Comments on _Réflexions_ so far

[Robert Keeble <RKeeble@xxxxxxxxx>]

Bonjour à tous,

In this email and future postings on _Réflexions_, I'll note the page number
with my comments so it's easier to refer to the text, and in case I have
misunderstood the original.

p.6 "Obliques are not italics"
Yes, I agree that obliques masquerading as italics is troublesome, and it's
high on the list of complaints here at Quark. Some users would like to see
the 'i' button disabled if an italic face is not available in the font
family. I wonder if the 'i' button should change to an oblique 'o' if a real
italic isn't available, and when the caret is in text with a new "oblique,
not italic" attribute?

p.7 Small caps
The idea of synthesizing these with multiple master fonts when *true* small
caps are not available is interesting, but how many people like and use MM
fonts? I realize the support for them in XPress is rudimentary. Something
that might help this small cap/expert font problem would be to extend the
scope of fontsets (for any who have used the East Asian XPress), which is
essentially a virtual font that allows specifying harmonious "real" fonts
for digits, alphabetics, symbols, kanji...for creating Japanse/English
documents, but the idea could also be useful for getting some of the basic
features in OpenType like small caps. And it would also preserve your
current investment in fonts, while we wait for more OpenType fonts with
advanced features to appear. It could also be useful for setting math, where
you might have fonts of symbols combined with compatible text fonts.

p.8 Superscript/subscript
It sounds like the conversation is mainly about typesetting formulas, and I
get the same feeling about the passage discussing arbitrary ranges of text
that can be freely repositioned -- that sounds like formulas to me. If it
is, I wonder if it wouldn't be better to work out an interface tailored to
typesetting formulas? But I guess that is coming later in « Typographie
complexe »..."Read on, read on!"

p.11 Crénage
The point is made that kerning doesn't work between different fonts, or even
different sizes in the same face. But why is automatic kerning still
desirable in this case? Aren't the size and font changes themselves
disrupting the appearance of the text?
Regarding algorithmic kerning, I have to agree with F.H. Villebrod -- I'm
not sure automatic kerning will ever be good enough to satisfy « l'oeil du
typographe ». I'm not even convinced we could improve the situation
significantly more than the approach OpenType is taking, that of allowing
kerning of 3+ characters. Oddly enough, Unicode will force this type of
kerning as well, since we need to be able to adjust stacked combining marks.
If you can kern a base character and two accents, then I guess you can do
two letters with a space in between. :^)

p.13 (F.H. Villebrod) On devrait même pouvoir enregistrer des tableaux de
crénage différents selon le corps utilisé...

So the kerning does not scale correctly and tracking doesn't help either?


--- I've read farther than this, but this email is already getting long, so
more later


Rob Keeble
Quark, Inc. | Desktop R&D | Text & Typography