Re: stupid question

[Graham Percival]

On Wed, Sep 22, 2010 at 9:02 PM, Jean-Charles Malahieude
<address@hidden> wrote:
> Before I go to bed, I would like to know what you mean with you commit
> 29bd6abd1 (Build: dist fix)
> TRANSLATORS: would stop stop screwing up like this?  :/
>
> As far as I remember, the extending manual has never been compiled since the
> docs have been revamped, and there is just one translated node in
> "programming interface" (Using Scheme code instead of tweak) that I have not
> checked for long.

This commit:
http://git.savannah.gnu.org/gitweb/?p=lilypond.git;a=commit;h=2fd29d6224d27c989d9bfab126dd7220fa4f0647

Put a file into: Documentation/fr/extending/

This was the first time a file had been put into that directory (I'm
sure you had to create the directory, but git doesn't track
directories).  When you did this, you:
- did not create a GNUmakefile.
- did not add that directory to the SUBDIRS of the GNUmakefile of the
higher-level directory.

As a result, the file was not included in the "make dist" targets,
which caused the dist-check to fail, which meant that I had to spend
another 30 minutes recompiling all the binaries before I could release
.34.


I was a bit inaccurate in that commit message -- I should have blamed
whoever merged the lilypond/translation branch, not necessarily the
translator.  I don't expect translators to know (or care) about the
odd aspects of our build system, but I *do* expect the translator to
know about this (at least, after I've yelled at him a few times).

Whenever a merge from lilypond/translation adds a new file to master,
be careful.  If it's adding another .itely file to a directory that
already has some, then there's no problem.  If it adds a .itely (or
.itexi) to a directory that *doesn't* have them, then you **must**
play around with GNUmakefiles.

Cheers,
- Graham