Werner LEMBERG wrote:
http://www.jonathankulp.com/stable-devel.pdf
BTW, to run stable and development versions in parallel, I do it a bit
differently:
. `stable': Configure and install as normal.
. `devel': Configure with
./configure --prefix=/usr/local/test
and install as normal.
Aha! An option at the configure stage. I never thought of that.
. If I want to use the development version, I then say
PATH=/usr/local/test/bin:$PATH lilypond ...
This part can be saved into a file, of course.
So it's the question what exactly you want to achieve if you say `use
the development version'. Leaving the binary directly in the
compilation directory probably makes it difficult to say `git pull'
without recompilation afterwards.
Yes, when I pull from git I check to see if any of the .cc files have
changed, and if so I recompile. This method of using two versions was
suggested to me by Carl Sorensen. I like it because after recompiling I
always have the latest changes without having to install or uninstall.
As long as I don't do "make clean" it works. I should probably explain
this in there. On my own machine I don't do this precise method (with
the "exec ~/lilypond..." file), I always call Lilypond using a personal
shell script, so I just use the command "~/lilypond/out/bin/lilypond" in
the script instead of "lilypond". My script has flags to set
showFirstLength and showLastLength at the command line, open the pdf in
a viewer, and so forth.
[Well, I always compile in a separate directory after cloning the
repository with `git clone -l -s . ../lilypond.compiled' so that `git
pull' in the original git repository is safe too -- if I want to
synchronize the compilation directory I simply say `git pull' there
too.]
Ah, I think I see why you do it this way now. The only development I do
is on the documentation, so I don't need to keep a separate repo of the
program files.
Maybe I should add your method as well? It would be easy enough to do.
Could you write a sentence or two that I can copy in there explaining
the advantage of this method from a dev's point of view? I think I
understand the advantage, but I don't want to get it wrong.
Thanks for looking at this, Werner.
Jon
--
Jonathan Kulp
http://www.jonathankulp.com
_______________________________________________
lilypond-devel mailing list
lilypond-devel@gnu.org
http://lists.gnu.org/mailman/listinfo/lilypond-devel
