[Fontinst] fontinst'ing type1 on miktex : next problem

[Lars Hellström]

At 05.30 +0200 2003-05-27, steven schkolne wrote:
>hi, my last problem was solved by adding missing .afm files.... i should
>make this public since i'm not the only one who ran into it....
>
>http://groups.yahoo.com/group/miktex/message/11527
>
>but anyway, i have moved one step further and am greeted by another
>confusing error (documented below).  can anyone help me?  it seems to
>not like my input files, which i got directly from adobe.  when i hit
>enter to bypass the error it comes up again and again.... hundreds of times
[snip]
>> INFO> run \installfont <pbhb7t><pbhb8r,latin><OT1><OT1><pbh><b><n>
>> (pbhb8r.mtx
>> ! Missing number, treated as zero.
>> <to be read again>
>>                    \g-Z
>> \scale ...val_expr {#2}\global \multiply \result b
>>                                                   y\a_count \global
>> \divide ...
>>
>> \eval_expr #1->\global \result =#1
>>                                   \relax

This is indeed a bad error, and probably not one which a user can be
expected to debug on one's own. OTOH, there is always the option to ask on
the fontinst mailing list, as you have done.

Right now, I can think of two possible reasons for this error:

1. Somehow your initial problems gave you a bunch of corrupt MTX files.
Delete all pbh*.mtx and pbh*.pl, then try again and see if things are
better.

2. The AFMs may not look as fontinst expects them to (if this is the error,
then chances are about 50-50 whether the "bug" is with fontinst or with the
foundry). In particular, if the block of CC (composite character) lines
comes before the block of C (base character) lines then fontinst's AFM
parser may be too stupid to generate correct MTX from it. In this case,
hand editing the AFM files should solve the problem.

Lars Hellström