Some glyph-shapes look weird to me...

[Simon Budig <Simon Budig unix-ag org>]

Hi all.

Ok, this is most probably too late, but I thought I would speak up - at
least this prevents me from being annoyed at myself for not speaking
up...  :-)

I recently discovered the beta version of the Vera Fonts and am quite
happy with what I have seen. I recently did a talk and used them for
the slides...   :-)

However, at the very first slide I noted my email address
"<simon gimp org>" and found that the glyphs ">", "<" and "@" look a bit
weird to me. The following comments apply to Vera Sans Roman, most also
apply to other fonts of the family.

The greater and smaller sign are very pointy and kind of stick out.
Compared to other fonts they do not seem to "embrace" the email-adress
itself. They might work pretty OK for typesetting of math, but I'd guess
that they are more commonly used for highlighting stuff like URLs or
EMail-adresses. Note that the math typesetting thingie would not break
if the glyphs would change their appearance to be a bit more open...

The AT-glyph is weird, because it uses a totally different "a" shape
than the regular small letter a. The "a" used within "@"  basically
is an "o" with a vertical bar next to it, while the "a" in the font has
a more classical shape. Also IMHO it would not hurt if the baseline of
the internal "a" of the "@" would be a bit more close to the generic
baseline. Also en- or disabling hinting changes the appearance a lot,
without hinting there is a big variation in the stroke width where the
"o" meets the vertical bar.

The last thing I noted is, that (at least with freetype 2.1.2, which
comes with debian testing) the hinting influences the rendering pretty
much, even in very big sizes. The rightmost stem of a "m" displayed at
the size 64 points moves for more than one pixel, when hinting is
toggled. Sometimes this changes the inter-letter space and makes the
word look weirdly typeset, while in the unhinted view the spacing looks
perfect.
 The internal "a" of "@" even shrinks one pixel the bottom
*and* on the top. At 59 points the internal storeys (?) of the "g" and
the "o" becomes much more "circular" when hinting is active.

(I tested this using ftstring, "h" toggles the usage of hinting or not).

Sorry, that I did not discover the fonts earlier, since we are already
in march I do not expect you to listen to me, but maybe some of the
points mentioned above could be discussed and fixed in the font if you
agree to me.

Bye,
        Simon

-- 
      Simon Budig unix-ag org       http://www.home.unix-ag.org/simon/