Re: BSR CM type 1 arrows, StMaryRd, and RSFS

[bbeeton <BNB@MATH.AMS.ORG>]

i wrote:
    many of our authors have *very* *strong* opinions about the shape of
    script letters, and rsfs is closer to their concept of script than
    is anything else readily available.  (but it's a real beast as far
    as placement of subs, sups and diacritics; that's why both knuth's
    calligraphic and the euler script are significantly more vertical.)

berthold responded:
    Hmm, interesting makes it sound like it ought to be treated as a math font
    then so that one can use the bogus metrics used in math fonts to 
    position subscripts and superscripts.  Of course, that does mean
    using up yet another math family.  Does it make sense to squeeze it
    in with another math font (and then have to translate character codes
    from A-Z to wheever it has to fit in the remaining space)?

but it already *is* being treated as a math font with all its special
metrics.  the problem is the shapes of the letters and the excessive
slant (visually more, if i remember correctly, than the usual math
italic).  there's just no way to gracefully put a hat on a flowery
script cap L.  there have been some attempts to develop other, more
easily manipulated, script alphabets, but mathematicians are used to
the old wedding-invitation script and tend to rebel at anything else.
i've been involved peripherally in one such project, and must report
that
 - it's non-trivial
 - there are only so many ways to shape a script cap I to make it
   intelligible and still distinct from the italic cap I and not
   fade away in second-order superscripts.

i developed a set of desiderata for a math script alphabet that i'd be
willing to post if anyone is interested.
						-- bb