[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: newmath.dtx patches


At 11:56 AM 98/01/21 +0100, Ulrik Vieth wrote:

>I've noticed, however, that he forgot to add the size scaling code
>from mathtime.sty (lots of \DeclareMathSizes) and the correponding
>adjustments for the \delimiterfactor.

Because Times and MathTime are not optically scaled it is very
important that one not use the default 10/7/5pt scheme of CM.

>Some notes on the size scalings:  While mathptm goes 10/7.4/6 
>in \normalsize, it goes 9/7/5, 8/6/5, 7/5/5, for smaller sizes.
>In MathTime, however, the scaling is 10/7.6/6, 9/7.6/6, 8/6.8/6, 
>and it never goes below 6pt, even for a 5pt typesize.  

>For sizes below 10pt, e.g. \small and \footnotesize, the relative
>size of first-level subscripts and superscripts is about 85% while
>it is 76% at 10pt and goes down to 70% and 50% for the huge sizes.
>I really wonder if there is a principle behind all this, or is it
>just a set of good values derived by trial and error.

I made up sizes when I noticed that the CM scaling of 10/7/5 was
much too aggressive.  If you look at an optically scaled font like CM
you'll see that the horizontal scaling is much slower than the
vertical scaling.  I took as a guideline the square root of the 
~area~ ratio rather than the vertical size ratio (which is what
design size reflects) in order to partially compensate for this.

Beyond that, it was trial and error and numerology (i.e. I didn't really
want to have 9.345/7.523/6.189 :-)  Also, there is a limit below which
text becomes hard to read particularly on laser printers so I had
it bottom out at 6pt instead of CM's 5pt.

This was in the support for plain TeX, LaTeX 2.09 and LaTeX 2e that
we made up.  I believe Frank tweaked the numbers some more for
the poltiically correct LaTeX 2e support later.  If you find better numbers
that follow some magic system of equations all the power to you!

>- Should the mathpm, mathematica and mathtime sizes scalings 
>  be made all the same, since they are all based on Times Italic?

>- Should all layouts with nonstandard size scalings set 
>  \delimiterfactor, so that \big remains \big and doesn't
>  suddenly become \Big when switching the math layout?

Regards, Berthold.