Re: Issue 1933 in lilypond: Lilypond-book requires msvcrt again

[lilypond]

Comment #18 on issue 1933 by address@hidden: Lilypond-book requires  
msvcrt again
http://code.google.com/p/lilypond/issues/detail?id=1933

> Second, we could always just rewrite the python scripts.  And  
> by "rewrite", I mean "add maybe 10 lines of python to use os.popen3  
> instead of subprocess, as discussed at the top of  
> python/book_snippets.py".

It seems like the consensus above is that this is a bad idea.  I don't have  
an opinion either way, but if we are to do this, we need to have patchy  
automatically reject all patches that add an unsupported Windows module to  
a Python patch.

> Third, we could seriously examine Jan's patch to determine why it worked  
> at the time, and what has changed since then.

If this took not-too-much-time I'd agree, but it does not seem that we have  
the developer resources or desire to go through the fix-test-fix-test cycle  
that's needed to fix this.  And once it is fixed, we have to ask  
ourselves "when the same snag comes up later down the line, will the people  
be around who can get it fixed in a reasonable time frame."

I think that the question to ask is "will the inconvenience of asking users  
to download Python separately be outweighed by the gain in developer time  
and the diminishing of delay between stable resources?"  It is true that if  
the answer is "yes" we will have in effect given up on having Grand Unified  
Binaries of python, but there is little value in having them if the distro  
on the Python website is easily accessible and if writing those 10 lines of  
Python code has now taken us since September 25.