strange PK problem with a lilypond font

[Han-Wen Nienhuys]

address@hidden writes:
> 
> [CVS 2002-08-18 09:13]
> 
> There is a serious bug in the fetaXX fonts: The metrics are
> resolution-dependent!  As a consequence, feta16.600pk anda
> feta16.1800pk and its tfm files, to take an example, have different
> checksums.
> 
> A diff on the tfm files (using tftopl) starts like this:
> 
> --- feta16.600.pl       Sun Aug 18 12:06:05 2002
> +++ feta16.1800.pl      Sun Aug 18 12:05:55 2002
> @@ -4,7 +4,7 @@
>  (DESIGNSIZE R 16.0)
>  (COMMENT DESIGNSIZE IS IN POINTS)
>  (COMMENT OTHER SIZES ARE MULTIPLES OF DESIGNSIZE)
> -(CHECKSUM O 16014264425)
> +(CHECKSUM O 36011723524)
>  (FONTDIMEN
>     (SLANT R 0.0)
>     (SPACE R 0.0)
> @@ -32,12 +32,12 @@
>     )
>  (CHARACTER O 4
>     (CHARWD R 0.15)
> -   (CHARHT R 0.223644)
> +   (CHARHT R 0.223632)
>     (CHARDP R 0.262501)
>     )
>  (CHARACTER O 5
>     (CHARWD R 0.15)
> -   (CHARHT R 0.223644)
> +   (CHARHT R 0.223632)
>     )
> ...
> 
> 
> It looks like an epsilon quantity being added or subtracted
> somewhere...


I know there is something fishy going on (I noticed the checksum
errors), but I don't really know how to find the culprit. Can you
investigate?

We have



        fet_beginchar("longa rest", "-2", "longarest");
               set_char_box(0, breve_rest_x#, breve_rest_y#, breve_rest_y#);
               draw_block ((0,-floor (breve_rest_y)), (breve_rest_x, floor 
(breve_rest_y)));
        fet_endchar;

        fet_beginchar("breve rest", "-1", "breverest");
               set_char_box(0, breve_rest_x#, 0, breve_rest_y#);
               draw_block ((0,0), (breve_rest_x, floor (breve_rest_y)));
        fet_endchar;

        breve_rest_y# =  staff_space#;
        breve_rest_x# = 3/5 staff_space#;

        staffsize#:=20pt#;  % feta20.mf

or is there a problem with equalities vs. assignments? 

(So, how much do we get for a bug in MF? :-) 

--
Han-Wen Nienhuys   |   address@hidden   |   http://www.cs.uu.nl/~hanwen