the Chinese and Koreans for a moment), where Latin script dominated. The letter forms that were taken as a model were already heavily formalised, becoming even more so as printing faces. So, what one considers latinocentric (formalised in stead of chirographic forms), really is the essence of type (printing type, that is), not an exclusive particularity of Latin script.

Greek script never came loose from chirography, likely because it was introduced already too late in print, or because those who commissioned the first Greek types attached too heavily to the exotic Greek manuscripts of their days. Anyhow, the fact that for centuries Greek has been set with cursive types, makes that all Greek fonts look the same, i.e. there is but one style that is deemed proper for it, while Latin script has evolved over the centuries and can look back at a rich variety of styles. I strongly support type designers to give Greek (and Cyrillic) script a similar richness of style. This doesn't mean that proper glyph shapes are to be rejected (such as in Serbian or Bulgarian); chirographic idiosyncrasies (such as the pleade of Greek script ligatures), though, should be removed as unnecessary redundancies.

hrant's picture

> I’m not sure whether it’s just a trend or whether the real reasons
> behind it are more practical (e.g. saving up some space and
> simplifying letterforms when typesetting in very small sizes etc.)

Trends happen. Yes, it's important to be aware of them, but even more important is: recognizing and supporting the ones that are actually good; and seeing the limits of style in the domain of text faces.

Saving space is a good thing, usually. It's just another factor.

Simplfying: now there's the bugbear. Way too often it ends up resulting in empoverishment, and not much else. I wish it was just another factor, but usually it's too dominant. A symptom of Modernism.

And what usually gets dropped from the equation is cultural and functional insight and sensitivity. This can only lead to an icrease in the size of the pointless gray blob consuming (intentional choice of term, that) the world.

> what one considers latinocentric (formalised in stead
> of chirographic forms), really is the essence of type

This is exactly the sort of cultural hubris that leads to the demotion of minorities. And it is the disease of Modernism that has drawn a veil over the former and made the latter actually seem like a positive! :-/

In essence, anything not part of the dominant
culture is an "unnecessary redundancy"...


Seeing chirography and latinization as opposites may seem historically
valid, but in essence it's an illusion. I for one am firmly against both.


rcc's picture

Sorry, Vangelis. After solemn reflection, you're quite right. I was dreadfully mistaken to have supposed your wish was to create something durable. So, by all means, kick out those inconvenient mainstays of legibility and readability, demolish those ornery notions of underlying structural unity, string things together higgledy-piggledy if you like, and then foist your results upon masses of incognoscenti. (You've got nothing to lose. Most users will never know better anyway. Besides that, it's freeware, for God's sake, so they'll receive apposite quality and value for sums suitably unpaid.)

As a gesture of good will, I've even elided my previous posts. That's neither retreat nor retraction, mind you, but an attempt to assuage any perturbations triggered by the frictions of reality. I certainly shan't trouble you again.

Rhythmus.be's picture

Hrant — I suppose, you consider rational argument as belonging to the hybris of dominant Western culture, too? Claiming that latinisation and chirography aren't opposites, merely while you are both against them, suggests that either you suffer from hybris yourself, or that you're an inconsequent schizo. If you consent that my statement at least seems "historically valid", then could you please also explain why you think it's an illusion?

Rationality, formalisation and economic use are assets of progress, not the privilege of one particular culture. If Modernist Western culture has been the pioneer of progress and has spend numerous efforts in promulgating civilisation (amongst the fruits of which are print and type design), that sure is no reason to disdain it as being imperialistic. Can't one just acknowledge that at least a few things are universal goods, belonging to mankind in general, apart from whom invented or discovered them first?

hrant's picture

1) Seems can mean not really.
2) Historical validity can only ever be a means, not an end.

> ... apart from whom invented or discovered them first?

Weren't you just relying on things like "Printing was invented in Western Europe"? As if that's even true however. "Forget about the Chinese", he says. Pretty soon it's the Chinese who will be having us forget about you.

Rationality, formalization etc. are things that can have good or bad results, depending most of all on intent. And so-called "western civilization" can't be accused of having good intentions, historically speaking. Much of what you consider progress I consider not.


Rhythmus.be's picture

> Pretty soon it’s the Chinese who will be having us forget about you.

And I contentedly wait for that day when the Chinese will surpass our decadent culture, by being even more modernist. Still, they didn't invent printing as it is understood generally, and as it was (re)invented, refined, used and promulgated by the Europeans. The endeavours of the Chinese and the Koreans in printing with moveable types is an interesting anecdote of history, but nevertheless remains marginal in regard to the history of print and of present evolutions. The technology which the Chinese so eagerly are copying today (so that they soon will flood our markets) is the fruit of Western inventions.

> “western civilization” can’t be accused of having good intentions, historically speaking

By what standard do you measure the "goodness" of an intention? History and morality are two different things—and I am both against them ;-) I know, Western civilisation is but a euphemism for the pursuits of a bunch of capitalistic, macho warmongers. Nevertheless, and perhaps despite its intentions, it has brought us science without which most of the world's population would be starving even more, global communication, the abolishment of slavery and torture &c. You are free to not consider this as progress, of course.

Anyhow, as for type design, either you stay with tradition and the historical particularities of a given script, and so choose for calligraphy and chirography. Or you comply to the universal standards of a script system which uses moveable and reusable type forms that in order to set well must be economic and formal. In that case you will be designing type.

hrant's picture

The Koreans first invented printing with movable metal types. Period. And it's not impossible that Gutenberg (or whoever) got the idea from them. If that's true, should we start mostly making Latin type that looks Korean? Even if (or actually, especially) if Korea was trying to dominate the world, that would be stupid.

> it has brought us science

You did not invent science. You were living in trees when that happened.

> perhaps despite its intentions

In the end, nothing really happens despite intentions. World misery (you abolished slavery?! give me a break) is increasing under your watch. Under your directions. For most people the 20th century, your century, was the worst in human history.

> as for type design, either ... or

You should work for the White House.


William Berkson's picture

>Western civilisation ...it has brought us science

Modern science, correct.

The 'scientific revolution' (Copernicus, Galileo, Newton) drew on ideas particularly from ancient Greece (part of Western civilization) and Arabic mathematics (also part of Western civilization) which in turn drew from Indian mathematics (ideas developed before much Western influence, I believe). However, it was decisively new and different, advancing far beyond any of its antecedents.

hrant's picture

The Romans were the first true westerners.
The others, such as the Ancient Greeks and Arabs, were not western.


William Berkson's picture

In history of ideas, the Greeks are normally and correctly viewed as one of the founders of Western civilization.

This is particularly true in history of science. Archimedes is the closest to modern science of any thinker prior to the scientific revolution, and the translation of his works into Latin was a key ingredient in the new ideas of Galileo--who as I remember called him 'the divine Archimedes'--and Newton.

hrant's picture

No, you've simply co-opted the Ancient Greeks to make yourselves look better*. Referring to your own corrupt souces does not change the truth that the Ancient Greeks were fundamentally different in core philosophy than the Romans, and the Romans were fundamentally the same in that respect as the contemporary West.

* Just like you personally just tried to do with the medieval Arabs. Puhleez.


Rhythmus.be's picture

Hrant — What do you mean by continually referring to us as 'you'? Do you, US citizen from the west coast, not consider yourself as a westerner?

hrant's picture

I'm an Armenian mongrel.


William Berkson's picture

>personally just tried to do with the medieval Arabs

Yeah it's part of my personal plot against you to claim that our numerals are Arabic and that Algebra is Arabic. That's all a big lie I invented. And that Descartes combined the Arabic algebra and Greek geometry. And that Newton applied them both to Archimedes' problem of squaring the parabola and so invented the calculus. All part of my hopelessly Roman way of thinking.

And pigs fly.

hrant's picture

You tried to make the Arab's development of Math seem like part of western
civilization (go read what you wrote). Pigs don't fly, but they do try to tell
us what to think. You've chosen to believe them, because it validates you.


William Berkson's picture

>You tried to make the Arab’s development of Math seem like part of western civilization

It wasn't me; it's in every history of science. The development of Algebra from Hindu and *Greek* roots is during the Islamic era. And Islamic culture is Western. Islam itself was heavily influenced by Judaism and Christianity. And Islamic scholars studied and developed the ancient Greek philosophers when Europe had lost most of their writings. Those who don't understand history of ideas, like you, think that Islamic cultures are not part of Western tradition, but they are--they drew from both the Greek and Hebraic traditions in their earlier years, and in turn they heavily influenced Europe during the Renaissance.

Islam didn't participate in the European Renaissance, and so Islamic and European cultures started to diverge, but that's another story.

hrant's picture

History is written by the winners, and yours is written by you.
Neither Islam, nor Judaism, nor Christianity are Western.


hrant's picture

BTW William, any opinions on uuuh... Vangelis's font maybe?


Jongseong's picture

Vangelis, first of all it is great that you are contributing to the DejaVu project. A lot has been said in this thread that I could comment on, but let me just add my opinions about the font for now.

First the lowercase letters:

Of the variant forms of letters (gamma, zeta, nu, ksi, finial sigma, upsilon), I like the first ones in all the cases. Breaking the symmetry helps letter differentiation. I agree that the curves have to be touched up in alpha, and also in the zeta, ksi, and finial sigma. The upper-left serif of eta needs thinning.

I think I might prefer the theta opened up on the left, although it might be a challenge to keep a clean look in that form. The chi might work better with semi-serifs for the thicker diagonal. I am curious about the decision to shape the tau like you did, with the foot slightly dipping below the baseline and diverging from the forms of iota and pi.

Maybe you could make the angle of the acutos steeper when it combines with the trema. Right now the combination isn't completely happy.

As for the uppercase:

I agree that the crossbar of the alpha is too low, but I'm guessing that comes from the latin cognate forms from DejaVu Serif. Still, you could make the Greek forms subtly different from the latin. The serifs of the middle horizontal strokes of the theta and the ksi are a bit too prominent.

Overall, the text setting looks good, though I'll have to print it out to form a better opinion. Please keep posting improvements.

vank's picture

Good morning everybody,

many thanks for the great input Brian :-) This is just a fast post; I'll be back posting more later on. I just wanted to show you the already existing glyphs in the DejaVu Serif, designed sometime ago by another member of the project. The link is here, i've typeset exactly the same text for easier comparison and discussion:



Jongseong's picture

I've been thinking about the existing and new versions of Greek glyphs. Putting aside the issue of their design merits for a moment, I'm curious whether you (Vangelis) intend the new version to complement the existing one or to replace it.

I think both styles of Greek have their uses. In fact, I think maybe three different styles of Greek glyphs are necessary for full versatility.

Let me explain. Who needs the Greek glyphs in a multilingual open source font like DejaVu Serif? I think most users will fall under one of three categories:

1. Greek speakers.
2. Scholars who deal with Ancient or Mediaeval Greek.
3. Those who need the odd Greek glyphs now and then for technical symbols.

As you implicitly suggested, each category has its own rules of what is acceptable in the design of the Greek glyphs. It is possible to have a single set of Greek glyphs for all those purposes, but perhaps what is needed is a different approach in designing the glyphs for each case.

Many of the glyphs you designed will be inappropriate for use as symbols in technical contexts. In an English-language mathematics book , it would be useless to have an alpha that can be confused with a single-storey a, or a chi that is identical to x. International readers who have been taught just those canonical Greek letterforms (in most cases just a subset of the whole alphabet) will surely be left confused and clueless.

And as you mentioned, different styles of Greeks are appropriate for setting something like the Septuagint and the latest Greek newsmagazine.

So my thinking is that a multiscript font that aims for full versatility should offer up to three different styles of Greek to suit each need, maybe as Opentype stylistic sets.

hrant's picture

> I think both styles of Greek have their uses.

Exactly - and that applies to any non-Latin script. This is why Nour&Patria is a multilateral system, with a master design for each script, each with a subordinate in the other script (causing Modernists to consistently claim that it's "too complicated"...) On the other hand, it remains important to realize and admit which styles are better at what - for example that a Latinized Greek style is unsuitable for the dignified setting of a Greek book for example.


crossgrove's picture

I've been sketching a Latin lc set that is subordinate to Greek lc, trying (from my limited Latin POV) to reverse the approach. I'm pretending Greek is the standard and that I'm unfamiliar with Latin. Hrant, I wish I could see Nour/Patria (I can't find it on your foundry pages).

As Brian points out, there are different uses for Greek glyphs. There are different applications for type; if a Greek newspaper or book needs to refer to some Latin-based text, of course you don't want the Latin to visually clash with the Greek, but visual harmony is not the same as identical texture, color, etc. in every detail. Why would we consider it important to harmonize different scripts unless there were technical constraints? Hopefully we try to overcome those constraints whenever possible?

In setting multilingual documents in multiple Latin-based languages, the challenge is always to differentiate the material in the different languages, or at least organize the layout so a reader can follow one or the other language. Doesn't the different look of different languages, and different scripts, help this?

I can't think of a real functional reason to latinize a Greek typeface.

hrant's picture

> I’ve been sketching a Latin lc set that is subordinate to Greek lc

There are two (and really not any others that I
can think of right now) precedents I'd check out:
1) Cadmus, by Carter, from the 70s - on the left, with Skia, it's Latin:

2) Don't know the name, but Michail Semoglou's font. http://www.cannotnot.net/ _
I might also talk to Victor Gaultney, who has designed
multiple Greeks of the same design for different uses.
Here's something on N&P:
http://www.themicrofoundry.com/ad/CR.pdf _
BTW, William, if you're here, you might want
to read the bit about even color there...

crossgrove's picture

Oh, excellent, thanks for the PDF. I hadn't seen that before. I had seen Carter's earlier Greek, which I think he now has reservations about.

"In my desire to make a Greek face that was free of Latin influence, I succeeded in giving it an archaic look, unwelcomed in the westward-looking Greece of the '70s."

I suppose this is less a concern now. Clearly non-native readers should tread cautiously when attempting unfamiliar scripts. My attempts to adapt Latin to Greek conventions are pretty gratuitous; I'm just flipping the cultural [mis]appropriation, and probably not very well either; I'm too familiar with Latin. But it might have value elsewhere.

Jongseong's picture

I’ve been sketching a Latin lc set that is subordinate to Greek lc

I have toyed around with the idea. In fact, I've thought about a Greek uppercase set that is subordinate to Greek lowercase. No matter the weight of tradition and the lack of alternatives, I find the combination of latinate Greek uppercase and fully traditional Greek lowercase forms too strange to belong to the same script.

I can’t think of a real functional reason to latinize a Greek typeface.

One practical reason I've encountered for making Greek glyphs that are essentially part of the extended Latin alphabet is the International Phonetic Alphabet. Instead of having independent Latin and Greek scripts, one has a single script that uses Latin and Greek glyphs among others side by side. The question might then become whether that counts as designing Greek glyphs or extended Latin glyphs, but either way, one should make a 'latinized' Greek for the purpose.

hrant's picture

> I find the combination of latinate Greek uppercase and fully
> traditional Greek lowercase forms too strange to belong to
> the same script.

But this applies (almost as much) to Latin as well!

One key to this puzzle however might be... uncial.


Jongseong's picture

At least in Latin, the lowercase letters were given uppercase-style serifs, giving the two cases some visual harmony despite the disparity in letterforms.

On the other hand, I am still not fully comfortable with seeing Greek lowercase letters with added serifs, although this is exactly the kind of change that the Latin lowercase went through to be matched with the uppercase.

One key to this puzzle however might be… uncial.
That makes sense. As a theoretical exercise, we could go back to the point when the combining of combining uncial-derived minuscules with the traditional Roman majuscules was introduced, and think of a different solution to making the cases harmonize. Maybe I would have preferred that the minuscule letterforms dictate the combining majuscule forms, not vice versa, so that the resulting alphabet would have more of an uncial flavour.

hrant's picture

Concerning serifs:
1) Why not reduce them in the Greek UC?
2) Does "this is exactly the kind of change that the Latin lowercase went through to be matched with the uppercase" really make sense? The lowercase arose by writing the UC forms more quickly; the two cases didn't exist in parallel until much later, and the lc had serifs all along.


Jongseong's picture

1) Why not reduce them in the Greek UC?
An idea I had as well, and probably lots of others who have thought about this. I just wonder why we don't see that type of a solution more

2) Does “this is exactly the kind of change that the Latin lowercase went through to be matched with the uppercase” really make sense? The lowercase arose by writing the UC forms more quickly; the two cases didn’t exist in parallel until much later, and the lc had serifs all along.
The Latin minuscule had serifs, that is, in strokes and exit strokes that made sense when writing the forms quickly. But it gained inscriptional-style symmetric serifs to harmonize with majuscule forms. An uncial-based m might have just the in stroke and the one exit stroke, but an m with majuscule-style serifs has those foot serifs on all three feet. That's what I meant.

hrant's picture

> I just wonder why we don’t see that type of a solution more

Most non-Latin type development seems to be either highly Latinized or highly conservative. It's rare to see any understated innovation (which is actually the case with Latin as well).

Your serif history (Missgiggles, correct me if I'm wrong ;-) isn't exactly right, I don't think. Although one-sided serifs can be said to be more chirographic, two-sided serifs were also pretty much a result of writing. You're right that serifs later became "formalized" into totally symmetrical shapes, but that applies to the UC as well.


William Berkson's picture

>William, if you’re here, you might want
to read the bit about even color there…

Hrant, as you may remember, I do like your Harrier--which I believe is the latin component of Nour here, right?

Patria itself I think has problems. I believe you have written that you developed it by straightening Harrier, and I see you also greatly shortened the descenders. I think this approach created problems in Patria which were not fully solved, eg with the hook of the j crowding the baseline.

About the evenness of color of the two parts of each design I can't comment, not knowing Armenian script. As to color of the latin of Patria on its own, I would have to see text set at some length to judge it properly. But my impression looking just at the alphabet is, as I said, that there are unresolved problems.

How this relates to vank's effort on Greek, I'm not sure, but I think it does relate to the perils of two scripts--which in the case of Nour & Patria my impression is that one half of your project worked, and the other not so well, at least on the latin side.

hrant's picture

Harrier is the subordinate of Nour. Patria (which was indeed derived from Nour, but I admit has more issues) is the main Latin style shown in that PDF. I was just pointing out that at least some highly accomplished type designers have noted Patria's virtues with respect to even color, contrary to what you said about that yourself recently. But now you suddenly say "I would have to see text set at some length to judge it properly." As I've said before, your personal dislike of me causes you to lash out with semi-random negative criticisms, apparently not even bothering to remember them, as needed. This is not mensch.


William Berkson's picture

>But now you suddenly say “I would have to see text set at some length to judge it properly.”

I have said in the past both that I like Harrier and that I think Patria has a lot of problems with color. Since for once you actually addressed me as a human being instead of viciously attacking a bogy man of your own creation, I was responding in kind in a more gracious manner. Honestly I don't think I am going to change my view of Harrier and Patria by seeing more, but I am happy to see more text set in it, and am open to changing my mind.

Neither my admiration for Harrier nor my distaste for Patria have anything to do with my problems with your lack of civility on Typophile. I was just honestly reporting what my eyes tell me. Since you yourself think it has some issues, even though you may not use the same terminology, our eyes may be not that different ;)

hrant's picture

> I think Patria has a lot of problems with color.

And I was saying that the combination of you not really having seen it (by your own admission) and the positive opinion of the likes of Bruno Maag means that the integrity of your opinion on the color is suspect, especially in light of your obvious and often-stated personal problem with me, which has in fact caused lapses of judgement on your part in the past (again by your own admission, if not extending to the actual cause). It's OK, you're just a human being, like me - but your claim to mensch is arrogant.


hrant's picture

The main problem I myself see in Patria concerns certain UC serifs. There are others, like some of the spacing (although it's still better than average), the Pilcrow, etc. but the color I'm pretty happy with. Things like the weak top-right of the "k" and the crowded descender on the "j" are intentional, and I think -and the CR jury seemed to agree- mild enough. In contrast, your criticism of the color, especially in light of the fact that there's really no way to judge it without "really looking at it"*, is a self-validating reaction motivated by the fact that you don't like my opinions about even color. I see your stance as highly insular in the overall, so when you say "I don’t think I am going to change my view of Harrier and Patria by seeing more", you have no idea how much I believe you.

* Which is why type designers value super-hi-res output.

But, hope against hope, I still hope you see enough logic in my argument to let this thread get back on track. You could even offer opinions about Vangelis's design since, contrary to what you implied, it's very much possible to have useful opinions about fonts is unfamiliar writing systems (certainly in the realm of color for example).


William Berkson's picture

>you not really having seen it (by your own admission)

Have you forgotten that you sent me, at my request, the PDF you have now posted on Typophile? To me this sample is not enough to give a full and fair test, but as I said I doubt I will change my mind, especially as you yourself have problems with Patria.

I don't understand why you have to be so defensive. You got an award for Nour & Patria, and as I have said for a long time, I like Harrier. You said that J-F Porchez also has problems with Patria. So I agree with him and disagree with Bruno Maag. So what?

Why don't you make up your own mind, and finish the thing to your satisfaction and release it?

>but your claim to mensch is arrogant.

Hrant, when did I ever claim to be a 'mentsh'? I aspire to be a mentsh, and strive to be. It is up to others to judge how well I succeed.

Part of being a mentsh is what is called 'derech eretz', meaning having skill in personal relationships, including good manners and kindness. I don't claim to be a model in either respect. I don't know you, but frankly judging by your conduct on Typophile you are not the first person I would think of to consult about my progress in these matters.

edit: we cross-posted, Hrant. I'll just add that unfortunately I don't feel I know Greek script well enough to comment on vank's effort.

hrant's picture

> I don’t understand why you have to be so defensive.

It's annoying when I'm trying to focus on the untenability of your comments concerning color and you try to distract people with things like "I doubt I will change my mind, especially as you yourself have problems with Patria", even after I've stated that I think color is not one of the issues for me. You're weaseling.

> when did I ever claim to be a ‘mentsh’?

Dunno - when you bought the domain name?

> I don’t feel I know Greek script well enough to comment on vank’s effort.

But for some strange reason not knowing enough about Patria hasn't stopped you from commenting on its color, which also just happens to be one of the hardest thing to comment on confidently (or I should say capably, since self-confidence is not something you lack) in type design.

I remain unfooled.


William Berkson's picture

>Dunno - when you bought the domain name?

The first words on the site are "Welcome to Mentsh.com the site for those who aspire to be a 'mentsh' and to raise their children as 'mentshen'"

It is about ideals, with no claim that I personally always live up to them. It implies that I endorse these ideals and strive to live up to them, and that is true.

>for some strange reason not knowing enough about Patria hasn’t stopped you from commenting on its color

The latin part of Patria is a latin script, just as Harrier, the latin part of Nour is. I am not commenting on the Armenian, as I don't know Armenian script. I have seen both latin alphabets and as I have said like one and not the other. Get over it already.

vank's picture

Hi everybody and thank you for your contributions.

Hrant I totally agree with you, one should be aware of the trends and recognize and support the ones that actually are good. That’s why I did not choose a Latinized zeta or a Latinized final sigma on the design. Many thanks anyway. :-)

Brian, my sole intention has been to design a version for use as is. Whether the project members will eventually decide upon the replacement of the existing version or not is a whole other subject, yet I am definitely not designing something from which certain glyphs would be taken and compiled with existing to constitute the Greek version of DejaVu.

>Let me explain. Who needs the Greek glyphs in a multilingual open >source font like DejaVu Serif? I think most users will fall under one >of three categories:
>1. Greek speakers.
>2. Scholars who deal with Ancient or Mediaeval Greek.
>3. Those who need the odd Greek glyphs now and then for technical >symbols.

Now that’s great input! Sincere thanks for it. I have to admit that, so far, I was considering mainly the first category. It was good being reminded I should consider the third as well. Still, Greek glyphs to be used as technical symbols should almost always be italics, hence they should be separately designed. It's on purpose that I don’t refer to the second category by the way. There are already wonderful, freeware typefaces that perfectly serve such needs; the releases of the Greek Font Society. On the other hand, I, myself, was aiming on a general purpose contemporary Greek design that would be to some extend usable in a variety of applications, from on-screen reading to typesetting one’s thesis.

Regarding some other posts on the thread, I feel that there’s a tendency to forget that Greek language is not a museum exhibit; it’s a “living language” and, as such, subject to changes and novelties. Discussion on whether Greek glyphs should have serifs or not, is, to my eyes, equivalent to asking whether glyphs with accentuation marks (“tonos”) or lc should be included; were they used in classical Greece? No. Why use them now then, they were “novelties” introduced later on. In a similar manner, what’s the point of discussing on whether “eta” should have a descender? It could have yes, it could as well not have. Both forms can be seen in Greece nowadays with the latter (no descender) being way more common in typefaces designed by Greek designers. So, who knows if both alternatives will still be used in 30 or 50 years (I bet they will be)! So what? Today, is a an eta without descender equally or even more usable than an eta with descender in Greece? Most probably yes. Do we have open type? Can we include both etas? Why don't we and let one choose what they like?

The current version is the one shown

Quite some changes have taken place. I might have displeased someones with my "xi". I eventually chose a more "consistent" form, compared to the previous version. Some other glyphs have changed as well. The result is definitely far from being considered all smooth and fine-tuned, but i think it starts getting there.

hrant's picture

The Greek script is very much a living thing, and so is Latin. I myself have proposed reforms to the Latin script. So at least for me it's not about conservatism. But it is about sensitivity, to things beyond the whims of personal expression and mere style. So for example when you ask "what's the point of discussing on whether 'eta' should have a descender?" I would say that the point is to know the effects of such a decision, not least concerning the health of Greek culture as well as the readability of the resultant design. Also, you can't blindly do to Greek what people typically do to Latin - not just because there are structural differences but because the users of Greek and Latin need different things, hence the scripts need different things. The case with Greek isn't so dramatic, but for some scripts -like Armenian- the danger of eventual total assimilation into larger cultures is a very good reason to be extra careful and not give in to expressive whims, instead channeling one's expression to help one's culture, not to mention make something functional, not just abstractly pretty.


Jongseong's picture

Vangelis, I like the changes you made. Specifically in the lowercase, the chi is more dynamic, the descenders look better without serifs, and the letterform for beta is more in accordance with the rest of the alphabet.

The lowercase theta needs to resolve some dark spot issues, perhaps by making the middle stroke a touch thinner.

Hrant, you say it's not about conservatism but sensitivity. I share those reasons, so I guess our differences in opinion (I don't find the supposed latinization in the design of Greek minuscules as offensive) come from different ideas of cultural authenticity.

Attending to cultural needs is important, but most often such an intention finds the easy expression in the avoidance of any similarity with the product of another 'threatening' culture. A good topic that gives many fascinating examples is orthography reforms in European alphabets, for example in Czech where differences from Polish orthography were deliberately introduced to emphasize cultural difference. This tendancy is not a bad thing in itself, but if it becomes prescriptive so that any resemblance to a product from a threatening culture becomes reason enough for its rejection, then you're overdoing it.

I think Hrant won't accept what he perceives as latinization to be culturally authentic. For me, the fact that such a style of Greek type is widely accepted as standard for Greek speakers is enough to accept it as culturally authentic.

By the way, I've seen Greeks who don't like the 'latinized' style (especially oversees Greeks), but have never found a non-Greek who approved of it (not that I've seen that many people express an opinion on this matter). I think it just goes to show that Westerners judge Modern Greek culture with the huge expectations coming from idealized notions of Classical glory, and tend to label anything that fails to conform to such notions as devaluations. This is an unfair burden for Greeks. I'd rather trust them to do what they choose to do with what is in fact their culture and not hold them to unreasonable expectations based on what we think they should do.

hrant's picture

The issue of cultural authenticity is hyper-complex, and it's not surprising to find large differences of opinion even among people who care about nurturing it. In my book simply caring about it is already a big deal. Even people who violate it aren't necessarily bad, as long as they know they're doing it - it's the total oblivion to it, which is sadly very common, that makes me angry. For example in type design when somebody copies formalistic attributes from one script to another for the sake of "consistency" with no regard -or even disdain- for much else, that's Bad.

> most often such an intention finds the easy expression in
> the avoidance of any similarity with the product of another
> ‘threatening’ culture.

Indeed, and this is a poor substitute for the intelligent preservation of cultural identity. But your placing of "threatening" in quotes can be seen as dismissive of very real threats out there, sometimes in fact documented (such as with Benjamin Franklin, who actively thought of ways to assimilate Jews into US culture) but much more often covertly, or "subliminally".

> the fact that such a style of Greek type is widely
> accepted as standard for Greek speakers is enough
> to accept it as culturally authentic.

Conscious public perception is undoubtedly a huge element in cultural autheticity, but since so much (I would say half) of our reality is subconscious I think this attitude is superficial. It's really much more complex than that, and we shouldn't succumb to the allure of over-simplifying it.

Also, I personally think that authenticity can be over-played. As I said I'm not a conservative, so to me more important that what's been authentic is what should be authentic. Anything out there plays a part in the future - and that's where we're headed! In the end it's not a culture itself that's important, it's the users of that culture, and the richness they contribute to life.

> have never found a non-Greek who approved of it

Although I certainly agree that the "classical burden" is out there (especially for Greeks) and is unjust, this is still no excuse for knee-jerk reactions (e.g. blind Latinization) like we saw in the 90s against Modernism (reminding here that I can't stand Modernism). Also, there certainly are non-natives who approve of Latinization, like Ludwig right here and Jan van Krimpen more famously. And if you read Morison's intro to Schonfield's book about reforming Hebrew, you might share the shiver in the spine that I felt.

I also don't agree with the them-us dichotomy. We're all part of this world, and everybody is at least a tiny bit Greek. We should all care what happens.


Syndicate content Syndicate content