Jan Nieuwenhuizen wrote:
> On Saturday, 11 March 2000, Alain CULOS writes:
>
> You could try using postscript output:
>
> lilypond -f ps foo.ly
> gv foo.ps
That's not working at all (failing to display anything).
Whereas going the normal path ly->dvi then dvi->ps works beautifully, though
I don't read the dvi fully (some missing fonts), the resulting ps is perfect (or
appears to be so).
[ly2dvi -P xxx] works though.
Going direct to ps as advised above gets me the following error log (gv) :
=========
...
Displaying DSC file c:/Utilisateurs/Alain/Documents/Musique/Fanchon.ps without
pages
Can't find (or can't open) font file feta-nummer10.
Scanning c:\psfonts for fonts... 0 files, 0 scanned, 0 new fonts.
Can't find (or can't open) font file feta-nummer10.
Substituting font Courier for feta-nummer10.
Loading NimbusMonL-Regu font from
C:\PROGRAM\GSTOOLS\gs5.50\fonts\n022003l.pfb... 1977116
623474 1449714 161896 1 done.
Can't find (or can't open) font file cmti10.
Substituting font Courier for cmti10.
Can't find (or can't open) font file feta20.
Substituting font Courier for feta20.
...
Can't find (or can't open) font file feta-din10.
Substituting font Courier for feta-din10.
Can't find (or can't open) font file feta20.
Substituting font Courier for feta20.
Unrecoverable error: syntaxerror in %ztokenexec_continue
Operand stack:
--nostringval-- --nostringval-- --nostringval--
--- Begin offending input ---
--- End offending input ---
file offset = 65536
gsdll_execute_cont returns -18
=========
> > OK, but yap complains, here's part of the log book :
> That's not so difficult:
> > - C:\program\texmf\miktex\bin\makemf.exe --verbose feta-nummer10
> > Cannot find feta-nummer source file.
>
> Miktex can't find the feta-nummer10.mf file. Try setting something like
> this:
>
> set path=c:\texmf\miktex\bin;%path%
> set TEXINPUTS=%TEXINPUTS%;c:\lilypond-1.2.0\tex
> set MFINPUTS=%MFINPUTS%;c:\lilypond-1.2.0\mf
That was done and I re-checked it, it looks good. The problem remains ...
But I think the problem is actually there - I discovered yesterday that my
autoexec.bat requires too much environment space though it was fine the boot
before (I must have added a few lines) :-((( silly system. When I launch yap
from bash, there is no problem (after one interation with initexmf -u).
Thanks for your help. By the way, the problem was not space related (in the
directory name).
> You should check the path where 'feta20.mf' lives (MFINPUTS), and
> `lilyponddefs.tex' (TEXINPUTS).
Yes, I checked that and it looks OK too.
> > I'd appreciate if someone could help or point me to the right doc.
> Have a look over here:
>
>http://www.cs.uu.nl/~hanwen/lilypond/Documentation/programmer/out-www/README-W32.html
Geez, that's more of a helper for the programmer than the user, I don't really
want to go into that (I already overbooked myself).
> In DOS and
> UNIX, spaces are traditionally used to separate files in a list of
> filenames. Consider these two commands:
>
> c:\> copy foo.ly bar.ly c:\lilypond-1.2
> c:\> copy foo.ly bar.ly c:\program files\lilypond-1.2
>
> What does the latter mean?
Well, you can always argue about this one, so just in brief :
Why is there quoting in Unix if not to include special characters, e.g. spaces ?
DOS does not cope with much but Windows DOS copes with a little more and I use
bash anyway and quoting makes it work most often although not always.
> > Queries :
> > 1/
> > Format conversion :
> > from filename extensions .pmx .mtx .mus .mts
> > conversion to .ly and/or .midi ?
> > Preferred answer would point to portable source code and/or win32 binaries.
> > Other answers could point to documented standards for the above formats.
>
> Don't know. Extensions are ambiguous.
I know, Unix despite all its advantages is bad in this respect. I think good
naming conventions help users better organise and learn more about their
computer.
OK, it would be better still if recognising a format only relied on the contents
of the file (with some magick number or initial string) which is what is used in
some cases in the Unix world but not consistently enough, at least the windows
world is pretty consistent on the extensions because it relies on it.
> > 6/
> > Give sample for altering margins (to include longer instrument names on
> > the left for instance), I'd love that.
>
> try:
> \paper{
> indent = 6.0\cm;
> linewidth= 15\cm;
> }
>
> You should check out ly2dvi, it handles margins.
Well, I meant it would be nice to give an example in the docs.
Also my question was not accurate enough, there seems to be huge flexibility for
left and right margins, but what about the top and bottom ones ?
> > 9/
> > Also in midi2ly, would it be a big task to correct (depending on an
> > option) the length of notes since manual play is rather not exact and
> > assume less skips than detected ? And also display rests rather than
> > skips when these are large enough and comparable in length to notes
> > prior or following the skip/rest ?
>
> You can choose different types of quantisation, eg
> midi2ly --smallest=16
> will not produce shorter notes than sixteenths.
I have seen that alright, but. If you have a piece with notes like full notes
and half notes,
it does not make sense to me that it should convert (if in fool/bullet proof
mode, e.g. dummy
mode) to 19/32's of something or other followed by some wierd skip, just round
to the closest
match and forget about the skip and put proper rests when a real skip is
recognised. I know
this is not easy at all but it would be so nice.
Thanks again for all replies (very helpful),
Regards,
Alain.
