You're crazy. I like the idea, but I can't read it very well. Maybe tone it down a bit? I think tapering into the missing areas would help.

you might look up Miniscule Deux for some ideas.

> To me it could well be cuneiform.

Exactly, and we all know that the Sumerians used very little ink. Great work, Ruudi!

What does the date have to do with this? It's brilliant
on any day. I'm especially struck by the numerals: they
are the epitome of Good Design, reminding me of the "d"
in Peignot, which was wisely left in its original sorry state.

hhp

I don't think Gosselin analyzed numbers in his study, so there would be no scientific basis for redrawing the numbers.

At last, a true breakthrough in reading science!

Thank you all for your comments - I agree the font can get better with more science! I await the feedback of Microsoft Kevin Larson, and as Missink, based on Arial, I ask for review by Professor Allan Haley and Sir Rob Nicholas of Monotype LTD.

The initial image at the top, labelled "Data", shows in its first column a set of letters where the areas omitted from the typeface are more strongly blurred than the rest of the letter.

If I knew where that first image came from, and how it was generated, I could more seriously critique the specifics of this typeface.

Of course, the date of original posting is relevant, as the actual typeface is totally unreadable, thus wasting ink instead of saving it. (Although texts consisting solely of numbers would be readable, but then no savings in ink would be realized over unmodified Helvetica or Univers.)

But even if it's an April Fool's joke, one could perhaps produce something useful out of it. An ink-saving typeface made of dots was recently proposed. If, indeed, the areas left in this typeface are those most crucial to legibility according to some legitimate source, then one could just put the rest of the letter back in, but screened to a light grey - and, presto, the typeface isn't a joke any more.

Ruddi, a message to you Ruudi,
I’m a fan of the Gosselin work on which this is based, and I think this is a very promising first attempt on using that data. Your font is missing the low frequency information that is adding the location of the letters, but I think exclusively preserving the high spatial frequency is surprisingly successful. I’m looking forward to hearing about how well it reads. Since Missink is based on Arial, it would be interesting to compare it against Arial in a performance test, but since the peanut gallery will roast you for that, I recommend comparing it against a REfont.

This work is brilliant, but about 500 years too late.

— Frank U.

Mars: I wonder how that experiment would do on letterpress.

@hrant:
It's a DVD of whitespace clip art.

As long as you avoid audio files, I suppose you don't have to worry about John Cage's estate suing for copyright infringement...