Welcome Filippo! Even I am a beginner and I am actually confused by what you have said. I think you are contradicting yourself here

I don't know, I feels like Bezier curves are not properly really definable in every single smallest aspect.

Don't you think simple geometrical shapes are more constricting than bezier curves?

They seems...inaccurate compared to the strength of a circle, of a simple geometrical form.

What do you mean? Inaccurate is being used very loosely here. Circle is strong but there is very little which can be achieved with straight/regular shapes. Even the most geometric typefaces are not perfectly geometric. There is a slight variation in curvature or slant. Beziers are very powerful and the most complicated shapes can be achieved using beziers.(IIRC, I remember someone drawing a girl using fontlab bezier curves)
Cheers
Typo

Bezier curves are geometric.
In fact, all digital fonts are geometric.

Nick, that's true, but to me Filippo is essentially alluding to simplicity. A line is simpler than a circle, which in turn is simpler than an oval, which is simpler than a bézier - that sort of thing.

The problem Filippo is that simplicity is a matter of style, not function; the way humans perceive things (especially when in "deep reading") is not at all simple. Here's a very primitive but telling example: if you make a mathematically perfect square where all the sides are mathematically equal in thickness you'll end up
with something that looks a bit oblong and with horizontals that look thicker than the verticals.

An individual might personally prefer things to look simple, but that doesn't lead to Good Design (which is about helping many people, sometimes without them realizing :-).

hhp

You may find this helpful: http://www.typophile.com/node/96309

In these days I analyzed the situation and I think I arrived to a good theorem (but I want to know if you agree). In type design there are three ways of creating a single glyph:

1) Free hand (with a specific calligraphic pen, a normal pen or a pencil): it is a very versatile system but, on the other and, it is also very inaccurate.

2) Geometrical basic figures on a grid (before the computer with compass and rulers on graph paper): a system extremely precise but not so versatile (theoretically all glyphs can be projected only by a series of circles and lines but...in most cases this way is too long and complicated).

3) Bezier curves (before the computer with series of plastic or wood curves): I think this is the way in the middle, quite precise and quite versatile and it is probably for this reason that is the most widespread.

I personally prefer way number 2 but, recently, I realized that is not always applicable mainly because of problems of time. So sad!

@LexLuengas

You may find this helpful: http://www.typophile.com/node/96309

I was just talking about this...thank you very much! So I am not the only one wanting to give a strictness to Bezier curves and their nodes... :-)
The question is: is really possible? I hope to find the answer reading this article!
Thank you again!

Shouldn't this be in General Discussions?

For fonts based on a grid, you might want to check out Fontstruct:
http://fontstruct.com

Filippo, it's good you're thinking in such terms. But there are more ways! :-)

Here's a really strange one: http://typophile.com/node/31095

hhp

For example, yesterday I tried to project by hand a small set of glyphs with the method of plastic curves and pencil. In Adrian Frutiger's pencil drawings I saw he was used to specify the thickness of the lines of the glyph and so I tried.
This is the result (italic 15°): Photo

I think the final result is good but:
1) the thickness specified by numbers (as you can see from the photo) isn't exactly equal in every single point...it is not possible! Where I wrote 12 (in the top part of the "a") is not really every time 12 mm but can be 12.01, 12.1, 12.05, ecc...
2) to join two points with a set of plastic curves we can use millions of different curves. If I drew the glyphs by free hand would be the same thing but the lines have been more dirty.

So can the design of a glyph be considered geometric? I think, in this case and in a lot of other, no...because is more a matter of eye rather than of strict proportions and because geometry is unconditionally precise and this glyph is not. Don't you agree?

Sorry if I stress you with this sort of questions but you have to help me killing my mental block...and this can be done by reasoning. I'm a square and want to be a curve!

The secret of good design is this: Geometry is shit*.

If it looks good, it is good. All the Victorian analogies comparing the body to a machine and more modern ones comparing our minds and perceptions to computers, cameras, or – whatever else, have to be put aside. The body and the mind are a hopeless mess by the measure of such mechanical and digital models. Our perceptions are imprecise, evolved to make us really good at hunting and foraging and spotting tigers in the brush, although since we are social animals with brains, they aren’t nearly as good as the tigers. That we are even capable of distinguishing a circle from a square has to be a byproduct of those skills.

Our eyes are affected by all kinds of peripheral visual noise to the point that, in the context of a string of characters, circles don’t necessarily look round. You have to correct them. Vertical strokes look thinner than horizontal ones of the same thickness. You have to correct them too. Identical features can look different and different ones can look identical.

*I lied. Geometry isn’t strictly shit, but strictly glyphs made up of replicated un-modualted basic geometric shapes . IMAHO.

The other secret to good design is that being good at drawing and design is no big deal. It's all just just geometry after all. :o)

Urban Dictionary: Imaho
www.urbandictionary.com/define.php?term=Imaho
In My Absolutely Honest Opinion, Similar to IMO and IMHO, though it also has the second meaning: "I'm a ho".

In My (Always) Humble Opinion.

Get your mind outa the gutter. I'm very humble and proud of it.

Ha ha. Reminds me of Golda Meir:

“Don't be so humble, you're not that great.”

@russellm

IMAHO

Opinion is important for a person who has your point of view! :-)

All the Victorian analogies comparing the body to a machine and more modern ones comparing our minds and perceptions to computers, cameras, or – whatever else, have to be put aside.

This is the key. The eye is all.
Probably geometry in typography can be compared more often to a limitation rather than an aid. But can also be a refuge for uncertain people.

@Renaissance Man

“Don't be so humble, you're not that great.”

Haha, nice quote!!!

Hi Filippo, I think in typography you should focus on optical proportions rather than on geometrical ones.
I paste here a though I’ve written some times ago on my facebook page.

Typeface preferences VS degree of typographic skills

•••
ABSTRACT
“the less you’re experienced in graphic and/or type design, the more you love strictly geometrical fonts”
•••

It’s strange how the appreciation for a font features may vary with knowledge and this trend seems to be systematic.
Observing around me — not only among professionals — it seems the less you’re experienced in graphic and/or type design, the more you love strictly geometrical fonts with round shapes.
So younger designers and non-professionals will tend to love fonts like Futura, or other round-like geometric fonts; they would prefer Didot to Bodoni as more minimalistic, etc.
Me as first, I loved Didot but found a bit difficult to use on small sizes.
Today, analysing modern (=modern means 18th century) typefaces, I discover how Bodoni’s construction, balance and optical details are much more readable than Didot making of it a typographically better font the Didot itself.

But the fact persist: the more you dig in typography, the more your taste and expectations drive towards the opposite path than “common people”, to read as clients, opinion leaders, but also graphic designers without solid typo background.

Of course you can say there are two audiences and you must address each one in a different way, but when you’re working on a visual identity project and have to choose a typeface, it can find yourself in a paradoxical situation.

But what about all the laymen who love script fonts?

hhp

it's another case, my though focused only on typographic characters (not calligraphic)

I'm so experienced that my anti-geometric tastes have evolved to the point that I am finally starting to favor Papyrus.

… over Comic Sans?

It’s strange how the appreciation for a font features may vary with knowledge and this trend seems to be systematic.
Observing around me — not only among professionals — it seems the less you’re experienced in graphic and/or type design, the more you love strictly geometrical fonts with round shapes.
So younger designers and non-professionals will tend to love fonts like Futura, or other round-like geometric fonts; they would prefer Didot to Bodoni as more minimalistic, etc.
Me as first, I loved Didot but found a bit difficult to use on small sizes.
Today, analysing modern (=modern means 18th century) typefaces, I discover how Bodoni’s construction, balance and optical details are much more readable than Didot making of it a typographically better font the Didot itself.

But the fact persist: the more you dig in typography, the more your taste and expectations drive towards the opposite path than “common people”, to read as clients, opinion leaders, but also graphic designers without solid typo background.

Of course you can say there are two audiences and you must address each one in a different way, but when you’re working on a visual identity project and have to choose a typeface, it can find yourself in a paradoxical situation.

I do not totally agree with the thought according to which a young and unskilled graphic designer is used to chose a geometry based typography. I think instead that his choice would fall much more easily on unconventional ones, independently if based on geometry or not.

The point changes if we talk about the real design of a typeface. Probably the young type designer, that decides to create a typeface, opts for a geometrical one because, during the construction of single glyphs, he will rely not only on his eyes, of which probably he still do not totally trust, but also on well defined geometric rules.

The more he will practice in type design, the more he will trust in his eyes and the more he will leave geometrical path (IMHO of course).

For administrators: Not sure "Type ID Board" is the right place for this post. :-)

@Filippo: You can use building elements if it helps you get into the forms, even basic geometric figures, but in the end you’ll see quickly that a well-drawn form, no matter how constructed (rather than drawn, or casually handwritten), needs complex curves.
This does not mean complex curves can’t be produced simply: it just requires a visual training, attempts and hard work.

I realized this very early, as my very first experiments were alphabets built from modular forms, but I did not like them (right now I see that experiments with basic geometric parts were fascinating because it was difficult to attain them by hand: now that we are in the opposite situation, they are less interesting, although it depends a lot on *what* you are trying to convey with your alphabet).

Filippo, your observation on youth is bang on.

Novice type designers may well have sophisticated ideas about design in general, but it takes time to develop the special drawing skills needed to engage type design—in which the past and its masters are so evident— at a deeper level (although this is not necessarily more useful to typographers!)

As well as geometrical forms, other viable area for novice font makers are hand-drawn styles, distressed/filtered styles, and revivals.

The first digital font I made was Fontesque, in 1994. It was drawn during the Grunge/Deconstruction era; at the time I already had some skill at vector drawing, through 5 years using Illustrator, but knew dick all about how important hinting and alignment zones were to the Type 1 PostScript format. Fortunately, such knowledge was unnecessary in that kind of bouncy wonky typeface!

I think instead that his choice would fall much more easily on unconventional ones, independently if based on geometry or not.
You're right but I didn't even considered this case :p

As written, it's coming from the observation of design student works, vernacular designs, and my own past experience.
Maybe it's statistically irrelevant, but I feel in graphic design there is a trend for simplicity; geometry is an answer to that.
So it happens graphic designers, when looking for a font, search the same pure geometric simplicity. As it's really time taking to learn to look at letter shapes (I'm still a junior in the field but realise very well how hard is it) for an inexperienced (in typography) graphic designer, it's really difficult to feel the text-line rhythm is not perfect.
On the other side, as a typographer you know a purely geometric font can't be good. However it can "look as" geometric; if you compare lettershapes at 2000px size you'll see it isn't (BTW one of the reasons Gotham sells so much?)