chris@MIMSY.UMD.EDU.UUCP (03/12/87)
Not only is PK format more compressed than GF format, it is also easier to read (!). This is mainly because GF format does not guarantee that its bounding boxes are minimal, while PK does: gyre% ls -l *font.c -rw-r--r-- 1 bin 9104 Nov 18 02:17 font.c -rw-r--r-- 1 bin 21464 Nov 18 02:08 gffont.c -rw-r--r-- 1 bin 17222 Nov 18 02:18 pkfont.c -rw-r--r-- 1 bin 6400 Nov 18 02:18 pxlfont.c -rw-r--r-- 1 bin 5202 Nov 18 02:17 tfmfont.c TFM `fonts' are the easiest to `interpret', since they have no pixels. (The TFM `font' exists mainly for invisble fonts for SLiTeX.) PXL fonts also require practially no interpretation. Of course, the size of a C source file that interprets a font fomrat is not a perfect measure of the format's complexity, but I think this gives a pretty good `feel'. (`font.c' implements my generic in-core format, which is not directly related to any disk format. These fonts are simply data objects that support font operations, including a function that returns a pointer to a pixel array. The functions themselves are, of course, in the implementation-specific files.) -- In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7690) UUCP: seismo!mimsy!chris ARPA/CSNet: chris@mimsy.umd.edu