Re: Greek letters not slanted in -Tps eqn output

[joerg van den hoff]

On 07.08.22 14:52, Robert Goulding wrote:
> Using groff 1.23.0.rc1.2754-d5862
>
> Greek letters are not slanted in equations in Postscript output, but they
> are in -Tpdf output. Simple example attached.
>
> This is perhaps connected with recent activity about the SS font?
>
> Robert.

it seems the same principle issue to me. can you confirm that something like 
this

.LP
.EQ
alpha beta gamma delta rho sigma 1 over 2
.EN

when formatted with grops (-Tps) yields not only unslanted greek letters but 
also leads to
misalignment in the 1 over 2 fraction?

the problem really seems to lie with grops rather than gropdf:

when formatting the above fragment with grops and using `pdffonts' to list the fonts used in the 
resulting pdf (after pstopdf...) I get

YDINBV+Times-Roman                   Type 1C           WinAnsi          yes yes 
no       9  0
ATLOAE+Symbol-Slanted                Type 1C           Custom           yes yes 
yes      7  0

(where times-roman is the font used for the digits in that "equation"). the 
issue here is, that
symbol-slanted is listed, but the pdf shows unslanted greek letters (and 
misaligned 1/2 fraction).

for the gropdf formatted output it reads

Symbol                               Type 1            Custom           no  no  
no       5  0
Times-Roman                          Type 1            Custom           no  no  
yes      8  0

which is OK since gropdf in fact *does* use the symbol font (but explicitly 
slants the letters).

my (limited) understanding is that for grops, too, the listed fonts are as they should be (grops 
*is* supposed to use SS for the equations) but that the produced pdf output is "wrong" (as in:
unslanted letters, wrong metric information for greek letters used (I guess) 
and consequent
misalignment of elements to the right of the greek letters (the 1/2 fraction in 
the example).

this seems badly wrong, no?

best,
joerg