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

Re: mathfont 0.53

Matthias Clasen:

> I'm not sure if my mail from this morning came through, so I say it again:
> I have put another version at

> ftp://logimac.mathematik.uni-freiburg.de/logimac/Dokumente/www/mixed/mfnt0_53.tgz
> http://logimac.mathematik.uni-freiburg.de/mixed/mfnt0_53.tgz

No, it didn't come through this morning, and I also got one message
bounced from your address a few minutes ago, so I wonder if there's
anything wrong at your side.

> The major new additions are:

> - concrete and mathptm versions (due to Ulrik Vieth)
> - monolithic documentation in bigdoc.dvi
> - new fontdimensions for script and scriptscript sizes and 
>  for math spacing together with macros to use them. This is the part
>  that causes the rounding errors mentioned above. It would be good if
>  someone with better NFSS knowledge would have a look at that code.

If you look through the mail archives, you may find that it was
discussed to death back in 1993 that you cannot communicate such
constants as the script and scripscript sizes through fontdimens, 
because the fontdimens get scaled along with the fonts.  
Another point is that the math symbol and extension fonts already 
have 22 or 13 fontdimens with a fixed meaning, so if you try to
encode the math spacing in, say, fontdimen 23, TeX may incorrectly 
assume that there are valid values for the first 22 fontdimens,
even if those weren't set.  I'll have to see the actual code, 
I guess.

Cheers, Ulrik.