(Sorry for double-posting, but my first message didn't show up yet, maybe it's lost. This one is corrected anyway.) As far as I skimmed the discussions of the last days, this didn't show up yet: In my LilyPond module I call the program like so: \def\LP{mtxrun --ifchanged="\lily!filename.tmp" --exec bin:lilypond - dbackend=eps -dno-gs-load-fonts -dinclude-eps-fonts -ddelete- intermediate-files "\lily!filename.tmp"} \def\PDF{mtxrun --ifchanged="\lily!filename.eps" pstopdf "\lily! filename.eps"} \executesystemcommand{\LP} \executesystemcommand{\PDF} The former definition results in this call: mtxrun --ifchanged="c_whatshallwedo-lilypond-1.tmp" --exec bin:lilypond -dbackend=eps -dno-gs-load-fonts -dinclude-eps-fonts - ddelete-intermediate-files "c_whatshallwedo-lilypond-1.tmp" The .tmp file exists and contains the right LilyPond code. But I get an error like this from the LP call: ############################################## .../texmf/tex/texmf-osx-intel/bin/mtxrun:1953: attempt to call field 'HEXsum' (a nil value) ############################################## Line 1953 is: return md5.HEXsum(data) Looks like a bug in mtxrun to me. Lua MD5 docs at http://www.keplerproject.org/md5/manual.html#lua_api say, there'd be only "sum" and "sumhexa" methods. Replacing "HEXsum" with "sum" works (.md5 files are written), "sumhexa" gives the same error as above. But it seems LilyPond is called now every time, ignoring the checksum, so simply "sum" at that one place isn't enough. Further, the PDF call gives: MTXrun | file 'c_whatshallwedo-lilypond-1.eps' is unchanged Even if the .eps file doesn't exist at all. Greetlings from Lake Constance! Hraban --- http://www.fiee.net/texnique/ http://wiki.contextgarden.net https://www.cacert.org (I'm an assurer)