[Fontinst] A classic: SC & OSF sets

Lars Hellström Lars.Hellstrom@math.umu.se
Mon, 13 Jan 2003 12:03:35 +0100

At 14.41 +0100 2003-01-12, Philipp Lehman wrote:
>On Saturday 11 January 2003 18:45, Lars Hellström wrote:
>Lars, thanks for you reply.
>> Getting everything right with SC&OsF fonts is usually a job for
>> \reglyphfont.
>The problem is that I'm trying to write a recipe that will work for
>version 1.8 and \reglyphfont is in the prerelease branch only.

_Hopefully_ it won't stay prerelease that much longer, but since I haven't
been that good at getting things out in the past this is probably a
reasonable precausion.

>So I
>guess this means that there is no way of getting a SC & OSF set right
>with 1.8 short of renaming each and every glyph individually and
>adjusting the kerning tables?


>Let me check if I understand how \reglyphfont would be employed in
>this case, though. Instead of:
>  {T1}{psbj}{m}{n}{}
>I'd go like this:
>  \input csc2x
>  \reglyphfont{psbrc8x}{psbrc8r}

Right so far.

>  {T1}{psbj}{m}{n}{}

That should rather be


since the glyph definitions in psbrc8x is for zerooldstyle, oneoldstyle,
and so on.

>Is this correct? What's the difference between csc2x.tex and
>csckrn2x.tex anyway?

csckrn2x.tex throws away ("kills") the glyph definitions -- it only keeps
the kerns (where the glyph names are converted). It's for cases where you
can take the SC glyphs from an expert font, but need to get the
capital--small capital kerning pairs from an SC font (in the same family).

>But actually, the mechanism used under the hood still involves
>renaming each and every glyph, right? The difference is that
>csc2x.tex will do that for me and the \reglyphfonts supports 'batch'
>jobs, so to speak.

There is a difference in that the other method creates "copies" of glyphs
when fontinst is building a glyph base. \reglyphfont creates a new MTX file
which originally defines the glyphs under different names. The VF produced
may however be the same in both cases.

Lars Hellström