[Fontinst] T2A Minion

[Lars Hellström]

At 17.43 +0200 2003-09-04, Ulrich Dirr wrote:
>I thought I've done the following steps:
>
>1. transform Minion-Regular to raw font with 8r encoding:
>   \transformfont{pmnr8r}{\reencodefont{8r}{\fromafm{pmnr8a}}}
>2. transform MinionCyr-Regular to raw font with T2A encoding:
>   \transformfont{pmnrz}{\reencodefont{T2A}{\fromafm{pmnr6w}}}
>3. install font with metrics of raw T2A encoded font, cyrillic.mtx,
>   raw 8r Minion-Regular, raw Minion Expert font, latin:
>\installfont{pmnr6a}{pmnrz,cyrillic,pmnr8r,pmnr8x,latin}{T2A}{T2A}{pmn
>x}{m}{n}{}
>
>Is this correct? Maybe the sequence isn't ok?

You probably should make up your mind as to whether you want to make a
virtual font or not.

If you choose to not make a virtual font, you have to reencode the raw font
to the desired encoding and that font furthermore has to contain all the
glyphs you want to use. Point 2 above works in that direction.

If you choose to make a virtual font, then you can combine as many base
fonts as you like into it. Point 3 above works in this direction, combining
the base fonts pmnrz, pmnr8r, and pmnr8x.

As there is no need to match base font encodings to the final encoding when
making a virtual font, one instead usually tries to expose as many glyphs
as possible from the base font, even if that means the glyphs aren't in
such positions that they are immediately useful. Compare 8r (raw encoding)
with T1 and TS1 (the recommended encodings for typesetting). In your case
here, it would probably be more useful to reencode pmnr6w to 6r or the X2
encoding than to T2A, as that could also be used as base for T2B and T2C
encoded virtual fonts.

Lars Hellström