Re: make doc

[Carl Sorensen]

On 5/30/11 12:56 PM, "Phil Holmes" <address@hidden> wrote:

> ----- Original Message -----
> From: "Graham Percival" <address@hidden>
> To: "Phil Holmes" <address@hidden>
> Cc: "Carl Sorensen" <address@hidden>; <address@hidden>
> Sent: Saturday, May 28, 2011 6:22 PM
> Subject: Re: make doc
> 
> 
>> On Sat, May 28, 2011 at 05:48:39PM +0100, Phil Holmes wrote:
>>> 
>>> So the command line would need to be something like
>>> lilypond filename.ly > filename.log?
>> 
>> Something like that, probably.  Dunno.  What happens when you try
>> that command on the command-line?
>> 
>> Remember the first rule of build systems: a build system just sets
>> up the environment and then runs command-line commands.  (no wait,
>> that's the second rule.  The first rule is that all build systems
>> suck; most of them suck exponentially.)
>> 
>>> Will have to look at how that could be done, since the times lily is
>>> called that I've so far found are initiated by lilypond-book.
>> 
>> Well, figure out how to do it with lilypond by itself, then figure
>> out how to do it with lilypond-book, then figure out how to do it
>> inside the build system.
>> 
>> Cheers,
>> - Graham
>> 
> 
> OK - so I now know how to redirect the output from lilypond-book.py when
> it's making, for example, collated-files.tely - I can redirect it to
> collated-files.log.  FWIW simply doing the redirect on
> /input/regression/out-www/collated-files.tely creates a logfile of 30 Meg
> and 68,000-odd lines!  I think this is a good first step.  For those who are
> interested in the build system, I attach my notes to date of how make doc
> works - slightly stream-of-consciousness, I'm afraid.

Phil, I'd like to see this dumped in the CG, even in its current
stream-of-consciousness form.  I'll put it in if you'd like.

Thanks,

Carl