On Oct 16, 2010, at 3:02 AM, Richard Shann
<address@hidden>
wrote:
Jeremiah, Nils
I see that you had to re-instate lyparserfuncs.h, presumably
because I
forgot to delete includes of it from the various files when I
removed
the old parser.
Yeah. I added into the tarball. I could not compile the tarball
without including it. If it's supposed to left out I will change
that
(making sure it compiles) then and remove it from git.
The puzzle then is, how did clean builds work - in
particular windows gub?
Pulling from git provides the file. A tarball created with make dist
does not because the file was not mentioned in src/Makefile.am.
Jeremiah
I would like to avoid re-building the windows
executable, as it will need re-testing. I am not proposing that
we be
impossibly fussy here - there are no actual changes of source code
here,
just how you build from source. Executables are unaltered.
So if you can verify that the sources labelled 0.8.20 do build
correctly
on both platforms we are ready to release on Wednesday next, with
the
already-built-and-tested windows executable. (This executable
will not
be different if re-built, since the include file must have been
present,
and it would in any case not affect the resultant binary).
I have some very good progress to check in now if you can move
master on
to 0.8.21, Anacrusis displaying properly, Breve, Longa and whole
measure
rests, even plain chant all doing the right thing in the Denemo
display
*and* the LilyPond!
Richard
_______________________________________________
Denemo-devel mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/denemo-devel